Prepare a solution containing 20-50 pg/pL of the appropriate probe, 50% deionized formamide, 2X SSC buffer, 10X Denhardt's solution, 0.1% sonicated ssDNA, and 0.1% SDS. The following is a convenient recipe for a total volume of 100 pL (see Note 20):



2 pL


Deionized formamide

50 pL



10 pL


50X Denhardt's solution

20 pL


10 mg/mL ssDNA

10 pL


10% SDS

1 pL



7 pL

1. Add 10 pL of hybridization mixture to each well and add cover slips.

3. Incubate slides at 48°C for 2 to 4 h in a humidified atmosphere.

3.7. Posthybridization for Peroxidase-Based Color Development

1. Wash slides in IX PBS twice for 5 min each time.

2. Add 10 pL of streptavidin-peroxidase complex (100 pg/mL in PBS, pH 7.2). Gently apply the cover slips.

4. Remove cover slip, wash slides with 1X PBS twice for 5 min each time.

5. Add to each well 100 pL of 3'-amino-9-ethylene carbazole in the presence of 0.03% hydrogen peroxide in 50 mM acetate buffer, pH 5.0.

6. Incubate slides at 37°C for 10 min to develop the color—this step should be conducted in the dark. After this period, observe slides under a microscope. If color is not strong, develop for another 10 min.

7. Rinse slides with tap water and allow to dry.

8. Add one drop of 50% glycerol in PBS and apply the cover slips.

9. Analyze with optical microscope—positive cells will be stained a brownish red.

3.8. Posthybridization for Alkaline Phosphatase-Based Color


1. After hybridization, remove cover slip, wash the slides with two soakings in 2X SSC at room temperature for 15 min.

2. Cover each well with 100 pL of blocking solution, place the slides flat in a humidified chamber at room temperature for 15 min.

3. Prepare a working conjugate solution by mixing 10 pL of streptavidin-alkaline phosphatase conjugate (40 pg/mL stock) with 90 pL of conjugate dilution buffer for each well.

4. Remove the blocking solution from each slide by touching a paper towel to the edge of the slide.

5. Cover each well with 100 pL of freshly prepared working conjugate solution and incubate in the humid chamber at room temperature for 15 min. Do not allow the tissue to dry out after adding the conjugate.

6. Wash slides by soaking in buffer A for 15 min at room temperature twice.

7. Wash slides once in alkaline substrate buffer at room temperature for 5 min.

8. Prewarm 50 mL of alkaline-substrate buffer to 37°C in a Coplin jar. Just before adding the slides, add 200 pL of NBT and 166 pL of BCIP. Mix well.

9. Incubate slides in the NBT/BCIP solution at 37°C until the desired level of signal is achieved (usually from 10 min to 2 h). Check the color development periodically by removing a slide from the NBT/BCIP solution. Be careful not to allow the tissue to dry out.

10. Stop the color development by rinsing the slides in several changes of deionized water. The tissue may now be counterstained.

3.9. Posthybridization for Digoxigenin-Labeled Probe

1. Use antidigoxigenin-peroxidase solution (1:250 dilution in PBS).

3. Wash three times with PBS.

4. Develop color with AEC, as described in Subheading 2., step 17.

3.10. Posthybridization for Fluoresent Color Probes

1. Wash the slides with PBS three times.

2. Mount the slides with 50% glycerol/PBS with antifade reagent.

3. Observe under UV microscope with appropriate filter range.

3.11. Counterstaining and Mounting

An appropriate counterstain can be used for the cells that are not carrying the amplified signals. This step is critical, and one has to work out the fine detail of the counterstains before actually applying on the slides that were used for the aforementioned protocols.

1. Rinse in several exchanges of a buffer.

2. Dehydrate the sections through graded ethanol series (50%, 70%, 90%, and 100% (v/v) for 1 min each).

3. For permanent mounting, a water-based mounting medium can be used (see Subheading 2, step 38).

4. Apply one drop of mounting medium per each 22 mm to cover slip area.

5. The slides may be viewed immediately, if you are careful not to disrupt the cover slip. The mounting medium will dry after sitting overnight at room temperature.

3.12. Validation and Controls

1. Run two or three sets of experiments in multiwelled slides simultaneously, for not only must one validate amplification, but one must also confirm the subsequent hybridization/detection steps as well.

2. In all amplification procedures, use one slide as a control for nonspecific binding of the probe. Here we hybridize the amplified cells with an unrelated probe.

3. We also use HLA-DQa, or P-actin probes and primers with human peripheral blood mononuclear cells as positive controls to check various parameters of our system.

4. If you are using tissue sections, a cell suspension lacking the gene of interest can be used as a control. Then retrieved following the amplification procedure.

5. The cell suspension can then be analyzed with the specific probe to see whether the signal from the tissue leaked out and entered the cells floating above.

6. In the case of RT in situ amplification, one can use P-actin, HLA-DQa, and other endogenous-abundant RNAs as the positive markers.

4. Notes

1. If one chooses to use special slides that have Teflon coatings that form individual "wells," then they can be purchased from one of the following sources. These slides are not necessary to purchase, and we no longer use them and prefer to use

Frame-Seal Incubation Chambers, instead. However, the source of the material for both kinds of chamber system is MJ Research, Inc., Waltham, MA ([email protected]).

2. Sterile technique must be practiced at all times during the culturing and prefixation parts of this procedure. Sterility is especially important in handling cell cultures, both to protect the investigator and to avoid introducing microbial contamination of the cell culture system. Such contamination is often the cause of test failure. The human peripheral blood used in this procedure may be infectious or hazardous to the investigator. Proper handling and decontamination and disposal of waste material must be emphasized.

3. If one uses two slides with two Teflon-coated sides facing one another rather than a slide and a cover slip, this allows for a double-thick reaction chamber that can hold a bit more amplification cocktail.

4. It is possible to use frozen sections for in situ amplification; however, the morphology of the tissue after the amplification process generally is not as good as with paraffin sections. Apparently, the cryogenic freezing of the tissue, combined with the lack of paraffin substrate during slicing, compromises the integrity of the tissue. Usually thicker slices must be made, and the tissue can "chatter" in the microtome. As any clinical pathologist will relate, definitive diagnoses are made from paraffin sections, and this rule-of-thumb seems to extend to the amplification procedure as well. The exception to the rule is when one wishes to use immunohistochemical techniques to detect additional signals in the cells. Some of these techniques require frozen sections and, in such a circumstance, the use of frozen sections is appropriate. However, there are new fixatives—such as "Permiofix" from Ortho Diagnostics—that preserve cell surface antigens in fixed tissue. These matters are explained further in a later section on immunohis-tochemistry.

5. Difficulties will often be experienced in slicing sections thinly because the tissue is either too cold or it is insufficiently frozen. This is remedied by use of pathologist's freezing spray—merely blast the central area with a few quick bursts of spray, wait a few moments and proceed. If instead the tissue will not slice at all, it could be that the tissue is too solidly frozen. To remedy this problem, allow the tissue to equilibrate overnight at -70°C while it is mounted on the disk in the cryostat.

6. Some RNA signals may not be very stable at high temperatures. Therefore, we use the shorter incubation times (5-10 s) with RNA targets and longer times (90-120 s) with DNA signals. One may need to experiment with different periods to find the best heat treatment for the specific tissue and target.

7. The time and temperature of incubation should be optimized carefully for each cell line or tissue-section type. With too little digestion, the cytoplasmic and nuclear membranes will not be sufficiently permeable to primers and enzyme, and amplification will be inconsistent (or nonexistent). With too much digestion, the membranes will lose integrity and leak amplicons, making surrounding cells falsely positive, or high background or poor morphology will result. Often with excessive digestion, many cells will show peri-cytoplasmic staining, which represents leaked signal contaminating cells in which no positive signal actually exists. Attention to detail with the proteinase K digestion often can mean the difference between success and failure in an experiment, and this digestion should be practiced on extra sections by anyone attempting to conduct this protocol for the first time. In our laboratory, proper digestion parameters vary considerably with tissue type. Typically, lymphocytes will require 5 to 10 min at 25°C or room temperature, central nervous system tissue will require 12 to 18 min at room temperature, and paraffin-fixed tissue will require between 15 and 30 min at room temperature (these times can be accelerated by using higher temperatures of incubation, up to 55°C). However, the periods can vary widely, and one has to optimize the conditions by carrying out careful reactions with control cells.

8. Avian myoblastosis virus reverse transcriptase (AMVRT) and Moloney murine leukemia virus reverse transcriptase (MuLVRT) give comparable results in our laboratory. Other RT enzymes also will probably work. However, it is important to read the manufacturer's descriptions of the RT enzyme and to make certain that the proper buffer solution is used (which may be different from that recommended here). One can use an oligo(dT) primer, random primer such as pd(N)6, or a specific primer that only anneals to the mRNA of interest.

9. The preservation of intact mRNA is of primary importance and the success of first-strand cDNA synthesis depends upon the integrity of the mRNA of interest.

10. All reagents for RT in situ PCR should be prepared with RNase-free water (i.e., DEPC-treated water). In addition, the silanized glass slides and all glassware should be RNase-free, which we insure by baking the glassware overnight in an oven at 250 to 300°C before use in the RT procedure.

11. Choices of primers include approx 2 |g of pd(N)6, 1 ng of pd(T)12_18, or 15 pMof mRNA-specific primers.

12. The RT-PCR cocktail must be preserved from degradation and should be kept on ice to minimize the formation of nonspecific first-strand products.

13. In all RT reactions, it is desirable to reverse-transcribe only relatively small fragments of mRNA (<1500 base pairs). Larger fragments may not completely reverse-transcribe because of the presence of secondary structures. Furthermore, the RT enzymes—AMVRT and MuLVRT, at least—are not very efficient in transcribing large mRNA fragments. For DNA amplification, we routinely amplify genes of 300 to 500 base pairs, and we find this size works well. Recently, many articles have been published with new combinations of polymerase enzyme and buffers that allow efficient amplification of much larger fragments of DNA or cDNA (as much as 50 kb or more). We see no reason to prevent these new techniques from being adapted to in situ amplification.

14. Regarding the design of primers, the following are several additional points one should keep in mind: use of a computer-aided design program such as: http:// www.bioinformatics.vg/biolinks/bioinformatics/PCR%2520and% 2520Primer%2520Design.shtml combined with the data resources offered by GENBANK (http://www.ncbi.nlm.nih.gov/entrez/), can often lead to superior primers. The length for both sense and antisense primers should be 18 to 22 base pairs. At the 3'-ends, primers should contain GC-type base pairs (e.g., GG, CC, GC, or CG) to facilitate complementary strand formation (a GC-type bond will in other words, the primer can grab on and anneal more readily). The preferred overall GC content of the primers is from 45 to 55%. Try to design primers so they not form intrastrand or interstrand base pairs. Furthermore, the 3'-ends should not be complementary to each other, or they will anneal to one another and form primer-dimers. One can design an RT-primer so that it does not contain secondary structures, and it is not complementary to RNA that will form secondary structures.

15. Other thermostable polymerase enzymes also have been used quite successfully.

16. Some RT enzyme work best at 72°C and or other temperatures, and one should check this before setting the cycles in the thermocycler.

17. Annealing temperatures for reverse transcription and for DNA amplification can be chosen according to the following formula:

Tm of the primers = 81.5°C + 16.6 (Log M) + 0.41 (G + C%) - 500/n Where n = length of primers, M = molarity of the salt in the buffer, usually 0.047 M for DNA reactions and 0.070 Mfor RT reactions.

If using AMVRT, the value will be lower according to the following formula:

Tm of the primers = 62.3°C + 0.41 (G + C%) - 500/n

Usually, primer annealing is optimal at 2°C greater than its Tm. However, this formula provides only an approximate temperature for annealing because base-stacking, near-neighbor effect, and buffering capacity may play a significant role for any particular primer. Optimization of the annealing temperature should be conducted first with solution-based PCRs. It is particularly important to know the optimal temperature before attempting to conduct in situ amplification, for several reasons. First, the highest-fidelity annealing occurs at relatively high temperatures for any primer, and maximum specificity goes hand-in-hand with high fidelity. If the annealing temperature is two low, spurious priming can occur, where the match between primer and template is not exactly homologous. Nonetheless, annealing occurs anyway, because DNA has an enormous affinity for being double stranded at lower temperature—even when there are base-pair mismatches. Then, these false primes get extended by polymerase enzymes, and spurious amplification products result. However, if the annealing temperature is too high, there is very little or no annealing of the primers to the template. Then, there is no place for the polymerase enzyme to grab on for extention, and very little or no amplification of any kind results. Second, annealing temperatures are important because in situ reactions, in general, are neither as robust nor as efficient as solution-based ones. We hypothesize this is because primers cannot easily access DNA templates inside cells and tissues because numerous membranes, folds, the tissue matrix, and other small structures can prevent primers from binding homologous sites as readily as they do in solution-based reactions. Thus, the temperature of annealing should be just right to make the best of a difficult situation. Last, but not least, many researchers will go to great efforts to develop protocols that include a "HOT start," which can help prevent false priming on the on the first cycle (3-6). Similar techniques have evolved using dimethyl sulfoxide, formamide, and anti- Taq antibodies (i.e., Clontech's TaqStart). However, it has been our experience that better results can be obtained by instead concentrating on the optimization of annealing temperature because this optimization can minimize false priming throughout the thermal cycling regime, not just on the first cycle. Thus, it is worthwhile taking the trouble to run reactions in solution-based reactions with extracted DNA/RNA to optimize the annealing temperatures. What one seeks is an annealing temperature that results in thick bands in electrophoresis gels without an abundance of nonspecific amplicons. However, bear in mind that one is looking for the highest temperature that will give these sorts of results—as little as 1°C can make a substantial difference. Alternatively, if one does not have the time to fully optimize the annealing temperature, one can use a "touchdown" protocol, in which the annealing temperature is initially set rather high, but it ratchets down by approx 0.5°C with each subsequent annealing step for the first 10 to 20 cycles. The idea here is to first create a number of high-fidelity amplicons that get geometrically amplified in the subsequent cycles—in other words, one increases the signal-to-noise ratio for better results, even though the final annealing temperature might be substantially below the optimum.

18. There is much debate as to whether a hot start helps to improve the specificity and sensitivity of amplification reactions. In our laboratory, we find the hot start adds no advantage in this regard; rather, it adds only technical difficulty to the practice of the in situ technique. However, if one prefers to use a hot start, we recommend trying a variation suggested by Stuart Isaacson. Dr. Isaacson—whose laboratory specializes in archival brain tissue amplified and probed for RNA virus—suggests the following:

"It is our experience that a hot start is helpful when the amplification reaction does not work at all, or when the efficiency of recovery is low. By keeping separate the template from the amplification reaction mixture at temperatures below annealing, less stringent 'competitive' binding of primers to undesired sequences is minimized. This can allow a higher yield of amplification product. Thus, a hot start is not necessary routinely, but it is a way to optimize amplification efficiency and consequent signal intensity."

19. Sometimes it is desirable to recover samples of the amplified products from in situ reactions. For example, perhaps a sample is needed to sequence amplicons so that various alleles of a gene—or even various expressions of a gene—can be distinguished. Perhaps an investigator wishes to clone the gene in question to obtain larger quantities for future study. But perhaps the most exciting application is in the field of developmental biology, for amplification products can help in determining what proteins are involved in the processes of cellular differentia tion and organogenesis. For example, sections of embryos can be taken at various stages of development, and either specific primers or oligo(dt) primers can be used to identify the gene expression that is occurring in particular index cells. If one reserves and stores the supernatant after amplification, then one go back and re-amplify the gene-of-interest once the specific phase of differentiation has been identified in the index cells following hybridization of the whole tissue. In this circumstance, one might even use a micromanipulator to recover the specific cells, in order to study gene expression even more closely. By simply reserving the amplification cocktail after the thermal cycling procedure, recovery of the amplicon usually can be achieved. It has been our experience that there is usually a small amount of leakage out from the cells into the amplification cocktail during amplification, but with proper proteinase K digestions, there is very little or no leakage into other cells in the tissue sample. We have collected abundant data on this matter, in particular by running samples of the supernatant in electro-phoretic gels after amplification. We have found that amplified products usually can be detected by Southern blotting but not by ethidium bromide labeling because the signal is very weak. However, this product usually can be reamplified in a subsequent solution-based reaction using the same primers as the in situ reaction (or a nested primer pair), and this provides sufficient quantities of the amplicon for subsequent cloning or analysis. We hypothesize that the leakage of signal in the original in situ reaction tends to occur late in thermal cycling, as the signal recovered from the supernatant is almost invariably quite weak (at least when the proteinase K digestion was properly optimized). In the latter cycles of the PCR procedure, the geometric nature of the amplification makes the concentration of the amplified product very high at the original locus of the target, and it is reasonable to assume that some small fraction of the amplified signal could drift away from this locus and diffuse out of the cell. This would be particularly true for those positive cells whose nuclear or cytoplasmic membranes were sliced open by the blade of the microtome, such that a primary containment "vessel" of the signal was violated. However, if this diffusion were occurring in significant quantity or early in the amplification process, we would expect to see strong signals in the supernatant—manifested both in the agarose gel and in inter-cyto-plasmic staining after hybridization—because subsequent amplification would have occurred in the supernatant. In fact, we do not see strong signals, except in those circumstances in which there was excessive digestion with proteinase K. Last but not least, we have encountered no evidence to indicate that leaked signals enter into other cells to deliver false-positives upon subsequent hybridization, as we believe this would require fairly large membrane pores. Rather, we hypothesize that during the optimized proteinase K digestion, there is only partial digestion of certain transmembrane proteins which results in the formation of semi-permeable pores. These pores selectively allow positively charged molecules to pass, but they tend to block negatively charged molecules. Therefore, the Taq enzyme and single-stranded small primers can enter the cell, but double-stranded amplicons (which are highly negatively charged) cannot readily pass. Therefore, false signals do not enter the cells. To recover the supernatant, one simply can pry open a corner of the plastic sealant and siphon off the solution. This solution can be used for analyses or re-amplification. Ideally, one wishes to recovery as much of the supernatant as possible, and that it would comprise 10 to 15 |L of cocktail.

20. The salmon sperm should be denatured at 94°C for 10 min before it is added to the hybridization buffer. 2% BSA can be added if one is observing nonspecific binding. One can add 10 ||L of 20% BSA solution and reduce the amount of water.


1. Bagasra, O. and Hansen, J. (1997) In Situ PCR Techniques. John Wiley and Son, New York, NY.

2. Bagasra, O., Hauptman, S. P., Lischner, H. W., Sachs, M., and Pomerantz, R. J. (1992) Detection of human immunodeficiency virus type 1 provirus in mononuclear cells by in situ polymerase chain reaction. N. Engl. J. Med. 326, 1385-1391.

3. Rishi, I., Baidouri, H., Abbasi, J. A., et al. (2003) Prostate cancer in African American men is associated with downregulation of zinc transporters. Appl. Immunohistochem. Mol. Morphol. 11, 253-260.

4. Patterson, B., Till, M., Otto, P., et al. (1993) Detection of HIV- I DNA and messenger RNA in individual cells by PCR-driven in situ hybridization and flow cytometry. Science 260, 976-979.

5 Embretson, J., Zupancic, M., Beneke, J., et al. (1993) Analysis of human immunodeficiency virus-infected tissues by amplification and in situ hybridization reveals latent and permissive infections at single-cell resolution. Proc. Natl. Acad. Sci. USA 90, 357-436.

6. Nuovo, G. J. (1994) PCR in Situ Hybridization Protocols and Applications (2nd ed). Raven Press, New York.

Research and Clinical Applications

Was this article helpful?

0 0

Post a comment