We perform RNA ISH (mRNA or miRNA) on a high-throughput platform (a Tecan EVO liquid handling robot modified to handle slides). This is a semi-automated liquid handling system modified with a flow-through technology for slide handling and sophisticated temperature control providing automation of approximately 90% of all manual steps in the ISH protocol. All steps of ISH – from pre-hybridization to additions of probes, hybridization, stringency washes and antibody-mediated chromogenic or fluorescent reaction – are carried out by the robotic system. This gives high-quality and reproducible results. We have a 196-slide capacity per run, but most often do 40-96 slides per experiment.
We are using a dual signal amplification system including digoxigenin (DIG)- or fluorescein (FITC)-labeled probes, anti-DIG or anti-FITC horseradish peroxidase labeled antibodies and tyramide amplification. This method achieves the sensitivity of radioactivity with added cellular resolution and it can be viewed with brightfield or fluorescence microscopy.
Using a single probe [view schematic for single labeling], you can choose to develop the signal with a chromogenic method (alkaline phosphatase with BCIP/NBT, a purple precipitate) or fluorescence (Cy3, FITC or Cy5).
If you want to look at two genes at the same time [view schematic for dual labeling], we do dual ISH and develop with fluorescent systems (Cy3/FITC, Cy3/Cy5 or FITC/Cy5).
We can look at three genes at the same time, using DNP-labeled probes and suitable development systems, but find that this works less robustly than the two-gene protocol we use routinely – contact the core if you need this to discuss the feasibility for your project.
We can do ISH in combination with immunofluorescent antibody staining, but it depends on your antibody. The ISH technique uses proteinase K digestion which is detrimental to many antigens. In our experience antibodies recognizing structural proteins (for example glial fibrillary acidic protein or neurofilament H) or nuclear proteins (for example BrdU) work ok. We recommend trying your antibody on a few slides before committing to a larger number.
FAQs: In Situ Hybridization
The core makes and uses hapten (digoxygenin, fluorescein or DNP) -labeled mRNA probes. We make them by in vitro transcription from DNA templates provided by the customer or purchased from a company.
Design: for many genes you can find published probe sequences in the literature or in databases with ISH information such as Allen Brain Atlas, GenePaint or GUDMAP. Many of these are great – just blast the sequence to make sure that it fits your purpose (some of the data is 15+ years old by now).
You can also design a new template for your gene of interest. If designing from scratch, be aware that it needs to be 250-900bp long. Talk to the core director if you need advice.
Generation: the easiest way to make the template is by PCR or cloning. For PCR instructions view our DNA template generation by PCR or email the core director at atc-rnacore@bcm.edu to discuss. Make sure to include T3 and T7 RNA polymerase binding sites at the flanking ends of your PCR product.
A double stranded DNA template can also be purchased either from companies that sell DNA fragments (approx. $100-250 per fragment depending on size and amount), or from the core ($150-200). If purchasing from a company, make sure you include the T3 and T7 RNA polymerase binding sites properly. Email the core director at atc-rnacore@bcm.edu if you are unsure how to do this.
Yes, we routinely do dual ISH (two probes at the same time) and can in special circumstances do three (the third gene will need to be expressed at reasonably high levels as the assay is less sensitive than for gene one and two).
Usually the ISH procedure removes the fluorescence of the reporter gene. However, the mRNA is still there so we can use ISH to detect the expression of the reporter gene. The core has probes for EGFP, TdTomato and a few other reporters.
You can do IF after ISH, but it depends on the antibody. The ISH protocol includes a Proteinase K digestion step which destroys a lot of protein epitopes. In our experience antibodies recognizing structural proteins (e.g. glial fibrillary acidic protein or neurofilament H) or nuclear proteins (e.g. BrdU) work well, antibodies recognizing cytoplasmic antigens normally don’t.
We can offer the RNAScope HiPlex assay (up to 12 genes). Special conditions and pricing apply, so contact us at atc-rnacore@bcm.edu if you are interested.
Yes, but miRNAs are very short and therefore require both an adapted (less stringent) protocol and special probes. You will need to purchase probes from Exiqon/Qiagen and the minimum number of slides charged per experiment is 8. Contact the core director at atc-rnacore@bcm.edu before purchasing the miRNA probes to discuss the feasibility of the project.