Laboratory Investigation

Precision medicine, based on spatial biology, is crucial for accurately diagnosing cancer and predicting drug responses. Here, we introduce the Spectral Imaging-based Single-Step Multiplex Immunofluorescence (SISS-mIF) technique, utilizing hyperspectral imaging to capture fluorescence spectra simultaneously. This approach optimizes tissue autofluorescence spectra for each image automatically, allowing the use of fluorescent direct-labeled antibodies for multicolor staining in a single step. Unlike conventional methods, the images are generated as standardized intensity independent of capture conditions, enabling consistent comparisons under different imaging conditions. This technique allows the detection of CD3, CD5, and CD7 in T-cell lymphoma on a single slide. The use of fluorescent direct-labeled antibodies enables triple staining of CD3, CD5, and CD7 without cross-reactivity, maintaining the same intensity as single stains. Furthermore, we developed a joint Non-Negative Matrix Factorization-based Spatial Clustering Analysis (jNMF-SCA) with a modified spectral unmixing system, highlighting its potential as a supportive diagnostic tool for T-cell lymphoma.

Metastatic prostate cancer is a highly heterogeneous and dynamic disease and practicable tools for patient stratification and resistance monitoring are urgently needed. Liquid biopsy analysis of circulating tumor DNA and circulating tumor cells (CTCs) are promising, but due to the diversity of resistance mechanisms, comprehensive testing is essential. Previously, we demonstrated that CTCs can be characterized by mRNA-based in situ padlock probe hybridization. Now, we have developed a novel combinatorial dual-color (CoDuCo) approach with increased multiplex capacity of up to 15 distinct markers, complemented by semi-automated image analysis and machine learning-assisted CTC classification. Here, we present three exemplary cases of patient samples in which the CoDuCo assay visualized diverse resistance mechanisms (AR-V7, neuroendocrine differentiation (SYP, CHGA, NCAM1)), as well as druggable targets and predictive markers (PSMA, DLL3, SLFN11). The combination of high multiplex capacity and microscopy-based single-cell analysis is a unique and powerful feature of the CoDuCo in situ assay. This synergy enables the identification and characterization of CTCs with epithelial, epithelial-mesenchymal, and neuroendocrine phenotypes, the detection of CTC clusters, and the visualization of CTC heterogeneity. In conclusion, the assay is a promising tool for monitoring the dynamic molecular changes associated with drug response and resistance in prostate cancer. O_FIG O_LINKSMALLFIG WIDTH=200 HEIGHT=134 SRC="FIGDIR/small/592946v1_ufig1.gif" ALT="Figure 1"> View larger version (50K): org.highwire.dtl.DTLVardef@46a186org.highwire.dtl.DTLVardef@116824dorg.highwire.dtl.DTLVardef@c4c842org.highwire.dtl.DTLVardef@1378b67_HPS_FORMAT_FIGEXP M_FIG C_FIG

Recent clinical observations highlight the importance of the spatial organization of immune cells into lymphoid structures for the success of cancer immunotherapy and patient survival. Sequential chromogenic immunohistochemistry (scIHC) supports the analysis of multiple biomarkers on a single tissue section thus providing unique information about relative location of cell types and assessment of disease states. Unfortunately, widespread implementation of scIHC is limited by lack of a standardized, rigorous guide to the development of customized biomarker panels and by the need for user-friendly analysis pipelines able to streamline the extraction of meaningful data. Here, we examine major steps from classical IHC protocols and highlight the impact they have on the scIHC procedure. We report practical examples and illustrations of the most common complications that can arise during the setup of a new biomarker panel and how to avoid them. We described in detail how to prevent and detect cross- reactivity between secondary reagents and carry over between detection antibodies. We developed a novel analysis pipeline based on non-rigid tissue deformation correction, Cellpose-inspired automated cell segmentation and computational network masking of low-quality data. The resulting biomarker panel and pipeline was used to study regional lymph nodes from head and neck cancer patients. We identified contact interactions between plasmablasts and plasmacytoid dendritic cells in vivo. Given that TLR receptors, which are highly expressed in plasmacytoid dendritic cells play a key role in vaccine efficacy, the significance of this cell-cell interaction decisively warrants further studies. In conclusion, this work streamlines the development of novel biomarker panels for scIHC, which will ultimately improve our understanding of immune responses in cancer.

The diagnosis of Mycosis Fungoides (MF) is difficult and often delayed, exacerbated by the constraint of conventional immunohistochemistry (IHC) to analyze only one antigen per tissue section, often necessitating repeat biopsies and extensive workups. We sought to validate a high-throughput Multiplex Immunofluorescence (MIF) method, coupled with computer-automated image analysis, to generate comprehensive immunophenotyping data from a single formalin-fixed, paraffin-embedded (FFPE) biopsy. We applied an 11-biomarker MIF panel across 18 archived skin specimens (9 MF/TCR clonality positive and 9 control/TCR clonality negative). Initial validation confirmed that MIF antigen expression and spatial localization were concordant with sequential IHC-stained sections. Whole slide image stacks were analyzed using both computer-assisted and fully computer-automated pipelines. Both methods successfully delineated immunophenotypic differences. MF specimens showed a significant expansion of hematopoietic cells and proliferative T-lymphocytes compared to controls. Crucially, MF tissues also exhibited a significant increase in the percentage of atypical T-lymphocytes. Our results validate the potential of MIF to obtain comprehensive, high-dimensional diagnostic information from a single tissue section. Integration with computer-automated analysis offers a scalable, high-throughput platform that can significantly aid in the timely and accurate diagnosis of cutaneous lymphomas.

Non-destructive 3D pathology methods have emerged in recent years with the potential to enhance standard 2D histopathology by greatly increasing the amount of tissue sampled by imaging and by providing volumetric morphological context. Another key advantage is that tissues remain intact, allowing re-embedding after imaging for potential long-term storage and future histological or molecular analyses. However, the impact of 3D pathology protocols on biomolecules -- including DNA, RNA, and proteins -- and their compatibility with downstream assays, has not been systematically evaluated. Here, we applied a previously optimized 3D pathology protocol -- involving deparaffinization, fluorescent H&E-analog staining, optical clearing, and open-top light-sheet microscopy -- to formalin-fixed paraffin-embedded (FFPE) specimens of breast, prostate, and head and neck cancer. Following the protocol, tissues were re-embedded in paraffin and compared with paired FFPE controls that did not undergo 3D pathology processing. DNA and RNA were extracted and subjected to quality assessments. Amplifiability was tested by PCR and reverse transcription quantitative PCR (RT-qPCR) of housekeeping genes. Although the results showed a slight decrease in the average yield and increased fragmentation of both DNA and RNA, amplifiability was largely preserved. Sanger sequencing of the PCR products confirmed accurate sequence determinations, while total RNA sequencing indicated that the global transcriptomic profile was largely unchanged. IHC staining of common biomarkers produced comparable signals, suggesting those proteins are well preserved after the 3D pathology workflow. These results demonstrate the feasibility of combining 3D pathology with downstream molecular applications.

The multiplexed immunofluorescence (mIF) platform enables biomarker discovery through the simultaneous detection of multiple markers on a single tissue slide, offering detailed insights into intratumor heterogeneity and the tumor-immune microenvironment at spatially resolved single cell resolution. However, current mIF image analyses are labor-intensive, requiring specialized pathology expertise which limits their scalability and clinical application. To address this challenge, we developed CellGate, a deep-learning (DL) computational pipeline that provides streamlined, end-to-end whole-slide mIF image analysis including nuclei detection, cell segmentation, cell classification, and combined immuno-phenotyping across stacked images. The model was trained on over 750,000 single cell images from 34 melanomas in a retrospective cohort of patients using whole tissue sections stained for CD3, CD8, CD68, CK-SOX10, PD-1, PD-L1, and FOXP3 with manual gating and extensive pathology review. When tested on new whole mIF slides, the model demonstrated high precision-recall AUC. Further validation on whole-slide mIF images of 9 primary melanomas from an independent cohort confirmed that CellGate can reproduce expert pathology analysis with high accuracy. We show that spatial immuno-phenotyping results using CellGate provide deep insights into the immune cell topography and differences in T cell functional states and interactions with tumor cells in patients with distinct histopathology and clinical characteristics. This pipeline offers a fully automated and parallelizable computing process with substantially improved consistency for cell type classification across images, potentially enabling high throughput whole-slide mIF tissue image analysis for large-scale clinical and research applications.

High-resolution whole-organ imaging of cleared tissues captures cellular and molecular insights within the intact tissue and tumour microenvironments. However, current immunolabelling and clearing methods are complicated and time-consuming; extending to several weeks. Here, we developed Simple Ultrafast Multicolor Immunolabelling and Clearing or SUMIC, a method that enables multicolor immunolabelling and clearing of whole murine organs and human tissues within 2 to 2.5 days. Moreover, SUMIC is simple, robust, non-hazardous and versatile comprising antigen retrieval, permeabilization, collagenase-based digestion, immunolabelling, dehydration, and clearing. SUMIC permits quantitative and singlecell resolution analysis and detection of rare cells in whole organs, for example, round SMA positive cells in the thymus. Upon volumetric imaging, SUMIC-processed samples retain normal tissue architecture and can be used for paraffin-embedding and histology. We employed the SUMIC method for whole-organ mapping of lymphatic vessels across different ages and organs. This analysis revealed the expansion of lymphatic vessels in endocrine tissues but not in any other organs with aging. Hence, SUMIC will accelerate discoveries compared to other whole organ imaging pipelines.

Multiplexed methods for the detection of protein expression generate extremely data-rich images of intact tissue sections. These images are invaluable for the quantification and analysis of complex biology and biomarker development. However, their interpretation presents a considerable analytical challenge. Cell segmentation from images is a key bottleneck and a major focus of research activity in artificial intelligence. Most current methods depend initially on the use of a nuclear counterstain to identify nuclear boundaries, which is a relatively straightforward task. The cellular boundary is then assigned either by expansion of the nuclear outline, or by the use of membrane or cytoplasm-specific stains to delineate cell boundaries, or by some combination of the two. The task is critical, as inaccurate segmentation leads to information loss and data contamination from neighbouring cells. Increasingly sophisticated methods are being developed to address these issues, but each has its own shortcomings. We present an alternative method which is inspired by the fact that the assignation of a cellular phenotype by eye does not depend upon the accurate identification of cell boundaries. We present an easy-to-use deep learning-based cellular phenotyping method which leverages this human capacity to assign phenotypes without segmenting the entire cell, and which can accurately phenotype cells based on nuclear segmentation alone. Using human ground truth annotations of entire cellular regions, we developed a classifier leveraging the U-Net architecture within a commercially available deep learning image analysis platform, but the principle is transferrable to any deep-learning framework. Crucially, training requires only a single example of each compartmental stain (nuclear/cytoplasmic/membranous). The resulting algorithm assigns class identities to cells with nuclear labelling alone, without the need for whole cell expansion. The method is highly novel, broadly generalisable, and comparable in accuracy to intensity-based phenotyping methods, bridging the gap between inaccurate cellular segmentation and accurate phenotype generation.

BackgroundSpatial transcriptomic technologies are powerful tools for resolving the spatial heterogeneity of gene expression in tissue samples. However, little evidence exists on relative strengths and weaknesses of the various available technologies for profiling human tumour tissue. In this study, we aimed to provide an objective assessment of two common spatial transcriptomics platforms, 10X Genomics Visium and Nanostrings GeoMx DSP. MethodThe abilities of the DSP and Visium platforms to profile transcriptomic features were compared using matching cell line and primary breast cancer tissue samples. A head-to-head comparison was conducted using data generated from matching samples and synthetic tissue references. Platform specific features were also assessed according to manufacturers recommendations to evaluate the optimal usage of the two technologies. ResultsWe identified substantial variations in assay design between the DSP and Visium assays such as transcriptomic coverage and composition of the transcripts detected. When the data was standardised according to manufacturers recommendations, the DSP platform was more sensitive in gene expression detection. However, its specificity was diminished by the presence of non-specific detection. Our results also confirmed the strength and weakness of each platform in characterising spatial transcriptomic features of tissue samples, in particular their application to hypothesis generation versus hypothesis testing. ConclusionIn this study, we share our experience on both DSP and Visium technologies as end users. We hope this can guide future users to choose the most suitable platform for their research. In addition, this dataset can be used as an important resource for the development of new analysis tools.

BackgroundCollagen is the major component of the extracellular matrix (ECM). Collagen structural organization undergoes significant transformation during tumorigenesis. The visualization of collagen in histological tissue sections would aid in the study of tumor growth, encapsulation, and invasion. However, such visualization requires the use of special stains such as Picrosirius Red (PSR) or Massons Trichrome (MT), or more recently, second-harmonic generation imaging (SHG) in unstained tissue sections. However, PSR and MT both suffer from significant inter- (and intra-) lab stain variabilities, and SHG, while considered a ground truth by many, suffers from issues of system complexity/reliability, cost, and speed/throughput. These technical hurdles limit more widespread assessment of collagen in tissue samples. MethodsUsing high-contrast, high-throughput polarization imaging on PSR-stained slides to generate ground truth training polarization images, we developed a deep learning model (iQMAI) to infer the presence of collagen directly from hematoxylin and eosin (H&E)-stained whole-slide images (WSIs) with high specificity. After iQMAI inference across WSIs, individual collagen fibers were extracted, and features describing overall collagen intensity and fiber morphology were computed. iQMAI pixel-and feature-wise outputs were compared to ground truth polarization imaging to assess model performance. The trained iQMAI model was deployed on H&E-stained WSI from the TCGA LUAD, LUSC, LIHC, and PAAD datasets for evaluation. iQMAI-derived collagen features were compared to tissue composition, gene expression, and overall survival. ResultsThe iQMAI model shows significant generalization across multiple indications. iQMAI collagen predictions were similar to polarization imaging measurements of the same sample, with a mean structural similarity index (SSIM) of 0.84 (95% CI 0.69-0.93), a mean patch-wise RMSE of 0.04 (95% CI 0.02-0.08), and a linear correlation (R2=0.93). Comparing features of the collagen fibers extracted from iQMAI vs. polarization images yielded similar linear correlations between computed fiber tortuosity, length, width, and relative angle. The relationship between collagen fiber density and fibroblast density was distinct in non-small cell lung cancer (LUAD and LUSC), hepatocellular carcinoma (LIHC), and pancreatic ductal adenocarcinoma (PAAD). In PAAD, fiber density and fiber width were both negatively associated with the LRRC-15 gene expression signature, and increased fiber width was associated with longer overall survival. ConclusionsiQMAI is a deep learning model that accurately predicts collagen from an H&E-stained WSI, allowing for spatially resolved quantification of collagen morphology and enabling investigation of the interplay between collagen and other TME components. We demonstrate an example of the utility of iQMAI-based collagen assessment in PAAD, where collagen features are associated with immunosuppressive cancer-associated fibroblasts and overall survival. Understanding the relationship between collagen, the tumor microenvironment composition, and disease progression may aid the development of effective immunotherapies in PAAD and other cancer types.

Image-based analysis as a rapid method for mutation detection can be advantageous in research or clinical settings when tumor tissue is limited or unavailable for direct testing. Here, we applied a deep convolutional neural network (CNN) to whole slide images of melanomas from 256 patients and developed a fully automated model that first selects for tumor-rich areas (Area Under the Curve AUC=0.96) then predicts for the presence of mutated BRAF in our test set (AUC=0.72) Model performance was cross-validated on melanoma images from The Cancer Genome Atlas (AUC=0.75). We confirm that the mutated BRAF genotype is linked to phenotypic alterations at the level of the nucleus through saliency mapping and pathomics analysis, which reveal that cells with mutated BRAF exhibit larger and rounder nuclei. Not only do these findings provide additional insights on how BRAF mutations affects tumor structural characteristics, deep learning-based analysis of histopathology images have the potential to be integrated into higher order models for understanding tumor biology, developing biomarkers, and predicting clinical outcomes.

IntroductionOver 50% of breast cancer cases are "Human epidermal growth factor receptor 2 (HER2) low breast cancer (BC)", characterized by HER2 immunohistochemistry (IHC) scores of 1+ or 2+ alongside no amplification on fluorescence in situ hybridization (FISH) testing. The development of new anti-HER2 antibody-drug conjugates (ADCs) for treating HER2-low breast cancers illustrates the importance of accurately assessing HER2 status, particularly HER2-low breast cancer. In this study, we evaluated the performance of a deep learning (DL) model for the assessment of HER2, including an assessment of the causes of discordances of HER2-Null between a pathologist and the DL model. We specifically focussed on aligning the DL model rules with the ASCO/CAP guidelines, including stained cells staining intensity and completeness of membrane staining. MethodsWe trained a DL model on a multi-centric cohort of breast cancer cases with HER2- immunohistochemistry scores (n=299). The model was validated on 2 independent multi- centric validation cohorts (n=369 and n=92), with all cases reviewed by 3 senior breast pathologists. All cases underwent a thorough review by three senior breast pathologists, with the ground truth determined by a majority consensus on the final HER2 score among the pathologists. In total, 760 breast cancer cases were utilized throughout the training and validation phases of the study. ResultsThe models concordance with the ground truth (ICC = 0.77 [0.68 - 0.83]; Fisher P = 1.32e-10) is higher than the average agreement among the 3 senior pathologists (ICC = 0.45 [0.17 - 0.65]; Fisher P = 2e-3). In the two validation cohorts, the DL model identifies 95% [93%- 98%] and 97% [91% - 100%] of HER2-low and HER2-positive tumors respectively. Discordant results were characterized by morphological features such as extended fibrosis, a high number of tumor-infiltrating lymphocytes, and necrosis, whilst some artifacts such as non- specific background cytoplasmic stain in the cytoplasm of tumor cells also cause discrepancy. ConclusionDeep learning can support pathologists interpretation of difficult HER2-low cases. Morphological variables and some specific artifacts can cause discrepant HER2-scores between the pathologist and the DL Model.

The demand for genetic testing for breast cancer susceptibility genes is increasing for both breast cancer patients and healthy individuals. Here we established a novel high-throughput assay to detect germline pathogenic variants in breast cancer susceptibility genes. In general, up 10 to 50 individual genomic DNA samples were mixed together to create a mixed DNA sample and the mixed DNA sample was subjected to a next-generation multigene panel. Germline pathogenic variants in breast cancer susceptibility genes could be found in the mixed DNA sample; next, site-specific Sanger sequencing was performed to identify individuals who carried he pathogenic variant in the mixed samples. We found that the recall and precision rates were 89.9% and 92.9% when twenty individual genomic samples were mixed. Therefore, our new assay can increase an approximately 20-fold of efficacy to identify the pathogenic variants in breast cancer susceptibility genes in individuals when compared with current assay.

The application of deep learning methods to spatial transcriptomics has shown promise in unraveling the complex relationships between gene expression patterns and tissue architecture as they pertain to various pathological conditions. Deep learning methods that can infer gene expression patterns directly from tissue histomorphology can expand the capability to discern spatial molecular markers within tissue slides. However, current methods utilizing these techniques are plagued by substantial variability in tissue preparation and characteristics, which can hinder the broader adoption of these tools. Furthermore, training deep learning models using spatial transcriptomics on small study cohorts remains a costly endeavor. Necessitating novel tissue preparation processes enhance assay reliability, resolution, and scalability. This study investigated the impact of an enhanced specimen processing workflow for facilitating a deep learning-based spatial transcriptomics assessment. The enhanced workflow leveraged the flexibility of the Visium CytAssist assay to permit automated H&E staining (e.g., Leica Bond) of tissue slides, whole-slide imaging at 40x-resolution, and multiplexing of tissue sections from multiple patients within individual capture areas for spatial transcriptomics profiling. Using a cohort of thirteen pT3 stage colorectal cancer (CRC) patients, we compared the efficacy of deep learning models trained on slide prepared using an enhanced workflow as compared to the traditional workflow which leverages manual tissue staining and standard imaging of tissue slides. Leveraging Inceptionv3 neural networks, we aimed to predict gene expression patterns across matched serial tissue sections, each stemming from a distinct workflow but aligned based on persistent histological structures. Findings indicate that the enhanced workflow considerably outperformed the traditional spatial transcriptomics workflow. Gene expression profiles predicted from enhanced tissue slides also yielded expression patterns more topologically consistent with the ground truth. This led to enhanced statistical precision in pinpointing biomarkers associated with distinct spatial structures. These insights can potentially elevate diagnostic and prognostic biomarker detection by broadening the range of spatial molecular markers linked to metastasis and recurrence. Future endeavors will further explore these findings to enrich our comprehension of various diseases and uncover molecular pathways with greater nuance. Combining deep learning with spatial transcriptomics provides a compelling avenue to enrich our understanding of tumor biology and improve clinical outcomes. For results of the highest fidelity, however, effective specimen processing is crucial, and fostering collaboration between histotechnicians, pathologists, and genomics specialists is essential to herald this new era in spatial transcriptomics-driven cancer research.

Assay for transposase-accessible chromatin using sequencing (ATAC-seq) is a cornerstone for epigenomic profiling, yet its potential for genomic characterization remains poorly explored. Here, we systematically benchmarked bulk ATAC-seq against whole-genome sequencing (WGS) to assess its capacity for detecting small variants, copy number variations (CNVs), telomere-associated repeat content, and mitochondrial single nucleotide polymorphisms in cancer cells. Using paired datasets from patient-derived melanoma cell lines and from TCGA primary brain tumors, we demonstrated that ATAC-seq achieves high precision in small variants detection within accessible regions supporting cohort-scale genotyping and genetic stratification, robustly resolves CNVs in the nuclear genome, and support high-coverage mitogenome profiling, with strong concordance to WGS at standard sequencing depths. Notably, we present the first systematic evaluation of telomere-associated repeat content by ATAC-seq, revealing its untapped potential for studying genome stability. By bridging genomic and epigenomic insights into a single genome-wide approach, bulk ATAC-seq emerges as a cost-effective and versatile tool poised to transform cancer research and to support integrative molecular profiling in clinical settings.

Spatial proteomics techniques have revolutionized our understanding of tissue architecture, but are frequently limited by detection sensitivity, bioconjugation limitations, multiplexing capacity, and multi-modal integration. Here we present Protein and nucleic Acid Serial Tyramine Amplification (PASTA), a novel signal amplification approach that significantly enhances detection sensitivity while maintaining compatibility with diverse spatial profiling methodologies. PASTA utilizes horseradish peroxidase (HRP) recruitment pathways to generate tyramine radicals that deposit oligonucleotides, enabling adaptable signal amplification across multiple biomarkers at high-plex via cyclical imaging using complementary fluorophore-labeled oligonucleotides. We demonstrate that PASTA achieves up to 100-fold signal enhancement for markers with minimal background in blank controls. The method is compatible with in situ hybridization for DNA/RNA detection, proximity ligation assays for protein-protein interactions, sequential antibody staining protocols, or any modular combination thereof. PASTA enables antibody rescue of markers with suboptimal signal-to-noise ratios and is versatile in its applications to unconjugated antibodies, and multi-round probe-based RNA detection systems beyond current capabilities. This technique addresses key limitations in spatial-omics by enhancing sensitivity for challenging targets while maintaining compatibility with established multiplexing strategies, providing a versatile, cost-efficient, and valuable tool for comprehensive spatial tissue analysis in both research and clinical applications.

Owing to the high demand for molecular testing, the reporting of tumor cellularity in cancer samples has become a mandatory task for pathologists. However, the pathological estimation of tumor cellularity is often inaccurate. We developed a collaborative workflow between pathologists and artificial intelligence (AI) models to evaluate tumor cellularity in lung cancer samples and prospectively applied it to routine practice. We also developed a quantitative model that we validated and tested on retrospectively analyzed cases and ran the model prospectively in a collaborative workflow where pathologists could access the AI results and apply adjustments (Adjusted-Score). The Adjusted-Scores were validated by comparing them with the ground truth established by manual annotation of hematoxylin-eosin slides with reference to immunostains with thyroid transcription factor-1 and napsin A. For training, validation, retrospective testing, and prospective application of the model, we used 40, 10, 50, and 151 whole slide images, respectively. The sensitivity and specificity of tumor segmentation were 97% and 87%, and the accuracy of nuclei recognition was 99%. Pathologists altered the initial scores in 87% of the cases after referring to the AI results and found that the scores became more precise after collaborating with AI. For validation of Adjusted-Score, we found the Adjusted-Score was significantly closer to the ground truth than non-AI-aided estimates (p<0.05). Thus, an AI-based model was successfully implemented into the routine practice of pathological investigations. The proposed model for tumor cell counting efficiently supported the pathologists to improve the prediction of tumor cellularity for genetic tests.

Improved methods are needed to gain insights in how proteins exert their myriad roles in cells and organs. Multiplex in situ proximity ligation assay (misPLA), described herein, can provide a window into the functional states of proteins in cells and tissues by applying pairs of antibody-oligonucleotide conjugates to generate amplifiable DNA circles upon proximal binding. The analysis reveals interactions and modifications among sets of proteins, read out by recording the identity and location of the resulting localized DNA amplification products. We applied misPLA to both primary and cultured cells and to formalin-fixated paraffin-embedded (FFPE) tissues, to map dynamic changes in protein localizations, phosphorylations and interactions across surface markers, MAPK, immune-checkpoints, T- and B-cell receptors, and adhesion panels. Comparisons of single-plex versus nine-plex assays confirmed that misPLA maintains sensitivity and specificity while increasing throughput and spatial context. Across breast cancer, lymphomas and chronic myeloid leukemia (CML) misPLA uncovered shared and disease-specific signaling patterns, underscoring convergence of oncogenic networks. By preserving tissue architecture and enabling high-content functional spatial proteomics at single-cell resolution, misPLA offers a versatile platform for dissecting signaling heterogeneity, pathway crosstalk, and therapeutic responses, with broad applications in cell biology, biomarker discovery and in precision oncology.

Melanocytic atypia, ranging from benign to malignant, often leads to diagnostic discordance, complicating its prediction by machine learning models. To overcome this, we paired H&E-stained histology images with contiguous or serial sections immunohistochemically (IHC) stained for melanocytic cells via antibodies for MelanA, MelPro, or SOX10. We developed a deep-learning pipeline to identify melanocytic atypia by digitizing a real-world archival dataset of 122 paired whole slide images from 61 confirmed melanoma in situ (MIS) cases at two institutions. Only 37.7% of the cases contained tissue pairs that matched well enough for deep learning. Nonetheless, the MelanA+MelPro models achieved an average area under the receiver-operating characteristic (AUROC) of 0.948 and an average area under the precision-recall curve (AUPRC) of 0.611, while the SOX10 models had an average of 0.867 AUROC and 0.433 AUPRC. Despite learning from biologically different IHC stains, the convolutional neural network (CNN) models independently exhibited an intuitive convergent rationale by explainable AI saliency calculations. Different antibodies, with nuclear versus cytoplasmic staining, provided complementary yet consistent information, which the CNNs integrated effectively. The resulting multi-antibody virtual stains identified morphologic cytologic and small-scale architectural features directly from H&E-stained histology images, which can assist pathologists in assessing cutaneous MIS.

The characterization of tissues using multiple different primary antibodies detected by secondary antibodies, each possessing a different colored fluorophore (multiplex immunofluorescence), is a powerful technique but often impaired by endogenous autofluorescence present in the specimen. Our current research involves the use of multiplex immunofluorescence to identify specific cell phenotypes within the tumor microenvironment in archival formalin-fixed paraffin-embedded human prostate cancer tissue specimens. These specimens frequently possess high levels of autofluorescence, in part due to the biological age of the tissues and long storage times. This autofluorescence interferes with and, in the worst cases, completely obscures the desired immunofluorescent signals, thus impeding analyses by decreasing signal-to-noise. Here, we demonstrate that a recently published protocol for photochemical bleaching significantly decreases autofluorescence (80% average decrease of the brightest autofluorescent signals), across the visible spectrum, in fixed, archival prostate tissue specimens from aged men, that have been sectioned onto glass slides and stored for several months. Importantly, the method is compatible with subsequent immunofluorescence staining and yields markedly improved signal-to-noise. Inclusion of this method should facilitate studies employing multiplex immunofluorescence in sections cut from archival fixed human prostate tissues.