In situ hybridization protocol for enhanced detection of gene expression in the planarian Schmidtea mediterranea

Background The freshwater planarian Schmidtea mediterranea has emerged as a powerful model for studies of regenerative, stem cell, and germ cell biology. Whole-mount in situ hybridization (WISH) and whole-mount fluorescent in situ hybridization (FISH) are critical methods for determining gene expression patterns in planarians. While expression patterns for a number of genes have been elucidated using established protocols, determining the expression patterns for particularly low-abundance transcripts remains a challenge. Results We show here that a short bleaching step in formamide dramatically enhances signal intensity of WISH and FISH. To further improve signal sensitivity we optimized blocking conditions for multiple anti-hapten antibodies, developed a copper sulfate quenching step that virtually eliminates autofluorescence, and enhanced signal intensity through iterative rounds of tyramide signal amplification. For FISH on regenerating planarians, we employed a heat-induced antigen retrieval step that provides a better balance between permeabilization of mature tissues and preservation of regenerating tissues. We also show that azide most effectively quenches peroxidase activity between rounds of development for multicolor FISH experiments. Finally, we apply these modifications to elucidate the expression patterns of a few low-abundance transcripts. Conclusion The modifications we present here provide significant improvements in signal intensity and signal sensitivity for WISH and FISH in planarians. Additionally, these modifications might be of widespread utility for whole-mount FISH in other model organisms.

Step-by-step FISH protocol for planarians *Unless otherwise mentioned, steps are performed at room temperature. *Planarians are gently agitated throughout the protocol either on a nutator/rocker or by intermittent manual shaking unless noted otherwise. *For wash/incubation steps remove >90% of previous solution before adding new or fresh solution. Day 1: Animal fixation 1.1 Collect, wash, and transfer ~100 1-5 mm asexual planarians starved 1 week to a 15 mL conical tube. (For fewer animals volumes can be scaled down. For larger planarians, fewer animals can be processed or volumes can be scaled up) 1 1.2 Remove excess Planarian salts, add 10 ml of 7.5% NAC solution, and gently rock for 5-10 min.
1.3 Remove NAC solution and fix for 15-20 min in 9 ml 4% Fixative with gentle agitation.
1.5 Dehydrate in 10 ml 50% methanol for 5min, then in 10 ml 100% methanol 2 times for 5 min each and store animals at -20°C for 1 hr to a few months.

Day 2: Animal pretreatment and hybridization
2.1 Rehydrate samples by incubating in 10 ml 50% methanol for 5 min, then in PBSTx for 5 min.
2.3 Bleach animals by incubating in 15 ml freshly prepared Formamide-bleaching solution for 2 hrs under bright light 2 .
2.5 Wash in 10 ml PBSTx 2 times for 5 min each.
2.6a For intact planarians incubate in 10 ml ProteinaseK solution for 10-15 min with gentle agitation (see below for permeabilization of regenerates using HIAR) 3 .
2.8 Wash in 10 ml PBSTx 2 times for 5 min each.
2.9 Transfer 3-5 animals each to 2.0 ml screw cap tubes or to small in situ baskets (Intavis) in either an Intavis InSitu Pro robot or a 48-well plate.
2.11 Hybridize in 300 µl Riboprobe mix for >16 hrs at 56°C. Iterative TSA for enhanced signal intensity -For a single gene, develop first with DNPx-tyramide, wash for 24-hours (no need to quench peroxidase activity), incubate with anti-DNP-HRP antibody, wash, and perform second amplification with a fluorophore-conjugated tyramide.
-For iterative TSA with multicolor FISH, wash time can be reduced by performing detection for the more strongly expressed genes between iterative TSA reactions. For example, detection of a low abundance transcript, geneA, with DNP labeled riboprobe and a higher abundance transcript, geneB, with DIG labeled riboprobe could be performed as follows: first, perform TSA for geneA with DNPx-tyramide followed by inactivation of peroxidase activity with azide; then, following a few short washes, incubate with anti-DIG-POD antibody and detect geneB expression by TSA using a fluorophore-conjugated tyramide; finally, following inactivation of peroxidase activity for geneB and a few short washes, incubate with anti-DNP-HRP antibody and perform the iterative TSA reaction with a second fluorophore-conjugated tyramide. 1X SSC: 20X SSC stock diluted to 1X with deionized water 2X SSCx: 20X SSC stock diluted to 2X + 0.1% Triton X-100 0.2X SSCx: 20X SSC stock diluted to 0.2X + 0.1% Triton X-100
Yeast RNA: yeast RNA is prepared by dissolving yeast RNA in DEPC-treated water for three days, extracting through phenol, phenol-chloroform, and chloroform. RNA is precipitated and resuspended in formamide to a concentration of 50 mg/ml, stored at -20°C.
Blocking solution: 5% horse serum and 0.5% Roche Western Blocking Reagent (RWBR stock solution is at 10%) diluted in TNTx.  **Caution** at high concentrations formamide and H2O2 undergo a violent reaction. Always dilute these reagents into the water before mixing.

Tyramide Signal Amplification Buffer Cheat Sheet
Incubate for 10 min with tyramide solution. FAM, TAMRA, and DNPx tyramides can be used at ~1:500 dilution Dye Light 633 diluted 1:250 seems to give the best signal to noise for most applications.

Synthesis of Tyramide conjugates
The following protocol is adapted from [1]. Note: N-hydroxy-succinimydyl-esters are light sensitive and prone to hydrolysis, therefore it is best to use fresh, high quality anhydrous N,N-Dimethylformamide (DMF) and protect from light. 1) Prepare tyramine stock by dissolving tyramine hydrochloride to 10 mg/ml in DMF containing 10 µl/ml triethylamine. 2) Prepare fluor-conjugated NHS ester by dissolving to 10 mg/ml in DMF.
3) Add tyramine stock to fluor-conjugated NHS ester (see below). 4) Incubate at room temperature protected from light for 2 hours. 5) Dilute fluor-conjugated tyramide with 100% ethanol to 1 mg/ml (see below). 6) Aliquot and store at -20°C. (FAM-and TAMRA-tyramides are stable for at least a few years, DyLight 633 is stable for at least six months).

Riboprobe synthesis
The following protocol is adapted from the manufacturer's suggestions (Roche). DNA template for the in vitro transcription reaction was generated by PCR amplifying target sequence from cDNA that includes either T3 or SP6 promoter sequences (T7 could also be used). Unincorporated primers and nucleotides were removed from the PCR product using a DNA clean and concentrator kit (Zymo research).

DNAse Treatment
Add 5 µl DNAse mix and incubate at 37°C for 15min  [2]. 5. Planarians often become translucent during heating. 6. Do not post-fix samples following HIAR. Proceed directly to wash and prehyb incubation steps. 7. Prehyb, hyb, and wash hyb solutions in other protocols contain 1 mg/ml yeast RNA. We have not noticed any issues for several probes tested when yeast RNA is reduced to 100 µg/ml or substituted with 100 µg/ml Salmon Sperm DNA. 8. Riboprobe concentration is a critical variable to optimize. We have noticed interesting changes in gene expression patterns when probe concentration is varied. Typically, lower probe concentration results in cleaner signal for both low abundance transcripts and robustly expressed genes. Additionally, we have had better success with longer riboprobes. We have not noticed issues with probe penetration for probes under 1.5 kbp, as direct comparison between full-length probe, hydrolyzed probe, or smaller subcloned fragments for a gene yielded similar staining patterns. 9. 4IPBA is essential for Alexa 568 and Dye Light 633 tyramides to work. While it does not seem to be required for Alexa 488, FAM, or TAMRA tyramides, it also does not inhibit the reaction at the above concentrations, and at least for TAMRA yields slight enhancement of signal 10. Addition of Tween 20 (0.1% final) and dextran sulfate (2% final) to the TSA base buffer yields a noticeable increase in signal for several tyramides tested, but also increased background, resulting in poorer signal-to-noise ratio, similar to use of commercial TSA systems, such as the TSA plus system from PerkinElmer. Despite the slightly reduced signal, we prefer the signal-to-noise ratio achieved with the TSA buffer described. 11. For FAM-labeled riboprobes, LiCl/ethanol is very effective at precipitating small products and unincorporated FAM-UTP. We have used sephadex G50 quick spin columns (Roche) to further purify full-length probe from unincorporated nucleotides and shorter probe fragments, prior to probe precipitation. For the FAM-labeled probes we have examined, signal has been strong with minimal background whether probes were further purified with quick spin columns or not. However, for some probes yielding significant background, further probe purification may be beneficial. 12. We have often noticed that DNP-riboprobes yield very weak bands on a gel, despite similar concentrations as DIG or FAM riboprobes when quantified on a spectrophotometer. We have had success using DNP probes even when a band is not visible on a gel.