Tyramide signal amplification coupled with multiple immunolabeling and RNAScope in situ hybridization in formaldehyde‐fixed paraffin‐embedded human fetal brain

Several strategies have been recently introduced to improve the practicality of multiple immunolabeling and RNA in situ hybridization protocols. Tyramide signal amplification (TSA) is a powerful method used to improve the detection sensitivity of immunohistochemistry. RNAScope is a novel commercially available in situ hybridization assay for the detection of RNA expression. In this work, we describe the use of TSA and RNAScope in situ hybridization as extremely sensitive and specific methods for the evaluation of protein and RNA expression in formaldehyde‐fixed paraffin‐embedded human fetal brain sections. These two techniques, when properly optimized, were highly compatible with routine formaldehyde‐fixed paraffin‐embedded tissue that preserves the best morphological characteristics of delicate fetal brain samples, enabling an unparalleled ability to simultaneously visualize the expression of multiple protein and mRNA of genes that are sparsely expressed in the human fetal telencephalon.


| INTRODUC TI ON
Immunohistochemistry (IHC) and RNA in situ hybridization (ISH) are very important techniques in the field of developmental neurobiology that allow us to assess the expression of protein and mRNA for a wide variety of transcription factors and discrete signaling molecules (e.g., SHH and FGF). However, the technical complexity, insufficient sensitivity, and specificity of the routine application of these two methods make their use very limited for some histological analyses, particularly post-mortem human embryonic and fetal tissue.
The development of appropriate multiple immunofluorescence labeling methods is one of the main obstacles in applying IHC. The selection of the appropriate combination of primary antibodies becomes problematical when the primary antibodies are raised in the same host species, which leads to cross-reactivity of secondary antibodies with each of the primary antibodies (Tóth & Mezey, 2007). A second general problem is the insufficient sensitivity of primary antibodies to target antigens that are present in small or barely detectable amounts, leading to poor signal visualization for these antigens (Van der Loos 2007; Warford et al. 2014). However, relatively novel hybrid detection strategies have been recently introduced to improve the practicality of multiple immunolabeling methods and circumvent some of these technical obstacles. Some of these strategies involve elution of antibodies between rounds of staining, using HRP polymer conjugated secondary antibodies and the tyramide signal amplification method which covalently binds fluorescent tags to the tissue at the site of antigen detection, increasing sensitivity, and allowing the washing away of antibodies while leaving staining behind before further immunostaining (for more detailed information on the principle and history of these methods

Abstract
Several strategies have been recently introduced to improve the practicality of multiple immunolabeling and RNA in situ hybridization protocols. Tyramide signal amplification (TSA) is a powerful method used to improve the detection sensitivity of immunohistochemistry. RNAScope is a novel commercially available in situ hybridization assay for the detection of RNA expression. In this work, we describe the use of TSA and RNAScope in situ hybridization as extremely sensitive and specific methods for the evaluation of protein and RNA expression in formaldehyde-fixed paraffinembedded human fetal brain sections. These two techniques, when properly optimized, were highly compatible with routine formaldehyde-fixed paraffin-embedded tissue that preserves the best morphological characteristics of delicate fetal brain samples, enabling an unparalleled ability to simultaneously visualize the expression of multiple protein and mRNA of genes that are sparsely expressed in the human fetal telencephalon.
Unlike other RNA analysis techniques (e.g., RT-PCR), in situ analysis of RNA biomarkers allows spatial descriptive analysis of RNA expression in the developing fetal brain and is important for verifying the results of single-cell RNA transcriptomics. For example, investigating the graded and compartmented expression of genes (transcription factors, morphogens, and receptors) in developing cerebral cortex permits us to understand how the dorsal pallium is stereotypically divided into functionally distinct areas even at the earliest stages (Alzu'bi et al. 2017a;Clowry et al. 2018;Molnár et al. 2019).
However, a long-standing criticism of routine RNA ISH techniques is the lack of sensitivity and specificity, especially for genes expressed at very low levels (e.g., genes for morphogens and their receptors); therefore, the use of these techniques in the field of developmental neurobiology has been challenging. Recently, Advanced Cell Diagnostics, Inc., (Hayward, California, USA) presented a novel RNA ISH technology (RNAScope in situ hybridization) that implement a unique strategy of probe formulation. Pairs of probes that bind adjacent RNA sequences must hybridize independently to their respective targets for subsequent signal amplification to occur. This pairing requirement greatly increases the signal-to-noise ratio and ensures high specificity because of the low probability of two different probes binding adjacent non-specific sequences (for more detailed information on the principle of this technology see Wang et al. 2012).
RNAScope technology has proved to be a reliable method in many areas of biological and pathological research giving reproducible and quantitative results Bingham et al. 2017;Chan et al. 2018;Jolly et al. 2019;De Biase et al. 2021). Further, the gentle tissue permeabilization technique used in this technology, and the possibility for using a fluorescent assay allow the combination of RNAScope in situ hybridization with immunohistochemistry, enabling simultaneous visualization of mRNA and protein antigens (Wang et al. 2012;Grabinski et al. 2015;Kann & Krauss 2019;Annese et al. 2020). However, the use of RNAScope technology, specifically in combination with immunohistochemistry, in human fetal samples has not been extensively evaluated. In this study, we have evaluated the feasibility of multiple immunolabeling (with TSA method) and RNAScope (including coupling with immunofluorescence) techniques in formalin-fixed, paraffin-embedded sections of human fetal forebrain.

| Human fetal brains and ethical approval
Human fetal tissue from terminated pregnancies was obtained from the joint MRC/Wellcome Trust-funded Human Developmental Biology Resource (HDBR, http://www.hdbr.org; Gerrelli et al. 2015) based in Newcastle University and the Institute of Child Health University College London. All tissue used in this study was collected by the Newcastle center with appropriate maternal consent and approval from the Newcastle and North Tyneside NHS Health Authority Research Ethics Committee (REC reference 18/NE/0290).
Both centers are licensed by the UK Human Tissue Authority (license numbers 12,534 and 1220). All embryonic and fetal samples used in this study were obtained from RU486 induced medical terminations and ranged in age from 8 to 19 post-conceptional weeks (PCW). Ages were estimated from foot, and heel to knee length measurements according to Hern (1984). In addition, samples were karyotyped. Full details are provided in Table 1.

| Tissue processing and sectioning
Fetal brains were rapidly isolated and fixed for at least 24 h at 4°C in 4% formaldehyde (from paraformaldehyde powder, Sigma Aldrich, Poole, UK) in 0.1 M phosphate-buffered saline (PBS). 8 PCW embryos were similarly fixed and sectioned whole. The exact post-mortem interval to fixation is unknown but is in the range of 1-12 h. After fixation, embryos, whole, or half fetal brains were dehydrated in graded ethanol (70% for 15 min, 100% for 45 min, 2 × 100% for 60 min) at room temperature (RT). The larger fixed brains were dissected into blocks of approximately equal size, with the number depending on the size of the brain. Blocks were then incubated in xylene for 2 h before embedding in paraffin (Excelsior AS Tissue Processor, Thermo Scientific). Brain tissue blocks were cut into 8 μm thick sections (Leica RM 2235 microtome) mounted onto Superfrost+ slides (Thermo Fisher Scientific) and used for immunostaining and RNAScope in situ hybridization.

| Multiple immunofluorescence
This was performed using the method employed by Tóth & Mezey (2007) in adult rodent models with slight modification. In the first round of staining, paraffin sections were first dewaxed in xylene (2 × 5 min) and then rehydrated via four changes of graded ethanol (100%, 100%, 95%, and 70%). Endogenous peroxidase activity was blocked by treatment with methanol peroxide (3 ml of 30% hydrogen peroxide +180 ml methanol) for 10 min. For antigen retrieval, sections were rinsed under tap water and boiled in a TA B L E 1 Details of embryonic/fetal samples  microwave in 10 mM citrate buffer pH 6 for 10 min. Sections were then incubated with the appropriate normal 10% blocking serum (species in which secondary antibody was raised) in Tris-buffered saline pH 7.4 (TBS) for 10 min at RT before incubation with the primary antibody (diluted in 10% normal blocking serum) overnight at 4°C. Details of all the primary antibodies used in this study can be found in Table 2

| RNAScope in situ hybridization
Manual RNAScope assays were performed using The BaseScope™ Reagent Kit v2-RED (Catalog No. 323600, ACD Bio Techne) according to the manufacturer's protocol. The routine RNAScope assay consists of four main steps, the first step is the sections pretreatment to allow the access of the target probe (conceptualized as a "Z"), In the second step, multiple tandem probes are hybridized in pairs ("ZZ") to multiple RNA targets. In the third step, the detection probes are pre-amplified by an adapter linked to several amplifiers containing multiple chromogenic labels. In the last step, signal detection is carried out by developing a chromogen to produce small punctate red dots (Wang et al. 2012). In this study, each sample was quality controlled for RNA integrity with a probe specific to the ubiquitously expressed reference gene GAPDH. We have previously demonstrated by qPCR that in human fetal samples GAPDH mRNA is expressed at the same high level across all regions of cortex (Harkin et al. 2017). Negative control background staining was evaluated using a probe specific to the bacterial DapB gene.
Sections were then covered with drops of 30% hydrogen peroxide solution for 10 min at room temperature and washed in distilled water by moving the slides rack up and down three to five times. For target retrieval, sections were boiled with the manufacturer's target retrieval buffer (ACD Biotechne) for 20 min at 90-100°C using a Cookworks vegetable steamer (Argos). Individual tissue sections were isolated on slides using a hydrophobic barrier pen. To increase target accessibility, protease digestion was then carried out by incubating sections with protease plus solution (ACD Biotechne) at 40°C for 30 min (any oven or incubator maintain a temperature between 37-40°C is also suitable). Sections were finally washed with distilled water before proceeding to probe hybridization.

| Signal detection
For signal detection, sections were incubated for 10 min at RT in red solution (RNAScope 2.5 HD Assay-RED, ACD Biotechne) freshly prepared following the manufacturer instructions (mixing 1 volume of RED-B to 60 volumes of RED-A to make the total volume needed to cover the sections). Slides were then rinsed in distilled water for 2 min, counterstained with 10%-20% hematoxylin (using a higher concentration of hematoxylin can obscure any positive signal). Slides should be dried for 10-15 min in a 60°C dry oven and mounted using DPX (Sigma-Aldrich). Positive signals are indicated by red chromogenic dots in the cytoplasm or nucleus (Figures 3 and 4).

| RNAScope fluorescent in situ hybridization coupled with immunofluorescence
To avoid RNA degradation during the immunofluorescence steps, in situ hybridization was carried out first using RNAScope Multiplex Fluorescent Reagent Kit v2 Assay (ACD). Sections pre-treatment and probe hybridization steps were performed as described above.
Unlike RNAScope chromogenic assay, the signal amplification in

| Image acquisition
RNAScope in situ hybridization images were captured using Leica SCN400 Slide Scanner. Fluorescent images were obtained with a Zeiss Axioimager Z2 apotome. Processing of images, which included only adjustment of brightness and sharpness, was achieved using the Adobe Photoshop CS6 software.

| Multiple immunofluorescence
In Figures 1 and 2  with the same buffer used for antigen retrieval (10 mM citrate buffer, pH 6.0) in the microwave for 10 min. The heat treatment using the citrate buffer did not significantly reduce fluorescent signal from the first round, confirming that HRP-activated tyramide binds covalently and efficiently to electron-rich amino acids of proteins at the site of the immunoreaction making it resistant to the citrate treatment (Bobrow et al. 1989;Hasui & Murata, 2005;Shojaeian et al. 2018).
The use of these three strategies represents a significant improvement in multiple immunofluorescence labeling, profoundly improving the practicality of this method by giving bright images of antigen labeling we previously struggled to detect using fluorescently tagged secondary antibodies. The development of compact HRP polymer-conjugated secondary antibodies, to introduce large number of peroxidase molecules, has remarkably enhanced the detection sensitivity of this method ( Shi et al. 1999;Shojaeian et al. 2018).
This system also significantly simplified the staining procedures; it is ready to use and provides faster staining steps compared with using biotinylated secondary antibodies followed by treatment with avidin-biotin complex (ABC). Additionally, this system lowers the cost by permitting further dilution of expensive primary antibodies.
The intermediary antibodies elution method in combination with TSA methods provided a remedy for one of the major problems in F I G U R E 3 (a-c) Chromogenic RNAScope in situ hybridization for nAChR subunits mRNA in the cortical plate of 12 PCW fetal brain. CHRNA4 was relatively highly expressed compared with moderate expression for CHRNA5 and low expression for CHRNA7. (d) Very strong expression was detected for the positive control reference gene GAPDH. (e) No expression was detected for the negative control dapB gene.
(f-k) show results from conventional in situ hybridization. Comparison of sense/anti-sense staining (f, g) suggests expression of CHRNA4, particularly in the CP and VZ but it would not be easy to localize expression to individual cells as staining is diffuse. A similar result was obtained for CHRNA7 (j, k) whereas for CHRNA5, sense stained more strongly than antisense suggesting the experiment has failed, or a transcript from the reverse strand was detected.  we propose dual RNAScope ISH-IHC is also an extremely powerful method to study human neurogenesis, because it allows researchers to visualize simultaneous co-localization of gene and protein expression, and changes in expression, to specific cell populations across the developing fetal brain.

| Single and multiplexed RNAScope with immunofluorescence
RNAScope represents a valuable addition to RNA ISH methodology in the field of developmental neurobiology. This technique has many advantages-it is highly compatible with routine formalinfixed, paraffin-embedded fetal brains; sensitive enough, with remarkable background suppression, to allow detection of genes that are expressed at low levels like receptors and morphogens. This means that it can be used both to confirm expression of gene expression patterns revealed by single-cell RNAseq analysis but also detect expression of genes revealed by whole tissue RNAseq for which single-cell RNAseq is not sufficiently sensitive. Finally, it is a time saving method that can be performed in 1 day compared with routine RNA ISH which is time-consuming and labor-intensive (3-4 day protocol).
There is one drawback to using the RNAScope approach. The reagents and technology for making the probes are copyrighted and are expensive to purchase. However, the apparently guaranteed success of the method so far, in our hands, saves us a lot of time and money when considering the high failure rates we have encountered in using conventional methods. In addition, as described above, most of the incubation steps for protease digestion, probe hybridization, and signal amplification are required to be performed under strictly controlled conditions of temperature and humidity in a special oven sold by the manufacturer. Although we have found the oven easy to use and to give excellent results, any oven or incubator that can maintain a temperature between 37 and 40°C was found to work perfectly well.

ACK N OWLED G M ENTS
We are grateful to the staff of the Human Developmental Biology Resource. The human fetal material was provided by the Joint UK

DATA AVA I L A B I L I T Y S TAT E M E N T
Data sharing is not applicable to this article as no new data were created or analyzed in this study.