Multiplexed Fluorescence Imaging: A Guide to High-Plex Biomarker Analysis for Drug Development and Clinical Research

Abstract

This article provides a comprehensive overview of modern multiplexed fluorescence imaging, a transformative technology enabling the simultaneous visualization of numerous biomarkers within their native tissue context. Aimed at researchers, scientists, and drug development professionals, we explore the foundational principles of fluorophores and imaging techniques, including novel time-resolved fluorescent proteins and cyclic immunofluorescence. The piece delves into cutting-edge methodological applications in spatial biology and clinical tumor immunology, offers practical guidance for troubleshooting and optimizing protocols and conjugation chemistries, and concludes with a comparative analysis of leading platforms to guide technology selection. By synthesizing recent advances, this review serves as a strategic resource for leveraging high-plex imaging to unravel complex biological systems and enhance therapeutic development.

Core Principles and Emerging Fluorophores in Multiplexed Imaging

Fluorescence forms the cornerstone of modern biological research and drug development, enabling the visualization and quantification of biomolecules in living systems with exquisite sensitivity and specificity [1] [2]. The fundamental photophysical process—where a fluorophore absorbs high-energy light and emits lower-energy light—provides a powerful mechanism for probing cellular environments, tracking therapeutic compounds, and understanding complex biological interactions [3] [2]. For researchers investigating complex biological systems, the ability to simultaneously track multiple targets through multiplexed fluorescence imaging has become increasingly vital [4] [5]. This application note details the core principles of fluorescence, quantitative parameters researchers must control, and advanced methodologies for successful multiplexed experiments, providing a foundational framework for applying these techniques in pharmaceutical development and basic research.

The Fluorescence Process: A Photophysical Journey

The fluorescence process occurs through a precise sequence of photophysical events that transform absorbed light into emitted light, typically represented by a Jablonski energy diagram [2].

The fluorescence cycle comprises three essential stages. In Stage 1 (Excitation), a photon of energy (hνEX) from an external source such as a laser or lamp is absorbed by the fluorophore, creating an excited electronic singlet state (S'1) [2]. This process distinguishes fluorescence from chemiluminescence, where the excited state is populated chemically rather than by light absorption [2].

During Stage 2 (Excited-State Lifetime), the excited state exists for a finite time (typically 1-10 nanoseconds), where the fluorophore undergoes conformational changes and interacts with its molecular environment. These interactions cause partial energy dissipation, yielding a relaxed singlet excited state (S1) from which fluorescence emission originates. Not all excited molecules return to the ground state via fluorescence; other processes like collisional quenching, fluorescence resonance energy transfer (FRET), and intersystem crossing may also depopulate S1 [2].

In Stage 3 (Fluorescence Emission), a photon of energy (hνEM) is emitted, returning the fluorophore to its ground state S0. Due to energy dissipation during the excited-state lifetime, this emitted photon has lower energy and longer wavelength than the excitation photon. This energy difference, known as the Stokes shift, is fundamental to fluorescence sensitivity as it allows emission photons to be detected against a low background, isolated from excitation photons [3] [2].

Visualizing the Process: Jablonski Diagram

The following diagram illustrates the energy transitions during fluorescence, including vibrational relaxation, fluorescence emission, and competing processes like non-radiative decay and intersystem crossing.

Diagram 1: Jablonski energy diagram illustrating the photophysical processes in fluorescence, including excitation, vibrational relaxation, fluorescence emission, and competing pathways.

Quantitative Parameters in Fluorescence Spectroscopy

The utility of fluorophores in research applications depends on several key quantitative parameters that determine their brightness, stability, and detectability in biological systems.

Core Photophysical Properties

Table 1: Fundamental Properties of Fluorescent Dyes [2]

Property	Definition	Significance in Research
Extinction Coefficient (EC)	Capacity for light absorption at a specific wavelength (units: cm⁻¹ M⁻¹)	Fluorescence brightness per fluorophore is proportional to EC × Quantum Yield
Fluorescence Quantum Yield (QY)	Number of fluorescence photons emitted per excitation photon absorbed	Determines fluorophore efficiency; values range from 0.1-0.9 for commercial dyes
Excitation Spectrum	Plot of excitation wavelength vs. number of fluorescence photons generated	Identifies optimal excitation wavelengths for maximum signal output
Emission Spectrum	Plot of emission wavelength vs. number of fluorescence photons generated	Enables spectral discrimination for multiplex detection and background reduction
Stokes Shift	Energy/wavelength difference between excitation and emission maxima	Allows emission isolation from excitation light, fundamental to detection sensitivity
Photobleaching	Irreversible destruction of excited fluorophores due to reactive oxygen species	Limits observation time; depends on exposure duration and intensity

These parameters collectively determine a fluorophore's effectiveness in specific applications. For instance, the brightness of a fluorophore—which determines its detectability—is calculated as the product of its extinction coefficient and quantum yield [2]. Environmental factors including solvent viscosity, ionic concentrations, pH, and hydrophobicity can profoundly affect both fluorescence intensity and excited-state lifetime [3].

Representative Fluorophore Properties

Table 2: Experimental Quantum Yields of Selected Fluorophores [3]

Compound	Solvent	Excitation Wavelength (nm)	Emission Wavelength (nm)	Quantum Yield
Acridine Orange	Ethanol	493	535	0.46
Benzene	Ethanol	248	300-350	0.04
Chlorophyll-A	Ethanol	440	685	0.23
Eosin	Water	521	544	0.16
Fluorescein	Water	437	515	0.92
Rhodamine-B	Ethanol	555	627	0.97

The data in Table 2 illustrates the substantial variation in quantum yields across different fluorophores, with Rhodamine-B and Fluorescein exhibiting particularly high efficiency. This makes them valuable for applications requiring high signal-to-background ratios. When designing experiments, researchers should note that quantum yields below 1 result from energy loss through non-radiative pathways like heat or photochemical reactions, rather than the re-radiative pathway of fluorescence [3].

Multiplexed Fluorescence Imaging: Challenges and Advanced Solutions

Multiplexed fluorescence imaging—using multiple fluorescent dyes to examine various elements within a sample—has become essential for characterizing complex biological processes and disease states [5]. However, implementing effective multiplexing strategies presents significant technical challenges that require sophisticated solutions.

The Spectral Overlap Challenge

A primary obstacle in multiplexed imaging is signal crosstalk caused by the broad emission spectra of most fluorophores [5]. When using two or more fluorophores, their emission spectra frequently overlap, causing signals to bleed into detection channels intended for other fluorophores. This can yield false positives or negatives and otherwise obscure data if not properly accounted for [5].

Traditional approaches to minimize crosstalk involve careful selection of fluorophore combinations with minimal spectral overlap and using optical filters with restricted bandpasses centered around peak emission wavelengths of individual fluorophores [5]. However, these approaches involve trade-offs—narrow filter bandwidths provide higher specificity but discard more of the desired signal, reducing detection efficiency [5].

Advanced Unmixing Methodologies

To overcome spectral overlap limitations, researchers have developed several computational unmixing techniques:

PICASSO (Process of ultra-multiplexed Imaging of biomoleCules viA the unmixing of the Signals of Spectrally Overlapping fluorophores) enables more than 15-color imaging of spatially overlapping proteins in a single imaging round without using reference emission spectra [4]. This blind unmixing approach uses information theory, iteratively minimizing mutual information between mixed images based on the principle that spectral mixing increases mutual information between channels [4]. PICASSO requires only an equal number of images and fluorophores, making it compatible with standard bandpass filter-based microscopy systems [4].

Bleaching-Assisted Multichannel Microscopy (BAMM) leverages photobleaching behavior as an identifying property, discriminating between fluorescent species by their characteristic photobleaching rates [6]. This approach uses non-negative matrix factorization to extract spectral and photobleaching characteristics from timelapse data of sample bleaching, potentially tripling multiplexing capabilities without additional hardware [6].

Phasor-based analysis provides a powerful method for unmixing multiple fluorophores without reference spectra by transforming spectral data into phasor plots for rapid separation of signals [5]. This approach simplifies workflow by eliminating extensive calibration needs and can separate background autofluorescence from specific signals [5].

The following diagram illustrates a generalized workflow for multiplexed fluorescence imaging, incorporating both sample preparation and computational unmixing steps.

Diagram 2: Experimental workflow for multiplexed fluorescence imaging, showing sample preparation and computational unmixing steps with algorithm options.

Experimental Protocols

Protocol: PICASSO for Ultra-multiplexed Imaging

This protocol enables 15-color multiplexed imaging of spatially overlapping proteins in a single staining and imaging round, adapted from the PICASSO methodology [4].

Research Reagent Solutions & Materials Table 3: Essential Materials for PICASSO Multiplexed Imaging

Item	Function/Application
Spectrally overlapping fluorophores	Must be excitable by the same laser source; select from commercial organic dyes
Primary antibodies	Target proteins of interest; from same host species permitted
Fab fragments of secondary antibodies	Conjugated to fluorophores; form complexes with primary antibodies
Assembly buffer	Compatible with antibody-Fab complex formation
Mounting medium	Preserves sample integrity during imaging
Reference standards	For calibrating fluorescence measurements between experiments

Procedure

• Fluorophore Selection: Choose 15+ spectrally overlapping fluorophores that can be strongly excited by the same laser source. Include large Stokes shift fluorophores to enhance multiplexing capability [4].

• Antibody Complex Formation: Use primary antibody-Fab complex preformation technique to bypass host species limitations [4]:

  • Assemble each primary antibody with Fab fragment of secondary antibody bearing specific fluorophore
  • Combine all assembled antibody complexes together

• Sample Staining:

  • Apply the mixture of antibody complexes to specimen simultaneously
  • Incubate according to standard immunohistochemistry protocols
  • Wash to remove unbound complexes

• Image Acquisition:

  • Acquire images at different spectral ranges, each containing emission peak of specific fluorophore
  • Use standard bandpass filter-based microscopy system
  • Ensure number of acquired image channels equals number of fluorophores used

• Computational Unmixing with PICASSO:

  • Initialize channel solutions as acquired mixed images
  • Iteratively update solutions by minimizing mutual information between channels
  • Apply non-negativity constraints to ensure physically plausible results
  • Continue iterations until convergence criteria met (minimal change in mutual information)

• Validation:

  • Verify unmixing accuracy through control samples with known staining patterns
  • Compare with sequential staining results if available

Technical Notes: PICASSO performs particularly well in highly heterogeneous specimens like brain tissue where emission spectra show significant regional variation [4]. The algorithm's effectiveness stems from the assumption that spatial distribution of different proteins is mutually exclusive, enabling accurate information unmixing through mutual information minimization [4].

Protocol: BAMM (Bleaching-Assisted Multichannel Microscopy)

This protocol uses photobleaching characteristics to distinguish fluorescent labels, enabling resolution of up to three fluorescent labels in a single spectral channel [6].

Procedure

• Sample Preparation: Label cellular targets with fluorescent species showing distinct photobleaching characteristics, even with similar emission spectra.

• Timelapse Acquisition:

  • Place sample on confocal or widefield fluorescence microscope
  • Record timelapse movie during continuous illumination
  • Use standard filter sets compatible with fluorophore excitation
  • Continue acquisition until significant photobleaching occurs

• Pre-processing:

  • Perform background subtraction
  • Apply drift compensation algorithms to correct for sample movement

• Unmixing Algorithm Application:

  • For samples with isolated fluorophores: Use Non-Negative Least Squares (NNLS) with manually identified bleaching fingerprints
  • For complex cellular samples: Apply Non-negative Matrix Factorization (NMF) or Non-increasing NMF (NI-NMF) to simultaneously estimate bleaching characteristics and abundances
  • Use principal components of photobleaching curves as initial estimates

• Validation:

  • Compare results with known spectral properties of fluorophores
  • Verify biological plausibility of unmixed distributions

Technical Notes: BAMM is particularly effective for organic dyes, autofluorescent biomolecules, and fluorescent proteins with distinct photobleaching rates [6]. The technique can be combined with spectral unmixing for enhanced multiplexing capability [6].

Reference Standards & Analysis Tools

QUEL-QAL Library: An open-source Python library designed to streamline quantitative analysis of fluorescence images using solid reference targets [7] [8]. This tool provides modular, reproducible workflows for key metrics including response linearity, limit of detection, depth sensitivity, and spatial resolution in alignment with AAPM TG311 and FDA guidelines [7].

Fluorescence SpectraViewer: Interactive online tools for plotting and comparing fluorescence excitation and emission spectra for over 250 fluorophores, essential for experimental planning and fluorophore selection [2].

Reference Standards: Fluorescent microsphere standards for microscopy and flow cytometry, plus ready-made fluorescent standard solutions for spectrofluorometry, enabling calibration of measurements across different instruments and timepoints [2].

Antifade Reagents for Photobleaching Mitigation

Table 4: Properties of Common Antifade Reagents [3]

Reagent	Application Effectiveness	Stability & Safety Considerations
p-phenylenediamine	Most effective for FITC; also works for Rhodamine	Light-sensitive (blackens with exposure); extremely dangerous skin contact
DABCO	Highly effective for FITC; slightly lower efficacy than p-phenylenediamine	More light-stable; higher safety profile
n-propylgallate	Most effective for Rhodamine; also effective for FITC	Prepare as 1% solution in glycerol/PBS
2-mercapto-ethylamine	Used for chromosome/DNA specimens stained with propidium iodide, acridine orange, or Chromomycin A3	Prepare as 0.1mM solution in Tris-EDTA buffer

Application in Drug Development & Research

Advanced fluorescence techniques have revolutionized drug visualization in living systems. At the subcellular level, super-resolution microscopy enables exploration of the molecular landscape within individual cells and cellular responses to drugs [1]. When integrated with optical near-infrared II imaging, researchers can study complex spatiotemporal interactions between drugs and their surroundings across multiple biological scales [1].

The combination of these visualization approaches provides supplementary information on physiological parameters, metabolic activity, and tissue composition, leading to a comprehensive understanding of drug behavior [1]. For pharmaceutical researchers, these techniques enable tracking of drug distribution, target engagement, and therapeutic effects from subcellular to organismal levels, potentially redefining our understanding of pharmacology through analysis of drug micro-dynamics in subcellular environments [1].

Multiplexed fluorescence imaging particularly benefits immuno-oncology research, where studying immune cell locations relative to key biomarkers is essential, and neuroscience applications, where understanding neuronal synapse complexity requires simultaneous visualization of multiple proteins [5]. By enabling parallel observation of related components and processes, multiplexed fluorescence adds crucial context to observations, providing more meaningful results and revealing interdependencies that might otherwise be missed [5].

Fluorescence imaging has moved far beyond the simple visualization of protein localization. The field is now defined by the sophisticated engineering of both synthetic dyes and genetically encoded reporters, a synergy that enables the simultaneous monitoring of multiple biological processes in live cells—a capability central to advanced multiplexed imaging research [9].

This expansion addresses a fundamental challenge in systems biology: biological processes are not isolated but are the result of coordinated networks of signaling proteins, second messengers, and metabolites. While individual biosensors have provided profound insights, they cannot capture the complex, dynamic interactions between different molecular players [9]. The drive to overcome the spectral limitations of conventional microscopy has fueled innovations that separate signals not just by color, but also by time, space, and sophisticated computational unmixing. This document details the key reagents, quantitative data, and experimental protocols that form the modern toolkit for multiplexed fluorescence imaging.

The Scientist's Toolkit: Research Reagent Solutions

The following table catalogues essential materials used in the development and application of advanced fluorescent probes.

Reagent/Solution	Function/Description
HaloTag/SNAP-tag [9]	Self-labeling protein tags that covalently bind to synthetic fluorophores, enabling the use of bright, photostable organic dyes with genetically targeted specificity.
EMcapsulins [10]	Genetically encoded nanocompartments that accumulate heavy metals for correlative electron microscopy, providing multiplexable genetic labels for ultrastructural analysis.
Time-Resolved FPs (tr-FPs) [11]	A family of fluorescent proteins engineered to exhibit distinct fluorescence lifetimes, enabling multiplexing in a single spectral channel via Fluorescence Lifetime Imaging (FLIM).
Chemigenetic Biosensors [12] [9]	Hybrid biosensors combining a genetically encoded protein scaffold (sensing unit) with a synthetic fluorophore (reporting unit), offering enhanced optical properties and design flexibility.
SEPARATE Computational Unmixing [13]	A machine learning method that unmixes signals from two proteins labeled with the same fluorophore based on their distinct 3D spatial expression patterns, effectively doubling multiplexing capacity.
NIR-II Protein-Seeking Dyes [14]	Synthetic dyes that covalently bind to endogenous proteins like albumin, creating biomimetic NIR-II fluorescent complexes for high-contrast, deep-tissue imaging.
14-Dehydrobrowniine	14-Dehydrobrowniine, MF:C25H41NO7, MW:467.6 g/mol
NADP disodium salt	NADP disodium salt, CAS:24292-60-2, MF:C21H26N7Na2O17P3, MW:787.4 g/mol

Quantitative Comparison of Fluorescent Reporters

The selection of an appropriate reporter requires careful consideration of its photophysical properties. The tables below summarize key performance metrics for synthetic dyes and genetically encoded reporters.

Table 1: Properties of Advanced Synthetic Dyes and Chemigenetic Systems

Dye/System Class	Excitation/Emission (nm)	Key Properties	Primary Applications
BODIPY Dyes [15]	500-700 (tunable)	High quantum yield (>0.8), strong extinction coefficients, exceptional photostability.	Targeted cellular imaging (e.g., folate-conjugated for cancer imaging).
NIR-II Dye (CO-1080) [14]	1044/1079	Requires protein binding for brightness.	Deep-tissue bioimaging, lymphography, angiography.
HSA@CO-1080 FPs [14]	1044/1079	22-fold fluorescence enhancement upon binding HSA; improved photostability & biocompatibility.	Multicolor deep-tissue imaging with 1064 nm excitation.

Table 2: Properties of Genetically Encoded Fluorescent Reporters

Reporter Class	Example	Key Properties	Readout Mechanism
Single-FP Biosensors [12] [9]	GCaMP8 (Ca2+), KTRs (Kinase Activity)	Excitation-ratiometric; large dynamic range; improved kinetics (GCaMP8: ms transients).	Change in fluorescence intensity or subcellular localization.
FRET-Based Biosensors [12] [9] [16]	AKAR (PKA Activity), RasAR (Ras Activity)	Sensitive to conformational changes; measures molecular interactions <10 nm.	Change in FRET efficiency between donor and acceptor FPs.
Time-Resolved FPs (tr-FPs) [11]	tr-mNeonGreen, tr-mScarlet	Fluorescence lifetime (1-5 ns) as a distinguishable parameter.	FLIM for multiplexing beyond spectral limits.
EMcapsulins [10]	1M-QtFLAG, 2M-MxFLAG	6 distinct classes; readable by TEM/SEM; automatic semantic segmentation possible.	Genetically encoded EM contrast for correlative microscopy.

Experimental Protocols

Protocol 1: Multiplexed Live-Cell Imaging with tr-FPs and FLIM

This protocol enables simultaneous imaging of up to nine cellular targets by leveraging the distinct fluorescence lifetimes of tr-FPs [11].

• Construct Preparation: Clone genes of interest (GOIs) as fusion proteins with selected tr-FPs (e.g., tr-mNeonGreen, tr-mScarlet) using standard molecular biology techniques. Ensure tr-FPs within the same spectral channel have well-separated lifetimes.
• Cell Preparation and Transfection:
  • Culture mammalian cells (e.g., HEK293T, HeLa) in appropriate media.
  • Transfect with tr-FP fusion constructs using a preferred method (e.g., lipofection, electroporation). For multiple constructs, optimize DNA ratios to achieve balanced expression.
  • Seed transfected cells onto glass-bottom imaging dishes and culture for 24-48 hours before imaging.
• FLIM Data Acquisition:
  • Use a confocal or multiphoton microscope equipped with a FLIM module (time-correlated single photon counting, TCSPC, is recommended).
  • Set excitation lasers to the appropriate wavelengths for the tr-FPs used.
  • Acquire images at a resolution sufficient for the biological question. Collect a sufficient number of photons per pixel (typically >1000) for robust lifetime fitting.
• Data Analysis and Unmixing:
  • Fit the fluorescence decay curve at each pixel to a multi-exponential model using the microscope's FLIM software or open-source tools.
  • Assign a false color to each lifetime value corresponding to different tr-FPs.
  • Generate a composite lifetime map where each color represents a different target protein, enabling visualization of multiple structures or activities within a single cell.

Protocol 2: Creating Biomimetic NIR-II Fluorescent Proteins

This protocol describes the creation of bright, biocompatible NIR-II fluorescent proteins through the covalent binding of synthetic dyes to a protein shell like Human Serum Albumin (HSA) [14].

• Reaction Setup:
  • Prepare a solution of HSA in phosphate-buffered saline (PBS), pH 7.4, at a concentration of 10 µM.
  • Add the protein-seeking NIR-II dye (e.g., CO-1080) from a concentrated DMSO stock to the HSA solution at a 1:1 molar ratio (10 µM final concentration). Ensure the final DMSO concentration is <1% to avoid protein denaturation.
• Incubation and Conjugation:
  • Incubate the reaction mixture at 60 °C for 2 hours in the dark with gentle agitation. This temperature and time have been optimized for maximum yield.
• Purification:
  • Allow the solution to cool to room temperature.
  • Use ultrafiltration centrifugal devices (e.g., 10 kDa molecular weight cut-off) to separate the HSA@Dye conjugate from unreacted free dye. Centrifuge according to the manufacturer's instructions.
  • Wash the retentate with PBS three times to ensure complete removal of free dye.
• Characterization and Storage:
  • Measure the absorption and fluorescence spectra of the purified HSA@Dye to confirm successful formation.
  • The conjugate can be stored in PBS at 4 °C for short-term use or at -20 °C for long-term storage.

Protocol 3: Volumetric Multiplexed Imaging with SEPARATE

This protocol reduces the number of staining cycles required for 3D multiplexed imaging by pairing proteins and unmixing their signals computationally [13].

• Feature Extraction and Pairing:
  • Acquire high-resolution 3D immunofluorescence images of each protein of interest individually stained in control tissue samples.
  • Process these images through the feature extraction network (a convolutional neural network trained with contrastive learning) to generate feature-based distances between all proteins.
  • Identify the optimal protein pairs by maximizing the minimum feature-based distance between paired proteins.
• Sample Staining and Imaging:
  • For each pair of proteins (α and β), stain the experimental tissue sample with three fluorophore-conjugated antibodies:
    • Fluorophore A: Anti-protein α
    • Fluorophore B: Anti-protein β
    • Fluorophore C: A cocktail of anti-protein α and anti-protein β
  • Perform a single round of volumetric imaging (e.g., using a confocal microscope) to capture three channels: the signal of protein α, protein β, and the mixed signal.
• Training Data Generation and Network Training:
  • Generate a synthetic training dataset by linearly combining the single-protein images (α and β) with random ratios to simulate the mixed signal under various experimental conditions.
  • Train the protein separation network using the synthetic mixed images as input and the corresponding single-protein images as targets.
• Signal Unmixing:
  • Input the experimentally acquired mixed channel image into the trained protein separation network.
  • The network will output two unmixed images, each containing the predicted signal for one of the paired proteins.

Visualizing Multiplexed Imaging Strategies

The following diagrams illustrate the core concepts and workflows behind advanced multiplexing techniques.

Multiplexing strategies move beyond color to use time and spatial patterns for separating signals.

Biosensor designs and hybrid probes combine genetic targeting with synthetic chemistry for enhanced performance.

The development of time-resolved fluorescent proteins (tr-FPs) represents a transformative advancement in fluorescence microscopy. A recent landmark study introduces a family of tr-FPs with rationally controlled fluorescence lifetimes, enabling unprecedented capabilities in live-cell imaging and quantitative biology [17]. This technology addresses a critical limitation in conventional fluorescence microscopy, which primarily relies on steady-state intensity signals, by unlocking the underexplored dimension of fluorescence lifetime [17].

For researchers engaged in multiplexed imaging and drug development, tr-FPs provide a solution to the spectral overlap problem that constrains conventional fluorescence techniques. By engineering proteins with distinct lifetimes despite spectral similarity, this approach enables simultaneous monitoring of multiple cellular components and processes, thereby facilitating a more comprehensive understanding of system complexity with quantitative accuracy [17].

Key Properties of Time-Resolved Fluorescent Proteins

Spectral and Lifetime Characteristics

The engineered tr-FPs cover the visible spectrum while exhibiting a wide range of fluorescence lifetimes. This orthogonal control over spectral and temporal properties is crucial for expanding multiplexing capability. Researchers can now design experiments combining multiple tr-FPs based on both color and lifetime separation, dramatically increasing the number of simultaneously imageable targets.

Quantitative Properties of tr-FPs:

Protein Type	Excitation Maximum (nm)	Emission Maximum (nm)	Fluorescence Lifetime Range	Primary Applications
Blue Fluorescent Proteins	~380 [18]	440-470 [18]	Not specified in results	FRET pairs, multicolor labeling [18]
Cyan Fluorescent Proteins	~436 [18]	470-500 [18]	Not specified in results	FRET biosensors with YFP [18]
Engineered tr-FPs	Visible spectrum coverage [17]	Visible spectrum coverage [17]	Wide range, rationally controlled [17]	Temporal-spectral resolved microscopy, multiplexed super-resolution [17]

Performance Metrics for Biological Imaging

The development of tr-FPs focuses on optimizing key performance parameters essential for biological applications. These include:

• Brightness and Photostability: Critical for time-lapse imaging and long-term observation of cellular dynamics.
• Maturation Efficiency: Ensures rapid and complete fluorophore formation for accurate temporal analysis.
• Monomeric Character: Essential for fusion protein construction without perturbing native protein localization or function.

Experimental Applications and Workflows

Temporal-Spectral Resolved Microscopy

The tr-FP technology enables simultaneous imaging of nine different proteins in live cells, allowing researchers to correlate multiple cellular activities with cell cycle progression [17]. This application is particularly valuable for drug development professionals studying complex signaling networks and pathway interactions.

Experimental Protocol: Multiplexed Live-Cell Imaging with tr-FPs

• Cell Line Preparation:

  • Select appropriate cell line (e.g., HEK293, HeLa, or primary cells).
  • Engineer cell lines to express tr-FP fusion proteins for targets of interest using lentiviral transduction or CRISPR/Cas9-mediated knock-in.

• Microscope Setup Configuration:

  • Utilize a fluorescence lifetime imaging microscopy (FLIM) system equipped with:
    • Pulsed laser source matching tr-FP excitation spectra
    • Time-gated or frequency-domain detection capability
    • Environmental chamber for live-cell maintenance (37°C, 5% COâ‚‚)
  • Configure appropriate filter sets for spectral separation of tr-FP signals.

• Image Acquisition Parameters:

  • Set temporal resolution based on biological process under investigation (seconds to minutes between frames).
  • Adjust laser power to balance signal-to-noise ratio with phototoxicity concerns.
  • Collect sufficient photons per pixel for accurate lifetime calculation (typically 1,000-10,000 photons).

• Data Processing Workflow:

  • Perform biexponential or multiexponential fitting of fluorescence decay curves.
  • Apply lifetime unmixing algorithms to distinguish proteins with spectral overlap but distinct lifetimes.
  • Generate spatial maps of protein localization and abundance.

Multiplexed Super-Resolution Microscopy

tr-FPs enable concurrent visualization of four proteins using lifetime signals in super-resolution microscopy [17]. This application pushes beyond the diffraction limit while maintaining multiplexing capability, essential for detailed structural studies in neurobiology and cell biology.

Protein Stoichiometry Quantification

The lifetime properties of tr-FPs enable quantification of cellular protein stoichiometry [17], providing crucial information about complex formation and protein-protein interactions that is inaccessible with conventional intensity-based measurements.

Advanced Implementation: SEPARATE Methodology

The SEPARATE (Spatial Expression PAttern-guided paiRing And unmixing of proTEins) methodology provides a complementary approach to multiplexed imaging that can be integrated with tr-FP technology [13]. This computational strategy enables volumetric multiplexed imaging by pairing proteins with distinct spatial expression patterns and unmixing their signals via machine learning.

Protein Pairing Strategy Based on Spatial Patterns

SEPARATE employs a feature extraction network trained through contrastive learning to quantify visual distinctiveness between protein spatial patterns [13]. The network calculates feature-based distances between protein clusters to identify optimal pairing for multiplexed imaging.

Quantitative Framework for Protein Pairing:

Metric	Mathematical Definition	Functional Role
Interclass Distance (d_inter)	(d_{inter} = \frac{1}{	{\bf P}	(	{\bf P}	-1)}\sum_{p1 \ne p2}		\,{\bar{f}}_{p1}-{\bar{f}}_{p2}		) [13]	Measures average distance between feature vectors of different proteins
Intraclass Distance (d_intra)	(d_{intra} = \frac{1}{	{\bf P}	}\sum_{p}\frac{1}{Np}\sum_{i=1}^{Np}		\,{\bar{f}}_p-{f}_p^{(i)}		) [13]	Measures average distance between feature vectors within the same protein
Optimal Grouping (G_opt)	({G}_{opt} = \mathop{\rm{argmax}}\limitsG\min ({d}{1},{d}{2},\ldots,{d}{n})) [13]	Identifies protein pairing that maximizes the minimum distance between paired proteins

Integration of SEPARATE with tr-FP Technology

The combination of SEPARATE's computational unmixing capability with the intrinsic lifetime discrimination of tr-FPs creates a powerful multimodal platform for high-dimensional biological imaging.

Research Reagent Solutions

Essential Materials for tr-FP Experiments:

Reagent/Material	Specifications	Primary Function
tr-FP Plasmid Library	Coverage across visible spectrum with varied lifetimes [17]	Genetic encoding of time-resolved fluorescent markers
Cell Line Engineering Tools	Lentiviral vectors, CRISPR/Cas9 knock-in systems	Stable tr-FP fusion protein expression in cellular models
FLIM-Compatible Microscope	Pulsed laser source, time-gated detector, environmental control [17]	Acquisition of fluorescence lifetime data in live cells
Lifetime Analysis Software	Biexponential fitting, phasor analysis, lifetime unmixing algorithms	Extraction of lifetime information and signal separation
Mounting Media (fixed samples)	Prolong Diamond, Fluoromount-G, or custom formulations	Preservation of fluorescence and lifetime properties

Protocol: Implementing tr-FPs for Live-Cell Multiplexing

Step-by-Step Experimental Procedure

• Molecular Construct Design:

  • Clone tr-FP genes in-frame with target proteins using Gibson assembly or traditional restriction enzyme methods.
  • Select linkers of appropriate length (typically 5-15 amino acids) to minimize steric interference.
  • Verify construct sequence integrity through Sanger sequencing before proceeding.

• Cell Line Development:

  • Transfect mammalian cells using polyethylenimine (PEI) or electroporation for initial validation.
  • Select stable clones using appropriate antibiotics (e.g., puromycin, G418) over 2-3 weeks.
  • Sort cells for uniform expression levels using fluorescence-activated cell sorting (FACS).

• FLIM Data Acquisition:

  • Seed tr-FP expressing cells in glass-bottom dishes 24-48 hours before imaging.
  • Switch to phenol-red free medium supplemented with HEPES buffer immediately before imaging.
  • Acquire reference lifetime measurements from cells expressing individual tr-FP fusions.
  • Perform multiplexed time-lapse imaging with temporal resolution appropriate for biological process.

• Lifetime Unmixing and Validation:

  • Fit fluorescence decay curves using maximum likelihood estimation.
  • Apply linear unmixing algorithms to distinguish signals from proteins with overlapping spectra.
  • Validate unmixing accuracy through control experiments with known expression ratios.

Troubleshooting Common Issues

• Poor Signal-to-Noise Ratio: Increase laser power or acquisition time; optimize tr-FP expression level.
• Incomplete Lifetime Separation: Select tr-FPs with more distinct lifetimes; check for environmental factors affecting lifetime.
• Cellular Toxicity: Reduce laser power; use more photostable tr-FP variants; shorten acquisition periods.
• Incorrect Protein Localization: Verify fusion does not disrupt target protein function; test multiple linker designs.

The integration of time-resolved fluorescent proteins with advanced computational methods like SEPARATE provides researchers with an unprecedentedly powerful toolkit for investigating complex biological systems. These technologies enable quantitative analysis of multiple cellular components simultaneously, opening new avenues for understanding disease mechanisms and developing targeted therapeutics.

Fluorescence imaging is an indispensable tool in biomedical research, enabling the visualization of biological processes, disease progression, and therapeutic effects. However, conventional fluorescence imaging in the visible (400-700 nm) and first near-infrared window (NIR-I, 700-900 nm) faces significant challenges that limit its effectiveness for deep-tissue and prolonged imaging studies. These limitations include strong light scattering, significant tissue autofluorescence, and limited penetration depth, typically constraining high-fidelity imaging to superficial tissues [19] [20].

The emergence of second near-infrared window (NIR-II, 1000-1700 nm) fluorescence imaging represents a paradigm shift in optical bioimaging. By utilizing light in the 1000-1700 nm range, NIR-II imaging capitalizes on reduced photon scattering, minimal tissue autofluorescence, and lower light absorption by biological components [19]. This translates to superior imaging performance with deeper tissue penetration (up to several centimeters), enhanced spatial resolution, and significantly improved signal-to-background ratios compared to conventional imaging modalities [21] [20]. These advantages are particularly valuable in the context of multiplexed imaging, where simultaneously tracking multiple biomarkers is essential for understanding complex biological systems and disease mechanisms.

The NIR-II Advantage: Quantitative Performance Metrics

The theoretical advantages of NIR-II imaging are demonstrated by quantifiable improvements in key performance metrics compared to NIR-I imaging. The following table summarizes these advantages based on experimental data from recent studies.

Table 1: Performance Comparison Between NIR-I and NIR-II Fluorescence Imaging

Performance Metric	NIR-I Imaging (700-900 nm)	NIR-II Imaging (1000-1700 nm)	Improvement Factor
Tissue Penetration Depth	1-6 mm [20]	Up to 20 mm [20]	~3-5x deeper
Spatial Resolution	Micrometer-level [19]	Micrometer-level [19]	2-3x higher clarity
Signal-to-Background Ratio (SBR)	Moderate [19]	High [19] [20]	Significantly enhanced
Tissue Autofluorescence	Significant [19]	Greatly reduced [19] [20]	Drastically reduced

The exceptional performance of NIR-II imaging stems from fundamental physical principles. Light scattering in biological tissues decreases monotonically as wavelength increases from 400 to 1700 nm. Furthermore, the optimal imaging window within the NIR-II spectrum is between 1000-1350 nm, where reduced scattering coincides with relatively low water absorption, maximizing photon penetration and signal fidelity [19].

Engineering Advanced NIR-II Fluorophores: Design Strategies and Properties

The development of high-performance NIR-II fluorophores has focused on several key structural families, each with distinct photophysical characteristics and design principles suitable for multiplexed imaging applications.

Organic Small-Molecule Fluorophores

Organic small-molecule dyes are particularly promising for clinical translation due to their superior biocompatibility, tunable optical properties, and potential for rapid body clearance [19] [22]. Major design strategies include:

• Donor-Acceptor-Donor (D-A-D) Architectures: These fluorophores feature an electron-accepting core (e.g., benzobisthiadiazole) symmetrically connected to electron-donating units via Ï€-conjugated linkers. The HOMO-LUMO gap can be precisely tuned by modifying the electron-donating/withdrawing strengths of these components, allowing precise adjustment of absorption and emission profiles [19].
• Cyanine Derivatives: Characterized by a central polymethine chain conjugated to terminal heterocycles, cyanine fluorophores offer strong absorption and emission tunability [19].
• BODIPY and Aza-BODIPY Derivatives: These fluorophores provide excellent photostability and modular structures for chemical modification. Recent work has produced NIR-II emissive lipophilic membrane probes based on Aza-BODIPY that enable wash-free labeling with large ON/OFF ratios (>115) and fast staining rates (within 2 minutes) [23].

Innovative Design Strategies for Enhanced Performance

• Coulomb Attraction Interaction Tuning: A groundbreaking strategy for achieving NIR-II emission involves reducing Coulomb attraction interactions in excited states rather than simply reducing the fundamental band gap. This approach enables the development of fluorophores with low molecular weight (<500 Da) while maintaining emission beyond 1000 nm, facilitating better blood-brain barrier penetration and more efficient clearance [24].
• Aggregation-Induced Emission (AIE): AIE-active fluorophores overcome aggregation-caused quenching by restricting intramolecular motion in the aggregated state, resulting in enhanced fluorescence quantum yields in nanoparticles or solid forms [19].

Table 2: Representative NIR-II Small-Molecule Fluorophores and Their Properties

Fluorophore	Class	Emission Maximum (nm)	Molecular Weight (Da)	Key Features	Reference
CH-1055	D-A-D	1055	>500	Early NIR-II dye, moderate QY	[25]
LS7	GFP-chromophore derivative	>1000	<449	Low MW, targets Aβ fibrils, 22.7x fluorescence increase	[24]
BMP1	Aza-BODIPY	NIR-II	N/A	Wash-free, fast membrane staining, large ON/OFF ratio	[23]
IR-FTEP	Cyanine derivative	1100	N/A	QY of 1.0% in water	[25]
Q4	Cyanine derivative	1100	N/A	Good brightness for bioimaging	[25]

Experimental Protocols for NIR-II Imaging Applications

Protocol: Labeling Lipid-Bilayer Structured Nanovectors with NIR-II Membrane Probes

This protocol describes the use of wash-free NIR-II membrane probes (e.g., BMP1) for labeling hybrid vesicles, adapted from methods reported for targeting colon cancer [23].

Materials:

• NIR-II membrane probe (BMP1 or BMP2)
• Hybrid vesicles (e.g., composed of probiotic outer membrane vesicles and anti-PD-L1 scFv-expressing cellular vesicles)
• Appropriate buffer (e.g., PBS or HEPES)
• NIR-II fluorescence imaging system

Procedure:

• Probe Preparation: Prepare a stock solution of the NIR-II membrane probe (BMP1) in DMSO at a concentration of 1 mM.
• Vesicle Incubation: Add the BMP1 probe directly to the hybrid vesicle suspension at a final concentration of 1-5 µM.
• Staining Reaction: Incubate the mixture at room temperature for 2 minutes. The probe rapidly intercalates into the lipid bilayer without requiring washing steps.
• Purification (Optional): If necessary, remove unincorporated probe via gel filtration or dialysis.
• Validation: Characterize labeled vesicles using NIR-II fluorescence imaging to confirm incorporation efficiency and functionality.

Technical Notes:

• The wash-free property of BMP1 significantly streamlines the labeling process compared to traditional membrane probes that require extensive washing.
• This rapid staining protocol (2 minutes) preserves vesicle integrity and functionality.
• The large ON/OFF ratio (>115) provides high signal-to-background for sensitive detection.

Protocol: NIR-II Fluorescence Imaging of Brain Tissue in Alzheimer's Disease Models

This protocol utilizes low-molecular-weight NIR-II dyes for real-time imaging of amyloid-β deposits in mouse models of Alzheimer's disease [24].

Materials:

• LS7 dye (GFP-chromophore based NIR-II fluorophore)
• Alzheimer's disease mouse model (e.g., APP/PS1 transgenic mice)
• Control wild-type mice
• NIR-II fluorescence imaging system with InGaAs camera
• Anesthesia equipment and reagents

Procedure:

• Dye Preparation: Prepare LS7 dye solution in physiological saline at a concentration of 100-500 µM.
• Animal Preparation: Anesthetize mice using isoflurane or ketamine/xylazine cocktail according to approved animal protocols.
• Dye Administration: Administer LS7 dye via intravenous injection (e.g., tail vein) at a dose of 2-5 mg/kg.
• Image Acquisition: Acquire NIR-II fluorescence images at various time points post-injection (e.g., 5, 15, 30, 60 minutes) using appropriate NIR-II filters (1000-1400 nm range).
• Data Analysis: Quantify fluorescence signals in specific brain regions and compare between Alzheimer's model and control animals.

Technical Notes:

• The low molecular weight of LS7 (449 Da) facilitates efficient blood-brain barrier penetration.
• LS7 shows a 22.7-fold fluorescence increase upon binding to Aβ42 fibrils in vitro, providing high specificity.
• Real-time imaging enables dynamic monitoring of amyloid deposition progression.

NIR-II Fluorescence Imaging Workflow

Advanced Applications in Multiplexed Imaging and Therapeutics

Multiplexed Imaging with Computational Unmixing

Conventional multiplexed imaging is limited by the number of spectrally distinct fluorophores that can be simultaneously detected. The SEPARATE (Spatial Expression PAttern-guided paiRing And unmixing of proTEins) method overcomes this limitation by pairing proteins with distinct spatial expression patterns, labeling them with the same fluorophore, and computationally unmixing their signals based on three-dimensional spatial expression patterns using neural networks [13].

This approach effectively doubles multiplexing capability by imaging two proteins using a single fluorophore channel, significantly reducing the number of staining cycles required for visualizing multiple targets. The method employs a feature extraction network to quantify spatial distinction between proteins, followed by a protein separation network that unmixes signals from protein pairs labeled with identical fluorophores [13].

NIR-II Phototheranostics: Integrating Diagnosis and Therapy

NIR-II fluorophores have enabled the development of "phototheranostics" that combine diagnostic imaging with therapeutic interventions:

• Photothermal Therapy (PTT): NIR-II dyes with high photothermal conversion efficiency (e.g., 41-50%) can generate localized heat under laser irradiation, enabling precise tumor ablation while monitoring treatment efficacy in real-time [23] [22].
• Targeted Immunotherapy: Hybrid vesicles labeled with NIR-II membrane probes can execute targeted photothermal-immunotherapy, combining spatiotemporally controlled PTT with immunogenic vesicles to induce tandem-amplified immune responses and reprogram the tumor microenvironment [23].
• Image-Guided Surgery: The superior penetration and resolution of NIR-II imaging enables real-time visualization of tumor margins, lymphatic mapping, and precise surgical resection guidance, particularly for deep-seated tumors [21].

NIR-II Phototheranostics Mechanism

The Scientist's Toolkit: Essential Research Reagent Solutions

Table 3: Key Research Reagents for NIR-II Fluorescence Imaging Applications

Reagent/Category	Function/Application	Examples	Key Characteristics
NIR-II Membrane Probes	Labeling lipid-bilayer nanovectors	BMP1, BMP2 [23]	Wash-free, fast staining (2 min), large ON/OFF ratio (>115)
Low-MW Brain-Penetrant Dyes	Blood-brain barrier crossing, brain imaging	LS series dyes [24]	MW <500 Da, emission >1000 nm, target-specific (e.g., Aβ)
Donor-Acceptor-Donor Dyes	High-performance NIR-II imaging	CH-1055, IR-FTEP [25] [19]	Tunable emission, high brightness, good photostability
Hybrid Vesicle Systems	Targeted drug delivery and immunotherapy	OMV + aPD-L1 scFv vesicles [23]	Immunogenic, actively tumor-targeting, combinable with PTT
Computational Unmixing Tools	Multiplexed imaging beyond spectral limits	SEPARATE method [13]	Neural network-based, doubles multiplexing capability
TDD	TDD, CAS:55062-34-5, MF:C16H17N3O2, MW:283.32 g/mol	Chemical Reagent	Bench Chemicals
Ro 46-8443	Ro 46-8443, MF:C31H35N3O8S, MW:609.7 g/mol	Chemical Reagent	Bench Chemicals

The development of NIR-II fluorophores and photostable probes represents a significant advancement in fluorescence imaging, directly addressing the critical limitations of penetration depth and photobleaching that have constrained traditional approaches. These innovations are particularly valuable for multiplexed imaging studies, where tracking multiple biomarkers simultaneously provides comprehensive insights into complex biological systems.

Future developments in this field will likely focus on creating fluorophores with emission further into the NIR-IIb (1500-1700 nm) region for even better performance, improving biocompatibility and clearance profiles to facilitate clinical translation, and integrating artificial intelligence for enhanced image analysis and unmixing capabilities [26]. As these technologies mature, they will undoubtedly expand our ability to visualize and understand biological processes at unprecedented depths and resolution, accelerating drug development and advancing our fundamental knowledge of disease mechanisms.

The integration of NIR-II imaging with targeted therapeutic interventions creates powerful theranostic platforms that promise to revolutionize both biomedical research and clinical practice. By providing real-time feedback on drug distribution, target engagement, and treatment efficacy, these approaches represent a significant step toward personalized medicine with enhanced precision and effectiveness.

Advanced Techniques and Translational Applications in Spatial Biology

Multiplexed fluorescence imaging represents a paradigm shift in spatial biology, enabling the simultaneous visualization of dozens to hundreds of biomolecules within their native tissue context. These methods provide unique insights into complex biological systems by preserving spatial information that is lost in dissociative analytical techniques. For researchers and drug development professionals, understanding the principles, applications, and practical implementation of these technologies is crucial for advancing biomarker discovery, drug target validation, and patient stratification strategies. This application note focuses on three prominent cyclic imaging methods—Cyclic Immunofluorescence (CycIF), Multiplexed Error-Robust Fluorescence In Situ Hybridization (MERFISH), and the novel SPECTRE-Plex approach—detailing their fundamental principles, optimized protocols, and research applications.

The critical advantage of cyclic imaging methods over conventional fluorescence techniques lies in their ability to overcome the physical limitations of spectral overlap. While conventional immunofluorescence is typically limited to four to seven markers due to fluorophore emission spectra overlap, cyclic methods achieve high multiplexing through sequential imaging and signal removal or inactivation [27]. This capability makes them particularly valuable for characterizing complex cellular microenvironments, such as the tumor microenvironment, where intricate cell-cell interactions determine disease progression and treatment response [28].

Core Principles of Cyclic Imaging

Cyclic imaging methods share a common fundamental strategy: the sequential application, imaging, and removal of fluorescent signals to build high-dimensional datasets from a single sample. This iterative process bypasses the spectral limitations of traditional fluorescence microscopy, wherein the number of simultaneously detectable markers is constrained by the available fluorophore spectrum and filter sets. The cyclic approach typically involves four key phases that repeat for each imaging cycle: (1) staining with fluorescently-labeled probes, (2) multi-channel image acquisition, (3) signal inactivation or removal, and (4) image registration and alignment using a consistent reference marker, such as a nuclear stain [29].

The implementation details, however, vary significantly between protein-based detection (e.g., CycIF, SPECTRE-Plex) and RNA-based detection (e.g., MERFISH) methods, each with distinct advantages for different research applications. Protein-based methods provide direct information about protein expression and post-translational modifications that often correlate with cellular function, while RNA-based methods offer comprehensive transcriptomic profiling at single-cell resolution within intact tissues [27].

Comparative Analysis of Cyclic Imaging Methods

Table 1: Comparison of Key Cyclic Imaging Technologies

Method	Target	Multiplexing Capacity	Resolution	Key Principle	Typical Duration
CycIF [29]	Proteins	Up to 60-plex	Subcellular	Sequential immunofluorescence with chemical dye inactivation	14-36 hours (for ~7 cycles)
MERFISH [30]	RNA	Hundreds to thousands of genes	Single-molecule	Sequential FISH with error-robust barcoding	Several days
SPECTRE-Plex [31]	Proteins	Demonstrated 22-plex	Subcellular (high-NA objectives)	Optimized dye inactivation with microfluidics	~8 hours (for 7 cycles)
CODEX [27]	Proteins	Up to 60-plex	Subcellular	DNA oligonucleotide barcoding with hybridization cycles	Varies

Cyclic Immunofluorescence (CycIF)

Principles and Workflow

Cyclic Immunofluorescence (CycIF) is an accessible, highly multiplexed immunofluorescence imaging method that utilizes conventional microscopes and commercially available reagents. The method was specifically developed to overcome the limitations of conventional immunohistochemistry and immunofluorescence, which are typically restricted to imaging a few markers simultaneously [29]. CycIF achieves multiplexing through sequential rounds of staining, imaging, and fluorophore inactivation, building a high-dimensional image dataset cycle by cycle.

The fundamental CycIF workflow consists of four key steps that repeat for each cycle: (1) immunostaining with directly conjugated antibodies targeting 3-4 protein antigens per cycle, (2) nuclear staining with Hoechst or DAPI for image registration, (3) multi-channel fluorescence imaging, and (4) chemical inactivation of fluorophores using a mild base with hydrogen peroxide under light exposure [29] [32]. This process preserves tissue integrity while progressively building a multiplexed dataset. A notable advantage of CycIF is that the signal-to-noise ratio often improves with successive cycles due to reduction of autofluorescence through repeated bleaching steps [29].

CycIF Workflow Visualization

Diagram 1: The CycIF workflow involves iterative cycles of staining, imaging, and bleaching to build high-plex images from standard FFPE tissue sections.

Detailed CycIF Protocol

Sample Preparation

• Tissue Sections: Cut 5μm sections from formalin-fixed, paraffin-embedded (FFPE) blocks using standard microtomy procedures [29].
• Deparaffinization: Incubate slides at 60°C for 15 minutes, followed by two incubations in 100% Histoclear (first for 5 minutes, then for 10 minutes) at room temperature [31].
• Rehydration: Perform sequential 5-minute incubations in decreasing ethanol concentrations (100%, 95%, 75%, 50%, and 0%) at room temperature [31].
• Antigen Retrieval: Incubate slides in antigen retrieval buffer (e.g., citrate-based buffer, pH 6.0), heat to 100°C for 20 minutes, then cool to room temperature for 20-40 minutes [31] [29].
• Blocking: Wash slides twice with PBS and incubate in blocking buffer (e.g., 3% BSA in PBS) for 1 hour to reduce non-specific antibody binding [31].

Pre-staining and Initial Bleaching

• Pre-staining: Incubate samples with secondary antibodies to reduce autofluorescence caused by non-specific antibody binding in subsequent cycles [33].
• Initial Bleaching: Treat samples with fluorophore oxidation solution (high pH hydrogen peroxide with light exposure) to inactivate fluorophores and reduce background autofluorescence [29].

Cyclic Staining and Imaging

• Primary Antibody Staining: For each cycle, incubate with 3-4 directly conjugated primary antibodies diluted in antibody dilution buffer for 1-2 hours at room temperature. Directly conjugated antibodies are preferred over indirect immunofluorescence to avoid species limitation issues [32].
• Nuclear Staining: Include Hoechst 33342 (1-2 μg/mL) or DAPI in each cycle to facilitate image registration across cycles [29].
• Image Acquisition: Acquire four-channel images using a conventional epifluorescence microscope or slide scanner. For large specimens, collect tiled images with consistent exposure settings across cycles.
• Fluorophore Inactivation: After imaging, incubate slides in bleaching solution (100mM Tris pH 8.0, 3% Hâ‚‚Oâ‚‚) under bright light for 1 hour to inactivate fluorophores [29].
• Cycle Repetition: Repeat the staining-imaging-bleaching cycle until all targets have been imaged. Typical CycIF experiments involve 5-15 cycles to generate 15-60-plex datasets [29].

Post-cycling Processing

• Optional H&E Staining: After the final CycIF cycle, stain with hematoxylin and eosin to enable correlation with conventional histopathology [33] [29].
• Image Processing: Register and stitch individual image panels across cycles using nuclear staining as a reference. Process images to generate a high-dimensional representation using tools like MCMICRO pipeline [33].

CycIF Research Reagent Solutions

Table 2: Essential Reagents for CycIF Experiments

Reagent Category	Specific Examples	Function	Application Notes
Fluorophore-Conjugated Antibodies	Alexa Fluor 488, 555, 647 conjugates [32]	Target protein detection	Prefer directly conjugated antibodies; Alexa Fluor dyes bleach efficiently
Nuclear Stains	Hoechst 33342, DAPI [29]	Nuclear segmentation & image registration	Include in every cycle for consistent registration
Bleaching Reagents	Hydrogen peroxide, Tris base [29]	Fluorophore inactivation	High pH Hâ‚‚Oâ‚‚ solution with light exposure
Antigen Retrieval Buffers	Citrate buffer (pH 6.0), Tris-EDTA (pH 9.0) [31]	Epitope exposure	Required for FFPE samples; optimal pH varies by antigen
Blocking Buffers	BSA (3-5%), serum, commercial blocking reagents [31]	Reduce non-specific binding	Critical for minimizing background in multiplexed imaging

Multiplexed Error-Robust FISH (MERFISH)

Principles and Workflow

Multiplexed Error-Robust Fluorescence In Situ Hybridization (MERFISH) is a single-molecule transcriptomic imaging method that enables the quantification and spatial mapping of hundreds to thousands of RNA species simultaneously in individual cells within intact tissues [30]. As an image-based approach to single-cell transcriptomics, MERFISH combines single-molecule FISH with combinatorial barcoding and sequential hybridization to overcome the spectral limitations of conventional fluorescence microscopy.

The core principle of MERFISH involves a two-step labeling process using encoding probes with readout sequences [30]. First, unlabeled DNA "encoding probes" are hybridized to cellular RNA. These probes contain a targeting region complementary to the RNA of interest and a barcode region comprising a series of custom binding sites ("readout sequences"). The specific combination of readout sequences determines the optical barcode for each RNA species. Second, fluorescently labeled "readout probes" complementary to these sequences are hybridized in successive rounds of imaging, with each round detecting one bit of the barcode. The combinatorial barcoding scheme incorporates error-correction mechanisms to ensure accurate RNA identification despite occasional imaging or hybridization errors [30].

MERFISH Workflow Visualization

Diagram 2: MERFISH utilizes encoding probes with combinatorial barcodes that are read out through successive hybridization and imaging cycles to profile hundreds to thousands of RNAs simultaneously.

Optimized MERFISH Protocol

Probe Design Considerations

• Target Region Length: Encoding probes with target regions of 30-50 nucleotides provide optimal binding efficiency and specificity. Systematic optimization has shown that signal brightness depends weakly on target region length within this range, with 40 nt providing a good balance of efficiency and specificity [30].
• Barcode Design: Implement error-robust barcoding schemes (e.g., Hamming distance-based codes) to minimize misidentification errors during decoding.
• Probe Set Size: Include approximately 30-100 encoding probes per target RNA to achieve high molecular detection efficiency through binding redundancy [30].

Sample Preparation and Hybridization

• Fixation: Fix cells or tissue sections with 4% paraformaldehyde for optimal RNA preservation and probe accessibility.
• Permeabilization: Treat samples with detergent solution (e.g., 0.1% Triton X-100) or protease digestion to enable probe access to RNA targets.
• Encoding Probe Hybridization: Hybridize encoding probes to the sample using optimized buffer conditions. Recent protocol improvements demonstrate that hybridization duration can be significantly reduced through buffer optimization without sacrificing efficiency [30].
• Hybridization Buffer Optimization: Screen formamide concentrations (typically 10-30%) with fixed hybridization temperature (37°C) to identify optimal stringency conditions for specific probe sets [30].

Sequential Imaging and Stripping

• Readout Probe Hybridization: For each round of imaging, hybridize fluorescent readout probes complementary to specific readout sequences in the encoding probes.
• Multi-channel Imaging: Acquire images for all fields of view using appropriate filter sets. Include fiducial markers for image registration if imaging large areas.
• Probe Stripping: After each imaging round, remove readout probes using chemical denaturation (e.g., formamide-based stripping buffers) or enzymatic degradation.
• Cycle Repetition: Repeat the readout hybridization-imaging-stripping cycle until all bits of the barcode have been read. Typical MERFISH experiments involve 8-16 rounds of hybridization to encode hundreds to thousands of RNA species [30].

Image Processing and Data Analysis

• Image Registration: Align images from successive hybridization rounds using fiducial markers or nuclear stains.
• Barcode Decoding: Identify RNA molecules by decoding the fluorescence on-off pattern across imaging rounds for each detected spot.
• Error Correction: Utilize the built-in error-correction capability of the barcoding scheme to correct for missing or extra bits in the fluorescence patterns.
• Cell Segmentation: Identify cell boundaries using nuclear stains or membrane markers to assign transcripts to individual cells.

Recent MERFISH Protocol Optimizations

Recent systematic optimization of MERFISH protocols has identified several key improvements:

• Hybridization Acceleration: Modified hybridization conditions can substantially enhance the rate of probe assembly, reducing hybridization time while maintaining or improving signal brightness [30].
• Buffer Composition: New imaging buffer formulations improve photostability and effective brightness for commonly used MERFISH fluorophores, enhancing signal-to-noise ratio throughout extended imaging sessions [30].
• Reagent Stability: Implementation of stabilization approaches mitigates the "aging" of reagents during prolonged experiments, maintaining consistent performance across multiple hybridization rounds [30].
• Background Reduction: Prescreening readout probes against sample types identifies and mitigates tissue-specific non-specific binding, reducing false-positive counts [30].

SPECTRE-Plex: An Advanced Successive Staining Approach

Principles and Innovations

SPECTRE-Plex (Spatial Photo-inactivation Enhanced Cyclic Target REsolved multiPlexing) represents a recent advancement in cyclic imaging technology that addresses several limitations of existing methods, particularly in terms of speed, resolution, and practical implementation [31]. This end-to-end system integrates novel chemical, microfluidic, and optical innovations to enable rapid, high-resolution multiplex immunofluorescence imaging at relatively low cost compared to existing platforms.

The key innovations of SPECTRE-Plex include:

• Advanced Dye Inactivation Chemistry: Implementation of meta-chloroperoxybenzoic acid (m-CPBA) as a dye inactivation agent that operates orders of magnitude faster than previously used reagents (LiBHâ‚„ or Hâ‚‚Oâ‚‚) and does not generate oxygen bubbles that can interfere with microfluidics [31].
• Refractive Index-Matched Microfluidics: Development of a PDMS-based microfluidic device ("invisi-slip") with an embedded fluorinated ethylene propylene (FEP) window that enables aberration-free imaging with high-NA water dipping objectives [31].
• Integrated Automation: Custom python and μ-manager software that coordinates imaging, positioning, and fluidics for efficient and rapid repetition of cycles [31].
• Kinetic Analysis Capability: The in-situ design allows simultaneous imaging and antibody labeling, enabling direct measurement of antibody binding kinetics on tissues under various conditions [31].

SPECTRE-Plex Workflow and Protocol

System Setup

• Microfluidic Chamber: Assemble the invis-slip system with FEP optical window to create a refractive index-matched imaging environment.
• Fluidics Integration: Connect temperature-controlled fluidics with bubble-free operation enabled by m-CPBA chemistry.
• Objective Selection: Utilize high-NA water dipping objectives (e.g., 16×/0.8 NA for large FOV or 60×/1.2 NA for subcellular resolution) [31].

Staining and Imaging Protocol

• Sample Preparation: Follow standard FFPE processing including deparaffinization, rehydration, antigen retrieval, and blocking as described in the CycIF protocol [31].
• Automated Cycling: Implement iterative cycles of staining, washing, imaging, and dye inactivation in situ using integrated fluidics and imaging control.
• Rapid Dye Inactivation: Apply m-CPBA solution for efficient fluorophore inactivation without bubble formation [31].
• High-Resolution Imaging: Capture tiled images using a highly repeatable motorized stage with automated image registration.

Binding Kinetics Applications

The SPECTRE-Plex system enables unique applications for directly measuring antibody binding kinetics on tissues:

• Temperature Optimization: Assess binding half-times at different temperatures (e.g., anti-Muc2 IgG t₁/â‚‚ = 50 min at 22°C vs. 13 min at 37°C) [31].
• Condition Screening: Evaluate effects of solution pH, ionic strength, or other parameters on antibody binding efficiency.
• Therapeutic Antibody Validation: Characterize binding behavior of therapeutic antibodies or small molecules directly in tissue contexts [31].

Performance Advantages

SPECTRE-Plex demonstrates significant practical advantages over existing cyclic imaging methods:

• Speed: A 7-cycle run can be completed in approximately 8 hours compared to 14-36 hours for most manual or semi-automated cyclical methods [31].
• Resolution: Compatibility with high-NA water dipping objectives enables subcellular resolution imaging not typically achievable with conventional cyclic imaging systems limited by air objectives or refractive index mismatches [31].
• Scalability: The automated, integrated system supports consistent and reproducible data acquisition across multiple samples and experiments.

Applications in Biomedical Research

Tumor Microenvironment Characterization

Multiplexed cyclic imaging methods have proven particularly valuable in immuno-oncology research, where understanding the spatial organization of immune cells within the tumor microenvironment (TME) has become crucial for predicting treatment response and developing new therapeutic strategies [28]. These technologies enable comprehensive profiling of immune cell populations, including:

• Immune Contexture Assessment: Simultaneous identification of multiple immune cell subsets (T cells, B cells, macrophages, dendritic cells) and their functional states within tissue architecture [28].
• Spatial Relationship Mapping: Analysis of cell-cell interactions and neighborhood relationships that influence therapeutic response, such as the proximity of cytotoxic T cells to tumor cells expressing PD-L1 [28].
• Tumor Heterogeneity Characterization: Resolution of intratumoral heterogeneity through high-plex protein or RNA expression profiling across different tumor regions [29].

Drug Development Applications

For drug development professionals, cyclic imaging methods offer powerful tools for:

• Biomarker Discovery: Identification of novel predictive or prognostic biomarkers through comprehensive tissue profiling [28].
• Pharmacodynamic Assessment: Evaluation of drug effects on multiple signaling pathways and cell populations simultaneously within intact tissues.
• Therapeutic Antibody Validation: Direct measurement of antibody binding kinetics and target engagement in relevant tissue contexts using systems like SPECTRE-Plex [31].
• Mechanistic Studies: Elucidation of drug mechanisms of action through spatial analysis of pathway modulation and cellular responses.

Integration with Complementary Technologies

Cyclic imaging methods are increasingly integrated with other spatial analysis platforms to provide comprehensive tissue characterization:

• Spatial Transcriptomics Correlation: Combination of protein-based cyclic imaging (CycIF, SPECTRE-Plex) with RNA-based methods (MERFISH) applied to adjacent tissue sections to correlate protein and RNA expression patterns [27].
• Single-Cell Sequencing Integration: Registration of multiplexed imaging data with single-cell RNA sequencing data from the same or similar samples to validate and spatialize cell type identities [29].
• Clinical Histopathology Correlation: Registration of high-plex cyclic imaging data with standard H&E staining and clinical immunohistochemistry to bridge discovery research with diagnostic practice [33] [29].

Technical Considerations and Challenges

Experimental Design Considerations

Successful implementation of cyclic imaging methods requires careful experimental planning:

• Panel Design: Strategic selection of markers based on biological questions, with consideration of antibody performance, expression levels, and cellular localization.
• Cycle Ordering: Placement of high-signal antibodies in early cycles and lower-signal antibodies in later cycles, as background autofluorescence tends to decrease with successive bleaching steps [33].
• Validation: Rigorous validation of antibody specificity and performance under cyclic conditions, including assessment of epitope stability through multiple cycles.
• Controls: Inclusion of appropriate positive and negative controls to ensure data quality and reproducibility.

Data Analysis Challenges

The high-dimensional data generated by cyclic imaging methods presents significant computational challenges:

• Image Processing: Efficient registration and stitching of large image datasets across multiple cycles and channels [34].
• Cell Segmentation: Accurate identification of cellular boundaries, particularly in dense tissue regions, which can be complicated by artifacts such as uneven illumination, antibody aggregates, and tissue folding [34].
• Cell Type Classification: Integration of high-dimensional marker data to define cell types and states, requiring specialized computational approaches [34].
• Spatial Analysis: Quantitative assessment of spatial patterns, cell-cell interactions, and tissue organization using appropriate statistical methods.

Artifact Identification and Mitigation

Cyclic imaging data can be affected by various technical artifacts that require detection and correction:

• Fluorescent Contaminants: Autofluorescent particles or debris that can be misidentified as specific signals [34].
• Uneven Immunolabeling: Regional variations in staining intensity due to fluidics or incubation inconsistencies [34].
• Image Blur: Out-of-focus regions that reduce segmentation and quantification accuracy [34].
• Antibody Aggregates: Non-specific clusters of antibodies that generate false-positive signals [34].

Emerging computational tools, including machine learning approaches, are being developed to automatically identify and correct these artifacts to improve data quality and reliability [34].

Cyclic imaging methods—including CycIF, MERFISH, and the novel SPECTRE-Plex approach—represent powerful and increasingly accessible technologies for high-dimensional spatial analysis of biological tissues. Each method offers unique advantages: CycIF provides an accessible entry point for high-plex protein imaging using conventional microscopes; MERFISH enables comprehensive transcriptomic profiling at single-molecule resolution; and SPECTRE-Plex delivers accelerated, high-resolution multiplexed imaging through integrated automation and optimized chemistry.

For researchers and drug development professionals, these technologies offer unprecedented capabilities to characterize complex biological systems, discover novel biomarkers, and validate therapeutic mechanisms within the native tissue context. As these methods continue to evolve through improvements in reagents, instrumentation, and computational analysis, they are poised to become increasingly integral to both basic research and translational applications, ultimately contributing to more personalized and effective therapeutic strategies.

The spatial organization of cells within healthy and diseased tissues is a critical determinant of physiological function and pathological states. For decades, traditional immunohistochemistry (IHC) and immunofluorescence have been limited to visualizing only one or two protein markers in a single tissue section, restricting our understanding of complex cellular ecosystems [35]. The emergence of multiplexed fluorescence imaging technologies has revolutionized our ability to characterize tissues at the single-cell level while preserving precious spatial context. Among these advanced methods, CODEX (Co-Detection by indEXing) and SABER (Signal Amplification By Exchange Reaction) represent two powerful approaches that leverage DNA barcoding to overcome the limitations of conventional microscopy [35] [36]. CODEX utilizes DNA-conjugated antibodies and cyclic hybridization to enable highly multiplexed protein detection, while SABER employs primer exchange reactions to create long DNA concatemers for signal amplification in nucleic acid detection. These complementary technologies are rapidly advancing biological discovery and therapeutic development by providing unprecedented views into cellular organization, cell-cell interactions, and tissue microenvironments in conditions ranging from cancer to autoimmune diseases [35] [36].

CODEX: Principles and Workflow

The CODEX platform is a multiplexed tissue imaging technology that uses DNA-conjugated antibodies and cyclic fluorescent detection to simultaneously visualize up to 60 protein markers in situ [35] [37]. This method involves conjugating purified antibodies to specific DNA oligonucleotides (barcodes), which are then applied to tissue sections in a single staining procedure. The tissue is mounted on a microscope equipped with a microfluidics system that iteratively delivers fluorescently labeled DNA probes complementary to the antibody barcodes [35]. Each cycle involves hybridization, imaging, and stripping of the fluorescent probes, with the process repeated for multiple markers. The current multiplexing capacity of CODEX is approximately 57 DNA-conjugated antibodies plus additional markers detected via biotin-streptavidin and nuclear stains, totaling 60 markers [35]. Recent advancements have expanded this capacity to 85+ markers through the introduction of novel oligonucleotide sequences [37].

SABER: Principles and Workflow

SABER is a molecular toolkit that enables programmable signal amplification for fluorescence in situ hybridization (FISH) applications through primer exchange reaction (PER) [36]. This method endows standard DNA or RNA FISH probes with long, single-stranded DNA concatemers that serve as scaffolds for aggregating multiple fluorescent imager strands. The core innovation lies in the in vitro synthesis of concatemers using PER, which allows precise control over amplification levels by varying reaction conditions such as polymerase concentration, magnesium concentration, and extension time [36]. SABER achieves signal amplification ranging from 5 to 450-fold compared to conventional FISH, enabling detection of low-abundance targets and reducing imaging time and cost [36]. The method supports highly multiplexed imaging through an "Exchange-SABER" approach, where imagers are stripped and reused in sequential rounds, allowing a single fluorophore to detect multiple targets across imaging cycles [36].

Comparative Technical Specifications

Table 1: Technical comparison between CODEX and SABER technologies

Parameter	CODEX	SABER
Primary Application	Protein detection in tissues	RNA and DNA detection in cells and tissues
Multiplexing Capacity	60 markers (current), 85+ markers (with new barcodes) [35] [37]	17 targets simultaneously demonstrated, potentially much higher with sequential imaging [36]
Signal Amplification	Currently lacks amplification system; limited by baseline autofluorescence [35]	Programmable amplification from 5 to 450-fold [36]
Key Mechanism	DNA-barcoded antibodies + cyclic fluorescent hybridization	Primer exchange reaction (PER) generating DNA concatemers
Tissue Compatibility	FFPE, fresh-frozen, fixed-frozen, cell spreads [35]	Fixed cells and tissues, including thick tissue sections [36]
Species Compatibility	Human, mouse, monkey [35]	Demonstrated in mouse and human samples [36]
Hands-on Time	~4.5 h (conjugation) + ~6.5 h (validation) + ~8 h (experiment prep) [35]	~1-3 hours for PER reaction [36]
Spatial Context	Preserved at single-cell level	Preserved at subcellular level
Current Limitations	No signal amplification system; antibody costs and validation requirements [35]	Limited simultaneous orthogonal amplifiers compared to RCA-based methods [36]

Research Reagent Solutions

Table 2: Essential research reagents and materials for CODEX and SABER experiments

Reagent/Material	Function	Technology
DNA-conjugated Antibodies	Target protein binding with unique DNA barcodes	CODEX [35]
Maleimide-modified DNA Oligonucleotides	Covalent conjugation to antibody scaffolds	CODEX [35]
Fluorescently Tagged DNA Oligos	Complementary reporters for barcode detection	CODEX [35]
CODEX Buffer	Optimal hybridization and imaging conditions	CODEX [38]
Primer-modified FISH Probes	Target nucleic acid binding with primer sequence for extension	SABER [36]
PER Reaction Components	Enzymatic synthesis of DNA concatemers (hairpin, strand-displacing polymerase, nucleotides)	SABER [36]
Fluorescent Imager Strands	Signal detection via concatemer hybridization	SABER [36]
Nuclear Stains (Hoechst, DRAQ5)	Cell segmentation and identification	Both [35]

Detailed Experimental Protocols

CODEX Protocol: Antibody Conjugation and Validation

The CODEX procedure involves four critical sections: antibody conjugation, antibody validation, multicycle reaction, and data analysis [35].

Section 1: Antibody Conjugation (Hands-on time: ~4.5 hours)

• Partial antibody reduction: Incubate 15 µg of purified antibody in 37.5 µL of conjugation buffer with 1.5 µL of 100 mM TCEP for 1.5 hours at 37°C [35].
• DNA oligonucleotide conjugation: Add 7.5 µL of 100 µM maleimide-modified DNA oligonucleotide to the reduced antibody and incubate for 1 hour at room temperature [35].
• Conjugate purification: Remove excess oligonucleotides using spin filters, recover conjugated antibodies, and quantify using spectrophotometry [35].

Section 2: Antibody Validation and Titration (Hands-on time: ~6.5 hours)

• Tissue preparation: Perform antigen retrieval on FFPE or fresh-frozen tissue sections using appropriate methods (heat-induced epitope retrieval for FFPE) [35].
• Staining procedure: Apply DNA-conjugated antibodies to tissues, incubate, then perform post-stain fixation with 4% PFA for 1 hour [35].
• Hybridization and imaging: Hybridize with corresponding fluorescent oligonucleotides, image using appropriate microscope settings, and validate staining pattern specificity [35].

Section 3: CODEX Multicycle Experiment (Hands-on time: ~8 hours)

• Instrument setup: Prepare the microfluidics device, mount the coverslip onto an acrylic plate, and set imaging parameters [35] [38].
• Multicycle imaging: Program the CODEX Instrument Manager software for the required number of cycles (typically 20-60), set Z-stack planes to 5 with 1.5 µm pitch, and configure fluorescence channels [38].
• Post-imaging processing: After run completion, preserve coverslip in storage buffer for future imaging or perform H&E staining for morphological correlation [35].

SABER Protocol: Nucleic Acid Detection with Signal Amplification

Probe Design and Concatemer Synthesis

• Probe design: Design target-specific FISH probes (e.g., using OligoMiner pipeline) with 9-nt primer sequence on 3' end [36].
• Concatemer synthesis: Perform primer exchange reaction (PER) using catalytic hairpin, strand-displacing polymerase, and nucleotides (dATP, dCTP, dTTP only) for 1-3 hours at 37°C [36].
• Concatemer length tuning: Adjust concatemer length by varying polymerase concentration, magnesium concentration, or extension time [36].

In Situ Hybridization and Signal Amplification

• Sample preparation: Fix cells or tissues using standard methods (4% PFA) and perform permeabilization [36].
• Hybridization: Hybridize PER-extended probes to targets under appropriate stringency conditions [36].
• Signal detection: Apply fluorescent imager strands complementary to concatemer repeats and incubate for 30-60 minutes [36].
• Exchange-SABER variant: For multiplexing, simultaneously hybridize all concatemerized probes, then sequentially image and strip imagers using DNA displacement [36].

Imaging and Analysis

• Microscopy: Image using widefield (cells) or confocal (tissues) microscopes with standard filter sets [36].
• Image processing: Quantify signal amplification and perform cell segmentation for spatial analysis [36].

Applications in Biomedical Research

CODEX and SABER have enabled significant advances across multiple domains of biomedical research by providing unprecedented views into cellular organization and gene expression patterns.

CODEX Applications:

• Cancer Immunology: Analysis of tumor microenvironments in colorectal cancer patients revealed that spatial organization of PD-1+CD4+ T cells correlates with clinical outcomes, highlighting the importance of cellular neighborhoods in disease progression [35].
• Autoimmune Disease: Investigation of splenic architecture in lupus mouse models demonstrated how cellular positioning is altered in diseased tissues [35].
• Tissue Atlas Construction: CODEX has been used to characterize cellular heterogeneity and spatial relationships in human, mouse, and monkey tissues including tonsil, spleen, lymph node, liver, pancreas, and brain [35].

SABER Applications:

• Enhancer Mapping: Identification of in vivo introduced enhancers with cell type-specific activity in mouse retina through 10-plex SABER-FISH [36].
• Chromosome Organization: Simultaneous imaging of 17 different chromosomal targets in cultured cells [36].
• Low-Abundance RNA Detection: High-efficiency detection of mRNA transcripts with improved sampling efficiency compared to other amplification methods [36].
• Combined RNA/DNA FISH: Co-detection of reporter RNAs and the plasmids from which they are expressed in the same samples [36].

Technical Considerations and Limitations

While both technologies offer powerful capabilities for multiplexed imaging, researchers should consider several practical aspects when implementing these methods.

CODEX Limitations:

• Cost Considerations: Purified antibodies for FFPE tissues cost $300-600 per 100 µg; maleimide-modified DNA oligonucleotides for ~20 conjugations cost ~$800; fluorescently tagged DNA oligonucleotides for ~100 reaction cycles cost ~$400 [35].
• Amplification Limitations: The platform currently lacks a signal amplification system, limiting detection of low-abundance targets [35].
• Antibody Validation: Each antibody requires individual conjugation and validation, and some clones may not be suitable for multiplexing due to incompatible antigen retrieval requirements [35].
• Tissue Autofluorescence: Baseline autofluorescence can be substantial in certain tissues like brain and lung, potentially limiting signal-to-noise ratio [35].

SABER Limitations:

• Orthogonal Amplifiers: While demonstrating 17 simultaneous targets, the number of orthogonal concatemers is currently limited compared to RCA-based methods [36].
• Detection Efficiency: While generally high, detection efficiencies for RNA transcripts can vary depending on target accessibility and probe design [36].
• Probe Design Complexity: Requires careful computational design of orthogonal sequences to minimize off-target interactions in highly multiplexed experiments [36].

Future Directions

The fields of DNA-barcoded multiplexed imaging are rapidly evolving, with several promising developments on the horizon. For CODEX, ongoing work focuses on expanding the repertoire of validated DNA barcodes beyond the current 60 markers, with recent studies already demonstrating panels of 74+ markers [37]. The implementation of signal amplification systems is also under active development, including tetramer-based staining, branched fluorescent oligonucleotides, and rolling circle amplification approaches [35]. For SABER, future directions include increasing the number of orthogonal amplifiers for higher levels of simultaneous multiplexing and improving tissue penetration for thick tissue sections [36]. Both technologies are likely to benefit from continued improvements in computational analysis pipelines, enabling more sophisticated spatial analysis of cellular neighborhoods and interactions. As these methods become more accessible and widely adopted, they will undoubtedly spur new discoveries in developmental biology, disease mechanisms, and therapeutic development.

Multiplexed Error-Robust Fluorescence in Situ Hybridization (MERFISH) is a massively multiplexed single-molecule imaging technology for spatially resolved transcriptomics that simultaneously measures the copy number and spatial distribution of hundreds to tens of thousands of RNA species in individual cells while preserving their native spatial context [39] [40]. As a targeted spatial transcriptomics method, MERFISH combines the single-molecule sensitivity and quantification accuracy of single-molecule FISH (smFISH) with combinatorial barcoding strategies to achieve unprecedented multiplexing capability [41] [42]. This powerful combination enables the construction of comprehensive cell atlases by directly mapping transcriptional identities to spatial positions within tissues, revealing complex patterns of tissue organization, cell states, and their interactions across multiple length scales [39] [43].

The fundamental principle underlying MERFISH is the assignment of a unique binary barcode to each RNA species, which is then read out through sequential rounds of single-molecule fluorescence imaging [39] [42]. This approach leverages error-robust barcoding schemes that not only enable massive multiplexing but also provide built-in error detection and correction capabilities, ensuring highly accurate transcript identification and quantification [40] [44]. The technology has evolved significantly since its initial introduction, with MERFISH 2.0 incorporating optimized chemistry for sharper resolution, enhanced signal-to-noise ratios, and improved performance across diverse sample types, including archival FFPE and frozen samples [39]. These advancements have positioned MERFISH as a powerful tool for building spatially resolved cell atlases across diverse tissues and organisms, contributing substantially to efforts such as the Human Cell Atlas [45].

Core Technological Framework

MERFISH employs a sophisticated two-step labeling process that enables highly multiplexed RNA imaging. In the first step, cellular RNAs are hybridized with encoding probes - unlabeled DNA oligonucleotides containing a region complementary to the target RNA (targeting sequence) and a barcode region comprised of custom binding sites (readout sequences) [41]. The specific combination of readout sequences associated with each RNA species determines its unique optical barcode. In the second step, these barcodes are read through sequential rounds of smFISH using fluorescently labeled readout probes complementary to the readout sequences [41] [44]. This strategic separation of encoding and detection steps allows for rapid barcode readout while maintaining the binding redundancy provided by many tens of probes per RNA molecule, resulting in very high molecular detection efficiencies [41].

The error-robust nature of MERFISH stems from the use of combinatorial barcoding schemes with built-in error detection and correction capabilities. The most commonly used encoding schemes are modified Hamming distance codes that allow for the identification and correction of barcode reading errors that may occur during the sequential imaging process [42] [40]. For example, the 16-bit modified Hamming distance 4 (MHD4) code includes 140 unique barcodes and can detect and correct single-bit errors, ensuring highly accurate transcript identification even when imaging hundreds to thousands of genes simultaneously [42] [40]. This error robustness, combined with high detection efficiency, makes MERFISH particularly valuable for quantitative spatial transcriptomics and cell atlas construction.

Performance Metrics and Benchmarking

Recent systematic benchmarking studies have evaluated MERFISH performance alongside other high-throughput spatial transcriptomics platforms. When compared to sequencing-based methods, MERFISH demonstrates superior sensitivity and reduced dropout rates while maintaining strong correlation with both bulk and single-cell RNA sequencing results [43]. The technology quantitatively reproduces bulk RNA-seq and scRNA-seq data with improved detection limits, enabling accurate cell type identification without mandatory computational integration with scRNA-seq atlases [43].

Table 1: Performance Comparison of High-Throughput Spatial Transcriptomics Platforms

Platform	Technology Type	Gene Panel Size	Spatial Resolution	Key Strengths
MERFISH	Imaging-based (smFISH)	Hundreds to thousands of genes	Subcellular (nanometer scale)	High detection efficiency, single-molecule sensitivity, accurate quantification
Xenium 5K	Imaging-based	5,001 genes	Single-molecule	High sensitivity, strong correlation with scRNA-seq
CosMx 6K	Imaging-based	6,175 genes	Single-molecule	Large gene panel, subcellular resolution
Visium HD FFPE	Sequencing-based	18,085 genes	2 μm	Whole transcriptome, high multiplexing capability
Stereo-seq v1.3	Sequencing-based	Whole transcriptome	0.5 μm	Highest spatial resolution for sequencing-based methods

In evaluations of molecular capture efficiency across entire gene panels, MERFISH demonstrates strong performance in sensitivity and specificity metrics [46]. The platform's high detection efficiency (typically 80-90% for appropriately sized gene panels) and single-molecule resolution make it particularly suitable for identifying rare cell types and detecting subtle transcriptional variations within tissues - critical capabilities for comprehensive cell atlas construction [42] [43].

Table 2: MERFISH Performance Characteristics for Cell Atlas Construction

Performance Parameter	Specification	Impact on Cell Atlas Construction
Detection Efficiency	80-90% for standard panels [42]	Comprehensive cell type representation
Spatial Resolution	Subcellular (nanometer scale) [39]	Precise cellular and subcellular localization
Multiplexing Capacity	Up to thousands of genes [39]	Deep molecular profiling of cell states
Sample Compatibility	FFPE, frozen, cultured cells [39]	Diverse tissue sources for atlas building
Transcript Quantification	Single-molecule accuracy [43]	Accurate measurement of gene expression levels
Cell Throughput	Tens of thousands of cells per measurement [42]	Statistical power for rare cell type identification

Experimental Design and Protocol Optimization

Sample Preparation and Handling

Proper sample preparation is critical for successful MERFISH experiments and high-quality cell atlas construction. The process begins with careful tissue collection and fixation to preserve RNA integrity while maintaining tissue morphology. For most applications, immediate fixation in freshly prepared 4% paraformaldehyde (PFA) in nuclease-free PBS for 24 hours at 4°C is recommended, followed by thorough washing and storage in PBS or ethanol at -80°C until use [43]. Different tissue types may require specific optimization; for example, bone marrow samples need extensive optimization to preserve tissue quality and concomitantly maintain RNA integrity [39], while pancreatic tissues with high RNase content require specialized handling to overcome technical constraints [39].

For archival FFPE samples, which are invaluable for constructing retrospective atlases from biobanked tissues, specific processing protocols have been established. Tissues should be fixed for precisely 24 hours, dehydrated through graded ethanol series, cleared in xylene, and embedded in paraffin using standard histological protocols [39]. Sectioning thickness typically ranges from 5-10 μm, with careful attention to minimizing wrinkles and tears that can complicate imaging. Sections are mounted on specially treated coverslips or slides that promote adhesion while minimizing background fluorescence [39] [44]. Prior to hybridization, FFPE sections require deparaffinization with xylene or alternative clearing agents, rehydration through graded ethanol series, and antigen retrieval using proteinase K or heat-induced epitope retrieval methods to expose target RNAs [39].

Protocol Optimization for Enhanced Performance

Recent systematic investigations have identified key parameters for optimizing MERFISH performance through modifications to probe design, hybridization conditions, and imaging buffers [41]. Signal brightness, which directly impacts detection efficiency, depends on multiple factors including target region length, hybridization conditions, and buffer composition [41].

Table 3: Optimized Experimental Parameters for MERFISH

Parameter	Optimal Condition	Effect on Performance
Target Region Length	30-50 nt [41]	Balanced assembly efficiency and specificity
Hybridization Duration	1 day at 37°C [41]	Maximized probe binding while minimizing degradation
Formamide Concentration	Length-dependent optimization [41]	Enhanced signal-to-noise ratio
Imaging Buffer Composition	Newly optimized buffers [41]	Improved photostability and effective brightness
Encoding Probe Design	80 probes per RNA [41]	High detection efficiency without significant improvement from further increases

For high-density RNA libraries containing numerous abundant transcripts, combining MERFISH with expansion microscopy (ExM) substantially increases the total density of measurable RNAs [42]. In this approach, mRNAs are anchored to an expandable polyelectrolyte gel via acrydite-modified poly(dT) locked nucleic acid (LNA) probes hybridized to the poly(A) tail, followed by physical expansion of the gel [42]. This technique increases the distance between neighboring RNA molecules, reducing signal overlap and increasing the detection efficiency from approximately 20% in non-expanded samples to near 100% in expanded samples for high-abundance libraries [42].

The matrix imprinting and clearing approach has proven particularly valuable for MERFISH measurements in complex tissues [44]. This method involves embedding samples in polyacrylamide, anchoring RNAs to this matrix, and then clearing cellular proteins and lipids - major sources of off-target probe binding and autofluorescence [44]. The process significantly improves the signal-to-background ratio, detection efficiency, and detection limit of MERFISH, enabling high-performance measurements in tissues such as the mouse brain [44].

The Scientist's Toolkit: Essential Research Reagents

Successful implementation of MERFISH for cell atlas construction requires careful selection of reagents and materials. The following table outlines the essential components of the MERFISH toolkit:

Table 4: Essential Research Reagents for MERFISH Experiments

Reagent Category	Specific Examples	Function and Importance
Encoding Probes	Custom-designed DNA oligonucleotides with targeting and readout sequences [41]	Impart unique binary barcodes to each RNA species for multiplexed detection
Readout Probes	Fluorescently labeled oligonucleotides (Cy3, Cy5, Alexa Fluor dyes) [41]	Bind to readout sequences in sequential rounds to read out barcodes
Anchoring Probes	Acrydite-modified poly(dT) LNA probes [42]	Anchor polyadenylated RNAs to polymer matrix for clearing or expansion
Embedding Matrix	Polyacrylamide or expandable polyelectrolyte gel [42] [44]	Provide support for RNA anchoring and enable clearing or expansion
Clearing Reagents	Proteinase K, detergents (SDS, Triton X-100) [44]	Remove proteins and lipids to reduce background fluorescence
Hybridization Buffers	Formamide-containing buffers with salts and crowding agents [41]	Control stringency of hybridization to balance specificity and efficiency
Imaging Buffers	Optimized buffers with oxygen scavenging systems [41]	Enhance fluorophore photostability and signal brightness during imaging
Cell Segmentation Markers	DAPI, membrane antibodies (Na+/K+ ATPase, β-catenin) [43]	Define nuclear and cellular boundaries for single-cell analysis
Tomaymycin	Tomaymycin, CAS:35050-55-6, MF:C16H20N2O4, MW:304.34 g/mol	Chemical Reagent
DL-Mevalonolactone	DL-Mevalonolactone, CAS:503-48-0, MF:C6H10O3, MW:130.14 g/mol	Chemical Reagent

Data Analysis and Computational Integration for Atlas Construction

Image Processing and Transcript Decoding

The computational pipeline for MERFISH data analysis begins with image processing to extract single-molecule information from the raw fluorescence images. This process involves several sequential steps: (1) image registration to align images from different imaging rounds and channels; (2) background subtraction to reduce autofluorescence and non-specific signals; (3) spot detection to identify potential RNA molecules in each imaging round; and (4) barcode decoding to assign transcriptional identities based on the temporal fluorescence pattern of each detected spot [43]. Specialized software packages such as MERlin have been developed to automate these processing steps while providing tools for quality control and parameter optimization [43].

A critical challenge in MERFISH data analysis is accurate cell segmentation, which defines cellular boundaries to assign transcripts to individual cells. This is typically achieved by combining information from nuclear stains (DAPI) and membrane markers (antibodies against membrane proteins) [43]. Advanced segmentation algorithms like CellPose are often employed to precisely delineate cell boundaries, especially in complex tissues where cells are tightly packed and exhibit irregular morphologies [43]. The output of this analysis is a single-cell spatial expression matrix containing transcript counts for each gene in each segmented cell, along with spatial coordinates for both cells and individual transcripts [40] [43].

Cell Type Identification and Spatial Analysis

Once transcript counts are assigned to cells, standard single-cell bioinformatics approaches are applied to identify cell types and states within the tissue. This typically involves normalization of expression counts, feature selection, dimensionality reduction (PCA, UMAP), and clustering to identify transcriptionally distinct cell populations [43]. The spatial context provided by MERFISH enables additional analyses not possible with dissociated single-cell RNA sequencing, including identification of spatially restricted cell types, characterization of cellular neighborhoods, and analysis of expression gradients across tissue architectures [43].

Computational integration with existing single-cell RNA sequencing atlases can provide additional context for cell type annotation, though studies have shown that MERFISH data alone can sufficiently resolve distinct cell types in many tissues [43]. For example, in mouse liver and kidney tissues, MERFISH with a 307-gene panel independently identified distinct cell types and revealed clear spatial patterning, such as the zonation of hepatocytes and endothelial cells in the liver and the organization of podocytes in the kidney [43]. Integration methods should be applied judiciously, as intrinsic differences in RNA statistics between MERFISH and scRNA-seq can sometimes lead to misclassification when integration is forced [43].

Multi-Slice Integration and Spatial Alignment

Construction of comprehensive cell atlases often requires integration of data from multiple tissue sections. Recent benchmarking studies have evaluated various multi-slice integration methods for spatial transcriptomics data, identifying several high-performing approaches [47]. Deep learning-based methods such as GraphST and SPIRAL, statistical methods including Banksy and MENDER, and hybrid approaches like CellCharter and STAligner have demonstrated capabilities to integrate spatial and expression data while mitigating technical artifacts across slices [47].

The performance of these integration methods is highly dependent on application context, dataset size, and technology [47]. For spatial clustering, methods like SpaDo, STAIG, and MENDER excel at preserving biological variance, while for batch effect correction, GraphST-PASTE shows particular effectiveness [47]. The selection of integration strategies should be guided by the specific analytical goals and tissue characteristics, with particular attention to the strong interdependencies between upstream integration quality and downstream analysis outcomes [47].

Applications in Cell Atlas Construction

MERFISH has been successfully applied to construct detailed cell atlases across diverse tissues and organisms, contributing significantly to our understanding of tissue organization and cellular heterogeneity. In neuroscience, MERFISH has been used to define and discover cell types in individual brain regions and entire brains of both mouse and human, revealing unprecedented diversity in neural cell types and their spatial organization [41]. These atlases have provided insights into the molecular architecture of neural circuits and the spatial patterns of gene expression that underlie brain function and development.

Beyond the nervous system, MERFISH has been extended to a broad range of tissues including liver, gut, heart, and retina, as well as other organisms such as bacteria and plants [41]. In many cases, the high sensitivity of MERFISH has revealed cell types and states not observed in previous studies. For example, a recent application of MERFISH in gut inflammation discovered a striking and unanticipated diversity of activated fibroblast states, revealing previously underappreciated cellular heterogeneity in the inflammatory microenvironment [41]. Similarly, in cancer biology, MERFISH has enabled detailed characterization of tumor microenvironments, uncovering spatial relationships between cancer cells, immune cells, and stromal components that inform our understanding of tumor biology and therapeutic responses [39] [46].

The construction of high-resolution cell atlases using MERFISH provides a foundation for investigating developmental processes, tissue homeostasis, disease mechanisms, and therapeutic interventions. By preserving spatial context while measuring gene expression at single-cell resolution, MERFISH enables researchers to map the precise spatial distribution of cell types, identify transitional states, characterize cellular neighborhoods, and uncover spatial patterns of gene expression that would be obscured in dissociated single-cell analyses. As the technology continues to evolve with improvements in multiplexing capacity, sensitivity, and throughput, its impact on cell atlas construction across diverse tissues and biological contexts is expected to grow substantially.

MERFISH represents a powerful platform for spatially resolved cell atlas construction, combining single-molecule sensitivity with high multiplexing capability to map transcriptional identities within native tissue contexts. The technology's robust performance, compatibility with diverse sample types, and continuous methodological improvements position it as an essential tool in the spatial transcriptomics arsenal. As optimization of probe design, hybridization conditions, and imaging buffers continues to enhance performance [41], and as computational methods for multi-slice integration and spatial analysis mature [47], MERFISH is poised to drive increasingly comprehensive and detailed cell atlas projects across tissues, organisms, and biological conditions. The integration of MERFISH with complementary spatial technologies and the development of standardized analytical frameworks will further strengthen its contribution to constructing reference atlases that illuminate the spatial architecture of tissues at single-cell resolution.

Decoding the Tumor Immune Microenvironment (TIME) for Immunotherapy Biomarker Discovery

The tumor immune microenvironment (TIME) is a complex and dynamic ecosystem composed of tumor cells, immune cells (such as T cells, B cells, macrophages, and dendritic cells), stromal cells, cytokines, and extracellular matrix components. The spatial organization and functional states of these elements critically influence tumor progression, immune evasion, and therapeutic responsiveness to immunotherapy [48]. Despite the transformative success of immune checkpoint inhibitors (ICIs) in treating various solid tumors, a significant proportion of patients do not respond or develop resistance due to the highly heterogeneous and immunosuppressive nature of the TIME [49] [50].

Traditional immunotherapy biomarkers, such as programmed death-ligand 1 (PD-L1) expression assessed by immunohistochemistry (IHC), have provided initial guidance for patient selection but demonstrate major limitations. These include tumor heterogeneity, sampling constraints, and an inability to reflect the dynamic interplay between the tumor and the host's immune system during treatment [49]. Consequently, there is a growing need for advanced multiplexed analytical approaches that can comprehensively map the spatial architecture of the TIME to identify more robust predictive biomarkers [48] [50].

Multiplexed fluorescence imaging has emerged as a powerful technological platform that enables simultaneous visualization of dozens of protein biomarkers at single-cell resolution within intact tissue sections, preserving critical spatial information. By decoding the intricate spatial relationships and functional states of immune and tumor cells, these methods provide unprecedented insights into mechanisms of therapy response and resistance, ultimately guiding the development of more effective immunotherapeutic strategies [48] [51].

Multiplexed Imaging Technologies for Spatial TIME Analysis

Technology Comparison and Selection

Cutting-edge multiplex imaging technologies have significantly advanced our understanding of the TIME by providing nanometer-scale spatial resolution that illuminates previously inaccessible cellular interactions and organizational patterns [48]. The table below compares the major multiplex imaging platforms used for spatial analysis of the TIME:

Table 1: Comparison of Multiplex Imaging Technologies for Spatial TIME Analysis

Technology	Multiplexing Capacity	Spatial Resolution	Key Advantages	Primary Limitations	Best Suited Applications
Imaging Mass Cytometry (IMC) [48]	~40 markers	Subcellular	Minimal spectral overlap, accurate quantification	Requires specialized instrumentation	Deep immune phenotyping, cell state identification
Multiplexed Ion Beam Imaging (MIBI) [48]	~40 markers	Subcellular	Superior specificity, minimal background	Low throughput, expensive	High-plex protein mapping, signaling studies
Cyclic Immunofluorescence (CycIF) [48] [50]	50+ markers	Single-cell	Broadly accessible, conventional microscopes	Labor-intensive, long imaging times	Large-area spatial profiling, whole-slide imaging
CODEX [48] [50]	60+ markers	Single-cell	High spatial precision, excellent tissue preservation	Technically demanding, specialized analysis	Immune-tumor spatial relationships, cellular neighborhoods
Digital Spatial Profiling (DSP) [48]	100+ targets (protein/RNA)	Region-based (10-600μm)	Targeted profiling, region-specific analysis	Lower spatial resolution	Biomarker discovery, region-of-interest analysis
NanoPlex [52]	21+ targets	Super-resolution (STED/dSTORM)	Universal applicability, signal erasure strategies	New methodology, limited validation	High-resolution multiplexing, standard lab equipment
Multiplex IHC/IF [51] [53]	6-8 markers	Single-cell	Clinically accessible, standardized protocols	Limited multiplexing capacity	Routine pathology, translational research

Implementation Workflows

The experimental workflow for cyclic multiplex immunofluorescence imaging typically begins with formalin-fixed paraffin-embedded (FFPE) tissue slides undergoing deparaffinization and antigen retrieval. Each cycle consists of primary antibody labeling, fluorescent detection, image acquisition, and fluorophore stripping or inactivation before proceeding to the next cycle [50]. The resulting images from multiple cycles are computationally aligned (registered) to generate a spatially resolved dataset containing dozens of protein markers [50].

Diagram: Cyclic Multiplex Immunofluorescence Workflow

For oligonucleotide-based technologies like CODEX, the workflow involves labeling antibodies with unique DNA barcodes rather than fluorophores, followed by sequential hybridization with fluorescently labeled complementary probes and imaging [48]. This approach enables higher multiplexing while maintaining excellent tissue preservation.

Key Spatial Biomarkers in the TIME

Functional Immune Cell States

Multiplex imaging allows precise characterization of immune cell subsets based on simultaneous detection of multiple markers, distinguishing activated, exhausted, regulatory, or effector immune phenotypes [48]. Key functional T-cell states include:

• Activated CD8+ T-cells: Identified by expression of granzyme B, perforin, or interferon-γ, typically associated with effective immune surveillance and favorable responses to immunotherapy [48].
• Exhausted T-cells: Characterized by co-expression of multiple inhibitory receptors such as PD-1, TIM-3, LAG-3, and TIGIT, often associated with progressive dysfunction and poor prognosis [49] [48].
• Regulatory T-cells (Tregs): Defined by expression of FoxP3, CD4, and CD25, these cells suppress antitumor immunity and are generally associated with immune suppression and resistance to immunotherapy [49] [48].

Spatial Distribution Patterns

The spatial organization of immune cells relative to tumor cells and stromal components provides critical insights into functional immune states and therapeutic vulnerabilities:

• Immune-Inflamed Pattern: Characterized by dense T-cell infiltration into tumor parenchyma with close proximity between CD8+ T-cells and tumor cells, generally associated with favorable responses to checkpoint inhibitors [48] [54].
• Immune-Excluded Pattern: Features T-cells accumulated at the tumor margin but limited infiltration into the tumor core, often predicting resistance to immunotherapy despite the presence of immune cells [48].
• Immune-Desert Pattern: Demonstrates minimal T-cell presence in either the tumor parenchyma or stroma, suggesting lack of immune recognition and generally poor response to single-agent checkpoint blockade [48].

Table 2: Key Spatial Biomarkers and Their Clinical Significance in Immunotherapy

Spatial Biomarker	Description	Assessment Method	Predictive Value	Associated Cancers
CD8+ T-cell Density & Proximity [48] [54]	Density of CD8+ T-cells within specific distances from tumor cells (e.g., <20μm)	Distance measurement in multiplex images	Favorable response to ICIs	Melanoma, NSCLC, esophageal
Tertiary Lymphoid Structures (TLS) [49] [48]	Organized clusters of B cells, T cells, and dendritic cells	Multiplex IHC/IF for CD20, CD3, CD21, PNAd	Positive predictor for ICIs and prognosis	Various solid tumors, sarcoma
Myeloid Cell Infiltrates [49] [51]	Density and polarization of macrophages (M1/M2) and MDSCs	IMC, MIBI, multiplex IF for CD68, CD163, ARG1	Generally associated with resistance	NSCLC, melanoma, MCC
T-cell Exhaustion Signature [49] [48]	Co-expression of multiple inhibitory checkpoints	Multiplex imaging for PD-1, TIM-3, LAG-3	Resistance to PD-1/PD-L1 blockade	Various solid tumors
Heterotypic T-cell Clusters [54]	Physical conjugates between CD8+ T-cells and tumor cells/APCs	Imaging flow cytometry, multiplex IF	Enriched for tumor-reactive T-cells	Melanoma, multiple other cancers

Emerging Soluble Biomarkers

Beyond tissue-based spatial biomarkers, soluble immune checkpoints in peripheral blood offer minimally invasive opportunities for monitoring dynamic changes during therapy. For instance, soluble Tim-3 (sTim-3) has been identified as a promising prognostic marker and mediator of resistance to PD-1 blockade [55]. Elevated serum sTim-3 levels correlate with non-response to anti-PD-1 therapy in non-small cell lung cancer (NSCLC) and various digestive tumors, promoting CD8+ T-cell exhaustion through CEACAM-1 interaction [55].

Detailed Experimental Protocols

Protocol 1: Multiplex Immunofluorescence for Human Tumor TME Analysis

This protocol adapts established methodologies for comprehensive immune profiling in human tumors, particularly Merkel cell carcinoma (MCC), utilizing tyramide signal amplification (TSA) for sensitive detection of multiple markers [51].

Materials and Reagents

Table 3: Research Reagent Solutions for Human MCC mIF Panel

Reagent	Specific Role	Example Vendor	Catalog Number	Recommended Dilution
Opal 6-Plex Manual Detection Kit	Fluorophore system for multiplex detection	Akoya Biosciences	N/A	As per manufacturer
Anti-CD163 antibody	Macrophage marker	Leica	NCL-L-CD163	1:200
Anti-FoxP3 antibody	Regulatory T-cell marker	Abcam	ab20034	1:250
Anti-CD8 antibody	Cytotoxic T-cell marker	Abcam	ab178089	1:100
Anti-PD-L1 antibody	Immune checkpoint marker	Cell Signaling	13684	1:200
Anti-Sox2 antibody	MCC tumor cell marker	Invitrogen	14-9811-82	1:1000
Anti-CD3 antibody	Pan-T-cell marker	Agilent Dako	A0452	1:100
DAPI	Nuclear counterstain	Multiple	Multiple	As per manufacturer

Step-by-Step Procedure

• Tissue Preparation: Cut 4-6μm sections from FFPE blocks onto charged slides. Air-dry overnight at room temperature and store at 4°C with desiccants until use [51].

• Deparaffinization and Antigen Retrieval:

  • Bake slides at 60°C for 30 minutes to improve adhesion.
  • Deparaffinize in xylene (3 changes, 5 minutes each) followed by ethanol gradient (100%, 95%, 70%; 3 minutes each) and rinse in distilled water.
  • Perform antigen retrieval using pH 6.0 citrate buffer for CD163, Sox2, and CD3 or pH 9.0 EDTA buffer for FoxP3, CD8, and PD-L1 [51].

• Antibody Staining Cycles:

  • Block endogenous peroxidase with 3% hydrogen peroxide for 10 minutes.
  • Apply protein block (e.g., 10% normal goat serum) for 10 minutes at room temperature.
  • Apply primary antibody diluted in antibody diluent for 1 hour at room temperature or overnight at 4°C.
  • Apply HRP-conjugated secondary antibody for 10 minutes at room temperature.
  • Apply Opal fluorophore reagent for 10 minutes at room temperature.
  • Perform microwave treatment (approximately 2 minutes at 100°C) in citrate buffer to strip antibodies before the next cycle [51].

• Counterstaining and Mounting:

  • After all staining cycles, apply DAPI solution for 5 minutes at room temperature.
  • Rinse with distilled water and mount with Prolong Diamond antifade mountant.
  • Cure slides overnight in the dark at room temperature before imaging [51].

• Image Acquisition and Analysis:

  • Acquire multispectral images using a system such as the Vectra Polaris.
  • Perform spectral unmixing to separate individual fluorophore signals.
  • Utilize image analysis software (e.g., InForm, HALO, QuPath) for cell segmentation and phenotyping [51].

Protocol 2: NanoPlex for High-Plex Fluorescence Imaging

NanoPlex provides a universal strategy for fluorescence microscopy multiplexing using nanobodies with erasable signals, enabling high-plex imaging on conventional microscopes without specialized hardware [52].

Materials and Reagents

• Primary antibodies against targets of interest
• Engineered secondary nanobodies (OptoPlex, EnzyPlex, or ChemiPlex variants)
• NbALFA conjugated to appropriate fluorophores
• Washing buffer (PBS with 0.05% Tween-20)
• Cleavage reagents: UV light source (365nm LED for OptoPlex), specific enzyme (for EnzyPlex), or commercial chemical (for ChemiPlex)

Step-by-Step Procedure

• Sample Preparation:

  • Culture cells on appropriate substrates or prepare tissue sections as described in Protocol 4.1.
  • Fix with 4% paraformaldehyde for 15 minutes and permeabilize with 0.1% Triton X-100 if intracellular targets are assessed.

• Pre-complex Formation:

  • For each target, pre-mix primary antibody with Opto-2.Nb and NbALFA-fluorophore in a 1:3:4 molar ratio.
  • Incubate for 15-30 minutes at room temperature to form complexes before application [52].

• Staining Cycle 1:

  • Apply pre-formed complexes to samples and incubate for 1 hour at room temperature.
  • Wash 3 times with washing buffer to remove unbound complexes.
  • Acquire images for all channels using appropriate microscope settings.

• Signal Erasure:

  • For OptoPlex: Illuminate entire field of view with 365nm LED light (~60 mW/cm²) for 5-15 minutes while perfusing with washing buffer to remove cleaved fluorophores [52].
  • For EnzyPlex: Apply specific enzyme solution for 15-30 minutes at room temperature.
  • For ChemiPlex: Apply commercial chemical cleavage reagent as per manufacturer instructions.

• Validation of Signal Removal:

  • Re-image the same field of view to confirm complete signal removal before proceeding to the next cycle.
  • Repeat steps 2-4 for each subsequent staining cycle with different target panels.

• Image Processing and Analysis:

  • Register all cycle images to generate a composite high-plex dataset.
  • Perform quantitative analysis as described in Protocol 4.1.

Diagram: NanoPlex Signal Erasure Workflow

Data Analysis and Interpretation

Quantitative Spatial Analysis

The rich datasets generated by multiplex imaging require specialized analytical approaches to extract biologically and clinically meaningful insights:

• Cell Segmentation and Phenotyping: Utilize nuclear (DAPI) and membrane/cytoplasmic markers to identify individual cells and assign phenotypic identities based on marker expression thresholds [51].
• Spatial Neighborhood Analysis: Define cellular neighborhoods based on clustering patterns of different cell types and analyze their association with clinical outcomes [48].
• Distance Measurements: Calculate minimum distances between cell types of interest (e.g., CD8+ T-cells to nearest tumor cell) to quantify immune-tumor interactions [54] [53].
• Interaction Analysis: Quantify the frequency of specific cell-cell interactions or the presence of heterotypic clusters that may indicate functional immune engagements [54].

Correlation with Clinical Outcomes

To establish the clinical utility of spatial biomarkers, rigorous correlation with treatment response and survival outcomes is essential:

• Response Stratification: Compare spatial features (e.g., CD8+ T-cell density, spatial proximity metrics) between responders and non-responders to immunotherapy [48] [54].
• Survival Analysis: Evaluate the prognostic significance of spatial biomarkers using Kaplan-Meier analysis and Cox proportional hazards models [49] [48].
• Multivariate Models: Develop integrated models that combine spatial biomarkers with established clinical and pathologic factors to improve predictive accuracy [49].

Clinical Applications and Translation

Predictive Biomarker Discovery

Multiplex imaging has revealed several spatially-resolved biomarkers with significant potential for predicting immunotherapy response:

• In melanoma, the presence of heterotypic CD8+ T-cell clusters containing tumor cells and/or antigen-presenting cells identifies enriched populations of tumor-reactive T-cells with enhanced cytotoxic capability [54].
• Across multiple cancer types, close spatial proximity between CD8+ T-cells and tumor cells (within 20-100μm) consistently predicts favorable responses to immune checkpoint inhibitors [48] [54].
• Tertiary lymphoid structures (TLS) at the tumor invasive margin serve as organized hubs for immune activation and correlate with improved outcomes following immunotherapy [49] [48].

Therapy Monitoring and Resistance Mechanisms

Longitudinal assessment of the TIME using multiplex imaging enables monitoring of dynamic changes during therapy and identification of resistance mechanisms:

• Spatial analysis of on-treatment biopsies can reveal the emergence of immune-excluded phenotypes or compensatory upregulation of alternative immune checkpoints (e.g., TIM-3, LAG-3) following PD-1/PD-L1 blockade [49] [48].
• Increased distances between CD8+ T-cells and tumor cells after treatment initiation may indicate early progression or adaptive resistance [54].
• Shifts in myeloid cell populations and their spatial localization can identify developing immunosuppressive networks that limit therapeutic efficacy [49] [51].

Multiplexed fluorescence imaging technologies have fundamentally transformed our ability to decode the complex spatial architecture of the tumor immune microenvironment, moving beyond static, single-parameter biomarkers to dynamic, multidimensional assessments of immune function and organization. The protocols and applications detailed in this document provide researchers and clinicians with powerful tools to identify novel predictive biomarkers, elucidate mechanisms of therapy response and resistance, and ultimately guide more effective immunotherapeutic strategies.

As these technologies continue to evolve, several exciting directions promise to further enhance their utility: integration with other omics modalities such as spatial transcriptomics, increased automation and standardization to support clinical implementation, and the development of sophisticated computational approaches including artificial intelligence and machine learning for automated pattern recognition [49] [50]. Through continued refinement and validation, multiplexed imaging of the TIME holds tremendous potential to realize the promise of precision immuno-oncology, enabling optimal selection of patients for specific immunotherapies and monitoring of treatment efficacy in diverse cancer types.

Multiplexed fluorescence microscopy is an indispensable tool in biomedical research, enabling the simultaneous observation of multiple cellular components. However, a fundamental limitation constrains its potential: the spectral overlap of fluorophores. This crosstalk obscures signals and severely restricts the number of biomarkers that can be imaged at once. While traditional linear unmixing methods have been widely used, their performance is heavily dependent on accurately calibrated reference spectra, which can be difficult to obtain and are susceptible to environmental variation [56] [4].

Computational unmixing represents a paradigm shift, leveraging advanced algorithms to disentangle mixed signals. Techniques incorporating machine learning (ML) and novel signal processing dimensions are dramatically expanding multiplexing capabilities. This Application Note details how methods like SEPARATE and other ML-driven approaches are overcoming traditional limits, effectively doubling the multiplexing capacity of standard fluorescence microscopes without the need for costly hardware modifications. We provide a foundational protocol for implementing these powerful techniques, framed within the broader context of accelerating drug discovery and systems biology research.

Quantitative Performance Comparison of Unmixing Techniques

The table below summarizes the key performance metrics of several advanced computational unmixing methods, providing a direct comparison of their multiplexing capabilities and reconstruction quality.

Table 1: Performance Comparison of Advanced Unmixing Techniques

Method	Core Principle	Maximum Demonstrated Multiplexing	Key Metric (SSIM)	Required Prior Knowledge
AutoUnmix [56]	Asymmetric Autoencoder	4 colors (validated)	0.99 (3- & 4-color imaging)	None (Blind)
PICASSO [4]	Mutual Information Minimization	>15 colors in a single round	N/A	None (Blind)
BAMM [6]	Photobleaching Kinetics	+3 labels/single spectral channel	N/A	Bleaching fingerprints (can be self-calibrated)
Linear Unmixing [57]	Linear Inverse Problem	~4-5 colors	Varies with reference accuracy	Reference Emission Spectra
Phasor Analysis [5]	Fourier Transform & Clustering	N/A	N/A	None

Detailed Experimental Protocols

Protocol 1: Implementation of AutoUnmix for Blind Unmixing

This protocol utilizes a deep learning model to unmix 3-4 color images in a blind manner, without needing reference spectra.

Research Reagent Solutions & Materials

Table 2: Essential Materials for AutoUnmix

Item	Function/Description
Fluorophores	Organic dyes or fluorescent proteins (e.g., Alexa Fluor series, GFP, RFP).
Synthetic Datasets	For initial network training; images with known ground truth.
Python Environment	With PyTorch/TensorFlow and scientific computing libraries (NumPy, SciPy).
Pre-trained AutoUnmix Model	Asymmetric autoencoder network as described in [56].
GPU Workstation	Recommended for accelerated training and inference.

Step-by-Step Procedure

• Network Training (on Synthetic Data):

  • Input Preparation: Generate or acquire a dataset of mixed images y and their corresponding ground-truth unmixed images x.
  • Model Architecture: Implement the asymmetric autoencoder comprising:
    • Unmixing Network: Features a Spectral Learning Module (SLM) with a Channel Attention Mechanism (CAM) to learn inter-channel features, followed by a U-Net for spatial feature extraction [56].
    • Mixing Network: A lightweight U-Net that remixes the unmixed output to reconstruct the original input.
  • Loss Function: Train the network using a composite loss function (Equation 4, [56]) that combines Mean Squared Error (MSE) and Structural Similarity Index Measure (SSIM) loss between both the unmixed images and the ground truth (L_xy), and the re-mixed images and the original input (L_yx).
  • Transfer Learning (Optional): For application to a new imaging system, fine-tune the pre-trained model with a small set of real images from the target system to adapt the network parameters [56].

• Inference (on Experimental Data):

  • Image Acquisition: Acquire the multiplexed fluorescence image stack y with C channels.
  • Unmixing: Process the mixed image y through the trained unmixing network. The network outputs the unmixed image x_hat, where each channel corresponds to a pure fluorophore signal.
  • Validation: Qualitatively assess the unmixed images for morphological integrity and absence of crosstalk. Quantitative validation requires known ground truth, which may be simulated.

The following diagram illustrates the asymmetric autoencoder workflow of the AutoUnmix method:

Protocol 2: Implementation of PICASSO for Ultra-Multiplexed Imaging

This protocol uses an information theory-based approach to achieve very high levels of multiplexing (>15 colors) without reference spectra.

Research Reagent Solutions & Materials

Table 3: Essential Materials for PICASSO

Item	Function/Description
Spectrally Overlapping Fluorophores	Fluorophores excitable by the same laser but with varying emission profiles.
Primary Antibody-Fab Complexes	Pre-formed complexes to stain multiple primary antibodies from the same host species [4].
Bandpass Filter-Based Microscope	Standard fluorescence microscope; does not require specialized spectral detectors.

Step-by-Step Procedure

• Sample Preparation and Staining:

  • Assemble primary antibody–Fab fragment complexes for each target, with each Fab bearing a distinct fluorophore [4].
  • Apply all assembled antibody complexes to the specimen simultaneously.

• Image Acquisition:

  • Acquire N images of the specimen at N different spectral ranges, each encompassing the emission peak of one of the N fluorophores used.

• PICASSO Unmixing Algorithm:

  • Model: The relationship between acquired images (IMG) and pure fluorophore images (F) is modeled as IMG = A * F, where A is a mixing matrix with diagonal elements of 1 and off-diagonal elements α_i,j representing leakage from fluorophore j into the detection channel of i [4].
  • Initialization: Initialize the solution for each channel X_i(0) as the acquired image IMG_i.
  • Iterative Minimization: Update the solution iteratively to minimize the mutual information between all channel pairs. The core update equation subtracts a scaled version of other channels (Equation 2, [4]): X_i(k+1) = X_i(k) - γ * [ Σ (α_hat_i,j(k) * X_j(k)) ] for all j ≠ i where γ is a normalization factor and α_hat is the estimated leakage.
  • Convergence: Iterate until the mutual information between all unmixed channels is minimized, indicating successful separation.

The workflow and core computational process for PICASSO is outlined below:

The Scientist's Toolkit: Implementation Workflow

Integrating computational unmixing into a research pipeline involves careful planning from experimental design through data analysis. The following workflow provides a practical guide for scientists.

Critical Considerations for Success

• Fluorophore Selection: For methods like BAMM, select fluorophores with distinct photobleaching kinetics, which can be independent of their emission spectra [6]. For spectral methods, ensure fluorophores have a low similarity index (<0.98) for reliable separation [58].
• Control Samples: The accuracy of many unmixing methods hinges on high-quality reference controls. Follow the "Like-With-Like" principle: the autofluorescence of positive and negative control cells must be identical. For viability dyes, use heat-killed, stained cells as positives and heat-killed, unstained cells as negatives [58].
• Algorithm-Specific Requirements:
  • BAMM/NI-NMF: Acquire time-lapse data under continuous illumination to capture complete bleaching curves. Use the Non-Increasing Non-negative Matrix Factorization (NI-NMF) algorithm to enforce physically plausible (monotonically decreasing) bleaching traces [6].
  • Linear Unmixing: Precisely measured reference spectra from control samples are mandatory. Even small errors in reference spectra can lead to significant unmixing artifacts [4] [57].
  • Deep Learning Models (AutoUnmix): These require a substantial dataset for training. Using synthetically generated data for pre-training, followed by transfer learning with a small set of real images, is an effective strategy to overcome the scarcity of experimental training data [56] [59].

Computational unmixing methods are fundamentally expanding the horizons of multiplexed fluorescence imaging. By treating spectral crosstalk not as an insurmountable barrier but as a solvable computational problem, techniques like SEPARATE, PICASSO, and AutoUnmix are enabling researchers to extract significantly more information from a single sample. The protocols detailed herein provide a roadmap for integrating these powerful approaches into existing workflows, empowering researchers in drug development and systems biology to achieve unprecedented multiplexing capacity, thereby accelerating the pace of discovery.

Optimizing Performance: Protocols, Conjugations, and Probe Design

Systematic Protocol Optimization for Enhanced Signal-to-Noise in RNA FISH

Within the broader scope of multiplexed fluorescence imaging research, achieving a high signal-to-noise ratio (SNR) is paramount for the accurate detection and quantification of RNA molecules. Techniques like single-molecule RNA Fluorescence In Situ Hybridization (smFISH) and its highly multiplexed derivatives, such as Multiplexed Error-Robust FISH (MERFISH), rely on the precise optimization of experimental parameters to maximize specific binding and minimize background fluorescence [41] [60]. This application note provides a systematic framework for optimizing RNA FISH protocols, consolidating recent findings on probe design, hybridization conditions, buffer composition, and image analysis to enhance SNR for more reliable single-cell and spatial transcriptomic data.

Systematic Optimization Strategies

Optimizing RNA FISH is a multi-faceted process. The table below summarizes the key parameters and their demonstrated impact on the signal-to-noise ratio.

Table 1: Key Parameters for Systematic Optimization of RNA FISH Signal-to-Noise Ratio

Optimization Parameter	Impact on Signal-to-Noise Ratio	Recommended Approach
Probe Design & Specificity	Increases on-target signal and reduces off-target binding and false positives [61].	Use genome-wide BLAST-based tools (e.g., TrueProbes) to assess off-target binding; employ 30-50 nt target regions for robust hybridization [41] [61].
Hybridization Conditions	Directly influences probe assembly efficiency and brightness of single-molecule signals [41].	Systematically screen formamide concentration (e.g., 10-30%) at a fixed temperature (e.g., 37°C) to identify the optimal stringency [41].
Buffer Composition & Reagent Stability	Affects fluorophore photostability and longevity, preserving signal brightness over multi-round imaging [41].	Utilize new, optimized imaging buffers to improve photostability; address reagent "aging" during long experiments [41].
Image Analysis & Spot Detection	Enables accurate, automated quantification of RNA spots while minimizing human bias and false positive/negative calls [62].	Implement automated 3D spot detection software (e.g., TrueSpot) that uses robust thresholding algorithms [62].
Evonimine	Evonimine, MF:C36H43NO17, MW:761.7 g/mol	Chemical Reagent
Xanthiside	Xanthiside, MF:C17H23NO8S, MW:401.4 g/mol	Chemical Reagent

Probe Design for Enhanced Specificity

Probe design is the foundational step for achieving high specificity and sensitivity. While traditional tools apply heuristic filters, newer platforms like TrueProbes leverage genome-wide BLAST analysis and thermodynamic modeling to rank probes by their predicted specificity [61]. This integrated approach minimizes off-target interactions by quantitatively considering factors such as:

• Expressed off-target binding: Evaluating the potential for probes to bind to other transcripts, optionally weighted by gene expression data from the target tissue.
• Binding energy differences: Selecting probes with strong on-target affinity ((\Delta G^\circ{on})) and weak off-target affinity ((\Delta G^\circ{off})).
• Probe self-interaction: Filtering out sequences prone to self-hybridization or cross-dimerization within the probe set [61].

Regarding probe length, empirical data shows that for target regions between 20 and 50 nucleotides, the signal brightness—a proxy for probe assembly efficiency—depends only weakly on length within the optimal formamide concentration range [41]. This suggests that for many applications, a length of 30-40 nt provides a practical balance between specificity and efficient hybridization.

Optimization of Hybridization and Imaging Conditions

The biochemical environment during hybridization and imaging is a critical determinant of SNR. A systematic exploration of protocol choices for MERFISH has yielded several key optimizations:

• Hybridization Stringency: The efficiency with which encoding probes assemble onto target RNAs is modulated by denaturing conditions. Testing a range of formamide concentrations (e.g., from 10% to 30%) at a fixed temperature of 37°C can identify the concentration that maximizes single-molecule signal brightness without compromising specificity [41].
• Buffer Composition: The performance of fluorophores across multiple rounds of imaging is highly dependent on the imaging buffer. Recent protocol modifications have introduced new buffer formulations that improve photostability and effective brightness for commonly used dyes, directly enhancing the SNR in prolonged experiments [41].
• Reagent Stability: Signal intensity can degrade over the course of a multi-day MERFISH experiment due to reagent "aging." Specific protocol modifications have been developed to ameliorate this effect, ensuring consistent performance from the first to the last imaging round [41].

Advanced Image Analysis for Accurate Quantification

Robust computational tools are essential for converting raw image data into quantitative biological insights. TrueSpot is an automated software tool designed for high-throughput batch processing of RNA-FISH images, which directly addresses the challenge of SNR quantification [62].

• 3D Spot Detection: Unlike many deep-learning tools that operate only on 2D maximum intensity projections, TrueSpot performs spot detection directly on 3D image stacks, preventing information loss and improving localization accuracy in three-dimensional space [62].
• Automated Threshold Selection: A major challenge in automated image analysis is selecting an intensity threshold that separates true signal from background. TrueSpot uses a flexible algorithm that combines an analysis of the spot count curve with a sliding window Fano Factor calculation to automatically determine an optimal, data-driven threshold for each image [62]. This eliminates user bias and ensures consistency across large datasets.
• Performance Validation: The tool's performance is benchmarked using metrics like recall (sensitivity), precision, and F-Score, demonstrating superior performance compared to other existing tools in automatically detecting true signal puncta while minimizing false positives [62].

The Scientist's Toolkit: Essential Reagents and Software

Successful implementation of a high-SNR RNA FISH protocol relies on a suite of specialized reagents and computational tools.

Table 2: Key Research Reagent Solutions and Software for RNA FISH

Item	Function / Application
Encoding Probes	Unlabeled DNA oligonucleotides containing a target-specific region and a readout sequence barcode; form the core of multiplexed methods like MERFISH [41] [60].
Fluorescent Readout Probes	Short, fluorescently labeled oligonucleotides that bind to readout sequences on encoding probes; enable sequential barcode readout [41].
Formamide	A chemical denaturant used to control stringency in hybridization buffers; concentration must be optimized for each probe set [41] [63].
Deionized Formamide	Used in hybridization buffers to reduce background fluorescence [63].
Dextran Sulfate	A volume excluder used in hybridization buffers to increase effective probe concentration and hybridization kinetics [63].
TrueProbes	A computational software platform for designing high-specificity smFISH probe sets using genome-wide off-target binding analysis [61].
TrueSpot	An automated, open-source image analysis tool for detecting and quantifying RNA-FISH signal puncta in 2D or 3D images with minimal manual input [62].
QuPath	An open-source platform for whole-slide image analysis, capable of cell detection and subcellular RNA dot quantification for assays like RNAscope [64].

Experimental Workflow and Signaling Pathways

The following diagram illustrates the logical workflow for a systematic approach to RNA FISH optimization, integrating the key strategies discussed in this note.

Detailed Protocol: smFISH in Yeast as a Model System

The following protocol, optimized for S. cerevisiae, outlines key steps for robust RNA FISH and can be adapted for other cell types [63].

Sample Preparation and Fixation

• Cell Culture and Harvest: Grow yeast cells to mid-log phase in appropriate medium (e.g., Synthetic Complete (SC) medium). Fix cells by adding 32% paraformaldehyde (PFA) directly to the culture to a final concentration of 4% and incubating for 45 minutes at room temperature with gentle rotation [63].
• Cell Wall Digestion: Pellet fixed cells and resuspend in 1x Buffer B supplemented with 1.2 M sorbitol. Add lyticase (e.g., from Arthrobacter luteus) to digest the cell wall. The digestion time (typically 5-15 minutes) must be optimized for each strain to achieve spheroplasting without causing excessive cell lysis [63].
• Adhesion to Slides:
  • Apply poly-L-lysine to clean coverslips to promote cell adhesion.
  • Spot the spheroplasted cells onto the prepared coverslips and allow them to settle.
  • Permeabilize the cells by immersing the coverslips in 70% ethanol for at least 1 hour at -20°C. Samples can be stored in 70% ethanol at -20°C for several months.

Hybridization and Washes

• Probe Hybridization:
  • Prepare a hybridization mixture containing the smFISH probes (e.g., Stellaris probes), 10% dextran sulfate, 10% deionized formamide, and 2x SSC [63].
  • Apply the hybridization mixture to the coverslip and incubate in a dark, humidified chamber at 37°C overnight (~16 hours). The formamide concentration and temperature can be adjusted as needed for optimal stringency.
• Post-Hybridization Washes:
  • After hybridization, wash the coverslips with pre-warmed Wash Buffer (10% formamide in 2x SSC) to remove unbound probes.
  • Perform a second wash with 2x SSC to further reduce background.
  • Counterstain nuclei with DAPI if desired.

Imaging and Analysis

• Microscopy: Image the samples using an epifluorescence or confocal microscope equipped with a high-numerical-aperture (NA) 40x or 60x oil-immersion objective. For 3D quantification, acquire z-stacks encompassing the entire volume of the cells.
• Image Analysis: Use automated analysis software like TrueSpot [62] or FISH-quant [63] to detect and count RNA spots in 3D. The automated thresholding in TrueSpot is particularly valuable for unbiased, high-throughput quantification. Alternatively, open-source platforms like QuPath can be used for analysis of whole-slide images [64].

Enhancing the signal-to-noise ratio in RNA FISH is not achieved by a single change but through the systematic optimization of a connected workflow. As detailed in this application note, this process encompasses in silico probe design with advanced computational tools, empirical optimization of hybridization and imaging buffers to maximize specificity and signal longevity, and finally, the application of robust, automated 3D image analysis for precise quantification. By integrating these strategies—probe design, wet-lab protocol tuning, and computational analysis—researchers can significantly improve the quality and reliability of their data, thereby strengthening discoveries in spatial transcriptomics and single-cell biology.

Evaluating Antibody-Oligonucleotide Conjugation (AOC) Strategies for Cost and Reliability

Antibody-oligonucleotide conjugates (AOCs) have emerged as pivotal reagents in multiplexed fluorescence imaging, enabling researchers to visualize dozens of protein targets simultaneously within biological specimens [65]. These conjugates form the foundation of advanced imaging platforms such as CODEX, Immune-SABER, and Exchange-PAINT, which are transforming spatial biology research by revealing complex cellular interactions and tissue architecture [65]. The performance of these multiplexed imaging technologies is fundamentally dependent on the quality and consistency of the AOCs employed, making the evaluation of conjugation strategies a critical step in experimental design.

Within the context of multiplexed fluorescence imaging with multiple fluorophores, AOCs provide a mechanism to transcend the spectral limitations of conventional fluorescence microscopy. By utilizing oligonucleotide barcodes on antibodies, imaging transitions from being limited by the number of spectrally distinct fluorophores to being limited only by the number of unique oligonucleotide sequences that can be designed and detected through sequential hybridization imaging [65] [66]. This approach has demonstrated capability for imaging 40-60 markers simultaneously, enabling comprehensive cellular phenotyping and spatial analysis in complex tissues like tumor microenvironments and brain structures [65].

This application note provides a systematic evaluation of AOC conjugation strategies, focusing on the critical parameters of cost efficiency and reliability that directly impact their performance in multiplexed fluorescence imaging applications. We present comparative data on different conjugation methodologies, detailed protocols for their implementation, and analytical frameworks for selecting optimal strategies based on specific research requirements.

Conjugation Chemistry Landscape

Antibody-oligonucleotide conjugation involves creating stable covalent or high-affinity linkages between antibody molecules and synthetic oligonucleotides. The selection of conjugation strategy significantly impacts both the cost structure and reliability of the resulting AOCs, which in turn affects their performance in multiplexed imaging workflows [65].

Conjugation Methodologies

Four primary conjugation strategies have been established for AOC production, each with distinct advantages and limitations:

Maleimide-based covalent conjugation utilizes thiol-maleimide chemistry to attach maleimide-modified oligonucleotides to reduced cysteine residues within antibody molecules. This approach represents one of the most widely used methods due to its relatively straightforward chemistry [65]. However, a significant challenge with this method is the chemical instability of maleimide groups, which are sensitive to degradation and require careful handling and activation prior to conjugation. The activation process itself is time-consuming and can result in substantial oligonucleotide loss, contributing to increased costs [65].

Enzymatic conjugation employs microbial transglutaminase to create specific linkages between oligonucleotides and antibody molecules. This strategy offers improved site-specificity compared to maleimide-based approaches, potentially leading to more consistent conjugates with preserved antibody functionality [65]. The site-specific nature of this conjugation minimizes heterogeneity in the resulting AOC population, which is particularly valuable for quantitative imaging applications where consistent labeling efficiency is critical.

Click chemistry utilizes bio-orthogonal reactions such as the copper-catalyzed azide-alkyne cycloaddition to create stable triazole linkages between modified antibodies and oligonucleotides. This approach has gained prominence due to its high efficiency and specificity, with recent innovations in click chemistry linkers demonstrating enhanced stability of the resulting AOCs [67]. The reliability of this method, coupled with its compatibility with various antibody formats, makes it particularly suitable for generating large AOC panels required for comprehensive multiplexed imaging studies.

Avidin-biotin affinity-based assemblies leverage the strong non-covalent interaction between biotin and avidin/streptavidin to link biotinylated antibodies to oligonucleotides. While this method offers simplicity and flexibility, it can present challenges related to the size of the resulting complexes and potential non-specific binding in certain applications [67].

Table 1: Comparative Analysis of AOC Conjugation Methodologies

Conjugation Method	Chemistry Principle	Advantages	Limitations
Maleimide-based	Thiol-maleimide covalent linkage	Established protocol, widely adopted	Maleimide degradation; oligonucleotide loss during activation
Enzymatic	Microbial transglutaminase-mediated linkage	Site-specificity; batch consistency	Enzyme cost; optimization required
Click Chemistry	Copper-catalyzed azide-alkyne cycloaddition	High efficiency; stable linkers	Copper removal required; linker optimization
Avidin-Biotin	High-affinity non-covalent interaction	Simplicity; modular assembly	Large complex size; potential non-specificity

Cost and Reliability Considerations

The economic implications of AOC production represent a significant consideration for research planning, particularly for large-scale multiplexed imaging studies requiring panels of 40-60 markers [65]. A conventional multiplexed imaging panel of approximately 60 antibodies conjugated via maleimide chemistry can cost at least $60,000 in materials alone, with additional expenses incurred from the need for multiple conjugation attempts due to variability in success rates [65]. This substantial investment underscores the importance of selecting conjugation strategies that maximize reliability and minimize failed conjugations.

Recent advances in conjugation technologies have focused on improving batch-to-batch consistency through site-specific techniques that yield more homogeneous AOC populations [67]. Methodological improvements in analytical characterization, particularly using mass spectrometry, are addressing historical quality control challenges and ensuring more robust AOC production [67]. These advancements are gradually reducing the failure rates that have traditionally plagued AOC synthesis, though significant variability in conjugation quality and efficiency remains a challenge across all methodologies [65].

Quantitative Comparison of Conjugation Strategies

Systematic evaluation of conjugation methodologies enables data-driven selection of AOC production strategies. The following comparative analysis focuses on the key parameters of reliability, fluorescence intensity, consistency, and cost-effectiveness that directly impact multiplexed imaging performance.

Table 2: Performance Metrics of AOC Conjugation Strategies

Evaluation Parameter	Maleimide-based	Enzymatic	Click Chemistry	Avidin-Biotin
Reliability (Success Rate)	Moderate (variable between batches)	High	High	Moderate
Signal Intensity (Relative)	High	High	Moderate-High	Variable
Batch-to-Batch Consistency	Low-Moderate	High	Moderate-High	Moderate
Cost per Reaction	$$	$$$	$$	$
Panel Scalability	Moderate	High	High	Low-Moderate
Protocol Duration	2-3 days	1-2 days	1-2 days	1 day
Specialized Equipment Needs	Standard	Standard	Standard	Standard

The reliability of conjugation methods is profoundly influenced by reagent stability. Maleimide-modified oligonucleotides are particularly sensitive to storage conditions and require activation before use, a process that contributes significantly to variability [65]. In contrast, enzymatic and click chemistry approaches demonstrate superior reliability profiles due to more stable reagent formulations and more predictable reaction kinetics [67].

From a cost perspective, while maleimide chemistry may appear economically favorable in direct reagent cost comparisons, its true expense emerges when considering the hidden costs associated with failed conjugations and repeated attempts [65]. Enzymatic conjugation, despite higher upfront reagent costs, may provide better long-term value for large panel generation due to its superior consistency and reduced failure rates [65] [67].

Experimental Protocols

Maleimide-Based Conjugation Protocol

Materials Required:

• Purified antibody (100-200 µg)
• Maleimide-modified oligonucleotide
• Tris(2-carboxyethyl)phosphine (TCEP)
• Zeba Spin Desalting Columns (7K MWCO)
• Conjugation buffer (PBSE: PBS with 1 mM EDTA, pH 7.4)
• Storage buffer (PBS with 0.1% BSA and 0.02% sodium azide)

Procedure:

• Antibody Reduction: Dilute antibody to 1-2 mg/mL in PBSE. Add TCEP to a final concentration of 5 mM and incubate at 37°C for 30-60 minutes to reduce interchain disulfide bonds.
• Buffer Exchange: Equilibrate a Zeba spin column with PBSE. Apply the reduced antibody to the column and centrifuge at 1,500 × g for 2 minutes to remove excess TCEP and cysteine residues.
• Oligonucleotide Activation: Simultaneously, resuspend the maleimide-modified oligonucleotide in PBSE. If the oligonucleotide contains a protected maleimide group, perform deprotection according to manufacturer's instructions.
• Conjugation: Combine the reduced antibody with activated oligonucleotide at a 1:3-1:5 molar ratio (antibody:oligonucleotide). Incubate at room temperature for 2-4 hours or overnight at 4°C with gentle mixing.
• Purification: Remove unconjugated oligonucleotide using size exclusion chromatography or filtration devices with appropriate molecular weight cutoffs.
• Quality Control: Analyze conjugation efficiency using SDS-PAGE, HPLC, or mass spectrometry. Determine concentration spectrophotometrically and aliquot for storage at -20°C.

Troubleshooting Notes:

• Low conjugation efficiency may result from incomplete antibody reduction or maleimide degradation.
• Oligonucleotide aggregation can occur during activation; remove aggregates by brief centrifugation before conjugation.
• Include a non-reduced antibody control to verify that reduction was necessary for conjugation.

Enzymatic Conjugation Protocol

Materials Required:

• Purified antibody (100-200 µg)
• Amine-modified oligonucleotide
• Microbial transglutaminase (MTG)
• Dimethylformamide (DMF) or DMSO
• Sucrose octasulfate
• Conjugation buffer (50 mM Tris, 150 mM NaCl, pH 7.5)
• Size exclusion purification columns

Procedure:

• Antibody Preparation: Dialyze or buffer exchange the antibody into conjugation buffer to a final concentration of 1-2 mg/mL.
• Oligonucleotide Modification: Dissolve the amine-modified oligonucleotide in ultrapure water. If necessary, further modify the oligonucleotide with a MTG substrate peptide (e.g., HQSYVDP).
• Enzymatic Conjugation: Combine antibody, modified oligonucleotide (3-5 molar excess), and MTG enzyme (1:20 enzyme:antibody weight ratio) in conjugation buffer. Add sucrose octasulfate to 1 mM final concentration to enhance conjugation efficiency.
• Incubation: Incubate the reaction mixture at 37°C for 2-4 hours.
• Reaction Termination: Add EDTA to 10 mM final concentration to stop the enzymatic reaction.
• Purification: Remove unconjugated oligonucleotide and enzyme using size exclusion chromatography.
• Quality Control: Assess conjugation efficiency and functionality as described in Section 4.1.

Optimization Considerations:

• Enzyme-to-antibody ratio may require optimization for different antibody subtypes.
• Reaction time can be extended to 16 hours at 4°C for improved yield with sensitive antibodies.
• Site-specificity is typically higher than maleimide-based methods, preserving antigen binding.

Workflow Integration for Multiplexed Imaging

The integration of AOCs into multiplexed imaging workflows requires careful consideration of both conjugation strategy and downstream application requirements. The following diagram illustrates the decision pathway for selecting and implementing AOC conjugation methods:

Diagram 1: AOC Conjugation Strategy Selection Workflow

This decision pathway highlights how research requirements should guide method selection, with iterative quality control ensuring final AOC quality before multiplexed imaging application.

The Scientist's Toolkit: Essential Research Reagents

Successful implementation of AOC-based multiplexed imaging requires access to specialized reagents and materials. The following table details essential components for AOC generation and application:

Table 3: Essential Research Reagents for AOC-based Multiplexed Imaging

Reagent Category	Specific Examples	Function in Workflow	Key Considerations
Antibodies	Validated primary antibodies	Target recognition	Species, clonality, and validation for specific applications
Oligonucleotides	Maleimide-modified, amine-modified, DBCO-modified	Barcode generation for multiplexing	Modification type, purity, and sequence uniqueness
Conjugation Enzymes	Microbial transglutaminase	Site-specific conjugation	Activity units, storage conditions, and buffer compatibility
Purification Systems	Size exclusion columns, filtration devices	Removal of unconjugated components	Molecular weight cutoff, recovery efficiency
Fluorophore Reporters	Cyanine dyes (Cy3, Cy5), Alexa Fluor dyes	Signal detection	Spectral properties, brightness, and photostability
Signal Amplification	Tyramide Signal Amplification (TSA) reagents	Sensitivity enhancement	Amplification factor and background optimization
Image Acquisition	CODEX, Ultivue, SignalStar	Multiplexed imaging	Compatibility with microscope systems

The evaluation of antibody-oligonucleotide conjugation strategies reveals distinct trade-offs between cost efficiency and reliability that must be carefully balanced based on specific research requirements. Maleimide-based chemistry, while cost-effective, presents significant challenges in reliability that may ultimately increase total project costs due to failed conjugations and repeated attempts. Enzymatic and click chemistry approaches offer superior consistency and are particularly advantageous for generating large AOC panels required for comprehensive multiplexed imaging studies, despite their higher initial reagent costs.

The field of AOC development continues to evolve rapidly, with emerging trends including site-specific conjugation techniques to improve batch consistency, advances in endosomal escape strategies for intracellular delivery, and expansion of AOC applications beyond oncology into rare genetic disorders [67]. The maturation of regulatory frameworks for these complex bioconjugates, coupled with strategic partnerships between established biologics developers and oligonucleotide specialists, is accelerating the translation of AOC technologies from research tools to clinical applications [67].

For researchers implementing multiplexed fluorescence imaging workflows, the selection of AOC conjugation strategy should be guided by the specific balance of cost constraints, reliability requirements, and technical expertise available. The protocols and analytical frameworks provided in this application note offer a foundation for making evidence-based decisions that optimize both economic and scientific outcomes in spatial biology research.

Multiplexed fluorescence imaging has emerged as a powerful tool for visualizing and quantifying multiple molecular targets simultaneously within their native biological context. The performance of these advanced techniques, which include methods like multiplexed error-robust fluorescence in situ hybridization (MERFISH) and various forms of immunofluorescence, hinges critically on two fundamental aspects: the strategic design of molecular probes and the precise optimization of hybridization conditions. Achieving the delicate balance between hybridization efficiency and binding specificity presents a significant challenge that directly impacts the sensitivity, accuracy, and reproducibility of experimental results. Within the broader context of multiplexed fluorescence imaging research, this application note provides detailed protocols and empirical data to guide researchers in navigating these critical experimental parameters, ultimately enabling more robust and reliable scientific discoveries.

Probe Design Fundamentals

Effective probe design requires careful consideration of multiple physicochemical properties that collectively determine hybridization performance. The strategic selection of probe sequences must simultaneously maximize target affinity while minimizing off-target interactions.

Table 1: Key Probe Design Parameters and Their Impact on Performance

Design Parameter	Optimal Range	Performance Impact	Considerations
Target Region Length	30-50 nt [41]	Weak dependence on brightness beyond 30 nt [41]	Shorter sequences (<20 nt) may reduce binding energy
GC Content	Moderate (varies by method) [68]	High GC causes non-specific binding; low GC reduces affinity [68]	Must be optimized for specific hybridization conditions
Melting Temperature (Tm)	Consistent across probe set [68]	Ensures uniform hybridization behavior [61]	Critical for multiplexed applications
Specificity Screening	Genome-wide BLAST [61]	Minimizes off-target binding and false positives [61]	Should include repetitive element filtering

Advanced computational tools have been developed to integrate these parameters into cohesive design strategies. TrueProbes, for instance, employs a genome-wide binding affinity analysis that ranks candidates by predicted specificity before assembling the final probe set, considering expressed off-target binding, self-hybridization potential, and cross-dimerization [61]. This represents a significant advancement over traditional approaches that often select probes sequentially from 5' to 3' without global ranking by experimental utility.

The visual framework above outlines the systematic approach required for effective probe design, emphasizing the critical importance of specificity assessment and empirical validation in developing reliable imaging reagents.

Hybridization Condition Optimization

Hybridization conditions serve as the experimental counterpart to thoughtful probe design, with temperature, chemical denaturants, and probe concentration collectively determining the stringency of target recognition. Systematic optimization of these parameters can dramatically enhance assay performance.

Temperature and Denaturant Optimization

Formamide concentration and hybridization temperature interact to determine the effective stringency of hybridization. Research indicates that for DNA probe-based mitochondrial DNA capture, optimal hybridization efficiency in fresh frozen tissue samples occurs at 60°C, while RNA probes achieve peak performance at 55°C [69]. This temperature differential reflects the inherent biochemical properties of different probe types, with RNA-DNA hybrids generally exhibiting greater stability than DNA-DNA duplexes.

Probe Concentration Effects

The quantity of probes used in hybridization reactions must be carefully calibrated to target abundance and sample type. In mitochondrial DNA capture experiments, optimal DNA probe quantities differed significantly between sample types—16ng per 500ng of library for fresh tissue versus 10ng for plasma samples [69]. This highlights the necessity of empirical optimization for different experimental contexts and sample preparations.

Table 2: Optimal Hybridization Conditions for Different Applications

Method	Sample Type	Optimal Temperature	Optimal [Formamide]	Key Performance Metric
smFISH (50nt probes) [41]	U-2 OS cells	37°C	Variable (protocol-dependent)	Single-molecule brightness
DNA probe mtDNA capture [69]	Fresh frozen tissue	60°C	Not specified	61.79% mapping rate
RNA probe mtDNA capture [69]	Fresh frozen tissue	55°C	Not specified	92.55% mapping rate
DNA probe mtDNA capture [69]	Plasma	55°C	Not specified	16.18% mapping rate

Research Reagent Solutions

The successful implementation of multiplexed fluorescence imaging depends on a comprehensive suite of specialized reagents, each serving distinct functions within the experimental workflow.

Table 3: Essential Research Reagents for Multiplexed Fluorescence Imaging

Reagent Category	Specific Examples	Function	Application Notes
Fluorescent Dyes	Alexa Fluor series, Cy3, Cy5, FITC [15]	Signal generation	Photostability varies; consider imaging duration
Targeting Probes	DNA probes, RNA probes, encoding probes [41] [69]	Target recognition	RNA probes show higher affinity but different specificity profiles
Buffers & Denaturants	Formamide, SSC, blocking agents [41]	Control hybridization stringency	Significantly impact signal-to-noise ratio
Antibodies & Nanobodies	Trastuzumab, Fab fragments [15]	Protein target recognition	Used in cyclic immunofluorescence approaches
Enzymatic Reagents	Ligases, polymerases [41]	Signal amplification	Critical for methods using rolling circle amplification

Experimental Protocols

Protocol 1: MERFISH Probe Hybridization and Imaging

This protocol outlines the optimized procedure for performing multiplexed error-robust fluorescence in situ hybridization based on recent methodological improvements [41].

Materials:

• Encoding probes (30-50nt target region length)
• Hybridization buffer (with optimized formamide concentration)
• Readout probes with fluorescent labels
• Fixed cells or tissue samples
• Fluorescence microscope with high numerical aperture objective

Procedure:

• Sample Preparation: Fix cells or tissue sections according to standard protocols. Permeabilize to allow probe access.
• Encoding Probe Hybridization:
  • Apply encoding probe set diluted in hybridization buffer.
  • Incubate at 37°C for 24 hours to ensure complete hybridization.
• Washing: Remove unbound encoding probes with multiple washes in buffer containing formamide at equivalent concentration to hybridization buffer.
• Readout Probe Hybridization:
  • Apply fluorescent readout probes complementary to encoding probe barcodes.
  • Hybridize for 30-60 minutes at room temperature.
• Imaging:
  • Acquire images using appropriate filter sets for each fluorophore.
  • For multiplexed rounds, remove readout probes and repeat with next set.
• Image Analysis: Identify RNA molecules via combinatorial barcode decoding.

Technical Notes: Recent optimizations have improved performance through modified encoding probe hybridization, buffer storage conditions, and imaging buffer composition [41]. Pre-screening readout probes against samples can identify non-specific binding and reduce false positives.

Protocol 2: DNA/RNA Probe Hybridization for mtDNA Capture

This protocol describes the systematic optimization of hybridization conditions for mitochondrial DNA capture using either DNA or RNA probes [69].

Materials:

• Custom-designed DNA or RNA probes (120nt, biotinylated)
• Target DNA (300-500bp fragments)
• Hybridization equipment with precise temperature control
• Streptavidin-coated capture beads

Procedure:

• Library Preparation: Fragment genomic DNA to 300-500bp using focused ultrasonication.
• Hybridization Condition Optimization:
  • For DNA probes: Test temperatures from 55-65°C with probe quantities of 4-16ng per 500ng library.
  • For RNA probes: Test temperatures from 55-65°C with probe quantities of 5-20ng per 500ng library.
• Capture Reaction:
  • Incubate probes with target library at optimal temperature for 16-24 hours.
  • Capture probe-target complexes using streptavidin beads.
• Washing: Stringently wash to remove non-specifically bound fragments.
• Elution and Sequencing: Elute captured DNA and proceed to next-generation sequencing.

Technical Notes: Optimal conditions were determined as 60°C with 16ng DNA probe for tissue samples, and 55°C with 5ng RNA probe for tissue samples [69]. Performance metrics include mapping rate, average depth, and duplication rate.

Advanced Applications and Integration

The principles of probe design and hybridization optimization find application in increasingly sophisticated imaging methodologies. For example, the SEPARATE method enables volumetric multiplexed imaging by pairing proteins with distinct spatial expression patterns, labeling them with the same fluorophore, and computationally unmixing their signals based on machine learning algorithms [70]. This approach effectively doubles multiplexing capacity by imaging two proteins per fluorophore.

Similarly, advanced multiplexing techniques like Exchange-PAINT can localize up to 13 different nuclear targets with nanometer precision through sequential imaging of orthogonal oligonucleotide-conjugated probes [71]. These methods push the boundaries of what's possible in spatial biology but remain fundamentally dependent on the precise probe design and hybridization principles outlined in this document.

The synergistic optimization of probe design and hybridization conditions forms the foundation of successful multiplexed fluorescence imaging experiments. While computational tools continue to advance probe selection strategies, empirical optimization of hybridization parameters remains essential for achieving the delicate balance between detection efficiency and binding specificity. The protocols and data presented here provide researchers with a practical framework for developing and implementing robust imaging assays that maximize signal-to-noise ratio while minimizing false positives. As multiplexed imaging methodologies continue to evolve toward higher plex capabilities and greater spatial resolution, the fundamental principles outlined in this application note will continue to guide experimental design and implementation across diverse biological applications.

Mitigating Background and Preserving Sample Integrity During Long Imaging Cycles

Multiplexed fluorescence imaging is an indispensable technique for visualizing complex biological structures and molecular interactions, particularly in neuroscience and cancer biology. However, extending these methods to volumetric imaging or a high number of cycles presents significant challenges, primarily related to technical variability and prolonged experimental timelines. These extended processes can compromise sample integrity through factors such as photobleaching, altered antigenicity, and the accumulation of background fluorescence. This application note details established protocols for mitigating these issues, leveraging recent advancements in computational normalization and experimental design to ensure the acquisition of high-fidelity, biologically meaningful data.

Core Challenges in Prolonged Imaging Cycles

• Technical Variation: Staining intensity can vary significantly from slide to slide and batch to batch due to differences in tissue fixation, antibody concentrations, and imaging conditions. This technical noise complicates the integration of data from multiple samples [72].
• Background Fluorescence: Dense and low-contrast samples can exhibit high background signals, which obscure the specific signal from the target of interest and reduce the signal-to-background ratio [73].
• Prolonged Staining Times: In 3D multiplexed imaging, a single staining cycle for a thick specimen can take over 12 hours. Repeating this over multiple cycles risks sample degradation, compromised antigen integrity, and impractical experimental durations [13].
• Photobleaching: Fluorophores can lose their ability to fluoresce after prolonged or repeated exposure to light, which is a critical issue in cyclic imaging [15].

Computational Strategies for Background Mitigation and Normalization

Universal Normalization Pipeline (UniFORM)

The UniFORM pipeline is a non-parametric method designed to normalize multiplex tissue imaging (MTI) data at both the feature and pixel levels without relying on assumptions of data distribution. It corrects for technical variation by aligning the biologically invariant negative populations across samples, thereby preserving true biological signals [72].

Key Assumptions of UniFORM:

• Minimal Biological Variability in Negative Populations: The negative population (cells not expressing the marker) serves as a stable baseline of technical variation.
• Consistent Distribution Patterns: The intensity distribution patterns for positive and negative populations remain consistent across samples.

Protocol: UniFORM Normalization

• Software Requirement: Python-based UniFORM pipeline.
• Input Data: Feature-level data (e.g., mean intensity values per cell) or pixel-level raw images.
• Procedure:
  • Automatic Landmark Registration: The pipeline automatically identifies the leftmost mode (peak) of the intensity histogram for each marker, which corresponds to the negative population.
  • Rigid Alignment: It performs a rigid landmark registration to align these negative peaks across different samples or batches.
  • Intensity Adjustment: The intensity distributions of all samples are shifted and scaled to align these baseline signals.
  • Output: Normalized data where technical variations are minimized, and biological differences are preserved.
• Application: This method has been benchmarked on CyCIF, ORION, and COMET platforms, showing superior batch effect correction and improved performance in downstream analyses like clustering and UMAP visualization [72].

Local Mean Suppression Filter for Background Identification

For direct background identification in fluorescence images, particularly those with dense foregrounds, a local mean suppression filter offers an easy-to-use, nonlinear filtering solution [73].

Protocol: Background Identification with Local Mean Suppression

• Software Requirement: A fast Python implementation is available.
• Input Data: A single-channel fluorescence microscopy image.
• Procedure:
  • For each pixel in the image, calculate the mean intensity within a local neighborhood. The neighborhood size can be varied.
  • Compare the intensity of the central pixel to the local mean. If the pixel's intensity is less than the mean, assign it a provisional background label.
  • Accumulate labels generated from multiple different neighborhood sizes.
  • Based on the accumulated labels, assign a final background or foreground label to each pixel.
• Output: A binary mask identifying background pixels.
• Use Cases: Effective background identification in complex tissue images; can be used as a pre-processing step for image segmentation in multiplexed imaging contexts [73].

Experimental Design Strategy to Reduce Cycle Time

The SEPARATE (Spatial Expression PAttern-guided paiRing And unmixing of proTEins) method addresses the challenge of long staining times in 3D multiplexed imaging by strategically reducing the number of required staining cycles [13].

Principle: Instead of using one fluorophore per protein per cycle, SEPARATE pairs two proteins with distinct spatial expression patterns to be labeled with the same fluorophore in a single cycle. A protein separation network, a type of convolutional neural network, then computationally unmixes the signals of the two proteins from the single acquired image.

Protocol: Implementing the SEPARATE Workflow

• Step 1: Identify Optimal Protein Pairs

  • Imaging: Acquire 3D immunofluorescence images for each protein individually.
  • Feature Extraction: Use a feature extraction network trained with contrastive learning to convert each protein's spatial expression pattern into a feature vector.
  • Distance Calculation: Compute the feature-based distance between all protein pairs. This distance quantifies the visual distinctiveness of their patterns.
  • Optimal Pairing: Select the protein pairing configuration that maximizes the minimum feature-based distance between all pairs. This ensures that even the most similar pair can be robustly unmixed [13].

• Step 2: Experimental Staining and Image Acquisition

  • For each selected protein pair (α and β), perform a single staining cycle using three fluorophores:
    • Fluorophore A: Labels only protein α.
    • Fluorophore B: Labels only protein β.
    • Fluorophore C: Labels both proteins α and β (mixed signal).
  • Acquire a three-channel volumetric image [13].

• Step 3: Train and Apply the Protein Separation Network

  • Generate Training Data: Create synthetic training images by linearly combining the images of single proteins (from Step 1) with random ratios. This makes the network robust to variations in relative antibody concentration and protein expression levels.
  • Network Training: Train the protein separation network using the synthetic mixed images as input and the corresponding single-protein images as targets.
  • Signal Unmixing: Apply the trained network to the real mixed-signal channel (Fluorophore C). The network will output two distinct images, each representing the signal from one of the paired proteins [13].

This approach demonstrably enables the imaging of six proteins using only three fluorophores in a single cycle, effectively halving the required staining and imaging time [13].

Quantitative Performance of Featured Methods

Table 1: Performance Metrics of Key Computational Methods

Method	Primary Function	Key Metric	Reported Performance	Platforms Validated
UniFORM [72]	Feature- and pixel-level normalization	Improved k-nearest neighbor batch effect test (kBET) and silhouette scores	Outperformed MxNorm, Z score, ComBat, and mean division in batch effect mitigation and biological signal preservation.	CyCIF, ORION, COMET
Local Mean Suppression Filter [73]	Background identification	Comparison to state-of-the-art methods	Performance favorably compared with advanced image processing, machine learning, and deep learning methods.	Fluorescence microscopy
SEPARATE [13]	Protein signal unmixing	Correlation between spatial pattern distinction and unmixing performance	High correlation confirmed; enabled volumetric imaging of 6 proteins with 3 fluorophores.	3D mouse brain imaging

The Scientist's Toolkit: Essential Research Reagents and Materials

Table 2: Key Reagents and Materials for Multiplexed Fluorescence Imaging

Item	Function/Description	Example Use Case
Antibodies (Monoclonal, e.g., Trastuzumab) [15]	High-specificity antibodies derived from a single B-cell clone, conjugated to fluorescent dyes for visualizing specific proteins or antigens.	Visualizing HER2-positive tumors in vivo or ex vivo.
Nanobodies [71]	Small antigen-binding fragments used as secondary reagents, conjugated with oligonucleotide docking strands for multiplexed imaging.	Used in Exchange-PAINT for highly multiplexed super-resolution imaging.
Fluorescent Dyes (e.g., Cy3, Cy5, Alexa Fluor) [15]	Synthetic dyes that emit fluorescence at specific wavelengths upon excitation.	Labeling antibodies or other probes for detection in fluorescence imaging.
Oligonucleotide Docking Strands [71]	Short DNA strands conjugated to antibodies or nanobodies. Complementary dye-conjugated "imager" strands bind for sequential imaging.	Enabling multiplexed Exchange-PAINT imaging.
Bright Long-Wavelength Probes (e.g., PbS/CdS QDs) [74]	Quantum dots with bright emission in the second near-infrared window (NIR-II, 900-1880 nm and beyond).	High-contrast in vivo fluorescence imaging with reduced scattering and autofluorescence.
BODIPY Dyes [15]	Versatile fluorescent probes with high quantum yields, photostability, and tunable emission.	Cellular imaging; can be modified with targeting moieties (e.g., folic acid) for targeted cancer imaging.

Workflow Diagrams

SEPARATE unmixing workflow

UniFORM normalization process

Technology Landscape: A Comparative Analysis of Multiplex Imaging Platforms

The advent of multiplexed imaging technologies has fundamentally transformed spatial biology research, enabling unprecedented visualization of complex cellular interactions within intact tissues. These platforms allow researchers and drug development professionals to simultaneously visualize dozens of molecular markers, providing critical insights into cellular heterogeneity, spatial relationships, and functional states in diseases such as cancer [48]. The drive toward personalized medicine, particularly in oncology, has intensified the need for technologies that can accurately map the tumor immune microenvironment (TIME) and identify spatial biomarkers predictive of treatment response [48]. While traditional immunohistochemistry (IHC) and immunofluorescence (IF) are limited to visualizing only a handful of markers simultaneously, modern multiplexed imaging platforms overcome these constraints through innovative cyclical staining, DNA barcoding, and mass spectrometry-based approaches [75]. This technological evolution provides scientists with powerful tools to decipher the spatial architecture of tissues at single-cell resolution, advancing both basic research and clinical translation.

Each platform offers distinct advantages and limitations regarding multiplexing capacity, spatial resolution, throughput, and required instrumentation. This application note provides a comprehensive technical comparison of five leading platforms—Imaging Mass Cytometry (IMC), Multiplexed Ion Beam Imaging (MIBI), Cyclic Immunofluorescence (CycIF), Co-Detection by Indexing (CODEX), and Digital Spatial Profiling (DSP)—to guide researchers in selecting optimal methodologies for their specific research objectives. We detail experimental protocols, data analysis considerations, and practical implementation strategies to ensure successful integration of these powerful technologies into your research workflow.

Technology Comparison Table

Table 1: Comprehensive comparison of major multiplexed imaging platforms

Platform	Multiplexing Capacity	Spatial Resolution	Key Principle	Tissue Compatibility	Key Applications	Major Strengths	Major Limitations
IMC	~40 markers simultaneously [75]	Subcellular [48]	Metal-tagged antibodies detected by mass cytometry [48] [75]	FFPE, fresh frozen [75]	Spatial proteomics, tumor-immune interactions [75]	Minimal spectral overlap, no autofluorescence [75]	Tissue destruction during ablation, lower throughput [48]
MIBI	~40 markers [48]	Subcellular [48]	Metal-tagged antibodies detected by time-of-flight SIMS [48]	FFPE [48]	Deep phenotyping of tumor microenvironment [48]	High-dimensional data from single scan [48]	Limited availability, specialized equipment [48]
CycIF	50+ markers through cycling [48]	High (140-280 nm with confocal) [76]	Sequential fluorescence staining, imaging, and inactivation [48] [75]	FFPE, compatible with 3D thick sections [76]	High-resolution 2D and 3D tissue imaging [76]	Broad accessibility, conventional microscopes [48]	Labor-intensive, potential epitope damage [48]
CODEX	60+ markers [48] [75]	Single-cell [48]	DNA-barcoded antibodies with fluorescent hybridization [48] [75]	FFPE [77]	Cellular neighborhoods, spatial immunophenotyping [77]	Unlimited multiplexing in theory [75]	Complex reagent design, specialized fluidics [48]
DSP	Protein and RNA from selected regions [48]	Region of interest (ROI) [48]	Photocleavable oligonucleotide barcodes [48]	FFPE [78]	Biomarker discovery, pathway analysis [78]	Combines spatial context with digital counting [48]	Lower spatial resolution, ROI selection bias [48]

Detailed Platform Methodologies

Imaging Mass Cytometry (IMC)

Experimental Protocol:

• Tissue Preparation: Cut 1-5 μm sections from FFPE tissue blocks and mount on specialized IMC slides. Deparaffinize and rehydrate through xylene and graded ethanol series [75].
• Antigen Retrieval: Perform heat-induced epitope retrieval using citrate or EDTA buffer (pH 6.0 or 8.0, respectively) at 95-100°C for 20-45 minutes [75].
• Metal-Conjugated Antibody Staining: Incubate tissue sections with a panel of antibodies conjugated to pure metal isotopes (e.g., lanthanides) for 2 hours at room temperature or overnight at 4°C. Wash thoroughly to remove unbound antibodies [75].
• DNA Staining: Apply intercalator (e.g., 191/193Ir) to label nucleic acids for cell identification and segmentation [75].
• Laser Ablation and Data Acquisition: Introduce stained slides into the Hyperion XTi Imaging System. Ablate tissue region of interest with a UV laser, one pixel at a time. Transfer ablated material to inductively coupled plasma for ionization. Quantify metal isotopes by time-of-flight mass spectrometry [75].
• Image Reconstruction: Correlate detected ion signals with spatial coordinates to reconstruct high-dimensional images [75].

Data Analysis Considerations: IMC data requires specialized preprocessing, including compensation for isotopic impurities and normalization across samples. Cell segmentation typically utilizes nuclear markers followed by expansion to cytoplasmic and membrane boundaries. The high-dimensional data is amenable to both manual gating and automated clustering algorithms similar to those used in mass cytometry [75].

Multiplexed Ion Beam Imaging (MIBI)

Experimental Protocol:

• Sample Preparation: Section and mount FFPE tissues on conductive slides. Deparaffinize and perform antigen retrieval as described for IMC [48].
• Antibody Staining: Incubate with metal-conjugated primary antibodies (similar to IMC) optimized for MIBI detection [48].
• MIBI Imaging: Place samples in the MIBI vacuum chamber. Expose to a focused primary ion beam (typically oxygen) that liberates secondary ions from the tissue surface. Detect released ions by time-of-flight secondary ion mass spectrometry (TOF-SIMS) [48].
• Image Generation: Map detected ions to spatial coordinates to generate quantitative images for each marker [48].

Technical Considerations: MIBI requires ultra-high vacuum conditions and specialized instrumentation. The technique provides exceptional subcellular resolution but has lower throughput compared to other modalities. Panel design must consider potential interference between metal tags and tissue elements [48].

Cyclic Immunofluorescence (CycIF)

Experimental Protocol:

• Tissue Preparation: Mount FFPE or frozen sections on glass slides. For 3D CyCIF, sections 30-50 μm thick can be used to preserve intact cells [76].
• Initial Staining Cycle: Incubate with first set of fluorophore-conjugated antibodies (typically 4-6 markers). Perform confocal microscopy to acquire high-resolution images [76] [48].
• Fluorophore Inactivation: Remove fluorophores through mild chemical inactivation (e.g., with Hâ‚‚Oâ‚‚ and base) or photobleaching. Validate removal through imaging before proceeding to next cycle [75].
• Subsequent Cycles: Repeat staining and imaging with new antibody panels for 8-18 cycles to achieve high multiplexing [76].
• Image Registration and Processing: Align images from all cycles using computational methods to account for potential tissue shift. Generate final high-plex composite image [76].

Technical Considerations: CyCIF is particularly valuable for 3D tissue imaging, as standard 5 μm sections contain few intact cells—less than 5% of nuclei remain intact in such thin sections compared to ~80% in 30-40 μm sections [76]. This has profound implications for accurate cell phenotyping and interaction analysis. The method requires rigorous validation of antibody stability across multiple cycles and careful monitoring of epitope integrity [76] [75].

CODEX (Co-Detection by Indexing)

Experimental Protocol:

• Tissue Staining: Incubate FFPE tissue sections with a master mix of DNA-barcoded antibodies [77].
• CODEX Instrument Setup: Mount sample on specialized CODEX fluidics instrument. Add fluorescently labeled complementary oligonucleotides (reporters) [77].
• Cyclic Imaging: Perform iterative rounds of (a) hybridization with fluorescent reporters, (b) image acquisition using a widefield or confocal microscope, and (c) reporter removal through denaturation [75] [77].
• Image Processing: Stitch tiles, compensate for drift, and deconvolve images using CODEX software. Perform cell segmentation using nuclear markers [77].

Data Analysis Pipeline:

• Preprocessing: Apply normalization techniques (Z-score normalization is recommended) to mitigate technical noise [77].
• Cell Segmentation: Use watershed-based algorithms implemented in CODEX Segmenter software [77].
• Cell Typing: Employ unsupervised clustering (Leiden, X-shift, or k-means) followed by spatial verification of cell identities [77].
• Spatial Analysis: Identify cellular neighborhoods and interaction patterns using spatial statistics packages [77].

Digital Spatial Profiling (DSP)

Experimental Protocol:

• Tissue Staining: Incubate FFPE sections with oligonucleotide-barcoded antibodies (for protein detection) or in situ hybridization probes (for RNA detection) [48] [78].
• Region of Interest (ROI) Selection: Use standard immunofluorescence markers (e.g., pan-cytokeratin, CD45) to select morphologically defined regions for analysis [48] [78].
• UV Cleavage: Expose selected ROIs to UV light, which releases oligonucleotide barcodes from the tissue [48].
• Barcode Collection: Aspirate released barcodes and transfer to a collection plate [48].
• Quantification: Quantify barcodes using nCounter technology or next-generation sequencing [48] [78].
• Data Integration: Map molecular data back to spatial regions of origin for analysis [78].

Application Considerations: DSP is particularly valuable for biomarker discovery and pathway analysis where high-plex quantification from defined tissue regions is more important than single-cell resolution. The platform supports both protein (~-100-plex) and RNA (whole transcriptome) analysis from the same tissue section [78].

Workflow Visualization

Diagram 1: Multiplexed imaging platform selection workflow

Research Reagent Solutions

Table 2: Essential research reagents for multiplexed imaging workflows

Reagent Type	Specific Examples	Function	Compatibility
Metal-Conjugated Antibodies	Pure metal isotopes (lanthanides)	Target detection for IMC/MIBI	IMC, MIBI
DNA-Barcoded Antibodies	CODEX antibodies, Immuno-SABER	Target detection with DNA barcodes	CODEX, DSP
Fluorophore-Conjugated Antibodies	Alexa Fluor dyes, OPAL dyes	Fluorescent target detection	CyCIF, conventional IF
Signal Amplification Reagents	Tyramide Signal Amplification (TSA)	Signal enhancement	CyCIF, multiplex IHC
Tissue Clearing Reagents	Matrigel, RIMS	Tissue optical clarity	3D imaging, thick sections
Cell Segmentation Reagents	Nuclear stains (DAPI, Ir intercalator)	Cell identification and segmentation	All platforms
Antigen Retrieval Buffers	Citrate buffer (pH 6.0), EDTA (pH 8.0)	Epitope exposure	FFPE tissues across platforms
Fluorophore Inactivation Reagents	Hâ‚‚Oâ‚‚ with base	Fluorophore removal between cycles	CyCIF, IBEX

Advanced Applications and Integration

3D Tissue Imaging

Recent advances in multiplexed imaging have enabled three-dimensional tissue characterization, revealing critical limitations of conventional 2D analysis. Studies using 3D CyCIF have demonstrated that standard 5 μm histological sections contain few intact cells and nuclei, with fewer than 5% of nuclei remaining intact compared to multiple layers of intact nuclei visible in 30-40 μm thick sections [76]. This cellular fragmentation leads to significant errors in cell phenotyping, with false negative rates of 30-40% for polarized proteins and approximately 5% for uniformly distributed proteins [76]. Thick-section 3D imaging preserves crucial cellular architecture and enables accurate identification of cell-cell contacts and juxtracrine signaling complexes within immune niches [76].

Multimodal Integration

The integration of complementary spatial technologies provides more comprehensive biological insights than any single platform alone. For example, Standard BioTools has developed a three-step workflow combining IMC with RNAscope in situ hybridization technology to simultaneously visualize key protein and RNA markers in tissue samples [75]. This multimodal approach leverages the high-plex protein detection capability of IMC (up to 40 markers) with the precise RNA localization of RNAscope, enabling deeper investigation of tumor-immune interactions and cellular functional states [75].

Spatial Biomarkers for Immunotherapy

Multiplexed imaging has identified critical spatial biomarkers predictive of response to immune checkpoint inhibitors (ICIs). Key determinants include:

• CD8+ T-cell Density and Location: Increased CD8+ T-cell density and spatial colocalization with tumor cells correlate with improved immunotherapy response across multiple cancer types [48].
• Tertiary Lymphoid Structures (TLS): Organized clusters of B cells, dendritic cells, and T cells detected through multiplex imaging serve as important predictive markers of immunotherapy response [48].
• Spatial Exclusion Patterns: Immune-excluded phenotypes, characterized by T cells accumulating predominantly at the tumor margin with limited infiltration into tumor cores, typically predict resistance to immunotherapy [48].

These spatial biomarkers provide valuable insights beyond traditional markers such as PD-L1 expression, enabling more precise patient stratification for immunotherapy [48].

Protocol Optimization Guidelines

Panel Design Considerations

Successful multiplexed imaging requires meticulous panel design:

• Validation: Antibodies must be rigorously validated for specificity in the intended application, particularly for cyclic methods where epitope integrity across multiple cycles is crucial [77].
• Cofactor Optimization: For arcsinh normalization, optimize cofactors (typically 150) based on your specific dataset to properly scale background-subtracted values [77].
• Signal Range Management: Use percentile normalization (1st-99th percentiles) to cap extreme values that may represent background fluorescence or artifacts [77].

Data Normalization Strategies

Normalization is critical for accurate cell type identification in multiplexed imaging data:

• Z-score Normalization: Effectively mitigates effects of technical noise and enables comparison across markers by equalizing signal intensities [77].
• Log (Double Z) Normalization: Appropriate for datasets where identifying positive markers with high probability is essential; applies Z normalization twice with negative log transformation [77].
• Marker Prioritization: Normalization choice has greater impact on cell-type identification accuracy than clustering algorithm selection [77].

Analytical Best Practices

• Granularity Considerations: Increasing cell-type granularity leads to decreased labeling accuracy; subtle phenotype annotations should be avoided at the clustering step [77].
• Spatial Verification: Always validate computational cell-type calls with spatial visualization to confirm biological plausibility of identified patterns [77].
• Segmentation Impact: Unsupervised clustering better accounts for segmentation noise during cell-type annotation than traditional hand-gating approaches [77].

Multiplexed imaging technologies have revolutionized spatial biology by enabling high-dimensional analysis of cellular organization and interactions within intact tissues. Each platform—IMC, MIBI, CyCIF, CODEX, and DSP—offers unique strengths tailored to specific research applications, from ultra-high-plex single-cell analysis to region-specific biomarker discovery. As these technologies continue to evolve, integration with artificial intelligence, spatial transcriptomics, and clinical outcomes will further enhance their utility in both basic research and translational medicine. By following the detailed protocols and optimization guidelines outlined in this application note, researchers can effectively leverage these powerful platforms to advance our understanding of complex biological systems and accelerate therapeutic development.

Multiplexed fluorescence imaging has become an indispensable tool in biomedical research, enabling the simultaneous visualization of multiple molecular targets within their native spatial context. This capability is crucial for understanding complex biological systems, such as the tumor microenvironment, where the spatial relationships between different cell types dictate disease progression and treatment response [79]. The rapid evolution of technologies, from conventional immunofluorescence to highly multiplexed spatial omics platforms, presents researchers with a broad spectrum of choices, each with distinct advantages and limitations. This application note provides a structured comparison of current multiplexed imaging modalities based on three critical metrics—spatial resolution, multiplexing capability, and clinical translational potential—to guide researchers in selecting appropriate methodologies for their specific applications. Furthermore, we present detailed protocols for implementing these technologies, along with visualization of their workflows and essential reagent toolkits.

Comparative Analysis of Multiplexed Imaging Platforms

The current landscape of multiplexed imaging technologies can be broadly categorized by their underlying detection principles and workflow patterns. Table 1 summarizes the key quantitative metrics of major platforms, providing a foundation for comparative assessment.

Table 1: Performance Metrics of Multiplexed Imaging Platforms

Technology Category	Representative Platforms	Spatial Resolution	Multiplexing Capacity (Targets)	Tissue Compatibility	Clinical Translational Potential
Single-shot Fluorescence	PhenoImager HT [79]	~0.25 µm/pixel [79]	5-7 [79]	FFPE, Frozen	High
Single-shot Mass Cytometry	MIBI [79], IMC [79]	0.4-1.0 µm [79]	40+ [79]	FFPE, Frozen	Medium
Multicycle Optical (Oligo-based)	PhenoCycler (CODEX) [79]	Subcellular [79]	50+ [79]	FFPE, Frozen	Medium-High
Multicycle Optical (Antibody-based)	t-CyCIF, MxIF, IBEX [79]	~0.25 µm/pixel [79]	10-60 [80] [79]	FFPE, Frozen	Medium
Spatial Transcriptomics	10x Visium [81]	1-55 µm (region-based) [81]	Whole transcriptome [81]	FFPE, Frozen	Medium
In Situ Sequencing	STARmap [81]	Subcellular [81]	Hundreds to thousands [81]	Fresh, Frozen	Low-Medium

Technology Workflow Diagrams

The following diagrams illustrate the core experimental workflows for the primary multiplexing technology categories, highlighting their key procedural differences.

Figure 1: Workflow comparison between single-shot and multicycle imaging technologies. Single-shot methods stain all targets simultaneously, while multicycle methods use iterative staining and signal removal.

Detailed Experimental Protocols

Protocol 1: Single-Shot Multiplexing with Pdots

This protocol adapts the methodology from the Pdots study for highly multiplexed imaging in a single round of staining and imaging [82].

Reagents and Equipment

• Pdots Conjugates: Semiconducting polymer dots with tunable Stokes shifts (20-450 nm), conjugated to secondary antibodies or specific ligands [82]
• Primary Antibodies: Validated panel targeting antigens of interest
• Tissue Sections: FFPE or frozen tissue sections (4-7 µm thickness) mounted on glass slides
• Blocking Buffer: PBS with 1-5% BSA and 0.1% Tween-20
• Washing Buffer: PBS with 0.05% Tween-20
• Antigen Retrieval Solution: Citrate buffer (pH 6.0) or Tris-EDTA buffer (pH 9.0)
• Mounting Medium: Commercial antifade mounting medium
• Imaging System: Widefield or confocal fluorescence microscope with multiple filter sets and a sensitive camera [83]

Staining Procedure

• Tissue Preparation and Antigen Retrieval

  • Deparaffinize FFPE sections in xylene and rehydrate through graded ethanol series to water
  • Perform heat-mediated antigen retrieval using appropriate buffer for 20 minutes at 95-100°C
  • Cool slides to room temperature for 30 minutes

• Blocking and Primary Antibody Incubation

  • Apply blocking buffer for 1 hour at room temperature to reduce nonspecific binding
  • Incubate with primary antibody cocktail diluted in blocking buffer overnight at 4°C
  • Wash slides 3 times with washing buffer for 5 minutes each

• Pdots Secondary Detection

  • Incubate with Pdots-secondary antibody conjugates diluted in blocking buffer for 2 hours at room temperature, protected from light
  • Wash slides 3 times with washing buffer for 5 minutes each

• Mounting and Imaging

  • Apply coverslips using antifade mounting medium
  • Acquire images using multiple excitation wavelengths and emission filters matched to the Pdots spectra
  • For quantitative analysis, acquire all images with identical exposure settings and minimal saturation [83]

Image Processing and Linear Unmixing

• Spectral Reference Collection: Acquire reference images from single-labeled control samples for each Pdot type
• Spectral Unmixing: Process raw images using linear unmixing algorithms (e.g., in MATLAB) to reassign bleedthrough signals to correct channels [82]
• Validation: Verify unmixing accuracy through control samples and comparison with known biological patterns

Protocol 2: Multicycle Imaging with t-CyCIF

This protocol outlines the tissue-based cyclic immunofluorescence method for highly multiplexed imaging using conventional fluorophores [79].

Reagents and Equipment

• Primary Antibodies: Panel validated for sequential immunostaining
• Fluorophore-conjugated Secondary Antibodies: Multiple species-specific secondary antibodies conjugated to spectrally distinct fluorophores
• Stripping Buffer: 0.2M Glycine-HCl (pH 2.0) or commercial antibody elution buffer
• Blocking Buffer: PBS with 5% normal serum and 1% BSA
• Mounting Medium: Commercial hard-set antifade mounting medium
• Imaging System: Automated fluorescence microscope with motorized stage and environmental control

Staining and Imaging Cycles

• Initial Staining Round

  • Perform standard immunofluorescence with first primary/secondary antibody pair
  • Image entire tissue section using appropriate filter sets
  • Document stage positions for precise relocalization

• Antibody Stripping

  • Immerse slides in stripping buffer for 15-30 minutes with gentle agitation
  • Wash extensively with PBS to completely remove eluted antibodies
  • Verify removal by reimaging with same acquisition settings

• Subsequent Staining Rounds

  • Repeat blocking step to prevent nonspecific binding
  • Apply next primary/secondary antibody pair
  • Reimage using precise stage coordinates to ensure spatial registration
  • Repeat stripping and imaging cycles for all antibody targets

• Image Processing and Registration

  • Align all image stacks using fiduciary markers or computational registration
  • Generate composite multiplexed images for analysis

Table 2: Research Reagent Solutions for Multiplexed Fluorescence Imaging

Reagent Category	Specific Examples	Function & Application	Key Considerations
Fluorescent Probes	Pdots [82], Alexa Fluor dyes [15], Cy dyes [15]	Target detection with high brightness and photostability	Match excitation/emission spectra to available filters; consider Stokes shift [84]
Signal Amplification	Tyramide Signal Amplification (TSA) [79], ELISA, western blot	Enhance weak signals for low-abundance targets	Can increase background; requires optimization [79]
Antibodies	Trastuzumab (anti-HER2) [15], primary antibodies, Fab fragments [15]	Specific molecular target recognition	Validate for multiplexing; optimize concentration [85]
Tissue Processing	FFPE protocols [80], frozen section methods [80]	Preserve tissue architecture and antigen integrity	FFPE removes lipids; crosslinking affects epitopes [80]

Technology Selection Guide

Resolution and Multiplexing Trade-offs

The choice between imaging platforms involves balancing spatial resolution against multiplexing capacity, while considering sample compatibility and analytical requirements.

Figure 2: Decision framework for selecting appropriate multiplexed imaging technologies based on research requirements, sample availability, and infrastructure access.

Clinical Translational Considerations

The pathway to clinical adoption varies significantly across platforms, with key practical considerations including workflow complexity, infrastructure requirements, and data analysis pipelines.

Infrastructure and Workflow Requirements:

• Single-shot fluorescence methods (e.g., PhenoImager HT) offer relatively straightforward workflows compatible with clinical pathology laboratory operations, facilitating adoption [79]
• Mass cytometry-based imaging (MIBI, IMC) requires specialized instrumentation not typically available in clinical settings, limiting widespread implementation [79]
• Spatial transcriptomics platforms provide discovery-level data but face challenges in standardization, cost, and analytical complexity for routine diagnostics [81]

Sample Compatibility and Throughput:

• FFPE compatibility is essential for leveraging archival clinical samples with associated outcome data [80]
• Analysis time varies from hours (single-shot methods) to days (multicycle and sequencing-based methods), impacting clinical utility
• Assay robustness and reproducibility must be demonstrated across sample batches and processing conditions [83]

Multiplexed fluorescence imaging technologies offer powerful capabilities for spatial biology research, with platforms spanning a spectrum of resolution, multiplexing capacity, and clinical applicability. Single-shot methods like Pdots imaging provide an optimal balance for studies requiring moderate to high multiplexing with standard microscopy equipment, while mass cytometry and oligo-based approaches enable extreme multiplexing for comprehensive cellular profiling. Multicycle fluorescence methods offer a accessible path to high-plex imaging using conventional reagents but require careful validation to minimize artifacts from repeated processing. As these technologies continue to evolve, considerations of infrastructure requirements, sample compatibility, and analytical complexity remain crucial for selecting appropriate methodologies for specific research applications and clinical translation.

The tumor immune microenvironment (TIME) is a complex ecosystem where the spatial organization of cellular communities profoundly influences disease progression and therapeutic response. The ability to decode this spatial context is critical for advancing precision oncology. Spatial signatures—quantitative models derived from the multiplexed spatial analysis of tissue samples—are emerging as powerful tools to bridge this translational gap. These signatures integrate data on cell phenotypes, their functional states, and their physical interactions to generate biomarkers with direct clinical relevance. This Application Note details the framework for developing and validating spatial signatures that correlate with key oncology outcomes, such as response to immunotherapy and patient survival, providing a structured protocol for researchers in drug development.

Key Quantitative Findings from Clinical Validation Studies

Recent multi-cohort studies have demonstrated the prognostic power of spatial signatures in predicting immunotherapy outcomes in advanced non-small cell lung cancer (NSCLC). The quantitative findings from these validation studies are summarized in the tables below.

Table 1: Cell-Type-Based Spatial Signatures Associated with Immunotherapy Outcomes in NSCLC [86]

Signature Type	Key Constituent Cell Types	Compartment	Hazard Ratio (HR) for PFS (Training Cohort)	Hazard Ratio (HR) for PFS (Validation Cohorts)
Resistance Signature	Proliferating tumor cells, Granulocytes, Vessels	Tumor	HR = 3.8, P = 0.004	HR = 1.8, P = 0.05 (UQ Cohort)
Response Signature	M1/M2 Macrophages, CD4+ T cells	Stromal	HR = 0.4, P = 0.019 (5-year PFS)	HR = 0.49, P = 0.036 (UQ Cohort)

Table 2: Cell-to-Gene Spatial Signatures Derived from Transcriptomics [86]

Signature Type	Predicted Outcome	Hazard Ratio (HR) across Validation Cohorts
Resistance Signature	Poor Progression-Free Survival	HR = 5.3 (Yale), 2.2 (UQ), 1.7 (Athens)
Response Signature	Favorable Progression-Free Survival	HR = 0.22 (Yale), 0.38 (UQ), 0.56 (Athens)

Experimental Protocol for Spatial Signature Development and Validation

This protocol outlines a robust pipeline for generating and testing spatial signatures linked to clinical endpoints, based on established methodologies [86].

Sample Preparation and Multi-Omic Staining

• Tissue Collection: Obtain FFPE (Formalin-Fixed, Paraffin-Embedded) or fresh frozen tissue sections from patients enrolled in clinical trials, ideally those receiving first-line therapy to avoid confounding effects of prior treatments [86].
• Multiplexed Staining:
  • For Spatial Proteomics: Utilize a validated antibody panel (e.g., a 29-marker panel for NSCLC) with technologies like CODEX (Co-Detection by Indexing) [86] [87]. This allows for high-resolution mapping of over 40 proteins in situ [87].
  • For Spatial Transcriptomics: Employ Digital Spatial Profiling (DSP) with the GeoMx Whole Transcriptome Atlas (WTA) to profile RNA from specific tissue compartments, such as tumor and stromal Areas of Interest (AOIs) [86].

Image Acquisition and Data Processing

• Image Acquisition: Acquire high-resolution images using a compatible fluorescence microscope. For highly multiplexed panels, leverage ultra-multiplexing techniques like PICASSO (Process of ultra-multiplexed Imaging of biomoleCules viA the unmixing of the Signals of Spectrally Overlapping fluorophores), which enables imaging of up to 15 proteins in a single round without reference spectra [4].
• Spectral Unmixing: Process acquired images using linear unmixing algorithms or advanced information theory-based approaches like PICASSO to decompose signals from spectrally overlapping fluorophores and generate pure images for each marker [4] [5].
• Cell Phenotyping and Segmentation: Apply machine learning-based cell segmentation algorithms to the unmixed images. Identify cell types based on marker expression (e.g., CD3+ for T cells, CD68+ for macrophages) and quantify their abundances within pre-defined tumor and stromal compartments [86].

Spatial Analysis and Signature Training

• Spatial Interaction Analysis: Model cellular interactions and organization using methods like Voronoi diagrams and cellular neighborhood analysis to identify recurrent multicellular communities [86].
• Signature Training with Machine Learning:
  • Define Clinical Endpoint: Choose a clear endpoint, such as 2-year or 5-year Progression-Free Survival (PFS).
  • Split Cohort: Designate a training cohort (e.g., Yale cohort) and at least one independent validation cohort (e.g., University of Queensland cohort).
  • Model Building: In the training cohort, use a LASSO (Least Absolute Shrinkage and Selection Operator)-penalized Cox regression model to identify the minimal set of cellular features most predictive of the clinical outcome.
  • Apply Constraints: Enforce non-negative coefficients for a resistance signature and non-positive coefficients for a response signature to ensure biological interpretability [86].
  • Final Model: Train a final Cox regression model using only the cell types consistently selected by LASSO across multiple data splits.

Clinical Validation

• Blinded Testing: Apply the trained model to the held-out validation cohort(s) without any further parameter adjustment.
• Statistical Analysis: Evaluate the signature's performance using Kaplan-Meier survival analysis and log-rank tests to calculate Hazard Ratios (HRs) and p-values [86].
• Association with Other Endpoints: Test the signature's association with additional endpoints like Overall Survival (OS) to assess its broader clinical utility.

Visualization of Workflows and Signaling Pathways

Spatial Signature Development Pipeline

The following diagram illustrates the end-to-end workflow for developing and validating a spatial signature, from sample processing to clinical correlation.

Multiplexed Imaging and Unmixing Logic

This diagram outlines the core logic of ultra-multiplexed imaging, where signals from multiple overlapping fluorophores are acquired and computationally separated to generate pure channel images.

The Scientist's Toolkit: Research Reagent Solutions

Table 3: Essential Reagents and Platforms for Spatial Signature Research

Item Category	Specific Examples	Function in Workflow
Spatial Proteomics Platforms	CODEX [86] [87], Opal Polaris [87]	Highly multiplexed protein imaging (40-100+ plex) in intact tissue sections.
Spatial Transcriptomics Platforms	GeoMx Digital Spatial Profiler [86] [87]	Compartment-specific whole transcriptome or targeted RNA profiling.
Ultra-multiplexing Imaging Techniques	PICASSO [4]	Enables >15-color imaging in a single round without reference spectra by minimizing mutual information between channels.
Fluorophores	Organic dyes (e.g., Cy3, Cy5), Fluorescent proteins (GFP, RFP) [88]	Label antibodies or oligonucleotides to visualize target biomarkers.
Validated Antibody Panels	Custom 29-marker panel for NSCLC TIME [86]	Target key immune, tumor, and stromal cell phenotypes for consistent phenotyping.
Analysis Software	Commercial instrument software, Open-source tools (e.g., CellProfiler)	For image analysis, cell segmentation, and spatial data quantification.

Integrating Spatial Proteomics with Transcriptomics for a Multi-Omic View

Spatial biology has revolutionized our understanding of complex tissues by enabling the precise localization of molecular events within their native architectural context. While single-omics spatial technologies provide valuable insights, they capture only a partial view of the complex biological landscape [89]. The integration of spatial transcriptomics (ST) and spatial proteomics (SP) represents a powerful approach to overcome this limitation, offering a more comprehensive perspective on cellular organization and function [90]. However, a significant challenge in spatial multi-omics has been the inability to simultaneously profile mRNA and protein expression at high spatial resolution within the same tissue section, until recently [89] [90].

This application note details an integrated framework for performing and analyzing spatially resolved transcriptomics and proteomics from the same tissue section, leveraging advances in multiplexed fluorescence imaging [4]. By maintaining the spatial context across molecular layers, this approach enables direct single-cell comparisons of RNA and protein expression, reveals segmentation accuracy, and facilitates correlation analyses within individual cells [89]. We demonstrate this workflow using human lung cancer samples, highlighting its utility for investigating the tumour-immune microenvironment and advancing drug development programs.

Key Principles and Biological Rationale

The fundamental premise for integrating spatial proteomics with transcriptomics lies in the complementary nature of these molecular readouts. While transcriptomics reveals gene expression states, proteomics provides direct information about functional effector molecules [91]. However, the relationship between mRNA and protein abundance is complex and influenced by post-transcriptional regulation, translation efficiency, and protein degradation [91].

Recent studies have consistently observed systematic low correlations between transcript and protein levels when measured at cellular resolution [89] [90]. This disconnect makes integrated analysis particularly valuable for understanding disease mechanisms and identifying robust biomarkers. For drug development professionals, this multi-omic view provides critical insights into therapeutic targets and mechanisms of action that might be missed when examining either modality alone.

The spatial context adds another dimension to this integration, enabling researchers to investigate how cellular neighborhoods influence molecular relationships and how cell-to-cell interactions drive disease progression [90]. This is particularly relevant in immuno-oncology, where the spatial distribution of immune cells relative to tumor cells has prognostic and predictive significance.

Experimental Workflow

The complete integrated workflow encompasses tissue preparation, sequential multi-omic profiling on the same section, computational data alignment, and joint analysis. The key innovation is performing both spatial transcriptomics and spatial proteomics on the same tissue section rather than adjacent sections, ensuring perfect morphological correspondence [89] [90].

Sample Preparation and Tissue Processing

The workflow begins with proper tissue collection and preparation:

• Sample Collection: Use formalin-fixed paraffin-embedded (FFPE) tissue sections (5 µm) from human lung carcinoma samples or other tissues of interest [89]. Ensure proper ethical compliance and patient consent.
• Sectioning: Mount sections within the reaction region of Xenium slides (12 mm × 24 mm) and process for spatial transcriptomics.
• Quality Control: Assess tissue quality through initial H&E staining of serial sections if needed, ensuring tissue integrity is maintained throughout the process.

Spatial Transcriptomics Protocol

The spatial transcriptomics analysis uses the Xenium In Situ Gene Expression platform:

• Deparaffinization and Decrosslinking: Process slides following manufacturer's instructions (10x Genomics, Document CG000582 Rev E) to prepare tissue for hybridization [89].
• Hybridization: Add DNA probes designed to hybridize to target RNA sequences in a 289-gene human lung cancer panel [89].
• Ligation and Amplification: Perform ligation and amplification of gene-specific barcodes to enable spatial detection.
• Imaging: Load slides into the Xenium Analyzer with Xenium In Situ Gene Expression Reagents v1. The system performs cycles of probe hybridization, imaging, and removal to generate optical signatures for each barcode [89].
• Cell Segmentation: Perform cell segmentation based on DAPI nuclear expansion using the 10x Genomics pipeline [89] [90].

Spatial Proteomics Protocol

Following Xenium analysis, the same slides undergo hyperplex immunohistochemistry (hIHC) using the COMET system:

• Heat-Induced Epitope Retrieval (HIER): Perform HIER with the PT module (Epredia) to expose antigenic sites [89].
• Antibody Staining: Mount slides with microfluidic chips with an acquisition region of 9 mm × 9 mm. Perform sequential immunofluorescence staining using off-the-shelf primary antibodies for 40 markers, fluorophore-conjugated secondary antibodies, and DAPI counterstain (Thermo Fisher Scientific) [89].
• Cyclical Staining and Imaging: The COMET system conducts cyclical staining, imaging, and elution, generating a final stacked fluorescence image with 41 channels, including DAPI [89].
• Background Subtraction: Perform background subtraction using Horizon software (v2.2.0.1, Lunaphore Technologies) before exporting images for analysis [89].
• Cell Segmentation: Use CellSAM, a deep learning-based method integrating both nuclear (DAPI) and membrane (pan cytokeratin) markers for segmentation [89] [90].

H&E Staining and Pathology Annotation

After completing the spatial molecular profiling:

• H&E Staining: Perform manual hematoxylin and eosin staining on the post-Xenium post-COMET sections [89].
• Imaging: Image slides using a high-resolution scanner (e.g., Zeiss Axioscan 7) [89].
• Pathology Annotation: Conduct manual pathology annotation on digitized H&E images using QuPath software before integration to Weave [89].

Computational Data Integration

The integration of proteomic and transcriptomic datasets is performed using Weave software (version 1.0, Aspect Analytics):

• Image Registration: Co-register DAPI images from corresponding Xenium and COMET acquisitions to the H&E image using an automatic, non-rigid spline-based algorithm [89] [90].
• Data Fusion: Apply cell segmentation masks to calculate the mean intensity of each COMET marker and transcript count per gene per cell, generating an integrated dataset of gene and protein expression within the same cells [89].
• Visualization: Create interactive web-based visualizations in Weave, incorporating full-resolution H&E microscopy images with pathology annotations, COMET images, Xenium gene transcripts, and respective cell segmentation results [89].

For studies requiring advanced multiplexing beyond 15 colors in a single round, the PICASSO algorithm enables ultra-multiplexed fluorescence imaging without reference spectra measurements [4]. This method iteratively minimizes mutual information between mixed images to unmix signals from spatially overlapping proteins.

Key Analytical Methods

Correlation Analysis

To assess the relationship between transcript and protein expression:

• Data Filtering: Exclude unmatched pixels across Xenium, COMET, and Axioscan datasets due to varying acquisition region sizes [89].
• Statistical Testing: Calculate Spearman correlation between transcript count and mean immunofluorescence intensity using SciPy for both cell segmentation approaches [89].
• Marker Selection: Focus analysis on the 27 molecular markers where genes have corresponding protein markers [89].

Dimension Reduction and Clustering

For cell type identification and population analysis:

• Data Preprocessing: Exclude cells with total count <20, then normalize gene expression data using total count normalization and log transformation [89].
• Dimension Reduction: Apply UMAP for dimensionality reduction and construct neighbor graphs using 15 nearest neighbours and cosine similarity to create a graph where nodes (cells) are connected based on gene expression similarity [89].
• Clustering: Employ Louvain clustering to identify distinct cell populations [89].

Cell Type Annotation

• Proteomics-based Annotation: Use a hierarchical gating strategy, beginning with broad markers to categorize cells into major groups (e.g., tumour based on PanCK expression, immune cells via CD45 expression), then refine into specific subpopulations [89].
• Transcriptomics-based Annotation: Annotate cell types from Xenium data using scArches—a transfer learning framework that maps single-cell profiles onto reference atlases like the Human Lung Cell Atlas (HLCA) [89].

Research Reagent Solutions

Table 1: Essential Research Reagents for Spatial Multi-omics Workflow

Reagent Category	Specific Examples	Function	Considerations
Primary Antibodies	Off-the-shelf antibodies for 40 markers [89]	Target protein detection	Validate specificity and compatibility with FFPE tissues
Secondary Antibodies	Fluorophore-conjugated secondary antibodies [89]	Signal amplification and detection	Use cross-adsorbed secondary antibodies to minimize species cross-reactivity [92]
Fluorophores	DAPI, FITC, Cy5 [89]	Visualizing different targets	Match brighter fluorophores to less abundant targets; ensure compatibility with imaging system [92]
Gene Probes	289-gene human lung cancer panel [89]	Transcript detection	Design for specific hybridization with minimal cross-reactivity
Blocking Agents	BSA, normal serum [92]	Reduce non-specific binding	Use normal serum (5% v/v) from same species as secondary antibody host
Segmentation Markers	DAPI, Pan cytokeratin [89]	Cell boundary identification	Combine nuclear and membrane markers for accurate segmentation

Data Interpretation and Analysis

The integrated analysis enables several critical applications in drug development and basic research:

Transcript-Protein Correlation Analysis

A key finding from applying this integrated approach is the systematic observation of low correlations between transcript and protein levels at cellular resolution [89] [90]. This aligns with previous studies showing only moderate correlation between mRNA and protein abundances in tumor profiles [91]. The table below summarizes quantitative findings from correlation analyses:

Table 2: Representative mRNA-Protein Correlation Findings from Spatial Multi-omics Studies

Study	Sample Type	Correlation Range	Key Influencing Factors
Chong et al., 2025 [89]	Human lung carcinoma	Systematic low correlations	Cellular resolution, technical variability
Cell Reports Methods, 2022 [91]	Tumor proteomic profiles	Moderate correlation	Measurement error, protein reproducibility
PICASSO, 2022 [4]	Mouse brain	Varies by target	Spectral unmixing accuracy, antibody quality

Region-Specific Analysis of Immune and Tumor Markers

The spatial context enables region-specific analysis that reveals how cellular neighborhoods influence molecular expression patterns. This is particularly valuable for:

• Immuno-oncology Applications: Examining immune cell populations within tumor regions to understand mechanisms of response and resistance to immunotherapy [89].
• Cell-Type Specific Signaling: Identifying spatially restricted signaling pathways that would be obscured in bulk analyses or single-modality approaches.
• Tumor Heterogeneity Mapping: Characterizing intra-tumoral heterogeneity and identifying distinct tumor subclones with differential protein and gene expression patterns.

Troubleshooting and Optimization

Successful implementation of this integrated workflow requires attention to several technical considerations:

Multiplex Fluorescence Optimization

• Fluorophore Selection: Carefully assign fluorophores to different targets, matching brighter fluorophores to less abundant analytes or lower affinity primary antibodies [92]. Use a Spectra Viewer to visualize excitation/emission maxima and Stokes shifts.
• Blocking Optimization: Be rigorous with blocking to prevent unwanted signals from non-specific antibody binding. For indirect detection, use normal serum (5% v/v) from the same species as the secondary antibody host [92].
• Validation Experiments: Optimize single staining conditions for each primary antibody or primary/secondary antibody pair before multiplexing. Titrate each antibody reagent to identify conditions with low background and strong positive signal [92].

Computational Pipeline Validation

• Segmentation Accuracy: Compare segmentation results from different methods (DAPI nuclear expansion vs. CellSAM) and validate against morphological features in H&E images [89].
• Registration Quality Control: Visually inspect registration results across modalities to ensure accurate alignment before proceeding with integrated analysis.
• Batch Effect Management: When processing multiple samples, implement normalization strategies to account for technical variability between runs.

The integration of spatial proteomics with transcriptomics from the same tissue section represents a significant advancement in multi-omics research, providing unprecedented insights into cellular organization and function within native tissue contexts. This application note has detailed a comprehensive workflow that enables researchers and drug development professionals to simultaneously profile gene and protein expression at single-cell resolution while maintaining spatial context.

The methodologies described here—from sample preparation through computational integration—provide a robust framework for investigating complex biological systems, with particular relevance for cancer research, neuroscience, and immunology. As spatial technologies continue to evolve, coupled with advanced computational integration methods [93] [94] [95], this multi-omic approach promises to deepen our understanding of disease mechanisms and accelerate the development of novel therapeutic strategies.

Conclusion

Multiplexed fluorescence imaging has unequivocally matured into a cornerstone of modern biomedical research, providing an unprecedented, high-dimensional view of cellular organization and interactions. The convergence of novel fluorophores—such as time-resolved FPs and NIR-II dyes—with robust cyclic and DNA-barcoded methods and sophisticated computational unmixing has shattered previous multiplexing barriers. As these technologies become more standardized and accessible, their integration into clinical trial workflows is poised to revolutionize patient stratification, drug efficacy assessment, and the understanding of complex disease mechanisms. Future progress hinges on the continued development of more photostable probes, streamlined and cost-effective protocols, and advanced computational tools for managing the vast, spatially resolved datasets, ultimately paving the way for their routine use in precision medicine and diagnostic pathology.

References

• 1. Enhanced optical imaging and fluorescent labeling for ... [https://www.sciencedirect.com/science/article/pii/S2211383524000364]
• 2. Fluorescence Fundamentals | Thermo Fisher Scientific - US [https://www.thermofisher.com/us/en/home/references/molecular-probes-the-handbook/introduction-to-fluorescence-techniques.html]
• 3. Fluorescence Excitation and Emission Fundamentals [https://evidentscientific.com/en/microscope-resource/knowledge-hub/lightandcolor/fluoroexcitation]
• 4. PICASSO allows ultra-multiplexed fluorescence imaging of ... [https://www.nature.com/articles/s41467-022-30168-z]
• 5. Multicolor Microscopy: The Importance of Multiplexing [https://www.leica-microsystems.com/science-lab/life-science/multicolor-microscopy-the-importance-of-multiplexing/]
• 6. Super-multiplexed fluorescence microscopy via ... [https://pmc.ncbi.nlm.nih.gov/articles/PMC6033574/]
• 7. Fluorescence Reference Target Quantitative Analysis Library [https://pmc.ncbi.nlm.nih.gov/articles/PMC12045395/]
• 8. Fluorescence Reference Target Quantitative Analysis Library [https://arxiv.org/abs/2504.15496]
• 9. Decoding Molecular Network Dynamics in Cells: Advances ... [https://www.mdpi.com/2079-6374/15/9/614]
• 10. Genetically encoded barcodes for correlative volume ... [https://www.nature.com/articles/s41587-023-01713-y]
• 11. Time-resolved fluorescent proteins expand ... [https://www.sciencedirect.com/science/article/abs/pii/S009286742501027X]
• 12. Next-generation Genetically Encoded Fluorescent Biosensors ... [https://pmc.ncbi.nlm.nih.gov/articles/PMC11786609/]
• 13. Doubling multiplexed imaging capability via spatial ... [https://www.nature.com/articles/s42003-025-08357-5]
• 14. Biomimetic NIR-II fluorescent proteins created from ... [https://www.nature.com/articles/s41467-024-47063-4]
• 15. An update on recent advances in fluorescent materials for ... [https://pubs.rsc.org/en/content/articlehtml/2025/ra/d5ra03102h]
• 16. Fluorescent Proteins as Genetically Encoded FRET ... [https://pmc.ncbi.nlm.nih.gov/articles/PMC4634415/]
• 17. Time-resolved fluorescent proteins expand ... [https://pubmed.ncbi.nlm.nih.gov/40987294/]
• 18. The Fluorescent Protein Color Palette [https://evidentscientific.com/en/microscope-resource/knowledge-hub/techniques/confocal/applications/fpcolorpalette]
• 19. Near-Infrared-II Fluorescence Imaging of Tumors with ... [https://www.mdpi.com/1424-8220/25/22/7080]
• 20. Recent Advances in Second Near-Infrared Region (NIR-II) ... [https://www.frontiersin.org/journals/chemistry/articles/10.3389/fchem.2021.750404/full]
• 21. Application of NIR Fluorescent Materials in Imaging and ... [https://www.mdpi.com/2079-4991/15/11/811]
• 22. Recent advances in small molecule dye-based ... [https://pmc.ncbi.nlm.nih.gov/articles/PMC9556702/]
• 23. NIR-II Fluorescence Membrane Probes for Rapidly ... [https://www.sciencedirect.com/science/article/abs/pii/S0142961225006891]
• 24. Breaking shackles of molecular weight and emission for ... [https://www.nature.com/articles/s41467-025-63353-x]
• 25. Advances on fluorescence bioimaging and photo-mediated ... [https://www.sciencedirect.com/science/article/abs/pii/S0010854525001572]
• 26. Inorganic and hybrid nanomaterials for NIR-II fluorescence ... [https://pmc.ncbi.nlm.nih.gov/articles/PMC12068291/]
• 27. High-multiplex tissue imaging in routine pathology—are we ... [https://pmc.ncbi.nlm.nih.gov/articles/PMC10156760/]
• 28. Multiplex imaging in immuno-oncology - PMC [https://pmc.ncbi.nlm.nih.gov/articles/PMC10619056/]
• 29. Highly multiplexed immunofluorescence imaging of human ... [https://elifesciences.org/articles/31657]
• 30. Protocol optimization improves the performance of multiplexed ... [https://pmc.ncbi.nlm.nih.gov/articles/PMC12399773/]
• 31. SPECTREPlex: an automated, fast, high-resolution ... [https://www.nature.com/articles/s42003-025-08052-5]
• 32. Cyclic Immunofluorescence (CycIF), A Highly Multiplexed ... [https://pmc.ncbi.nlm.nih.gov/articles/PMC5233430/]
• 33. About CyCIF [https://www.tissue-atlas.org/cycif-method]
• 34. Addressing persistent challenges in digital image analysis of ... [https://pmc.ncbi.nlm.nih.gov/articles/PMC10401923/]
• 35. CODEX multiplexed tissue imaging with DNA-conjugated ... [https://pmc.ncbi.nlm.nih.gov/articles/PMC8647621/]
• 36. SABER enables amplified and multiplexed imaging of RNA ... [https://pmc.ncbi.nlm.nih.gov/articles/PMC6544483/]
• 37. A Validation Workflow for Novel Oligonucleotide ... [https://www.sciencedirect.com/science/article/pii/S0023683725001102]
• 38. CODEX® Multiplexed Imaging | Microscope Setup and ... [https://www.protocols.io/view/codex-multiplexed-imaging-microscope-setup-and-tis-b7carise.html]
• 39. MERFISH Spatial Transcriptomics [https://vizgen.com/technology/]
• 40. MERFISH Data and Protocols [https://zhuang.harvard.edu/merfish.html]
• 41. Protocol optimization improves the performance of ... [https://www.nature.com/articles/s41598-025-17477-1]
• 42. Multiplexed imaging of high-density libraries of RNAs with ... [https://www.nature.com/articles/s41598-018-22297-7]
• 43. Concordance of MERFISH spatial transcriptomics with bulk ... [https://pmc.ncbi.nlm.nih.gov/articles/PMC9760489/]
• 44. High-performance multiplexed fluorescence in situ ... [https://pmc.ncbi.nlm.nih.gov/articles/PMC5167177/]
• 45. Agenda: 2025 Human Cell Atlas General Meeting [https://events.humancellatlas.org/2025gm/agenda]
• 46. Systematic benchmarking of high-throughput subcellular ... [https://www.nature.com/articles/s41467-025-64292-3]
• 47. Benchmarking multi-slice integration and downstream ... [https://genomebiology.biomedcentral.com/articles/10.1186/s13059-025-03796-z]
• 48. Multiplex imaging analysis of the tumor immune ... [https://www.frontiersin.org/journals/immunology/articles/10.3389/fimmu.2025.1617906/full]
• 49. Immune Biomarkers for Checkpoint Blockade in Solid Tumors [https://pmc.ncbi.nlm.nih.gov/articles/PMC12384685/]
• 50. rethinking drug efficacy evaluation in cancer immunotherapy [https://www.frontiersin.org/journals/immunology/articles/10.3389/fimmu.2025.1609606/full]
• 51. Development of a Multiplex Immunofluorescence Assay for ... [https://pmc.ncbi.nlm.nih.gov/articles/PMC11502254/]
• 52. NanoPlex: a universal strategy for fluorescence microscopy ... [https://www.nature.com/articles/s41467-024-53030-w]
• 53. Protocol for multiplex immunofluorescence imaging of ... [https://www.sciencedirect.com/science/article/pii/S2666166725001820]
• 54. Tumour-reactive heterotypic CD8 T cell clusters from ... [https://www.nature.com/articles/s41586-025-09754-w]
• 55. Article Soluble Tim-3 serves as a tumor prognostic marker ... [https://www.sciencedirect.com/science/article/pii/S2666379124004075]
• 56. AutoUnmix: an autoencoder-based spectral unmixing ... [https://pmc.ncbi.nlm.nih.gov/articles/PMC10545201/]
• 57. unmixing - an overview | ScienceDirect Topics [https://www.sciencedirect.com/topics/computer-science/unmixing]
• 58. Spectral Unmixing in Flow Cytometry: 7 Top Tips for Success [https://bitesizebio.com/70438/spectral-unmixing-in-flow-cytometry/]
• 59. Applications, promises, and pitfalls of deep learning for ... [https://www.nature.com/articles/s41592-019-0458-z]
• 60. Multiplexed RNA imaging and in situ profiling in living cells [https://pubs.rsc.org/en/content/articlehtml/2025/sc/d5sc06220a?page=search]
• 61. TrueProbes: Quantitative Single-Molecule RNA-FISH ... [https://elifesciences.org/reviewed-preprints/108599]
• 62. TrueSpot: a robust automated tool for quantifying signal ... [https://genomebiology.biomedcentral.com/articles/10.1186/s13059-025-03772-7]
• 63. Optimized protocol for single-molecule RNA FISH to ... [https://pmc.ncbi.nlm.nih.gov/articles/PMC8264745/]
• 64. QuPath RNA ISH Analysis | In Situ Hybridization ... [https://acdbio.com/qupath-rna-ish-analysis]
• 65. Comparative Evaluation of Antibody-Oligonucleotide ... [https://www.sciencedirect.com/science/article/pii/S0023683725001722]
• 66. Leading multiplex immunofluorescence chemistries for ... [https://www.zeiss.com/microscopy/en/about-us/newsroom/press-releases/2025/multiplex-immunofluorescence-chemistries.html]
• 67. Antibody-oligonucleotide Conjugates Market - Global ... [https://www.researchandmarkets.com/reports/6055232/antibody-oligonucleotide-conjugates-market?srsltid=AfmBOor7xyvQ9FSyybqOblBOM4AX1vDQ5L0q7spEL1L7m_UnWQDFVLex]
• 68. The impact of microarray probe design on detecting CNV [https://www.ogt.com/us/resources/array-resources-and-support/array-resources/the-impact-of-microarray-probe-design-on-detecting-copy-number-variants-at-exon-resolution/]
• 69. Comparative evaluation of DNA and RNA probes for ... [https://pmc.ncbi.nlm.nih.gov/articles/PMC12333292/]
• 70. Doubling multiplexed imaging capability via spatial ... [https://pmc.ncbi.nlm.nih.gov/articles/PMC12167378/]
• 71. Mapping the nuclear landscape with multiplexed super- ... [https://www.nature.com/articles/s41467-025-61358-0]
• 72. Toward universal immunofluorescence normalization for ... [https://pmc.ncbi.nlm.nih.gov/articles/PMC12539234/]
• 73. Local mean suppression filter for effective background ... [https://pubmed.ncbi.nlm.nih.gov/40375425/]
• 74. High-contrast in vivo fluorescence imaging exploiting ... [https://pmc.ncbi.nlm.nih.gov/articles/PMC12075662/]
• 75. The Rise of Spatial-Omics Technologies | FluoroFinder [https://fluorofinder.com/spatial-omics/]
• 76. Highly multiplexed 3D profiling of cell states and immune ... [https://www.nature.com/articles/s41592-025-02824-x]
• 77. Strategies for Accurate Cell Type Identification in CODEX ... [https://www.frontiersin.org/journals/immunology/articles/10.3389/fimmu.2021.727626/full]
• 78. GeoMx Applications [https://nanostring.com/products/geomx-digital-spatial-profiler/applications/]
• 79. Spatial analysis by current multiplexed imaging ... [https://www.nature.com/articles/s41416-024-02882-6]
• 80. Spatial Multiplexing and Omics - PMC - PubMed Central - NIH [https://pmc.ncbi.nlm.nih.gov/articles/PMC12382484/]
• 81. Clinical and translational values of spatial transcriptomics [https://www.nature.com/articles/s41392-022-00960-w]
• 82. Highly multiplexed fluorescence microscopy with spectrally ... [https://pmc.ncbi.nlm.nih.gov/articles/PMC11633751/]
• 83. A beginner's guide to rigor and reproducibility in fluorescence ... [https://pmc.ncbi.nlm.nih.gov/articles/PMC6080651/]
• 84. Basics of fluorescence guide [https://www.abcam.com/en-us/technical-resources/guides/fluorescence-guide]
• 85. Background in Fluorescence Imaging [https://www.thermofisher.com/us/en/home/life-science/cell-analysis/cell-analysis-learning-center/molecular-probes-school-of-fluorescence/imaging-basics/protocols-troubleshooting/troubleshooting/background-fluorescence.html]
• 86. Spatial signatures for predicting immunotherapy outcomes ... [https://www.nature.com/articles/s41588-025-02351-7]
• 87. Spatial Oncology: Translating Contextual Biology to the Clinic [https://pmc.ncbi.nlm.nih.gov/articles/PMC12051486/]
• 88. Fluorescent imaging for cancer therapy and cancer gene ... [https://pmc.ncbi.nlm.nih.gov/articles/PMC8531657/]
• 89. An integrated approach for analyzing spatially resolved ... [https://www.frontiersin.org/journals/molecular-biosciences/articles/10.3389/fmolb.2025.1614288/full]
• 90. An integrated approach for analyzing spatially resolved ... [https://pmc.ncbi.nlm.nih.gov/articles/PMC12304548/]
• 91. Experimental reproducibility limits the correlation between ... [https://www.sciencedirect.com/science/article/pii/S2667237522001709]
• 92. Considerations for Multiplex Immunofluorescence [https://www.jacksonimmuno.com/secondary-antibody-resource/immuno-techniques/considerations-for-multiplex-immunofluorescence/]
• 93. A fast, scalable and versatile tool for analysis of single-cell ... [https://www.nature.com/articles/s41592-023-02139-9]
• 94. Computational Methods for Single-cell Multi-omics ... [https://pmc.ncbi.nlm.nih.gov/articles/PMC10025765/]
• 95. A Guide to Multi-omics Integration Strategies [https://frontlinegenomics.com/a-guide-to-multi-omics-integration-strategies/]