|Hybridisation to RNA in Tissue Sections |
For a good account of the theoretical background and scope of in situ hybridisation refer to the pink Course Manual for the University of Leicester In Situ Hybridisation Course, by Paul Senior et al. (September, 1989).
32P is cheap, available every week (HFI) and gives good signal rapidly but with only fair resolution at the tissue level. Losses and contamination can be easily monitored during the transcription/hybridisation process. A good nucleotide to start with.
35S and 33P give better resolution but take twice as long to give a result on autoradiography. 33P apparently gives lower background than 35S (Bev. Faulkner-Jones, WEHI). I have never tried 33P but I have found relatively high backgrounds with 35S for my particular probes.
In the early stages of optimising this technique, it is extremely useful to have tissue that will act as a positive control, ie. a known site of target mRNA where you can rely on obtaining signal to check that the experiment has worked in a technical sense. I include positive controls with every experiment to gauge variation between experiments.
The DNA which will act as a template must be subcloned into a vector containing promoters for RNA polymerase, eg. Bluescript has T3 and T7 promoters. Ideally, the insert DNA should be cloned into Bluescript in both orientations so you can use the same RNA polymerase to generate both the sense and anti-sense strands (and eliminate variability between the T7 and T3 RNA polymerases - one way to try to match sense against anti-sense).
The template must be linearised when undertaking in vitro transcription. First, determine the orientation of your template and which strand will be sense and which anti-sense. This diagram shows the multi-cloning site region of Bluescript II KS:
If a DNA insert designed to be the template is cloned into Bluescript as shown, then transcription with the T3 RNA polymerase will generate sense mRNA. However, to prevent transcription of Bluescript sequences the template needs to be linearised in this case with BamHI or XbaI. Similarly, if the template is linearised with HindIII and transcribed with T7 RNA polymerase then anti-sense mRNA will be generated - this will give the positive result on tissue section. It is important to avoid linearising the template with restriction enzymes which generate a 3' overhang. Always try to use an enzyme that leaves a 5' protruding end. If there is no alternative then you can use the 3'-->5' exonuclease activity of Klenow DNA polymerase to convert the 3' overhang to a blunt end (see page 58 of the Promega Protocols and Applications Guide).
To linearise the template:
Decide how much cold UTP to use in the reaction. The concentration of any nucleotide must be at least 1.5 mM (and ideally much higher than this). UTP will be the limiting nucleotide because only a certain concentration can be achieved when the nucleotide is radiolabelled. With 32P, if 10ul (100 mCi) of a 20ul reaction is hot UTP32, then the final concentration will be 1.67mM. I usually add an equal concentration of cold UTP (to give a final concentration of 3.33 mM) when transcribing a 750bp probe. Full-length transcripts are what is desirable and these are more likely to be produced at higher nucleotide concentrations. When transcribing from a longer template you could try 3:1 cold:hot UTP, ie. add cold UTP to a final concentration of 5mM. The down side of adding cold UTP is that the specific activity of the resultant probe decreases.
I usually make up a cocktail for the total number of probes I am transcribing then aliquot this and add the appropriately lined arised template and RNA polymerase to each one.
Promega 5 x Transcription buffer is:
It is recommended not to chill this buffer on ice after adding the template DNA because the spermidine can precipitate the DNA.]. The in vitro transcription reaction is incubated at 37ºC for about 40 minutes, then a further 1ul of RNA polymerase is added and the reaction incubated for a further 35 minutes.
The next step is to digest the DNA template with DNase.
To each tube, add:
(NOTE: If using 35S UTP it is important to add 1ul of 1M DTT at this stage and to ensure that all aqueous solutions that the probe is dissolved in from this point onwards contain DTT to a final concentration of 10 mM -- some people say 100 mM).
Add 75ul of sterile water to make the volume up to 100ul. At this point I remove 1ul and spot it onto DE81 paper for Cerenkov counting (to enable calculation of probe yield later) and I aliquot 1ul into an ependorf tube for running on a polyacrylamide gel to assess the quality of transcription. This piece of paper is designated the "initial" count.
Next the probe needs to be phenol/chloroform extracted and precipitated:
Hydrolysis of the probe is desirable if the probe length is greater than 300 to 400 bp. Alkaline hydrolysis is carried out to reduce the average probe length to about 150 to 200 bp to improve access of the probe to target mRNA in the tissue section (see page 45 of the Leicester manual). The formula for calculating hydrolysis time (T) is T = Lo -Lf / 0.11.Lo.Lf where Lo is the original length in kb and Lf is the desired length in kb.
To the 100ul of probe from the section above add 100ul of hydrolysis buffer and incubate at 60 ºC for the appropriate time (for a 750 bp probe this is approximately 48 minutes). After that time has elapsed add 200ul of stop buffer.
The probe is then precipitated by adding:
and left on dry ice as above before spinning for 15 minutes at 13,000g.
The probe is resuspended in 20ul of sterile water.
I remove 1ul from my final 20ul reaction volume and add 4ul of sterile water to this. Of this 5ul, I spot 1ul onto a piece of DE81 paper and keep the rest for loading onto a polyacrylamide gel. This piece of paper will give the "final" count.
With 32P, I count my papers in the tritium channel of the scintillation counter.
The yield of each probe is calculated using the following formula:
ie. for UTP32 in the above protocol using 1:1 cold:hot UTP:
yield (in ng) = "final" count/"initial" count x 100 x 500 x 4.13 x 1/3000 x 2
= "final" count/"initial" count x 137.7
Therefore the maximum theoretic yield is 137.7 ng of probe.
I cast a 4% polyacrylamide gel and run the pre- and post-hydrolysis probes on it to assess the quality of the reaction product and to check that all has gone well with the hydrolysis step. The most important aspect is the amount of full-length transcript present.
I cast a narrow, long gel and use the shark's tooth combs.
For 50ml of 4% acrylamide solution (with running buffer of 1% TBE):
Make up to 50ml with 7 M urea solution.
Add 120ul of 25% ammonium persulphate and 45ul of TEMED and let set for 45 to 60 minutes.
To the 1ul aliquots of probe, add 4ul of sequencing stop buffer (containing xylene cyanol and bromophenol blue). Heat the tubes to 80ºC for 2 minutes and then load 2ul per well. I leave an empty well between samples.
Run the gel till the bromophenol blue is about to run off. The gel will be quite radioactive so take due care. It can be placed against XAR-5 film as is, sandwiched between glad wrap, or dried on the gel drier and then placed against film (the latter is preferable). An exposure of 1 hour is usually adequate.
The following protocol uses paraffin sections as opposed to frozen ones. In general, paraffin sections result in better morphology, allow the collection of serial sections and produce far better results for "watery" tissues like embryos. I have not observed any diminution in signal when comparing results obtained with paraffin as opposed to frozen sections.
The tissue is collected and placed immediately in 4% paraformaldehyde (PFA) dissolved in PBS (phosphate buffered saline). It is generally left immersed in the fixative overnight. The pieces of tissue should not be too large - about 5 mm x 5 mm x 5 mm is large enough. It is important for the fixative to gain access to any compartments or spaces within the tissue as easily as possible. For example, any embryos older than E14.5 should have their head and body cut apart with a razor blade and their peritoneal cavity opened. The amniotic sac should be opened for any embryo older than E9
5.Larger pieces of tissue should be transected to allow ready exposure to fixative. This can be achieved with the least disruption when the tissue has already been immersed in 4% PFA for several hours - the tissue goes quite hard allowing easier dissection.
After 12 to 16 hours of fixation, the tissue is transferred to a 0.5M solution of sucrose dissolved in PBS. It is stored at 4ºC until processing through to paraffin. As a general rule this should be done as soon as is possible, although I have successfully used blocks of tissues stored at 4ºC for several weeks.
Processing through to paraffin consists of taking the tissue through a series of graded alcohols to a clearing solution and then paraffin. It is performed overnight on an automatic processor located in the Histology section of the Department of Anatomy (phone 344-5752 for bookings, Helen Makin is in charge). Either short (1 hour per step) or long (2 hours per step) cycles are possible. Larger pieces of tissue require the longer infiltration time. After processing, I generally embed my own tissue into blocks of paraffin. In this way I can orientate the tissue according to how I want to section it. This is important because orientation can be crucial in navigating one's way through a tissue block on the microtome.
Tissue sections are cut on a microtome in the H.F.I. histology facility. I generally cut 5 to 10 mm sections. Thicker sections will improve signal per section but possibly at the expense of resolution. A dark field microscope on hand next to the microtome can be essential in checking one's progress through the block. Sections are mounted on glass slides coated in aminoalkylsilane, an agent which assists adhesion of the section to the slide (see below for protocol of how to "sub" slides). To mount each section, first delicately transfer the section (using two paint brushes) from the cutting stage of the microtome to float it in a container of distilled water mixed with a few mls of absolute ethanol. Then use a plain glass slide to transfer the section to a water bath maintained at about 50ºC. The traces of alcohol on the section cause a surface tension effect which immediately removes all wrinkles and after 10 seconds or so the section is ready to be manoeuvred into position onto a "subbed" slide and then dried on a heating rack overnight at about 45ºC. Once thoroughly dried, wax sections can be stored in a safe location at room temperature and used successfully for hybridisation histochemistry years later.
Protocol for coating slides in aminoalkylsilane:
Prior to hybridisation paraffin slides need to be pre-treated with a gentle protease digestion to improve access of the riboprobe to the target mRNA. The protease used is pronase E (Sigma P5147). Pronase E is autodigested and aliquotted prior to use. To do this, dissolve 40 mg/ml of pronase E in distilled water and allow it to autodigest at 37ºC for 4 hours to destroy contaminating nucleases. Aliquot the solution into sterile ependorf tubes at 125ul/tube and dry it down in the speedivac. Store at -20ºC. When using coplin jars, these contain ioslides in 40ml of appropriate solution.
First it is necessary to determine how many slides you will hybridise. This often depends on the yield of your probes. For practical purposes, 50 slides is the upper limit for one run. For the riboprobes that I have used, I find a probe concentration of 75 to 100 ng of probe per ml of hybridisation buffer is optimal. This appears to work for other people's probes as well.
The amount of hybridisation buffer used per section depends on the size of each section but I generally use an average of 50ul per section. Therefore if you have a yield of 75ng of probe you will have enough to hybridise 20 sections with that probe.
The basic recipe for 1 ml of hybridisation buffer is:
My approach is to calculate the total amount of hyb buffer required (for all probes, including "losses"), make this up excluding the water, aliquot this out and then add the appropriate amount of water (and probe) for each probe separately. This is done as follows:
Steps 4 to 6 can be complicated so let's consider an example:
Assume we are hybridising 30 slides with 2 probes (15 slides each). For the anti-sense probe, the yield was, say, 65 ng and this is dissolved in 20ul of sterile water (as per section II.D above, after probe hydrolysis and precipitation).
From step 4, for the anti-sense probe:
Step 5: 626ul of hyb buffer (minus water and probe) is aliquotted out for the anti-sense probe.
Step 6: 106.7ul of water is added to this and then 17.3ul of anti-sense probe is added. The tube is heated at 80ºC, vortexed, then pulse spun in the microfuge before applying about 45ul to each section (to cover losses) and placing coverslips completely over the hyb buffer and section.
Comment on temperature of hybridisation:
It is said that the hybridisation of riboprobes to mRNA in tissue sections cannot be adequately predicted by the Tm equations used for nucleic acids bound to solid supports. Thus the optimal temperature of hybridisation needs to be determined empirically. For my probes, and also for the PTHrP probe, the optimal temperature seems to be 50ºC. However, for an IGF II riboprobe used by Felix Beck's group, 60ºC appears to be optimum. The other aspect to this is that the temperature of the post-hybridisation washes appears to be far less significant in terms of the signal achieved, compared to the temperature of hybridisation.
I follow the protocol of post-hybridisation washes as detailed in the pink Leceister in situ manual. In brief, this is as follows:
The aim of the formamide washing step is to remove all excess probe and particularly the viscous dextran sulphate from the sections. Three washes (with agitation) at 50 to 55ºC, each
of 45 to 60 minutes duration, are performed in formamide washing buffer. Each wash uses 500 ml of buffer in a plastic tray which can fit 2 slide racks each containing 25 slides. One litre of formamide washing buffer consists of:
The cover-slips are first dislodged by sequentially immersing each slide in a beaker containing approximately 200 ml of this buffer. (This is the same buffer is as that placed on the bottom of the hybridisation chamber to provide humidification during hybridisation (see above). Do not use water or 2 x SSC for this purpose - it must be a solution containing 50% formamide and the appropriate concentration of salts).
I make up 3 litres of this buffer and use 500 ml for the subsequent RNase A digestion step. The remaining 2.5 l is used to thoroughly rinse the slides and their racks to remove all traces of formamide washing buffer. 3 litres contains:
75 mg of RNase A (Sigma R5503) is dissolved in 1 ml of RNase buffer and boiled for 2 minutes before adding it to the 500 ml of pre-warmed buffer (final RNase concentration = 150ug/ml). The slides are incubated with agitation in this solution for 60 minutes at 37ºC. It is important that the slide racks and tr