Recent advancements have improved the sensitivity of RNA detection on tissues or whole cells, down to single molecule level (Baysoy et al., 2023; Bressan et al., 2023; Xu et al., 2023; Bai et al., 2024; Monne Rodriguez et al., 2023; Pringle et al., 1989; Wilcox, 1993). Our ability to detect RNA or DNA in tissue sections and whole mount preparations through in situ hybridization (ISH) using complementary Watson-Crick base pair binding of labelled probes is now very well established (Monne Rodriguez et al., 2023; Pringle et al., 1989; Wilcox, 1993), originally described more than half a century ago (Gall and Pardue, 1969).

Initially situated in individual and specialist core facility labs staffed by highly experienced technicians, ISH used to be considered a long and difficult assay to set up in the lab, especially when radio-labelled riboprobes had to be designed and synthesized in-house. Photographic emulsion-coated slides had to be stored for days or weeks before autoradiographic detection of silver grains to detect signal. As well as potential safety concerns, these experiments resulted in relatively poor spatial resolution with noticeable background staining (Gall and Pardue, 1969).

These challenges lead to the incorporation of non-isotopic labeling of riboprobes including biotin or digoxigenin by nick translation that enabled some degree of amplification using conjugated secondary anti-hapten antibodies. The bound probe was ultimately visualized using a chromogenic or fluorescent substrate. Indeed, this also opened the potential to stain a much wider range of sample types including whole embryos, thick vibratome sections and organoids (Fleck et al., 2023; Velasco et al., 2019; Borrelli and Moor, 2020).

In the last decade, additional commercial non-isotopic ISH technologies such as RNAscope from Advanced Cell Diagnostics, ViewRNA from Affymetrix and HCR from Molecular Instruments have opened these techniques to many more labs. The relative ease of use, including probe design has made the RNAscope and related assays increasingly popular in academic as well as clinical and Pharma settings for the detection of mRNA and other RNA species in tissues, cells and whole mount preparations.

As shown in Figure 1, these commercial kits and probes use dual binding of two separate probes in close proximity (ZZ-probes) to create a template for signal amplification using a branched DNA process. The dual probe requirement and elimination of repetitive sequence elements improve specificity, while the branched DNA hybridization process enhances sensitivity. Additionally, the reactions can be visualized as either chromogenic signals, which are suitable for standard light microscopes, or fluorescent signals. Both technologies can be used manually or with automated staining platforms (Mahmood and Mason, 2008). Hybridization signals can be quantified either manually through a modified H-Score, a semi-quantitative combined measure of the percentage of positively stained cells and the number of individual punctate signals or using image analysis tools such as HALO and Visiopharm to compare relative expression levels. Non-isotopic ISH technologies are safer and sensitive, and work well on both sections as well as whole-mount samples including fruit fly, chick and mouse embryos (Koshiba-Takeuchi, 2018).

These advanced techniques facilitate the delivery of spatial data in tissue sections and serve as excellent complements to other quantitative molecular assays such as reverse transcription quantitative real-time PCR, next-generation sequencing, and spatial transcriptomic multiplexing approaches. Increasingly, ISH on tissue sections has been utilized to orthogonally validate antibody staining results obtained through immunohistochemistry (IHC) or to act as a substitute when validated antibodies for a target are unavailable, the target proteins are secreted, or the target is a non-coding RNA (Atout et al., 2022; Bingham et al., 2016). Furthermore, ISH has enabled the examination of less conventional species and sample types, including plants, insects, and cephalopods, especially when protein analysis via IHC is impractical or challenging due to the scarcity or high cost of developing antibodies (Solanki et al., 2020; Shao et al., 2024; Elagoz et al., 2022; Gattoni et al., 2025).

ISH is now recognized as a versatile and well-established method used in both research and clinical settings for detecting coding and non-coding RNA species as well as DNA. It provides routine diagnostic information that aids clinical decision-making. Furthermore, ISH assays are being increasingly applied to detect pathogenic viral target nucleic acid species in both the research and diagnostic settings including head and neck squamous cell carcinoma where RNAscope enabled the accurate and direct visualization of viral mRNAs encoding the E6 and E7 oncoproteins in FFPE samples detecting integrated and transcriptionally active High Risk human papillomavirus virus (HPV) in clinical samples (Bishop et al., 2012; Kravvas et al., 2025). The use of these ISH assays have demonstrated the potential for routine clinical patient screening of HPV E6/E7 mRNA in cancer samples (Musangile et al., 2021; Chen R. et al., 2023).

The same robust ISH assays have been used to detect Severe Acute Respiratory Syndrome Coronavirus 2 (SARS-CoV-2), which causes COVID-19 in humans, in patient tissue samples (Liu et al., 2020; Best Rocha et al., 2020; Pesti et al., 2023), as well as many other viral pathogens including HIV (Baharlou et al., 2022), Mpox (Hall et al., 2024), Ebola virus (Worwa et al., 2022) and Avian influenza (Gaide et al., 2023).

Recently other commercial ISH assays have been launched for detection of single to low-plex targets and one such example is the Fluorescent In Situ Hybridization Chain Reaction (HCR) (Choi et al., 2018). This is an enzyme-free thermally constant amplification method dependant on pairs of fluorescently conjugated DNA hairpin probes and target specific initiator sequence (Figure 1). This method continues to evolve to improve signal to noise achieving automatic background suppression by utilizing a pair of split-initiator probes that each carry half of the HCR initiator sequence that must co-localize to generate the specific signal. This assay has been especially useful for fluorescently staining up to five targets in whole-mount vertebrate embryos and thick vibratome tissue sections, generating qualitative and quantitative data (Choi et al., 2014). This method can also be used to fluorescently detect microRNA targets alone or in combination with endogenous mRNA targets and/or proteins (Zhuang et al., 2020).

New assays designed to fluorescently detect up to 12 mRNA targets in FFPE sections and 48 targets (in cryosections have also been developed, including the Hiplex assay from ACD, that requires iterative rounds of 4-plex staining imaging and gentle stripping of the bound probes (Hashikawa et al., 2020; Goodman et al., 2023). In addition, co-detection of proteins has now been facilitated using a protease-free pre-hybridization step in lieu of the original protease digestion step that preserves target protein epitopes, enabling IHC or immunofluorescence co-detection.

Spatial transcriptomics is a rapidly growing field encompassing many technological innovations allowing scientists to investigate the spatial distribution of thousands of mRNA transcripts within individual cells and tissues in FFPE tissue sections (Jain and Eadon, 2024; Sibai et al., 2025). This extremely rich gene expression data is quantitative and can be coupled with cellular segmentation to help confirm cell lineage and cell boundaries to add increased granularity to the data. Broadly, these can be divided into two categories, those based on sequencing (e.g., Visium and GeoMX) and those that are more image-based (e.g., Xenium, CosMX and MERFISH). The sequencing-based assays have enabled whole transcriptome mapping and identification of splice variants (Chen T. Y. et al., 2023; Li et al., 2025; Nishi et al., 2025). The data obtained from sequencing-based assays from most of these methods present as individual spots within in a grid format. However, one disadvantage is that due to the size of these spots which normally overlay several cells, these technologies fail to provide true single cell resolution.

These platforms employ dozens of primary probes that bind to specific RNA targets in order to achieve high sensitivity, however, these fluorescent methods are not able to detect small RNAs, nor oligonucleotides.

Existing sequencing-based techniques including Visium require the acquisition of polyadenylated transcripts, lacking the sensitivity to detect many species of non-A-tailed RNAs, such as microRNAs and newly transcribed RNAs. Newer single-cell RNA-sequencing protocols, Smart-Seq-Total and VASA-seq, have modified the enzymatic polyadenylation process to allow plate-based and microfluidic-based single-cell total RNA-sequencing, respectively, enabling the spatial profiling of different RNA species in complex tissues (McKellar et al., 2023).

These powerful techniques generate a huge amount of data; however, they require expensive equipment and need highly specialized staff to run these experiments to fully utilize the data they generate.

Here we have focussed on lower plex approaches for detecting mRNA, miRNA and oligonucleotides in tissues that do not require such large capital expenditure and experience to set up and run., the higher plex assays used as part of spatial transcriptomic approaches have been reviewed elsewhere (Xia et al., 2019; McKellar et al., 2023).

Whilst the ability to detect relatively long mRNA transcripts has been well established (Wang et al., 2012; Zhang et al., 2024; Hu et al., 2025), more recently assays have been developed that enable the spatial profiling of much shorter RNA species including microRNAs (miRNA) and unlabelled oligonucleotide drugs in single and low-plex options in both frozen and paraffin tissue samples (Robles-Remacho et al., 2025). These approaches include miRNAscope (Wang et al., 2012; Spencer-Dene et al., 2023), HCR (Choi et al., 2018), NanoSIMS (King et al., 2024; He et al., 2021), Mass Spectrometry Imaging (van der Vloet et al., 2025) and CARS approaches (Spencer-Dene et al., 2023; Shi et al., 2023).

miRNAs comprise a set of small, non-coding RNA species typically 18-22 bp long that negatively regulate gene expression post transcriptionally. They play roles in many diseases including cancer and are also involved during development. Previously these could not be detected in tissue sections using standard in situ hybridization techniques but there are now several non-isotopic approaches that can be used including miRNAscope (Figure 1) and HCR (Jorgensen et al., 2010; Zedan et al., 2017; Nielsen et al., 2021). Other methods including CLAMP-FISH and SABER-FISH have not yet been reported for short targets, however this may be achievable in future (Higo et al., 2023).

Shorter 50-150 bp mRNA targets, including point mutations and splice variants in sub-optimally fixed and preserved tissues including human brain bank samples, can best be detected by a modified version of the RNAscope assay called BaseScope. While RNAscope uses paired 18-25 base “ZZ” probes that hybridize to the target sequence, BaseScope utilizes shorter probe sets of 1-3 ZZ pairs, and an extra signal amplification step which significantly increases sensitivity without generating excessive background staining. This chromogenic assay is developed chromogenically with Fast Red and is now also available as a chromogenic duplex assay (Rifai et al., 2023; Gregory et al., 2020; Baker et al., 2017). A summary of the most common branched DNA methods for detecting mRNA, miRNA and oligonucleotides in tissue sections is shown in Table 1.

Oligonucleotides are short single or double stranded RNA or DNA molecules that are increasingly being used as therapeutic modalities to treat both rare and common diseases. Once these short stretches of chemically modified nucleotides arrive at their target organ or cell type, they bind to their specific target in a complementary manner altering the transcriptional fate of that mRNA target, for example, catalysing its rapid degradation by RNase H1 thus reducing the downstream translation of that target protein that may be involved in a pathological process.

Recent advances in the detection of single and double stranded oligonucleotides in tissues designed to improve sensitivity and spatial resolution include the use of chloroalkane-conjugated oligonucleotides (Deprey et al., 2022), or novel highly sensitive imaging techniques including nanoscale secondary ion mass spectrometry (NanoSIMS) to detect radio-labelled antisense oligonucleotides (ASOs) in tissue (He et al., 2021). These labelled oligos could be readily discerned in specific cellular compartments and organelles including within the endolysosomal vesicles in hepatocytes and non-hepatocyte populations including the Kupffer cells and the liver sinusoidal endothelial cells. There are downsides to these powerful approaches, such as the need to use radiolabelled probes and the NanoSIMS hardware itself is highly complex and expensive.

In addition, commercial mouse monoclonal antibodies have now been generated against phosphorothioate groups that are one of the most common modifications engineered into oligo backbone chemistries to protect against endonuclease digestion. These have been shown to detect unlabelled single stranded ASOs in FFPE and frozen tissue sections by IHC (Kordasiewicz et al., 2012; Hung et al., 2013) and immunofluorescence (Brown et al., 2022; Ly et al., 2025), but unlike the miRNAscope and RNAscope Plus smRNA-RNA ISH assays shown in Table 1, are not specific for individual oligonucleotide sequences.

Whilst the use of FFPE samples for ISH detection offers many advantages, especially when starting material is particularly scarce, alternative sample types including fresh frozen and paraformaldehyde-fixed frozen samples have been used very successfully with ISH. In recent years, single-molecule fluorescent ISH approaches have been successfully used to detect multiple transcripts on fresh frozen skin and brain samples to complement larger transcriptomic and cell atlas studies (Gopee et al., 2024; Bayraktar et al., 2020). However, some aspect of the sample preparation such as perfusion fixation and sucrose cryoprotection can lead to adverse shrinkage and poor signal (Asp et al., 2006).

More recent innovations to try to achieve improved tissue clearing, greater imaging depth and more consistent and reproducible three dimensional labelling of mRNA transcripts in larger samples incorporating fluorescent ISH HCR have been used in fresh frozen mouse and rat brain as well as post-mortem human brain samples to detect neuronal transcripts (Kumar et al., 2021).

The growing success of modern ISH based assays and their increasing versatility in both the research and clinical diagnostic sectors (Monne Rodriguez et al., 2023; Pringle et al., 1989; Wilcox, 1993) has hugely significant economic considerations as the technology becomes more sophisticated and continues to grow. The ISH market size has been valued at $1.9 billion in 2025 and is predicted to rise significantly to $5.2 billion by 2035 (In Situ Hybridization Market | Global Market Analysis Report - 2035).

The ability to analyse gene expression and spatial localization within tissues has opened new avenues for understanding complex biological processes and developing novel therapeutic strategies (Higo et al., 2023). The move towards ever increasing higher plex spatial transcriptomic approaches incorporating automated platforms from whole tissues down to sub-cellular spatial resolution have dramatically improved our ability to resolve entire transcriptomes on tissue sections. There are now dozens of different methods available using different approaches for detection of many thousands of transcripts and whole genomes (Moffitt et al., 2022; Goltsev et al., 2018; Orjalo and Johansson, 2016; Kishi et al., 2019; Xia et al., 2019; Ren et al., 2024), and these technologies are continuing to improve. However, many of these assays are not yet able to detect smaller RNA species such as miRNAs nor oligonucleotides, suggesting that the lower plex approaches still have significant utility in the field.

Some challenges remain to be resolved. Whilst mRNA transcripts can be easily quantified as punctate signals corresponding to individual transcripts using image analysis tools, absolute quantification of unlabelled oligonucleotide concentration in tissues at single cell resolution is currently a major challenge as the oligo is primarily concentrated within the endolysosomal compartment. Use of image analysis tools can try to solve this through fluorescent or chromogenic intensity measurements to complement bioanalytical whole tissue bulk oligo pharmacokinetic quantification approaches.

Despite limitations and challenges, these persistently evolving techniques will lead to an expanding and intriguing field of novel applications with a substantial impact on a wide range of scientific fields, from developmental and tissue maintenance to tumour and aging biology.