very interesting, but I think cost effectiveness is one thing for not using this tech widely. regards

We've used HCR as a substitute for conventional in situ hybridization, to detect RNA. We have it working in various whole-mount insect tissues (bees, crayfish, wasps). It doesn't seem to require as much optimization as conventional in situs, and is easier to set up (you order probes online, rather than constructing them in house).

The key advantages are:

- it is easier to get running than traditional in situs, and the protocol itself is simpler. There is not much need for optimization. 

- imaging up to 4 genes in one sample, plus nuclear marker (though 2-3 is much easier)

- I (and others) have used it for quantitative imaging: each RNA molecule is detectable as a single diffraction-limited dot, so in theory you can "count" these in ImageJ to get an absolute measure of RNA expression. I've done this, but unsure how robust my pipeline is.

- depending on sample type and experimental design, it can work out cheaper than conventional in situs.

- more sensitive in my experience than conventional in situs

- epic confocal images!

Disadvantages:

- potentially more expensive than traditional in situs

- imaging can be tricky, and is time consuming. Four-channel HCR is a pain in the backside (but I have it working).

- not very well established, so if you were trying to use it for a clinical application you may run into issues. Less support online.

I think the key disadvantage (compared to traditional in situs) is cost: probes are 350USD unless you have an academic discount. And the hairpins are expensive, so each replicate is a bit costly (depending on your tissue: my bee ovaries need 10uL of hairpin, but my wasp embryos only need 2uL because they're so much smaller). But if you compare that to the cost of ordering or constructing a clone in-house, it starts to look much cheaper. One thing to note is I don't pay for imaging time, which brings costs down!

Hi Shannon Elisa Taylor, glad to hear someone uses it successfully. I've actually since this post purchased the kit, both a positive control and the experimental, but unfortunately I can't seem to get any signal at all, even with the actin probe.

Do you deviate much from the example protocols they provide? Also, have you noticed any specific sensitivity to RNAse other than normal in-situ/rna work would necessitate?

Thanks!

Hi Brian Lin , sorry to hear it didn't work for you.

I've not noticed any RNase sensitivity- we actually don't routinely use RNase free treatments for in situs or HCR, and they seem to work fine. But I could totally be missing out on big signal increases :)

I don't deviate from published protocols much (but see my methods below- for insect tissues in suspension). I use the honeybee in situ protocol for sample preparation from Dearden et al. 2009 (here http://biochemwebserve.otago.ac.nz/DeardenLab/Protocols2.html)

Minor deviations from the protocol:

- the wash steps on the second day (after hybridization): I do the wash steps in the morning, then leave it in SSCT until it's time to do the amplification (4-4.30ish). (Instead of starting the amplification step at around 11am!) No clue if this is important.

- I generally do hybridization/amplification steps around an ordinary 9-5 workday- ie put probe/hairpin on at 5ish, take it out at roughly 9am. It seems to work fine if I leave it longer too.

Key pain points I've noticed are:

- snap cooling the hairpins: It's critically important the hairpins are heated individually, and not in amplification buffer. I use a PCR machine and strip tubes for this, heaps easier.

- imaging: HCR seems to require long scan times (20-40us/pixel, under our FV1000 confocal). Also note that the types of expression pattern you're looking for are quite different to immunos. If you get high enough resolution (or if your RNA molecules are far apart) you see HCR signal as single dots (corresponding to 1 RNA molecule). I routinely use our silicon oil lens, instead of a standard non-oil immersion one, as the silicon oil lens detects a lot more light, so it is easier to detect signal. I've got some files recording my microscope settings, if you want to see them. I've also found the Alexa488 and 546 fluorophores are more reliable than others, but that is also partly due to quirks of our confocal.

- there is some weirdness with hybridization temperatures I've been unable to work out. HCR only works reliably for me in one incubator, which is set to 38C instead of 37C. I think this is part superstition, and in part suggests that hybridization temperature is important for HCR. We've also had HCR not work well in a heat block, so if a full sized incubator (or water bath?) is available, try to use that. (we use our insect incubators for in situs etc).

- my Nasonia embryos bleach incredibly quickly in glycerol, so I'd suggest imaging quickly (same day, ideally) or using an antifade mountant (prolong Gold is recommended).

That's all I can think of at the moment. Have a look at the new molecular technologies protocols- they've been updated to include some other useful hints. I'm also starting work in a new lab tomorrow, who use a lot of HCR, so will ask if they have any hot tips. :)

The full methods I use are below:

Where no units are given, assume units are uL. The methods (plus reagent recipies etc) are also available at https://ourarchive.otago.ac.nz/handle/10523/8611

Tissue was prepared for hybridization as for conventional honeybee in situ hybridization (Dearden et al. 2009). Hybridization was performed as per generic samples (Choi et al. 2016). Reagents are given in Table x. In-depth protocols are as follows. Stored ovaries were rehydrated in successive five minute 75%, 50%, and 25% methanol washes. Ovaries were then washed 3x five minutes in PTw. Ovarioles were separated under a dissecting microscope in PTw, then were digested in 1 mL PTw and 1 20 mg/g Proteinase K for 20 minutes, washed once with PTw, and refixed in 4% formaldehyde for 15 minutes. Fixative was removed by washing 6x with PTw.

Ovaries were then pre-hybridized in 500 of probe hybridization buffer for 30 minutes at 37C. The probe solution, consisting of 1 of each pair of probes in 500 of probe wash buffer, prepared at 37C, was added to the ovaries. The probe solution and ovaries were incubated overnight at 37C. Excess probes were removed by washing four times, for 15 minutes each, in probe wash buffer. Samples were then washed three times, for 5 minutes each, with 5X SSCT, then left at room temperature until the end of the day. Samples were incubated in 500 of amplification buffer for 30 minutes at room temperature. The hairpin mixture was prepared by incubating 10 of each 3M hairpin stock at 95C for 30 seconds, then left at room temperature for 30 minutes. To prevent bleaching of the hairpins, all subsequent steps were performed in the dark. 500 of amplification buffer was then added to the hairpins. The amplification buffer in the ovaries was replaced with the hairpin mixture and incubated overnight at room temperature. The hairpin solution was then removed and stored in the freezer, and the ovaries washed with 5X SSCT, in the following sequence: 2x 5 minutes, 2x 30 minutes, 1x 5 minutes. Ovaries were then counterstained with 1 DAPI in 1 mL 5x SSCT for 30 minutes, before being washed once more and stored in 70% glycerol. Ovaries were mounted on slides and imaged under the Leica FV1000 confocal, using the UPLSAPO30X 30X silicon oil lens.

Shannon Elisa Taylor , thanks for your answer!

I did a double take when you mentioned your acquisition settings, because when you say you did the incubation with amplification probe overnight, that should be "relative mRNA, qHCR" mode instead of "dHCR", so that signal is much larger than single molecule resolution?

I guess what I'm getting at is that I would have expected the signal to be viewable at like, 10x or 20x by eye? Can you see your HCR signal that way? Or must you use the confocal and do high mag scans?

I noticed some differences in the protocol you used compared to mine, and the next time I have a chance to run it, I'll try it and see how it looks.

It's funny, I bought an experimental probe off the bat thinking it was going to be fine, given how simple the protocol seemed, that didn't work a couple of times, then I got a gapdh positive control and also couldn't see any signal. But maybe it's just because I'm not imaging at proper mag...which would be an issue because I wanted to use these to hand count cells :p

Thanks very much for your help!

Oh, using lower resolution (ie 10x) is totally fine. I just mostly use the 30x because my samples are so tiny (single wasp embryos that are like 0.3mm long...). I've also used epifluorescence once just fine, and my supervisor used epifluorescence to image his first attempt at HCR. (If you're doing multichannel HCR + epifluorescence watch out for spectral overlap between the Alexa546 and Alexa594 hairpins. It's a pain and I'm not sure how you'd solve it without a confocal.) Imaging with a lower magnification(/resolution) just means the HCR signal dots appear larger and so more likely to blend together.

I definitely see individual HCR foci, using overnight incubations, for most genes. Some are too highly expressed though. The silicon(30x) lens has an awesome diffraction limit, which helps. If you look at the sample images in Choi et al. (2016) you can also see the individual RNA molecules for some samples, so this isn't unusual. I don't really worry about it. If you're trying to quantify something, you can always score any cell with more than, say, 10 HCR foci as expressing the gene.

Ahh, gotcha. Alright, I'll have to give it another shot using your protocol, with modifications since I'm using a decently thick tissue.

Thanks very much for your help!

Dear Brain,

I am sorry to hear your first HCR experiments didn't work out. Would you mind letting me know what sample type(s) you are working with? I will follow up with an email to discuss the details of your experiments.

Best wishes,

Harry

Hi Harry Choi , thanks for the message!

I'm doing wholemount mouse trachea, fixed in 4% PFA, and cut so it opens up as a flat rectangle. For normal immunostaining, it works fine with antibody incubations overnight, imaging is fine on a standard confocal or epi (tissue is thin enough).

Much appreciated!

Brian

Dear Brian Lin and co,

I'm also trying the HCR on 1% PFA fixed mice embryo's 5um slides, but can't seem to get it working. Did you succeed in your HCR experiments after all? I'm interested in your approach.

Kind regards,

Boutaina

Dear Boutaina,

Thank you for your interest in HCR v3.0 technology! Please contact us by sending an email to [email protected] and our team will be happy to help get in situ HCR up and running in your lab. Did you purchase HCR reagents from us? I don't seem to be able to locate your account. Thanks!

Best wishes,

Harry

Hi Harry,

Thanks for the quick reaction! Yes my supervisor did purchase the HCR reagents from Molecular Instruments indeed. We did follow the protocol, but we left the probe solution and the hairpin solution on the slides for 24h instead of 12-16h as recommended, would this be a reason for the probe not to work? We also did an antibody staining after the FISH HCR, could this also kill the signal?

Thank you

Dear Boutaina,

The hybridization and amplification time are flexible so I don't believe that's an issue. I have several quick questions:

1. Did you perform a post-fixation after in situ HCR and before immunostaining?

2. Have you tried an experiment only with in situ HCR and without antibody staining?

3. Was the mouse embryo fixed with 4% PFA (after dissection) before sectioning?