A Fast And Versatile Method for Simultaneous HCR, Immunohistochemistry And Edu Labeling (SHInE)

Abstract Access to newer, fast, and cheap sequencing techniques, particularly on the single-cell level, have made transcriptomic data of tissues or single cells accessible to many researchers. As a consequence, there is an increased need for in situ visualization of gene expression or encoded proteins to validate, localize, or help interpret such sequencing data, as well as put them in context with cellular proliferation. A particular challenge for labeling and imaging transcripts are complex tissues that are often opaque and/or pigmented, preventing easy visual inspection. Here, we introduce a versatile protocol that combines in situ hybridization chain reaction, immunohistochemistry, and proliferative cell labeling using 5-ethynyl-2′-deoxyuridine, and demonstrate its compatibility with tissue clearing. As a proof-of-concept, we show that our protocol allows for the parallel analysis of cell proliferation, gene expression, and protein localization in bristleworm heads and trunks.


Introduction
Advancements in sequencing techniques have significantly increased the number of organisms for which molecular research can be conducted in. Moreover, single-cell RNA sequencing allows such investigations on the level of individual cell types (Fonseca et al. 2016;Stark et al. 2019). In turn, these high throughput methods have generated a need for validating digital expression data in the corresponding organism, specifically for the visualization of gene-expression on the threedimensional level.
In recent years, strategies complementary to traditional enzyme-coupled in situ hybridization (ISH) have been developed. One of them is in situ hybridization chain reaction (HCR). In situ HCR is faster to perform than traditional ISH, and works at lower temperatures, which aids maintaining tissue integrity. Moreover, in situ HCR can be multiplexed by using different fluorophores compatible with fluorescent imaging (Choi et al. 2016), making in situ HCR an attractive companion technique to single-cell sequencing.
For in situ HCR, short DNA probes are synthesized that are complementary to the target transcript. These probes can be designed using web-tools and ordered on bulk scale, similar to primers, which makes production fast and affordable. Amplifier oligonucleotides carrying fluorophores are then annealed to an overhang region on the probes, triggering the release of a hairpin struc-Advance Access publication March 2, 2023 C The Author(s) 2023. Published by Oxford University Press on behalf of the Society for Integrative and Comparative Biology. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (https://creativecommons.org/licenses/by/4.0/), which permits unrestricted reuse, distribution, and reproduction in any medium, provided the original work is properly cited.

Fig. 1
Experimental workflow for simultaneous in situ HCR, IHC, and EdU labeling. Following the initial treatment with EdU, animals are sampled and fixed for in situ HCR. Our protocol allows for the implementation of HCR amplification, EdU detection (click reaction), and incubation with primary antibody for IHC on day 2, resulting in a total of 4 days from sampling to imaging. ture, and thereby revealing the binding sequence for another amplifier molecule. This leads to a chain reaction of linear signal amplification. The latest version of the method, termed HCR 3.0 (Choi et al. 2018), employs split initiator sequences: Two DNA probes form a probe pair and have to successfully bind next to each other to form a shared initiator sequence. Off-target binding or trapping of individual probes therefore should be less likely to cause signal, improving sensitivity of the method.
Whereas, high sensitivity is an important feature for the use of in situ HCR to validate scRNAseq data, there are also challenges and open questions to the use of this technique: First, the acquisition of images from in situ HCR-stained samples by confocal microscopy bears potential problems in larger specimens, as opaqueness of tissue or pigmentation are known to prevent deep imaging (Tainaka et al. 2015). While various tissue clearing methods have been developed to overcome imaging problems in such specimens (Vieites-Prado and Renier, 2021), their compatibility with RNA detection has not been systematically assessed.
Second, while visualizing RNA is a relevant experimental aim, there are contexts where covisualization of proteins by immunohistochemistry (IHC) would yield additional insights. For example, such codetection could help to benchmark gene expression in cells or tissues for which protein markers are already established, or to directly assess ratios of RNA and corresponding protein (Darmanis et al. 2015;Albayrak et al. 2016). Combining the visualization of protein and RNA can therefore be used to study developmental processes or gene regulation in ways each technology on its own would not allow.
Few protocols have been published that combine the convenience, flexibility, and sensitivity of in situ HCR with immunostainings. Those describing a combined approach either rely on the availability of commercial antibodies and proprietary technology (Schwarzkopf et al. 2021) or follow a longer protocol that performs immunostaining after finishing in situ HCR (Ibarra- García-Padilla et al. 2021;Elagoz et al. 2022), which delays workflows and potentially lowers the achieved signal.
Here we present a protocol (termed Simultaneous HCR, IHC, and EdU labeling/SHInE; Fig. 1) that addresses both of these challenges: SHInE combines in situ HCR with IHC, allowing codetection of RNA and protein. By introducing the primary antibody incubation of the IHC simultaneously to the amplification step of in situ HCR, SHInE saves experimental time and keeps the number of washing steps following amplification to a minimum. Moreover, SHInE is compatible with tissue clearing using the recently developed DEEP-Clear method (Pende et al. 2020), extending the portfolio of labeling techniques for this method.
Our protocol also offers the flexibility of using noncommercial, self-designed HCR probes along with custom antibodies and home-made buffers. In addition, SHInE has been adapted to include additional molecular assays, such as the labeling of proliferating cells with 5-ethynyl-2'-deoxyuridine (EdU) (Salic and Mitchison, 2008;Zattara and Özpolat, 2020). This further enables the study of complex biological processes.
To demonstrate the functionality of our protocol, we provide results for different sets of HCR probes, different antibodies, and in different tissues of the marine bristle worm Platynereis dumerilii. This invertebrate has gained importance as a model organism in various biological fields, including evolution, development, regeneration, neurobiology, reproduction, chronobiology, and ecology [reviewed in (Özpolat et al. 2021)]. As in other model systems, conventional, riboprobebased ISH has been successfully established for gene expression studies in this species (Tessmar-Raible et al. 2005). This approach has recently been expanded to the use of single-color and multiplexed HCR 3.0 (Revilla-i-Domingo et al. 2021; Kuehn et al. 2022). By combining HCR labeling of RNA with IHC, EdU, and tissue clearing, SHInE provides a new and flexible tool for studying a broad range of biological processes in P. dumerilii, and can likely be adapted to other species.

Materials and methods
A schematized workflow of the SHInE protocol is presented in Fig. 1. A detailed version of the protocol, including all reagents and buffers, has been deposited on protocols.io (DOI: dx.doi.org/10.17504/protocols.io.5qpvobnyzl4o/v1). This version contains detailed steps with estimated time requirements, pause points, and steps that can be adjusted according to specific experimental needs (Supplementary File 1).

Animal husbandry
Platynereis dumerilii worms were kept in continuous culture at the Max Perutz Labs Vienna Marine Facility under a 16:8 light: dark regime. For details on animal culture, see (Hauenschild and Fischer, 1969;Kuehn et al. 2019). In short, adult worms were mated, and batches were reared in flat, transparent plastic containers. Depending on batch size, around 50 to 200 animals can be kept together this way, in volumes ranging from 0.5 to 1 l of sea water mix.
Animals were kept in a 1:1 mixture of sterile-filtered natural sea water and artificial sea water (Tropic Marin Classic) adjusted for salinity (to 34-35 parts per thousand).This mix is referred to as ASW/NSW throughout the manuscript. Animals were fed twice a week, once with organic spinach leaves cut to roughly 3 × 3 mm pieces in a blender, and once with a mix of ground Tetramin fish food and Spirulina powder.

Posterior amputations
Sibling worms aged 3-5 months and of a size between 40 and 50 segments were sampled. Trunk pieces were surgically amputated by anesthetizing animals in 7.5% MgCl 2 mixed 1:1 with ASW/NSW, then removing all segments posterior to the 30 th segment with a scalpel, cutting perpendicular to the animals' body axis, similar to previous reports (Planques et al. 2018). Amputated animals were then placed in fresh ASW/NSW, after which they were left for recovery and regeneration in a clean culture box of 0.5l for up to 30 animals. Animals used for posterior amputation experiments were not used for any additional experiments.
For each experimental condition, 3-5 animals were used. Representative samples were selected and imaged for figures of this manuscript. Statistics in Fig. S1 and Fig. S2 are based on evaluation of multiple biological samples.

Sampling of worm heads for in situ HCR
Wild-type worms of similar age and developmental stage were sampled at zeitgeber time (ZT) 22.

EdU pulse labeling and click reaction
The concept of EdU incorporation (Salic and Mitchison, 2008;Zattara and Özpolat, 2020) was used to label proliferating cells in the animals. To this end, worms were transferred to glass beakers and incubated with 10 μM EdU in ASW/NSW.
For EdU labeling worm heads, animals were incubated with EdU at the onset of darkness (ZT16) for 6 hr. Worm heads were subsequently sampled at ZT22.
For EdU labeling of worm blastemas, animals that underwent amputations (see above) were incubated for 1h.
The "click" reaction is the detection step during which fluorescent dyes are attached to EdU to label those cells that previously incorporated EdU during DNA synthesis. The reaction was performed according to the manufacturer's protocol using the Click-iT™ EdU Cell Proliferation Kit for Imaging, with Alexa Fluor™ 488 dye. In short, following HCR probe binding and washes, samples were incubated for 30 min with Alexa Dye azides to label EdU in a copper-catalyzed, enzyme-free stable reaction. For more details, refer to the manufacturer's protocol or the SHINE protocol, day 3, Step Case "HCR and EDU" or "HCR, EDU, and IHC," Step 12.

Microscopy parameters and image processing
Images were taken on a Zeiss LSM 700 inverse confocal microscope using Plan-Apochromat 20x/0.8, WD 0.55 mm and LD LCI Plan-Apochromat 25x/0.8 mm (oil immersion) lenses. Upon screening, the acquisition parameters were set according to the specimen with the strongest fluorescence and applied on one entire set of samples, thereby avoiding possible overexposure and ensuring comparability. Using FIJI (Schindelin et al. 2012), regions of interest (ROIs) were selected on every raw image and fluorescence intensity was determined by multiplying the measured mean gray value of a ROI with its corresponding area. Signal-to-noise ratio was calculated by dividing the fluorescence intensity of each sample by the average fluorescence intensity of the corresponding controls. Contrast (a linear adjustment) was enhanced in FIJI (Schindelin et al. 2012) for every channel separately. Equal settings were used within each experiment to ensure that samples could be quantitatively compared. Minimum and maximum displayed fluorescence values were set as follows: Fig. 2 and Fig. 3: 339-2465 for IHC (Cy3) channel and 0-2372 for EdU (AF488) channel; Fig. 4: t482-808 for HCR (B2 Amplifier-Alexa 647); Fig. S1:1237-2569 for HCR (B1 Amplifier-Alexa 546). Panels were arranged in Adobe Illustrator.

HCR Probes
HCR probes were designed against the listed target genes (Table 1) using a probe maker tool developed by Ryan Null (Kuehn et al. 2022) that was installed using Python v3.9.2 and jupyterlab 3.0.10.
The number of probe pairs depended on the length of each target sequence. As detailed in Supplementary File 2, we used 30 probe pairs for Platynereis hox3, 49 probe pairs for Platynereis pdp1, and 27 probe pairs for Platynereis per.

Antibodies
Details on the used antibodies are listen in Table 2.

SHInE allows codetection of HCR probes, protein, and nuclear labels in whole-mount samples
To address the ability of SHInE to covisualize RNA, protein and S-phase-labeled nuclei, we produced and imaged two sets of samples. On the one hand, we used Platynereis heads to correlate the expression of the circadian clock gene period with the photoreceptor protein L-Cryptochrome (L-Cry) and the proliferation marker EdU (Fig. 2). On the other hand, we assessed expression of the posterior stem cell marker hox3 with stabilized microtubules and EdU in regenerating tail samples (Fig.  3). In both cases, background staining in the respective channels was controlled by the use of HCR probes directed against an unrelated sponge gene (see Table 1 for probe targets, Table 2 for target proteins, and Supplementary File 2 for HCR probe sequences). Moreover, obtained patterns were compared against the result of published expression studies to validate the accuracy of our method.

Covisualization of L-Cry, period mRNA, and EdU in heads
One of the interesting aspects about P. dumerilii is its use as a functional molecular model system for chronobiological analyses (Raible and Tessmar-Raible, 2014;  Fig. 2, the respective negative controls exhibit little to no background signal; autofluorescence in glandular tissue is observed in the HCR negative control (A'), but lacks the granularity and intensity of HCR signal and can therefore be easily distinguished; a.c.: anal cirrus.   Özpolat et al. 2021). At least two endogenous timing systems-a ∼24hr (plastic circadian-circalunidian) and a monthly (circalunar) one-coexist in Platynereis and are influenced by ambient light conditions (Zantke et al. 2013) Zurl et al. 2022). Period (Pladu_per) is a key gene of the circadian clock (Glossop and Hardin, 2002). The Platynereis period ortholog has previously been shown to be expressed in the head using riboprobe-based whole-mount ISH (Zantke et al. 2013).
In accordance with the published pattern, HCR probes designed against Platynereis period demarcate cells in the posterior medial brain as well as in the eyes (arrows and asterisks, respectively, in Fig. 2A and D). In addition to this established pattern, we noted weaker, more ubiquitous period staining, which cannot be seen in the negative control (cf. arrowheads in Fig. 4B/B'). Whereas, this expression is not seen in the conventional ISH for period, it might be attributed to the higher sensitivity of HCR and its ability to detect single mRNA molecules (Choi et al. 2018). In order to coassess cell proliferation in the head during the night, we incubated worms in EdU solution from darkness onset to 2 hr before the lights turned (i.e., for a total of 6 hr). Visualization of EdU using click chemistry revealed positive cells in the anterior part of the head, in a dispersed, salt-and-pepper pattern with few cells also localizing to the posterior of the head (Fig.  2B and D). At the investigated stage, we did not observe many proliferating cells in the eyes, which exhibit continuous growth and noticeably increase in size before sexual maturation (Fischer et al. 2010;Pende et al. 2020;Özpolat et al. 2021a).
To test the compatibility of SHInE with IHC, immunostainings were performed on the same samples using antibodies raised against Platynereis L-Cry. L-Cry is a photoreceptor that shares regions of expression with period and is well described also on the sub-cellular level for P.dumerilii Zurl et al. 2022). In line with the published expression patterns, the acquisition of SHInE-processed heads revealed expression of L-Cry in the medial brain nuclei and in the eyes (arrows and asterisks, respectively, in Fig. 2C and D), in the same region as period RNA ( Fig. 2A and D).

Covisualization of acetylated alpha-tubulin, hox3 mRNA, and EdU in posterior blastemas
Regenerating tissue after caudal amputation has been used in Platynereis for various studies of cell identity, regeneration and the hormonal control of maturation (Rosa et al. 2005;Balavoine, 2015;Starunov et al. 2015;Özpolat and Bely, 2016;Planques et al. 2018;Özpolat et al. 2021). A key gene involved in both posterior growth as well as regeneration is the homeobox gene hox3. It is expressed in a distinct set of cells believed to be ectoteloblasts (ectodermal stem cells). These cells are found in a region of high proliferation and expression of genes belonging to the germline multipotency program, called the segment addition zone (Gazave et al. 2013;Planques et al. 2018). We chose this gene for testing our protocol as it has a highly distinct, well-studied expression pattern. Cells positive for hox3 also exhibit enlarged nucleoli, which makes it possible to identify them and validate the hox3 staining.
Detection of hox3 using HCR probes revealed a distinct region along the border between the pygidium and the regenerating segments (arrowheads in Fig. 3A and D). We colabelled samples using antiacetylated tubulin antibodies, which are established to visualize neurites of the nervous system. Antibody signal was found along the ventral nerve cord (VNC) and extended laterally into the animal's parapodia and posteriorly into the newly regenerated anal cirri (asterisk and arrowheads, respectively, in Fig. 3C and D). Both hox3 HCR and antiacetylated tubulin labeling match the established patterns, demonstrating the compatibility of these two techniques with the SHInE protocol.
As with head samples, we also performed an additional labeling of proliferative cells by pretreating animals with EdU for 30 min just before fixation, and visualized EdU incorporated in cells undergoing S-phase using click chemistry. The pattern of proliferation is broad and matches previous observations of regenerating tissue after posterior amputation ( Fig. 3B and D) (Planques et al. 2018). Taken together, the analyses performed in both head and posterior regenerates show that SHInE allows for codetection of nuclear EdU with both mRNA and protein.

SHInE is compatible with tissue clearing
We have previously presented a novel method on tissue clearing and depigmentation named DEEP-Clear. This method is compatible with various labeling techniques, including IHC, fluorescent proteins, and EdUlabeling. Enhanced transparency and refractive index matching allow for high-resolution imaging of specimens up to the centimeter range (Pende et al. 2020). While DEEP-Clear was shown to allow for the detection of RNA using conventional riboprobes, the employed chemistry includes an alkaline solution, in which RNA is expected to be degraded over time. As in situ HCR only requires binding to short pieces of RNA, rather than alignment with longer mRNA molecules, we reasoned that after clearing, in situ HCR should yield comparable or even superior results when compared to riboprobe-based ISH. As a suitable test case for assessing the compatibility of SHInE with tissue clearing, we focused on the aforementioned expression domains of Pladu-period in the eyes and the oval-shaped domain ( Fig. 4 B/B') of the posterior medial forebrain. Whereas, worm eyes are strongly pigmented, and the brain is opaque, DEEP-Clear had been shown to improve visualization of signals in both tissues (Pende et al. 2020).
Indeed, a comparison of untreated samples (Fig. 4B') and samples pre-processed using DEEP-Clear (Fig. 4B) revealed that tissue clearing was not only compatible with in situ HCR, but also led to a noticeable improvement of the signal. To quantify this effect, we compared Pladu-period signal in the medial brain with equivalent regions of samples hybridized with the unrelated sponge HCR probes (see Materials and methods section) carrying the same amplifier (Fig. 4A/A'). This quantification revealed a significant increase in signalto-noise ratios upon tissue clearing (see Supplementary  Fig. S1).
The degree to which tissue clearing improves signal might depend on the type of fluorophore used: When we investigated the expression of the PAR-domain protein 1 gene (pdp1)-that is expressed in a similar pattern-using a spectrally different HCR amplifier (B1 amplifier coupled to Alexa 546), tissue clearing did not improve HCR signal intensity to the same extent than it did for the B2 amplifier coupled to Alexa 647) (see Supplementary Fig. S2). For blastemal tissue that is neither pigmented nor particularly opaque, clearing did not improve the HCR signal noticably (Fig. 4C-D').
In summary, we find that SHInE is compatible with tissue clearing using DEEP-Clear, which, depending on the type of tissue and the fluorophore used for labeling, can improve the HCR signal quality and reduce background signal.

Discussion
Here we present a novel protocol that combines in situ HCR with IHC and EdU labeling of proliferating cells. The simultaneous visualization of RNA and proteins in the same tissue, as well as the high sensitivity of HCRmediated RNA visualization, make the protocol suitable for detailed studies on the level of organisms, tissues, and cells.
As we demonstrate, it is possible to successfully apply SHInE using home-made reagents, with HCR probes designed with a free webtool, home-made buffers and custom antibodies. This serves to keep the protocol affordable and flexible to many biological questions. The combination of multiple labeling methods within the same steps also greatly reduces the time required compared to similar protocols, further increasing the accessibility of these kinds of experiments, or allowing researchers to sample more replicates or experimental conditions. Additionally, SHInE gives the researcher full control over reagents. This opens the protocol up for future improvement or adaptation to other models and tissues. For example, dextran sulfate, which is part of the probe amplification buffer, has been reported to negatively impact some antibodies during IHC (Callahan et al. 1991), so changing the concentration of the reagent could help in such cases. At several stages of the protocol, we tested different HCR probe concentrations and amplification lengths and report our findings and suggestions in the main protocol (Supplementary File 1), providing a foundation for other researchers to adapt and modify the protocol to their requirements.
As we show, compatibility of SHInE with tissue clearing allows for improved HCR signal intensity in originally opaque and pigmented tissue. Not only do we provide evidence for the compatibility of DEEP-Clear with in situ HCR, but furthermore show an improvement in the signal-to-noise ratio of acquired images. However, the extent of this improvement may vary depending on tissue type and spectral range of the used fluorophores.
We believe that this protocol, with its flexibility and ease of use, combined with low cost and customization opportunities, will be a valuable resource. Whereas, we have focused on Platynereis as a model species, the broad applicability of in situ HCR and DEEP-Clear for various invertebrate and vertebrate species suggests that the combined protocol we present here can easily be adapted for other model systems, and thereby help to complement the powerful possibilities opened by scR-NAseq in a variety of model species.

Data availability statement
All data relevant to this study are either included or referenced to within the main manuscript or its supplemental material.