Omar Bagasra and Twaina Harris
Since the first publication on the method of in situ polymerase chain reaction (PCR), several thousand research papers have appeared in peer-reviewed journals describing various findings based solely on the application of this method or combined with other more robust methods, including solution-based PCR, immunohistochemistry, Southern blot, etc. A few years after the advent of PCR, several investigators developed in situ PCR methods that differed considerably from each other with regard to tissue preparations, fixation, mounting of slides, reverse transcription technique, primer design, target selection, size, and amplicon size, and thermocycler designs and the use, among many other fine details. This chapter describes the detail procedures that are used in the author's laboratory. It also discusses the variations and modification that can be used for the specific needs of an investigator. This protocol should serve as a primer for the investigators, and each researcher must use his or her variation according to their needs.
Key Words: In situ PCR; PCR; primer design; reverse transcription; target selection; tissue preparation; thermocycler.
The solution-based polymerase chain reaction (PCR) method for the amplification of defined gene sequences has proven to be a valuable tool not only for basic researchers but also for clinical scientists. By using even a minute amount of DNA or RNA and choosing a thermostable enzyme from a large variety of sources, one can enlarge the amount of the gene of interest, which can be analyzed, cloned, or sequenced in a very short amount of time. Thus, genes or segments of gene sequences present only in a small sample of cells or a small fraction of mixed cellular populations that can be examined (1).
The ability to identify individual cells expressing or carrying specific genes of interest in a latent form in a tissue section under the microscope provides a
From: Methods in Molecular Biology, vol. 334: PRINS and In Situ PCR Protocols, Second Ed. Edited by: F. Pellestor © Humana Press Inc., Totowa, NJ
great advantage in determining various aspects of normal, as opposed to pathological conditions. By using the proper primers for genes that are expressed by certain histological cell types, one can potentially determine the origin of metastatic tumors by performing reverse transcriptase (RT)-based in situ PCR (2).
Our laboratories have been using in situ PCR techniques for several years and we have developed precise, sensitive protocols for both RNA and DNA that have been proven to be reproducible in multiple double-blinded studies (312). One can use this method for the amplification of both DNA and RNA gene sequences. By the use of multiple labeled probes, one can detect various signals in a single cell (12,13). In addition, under special circumstances, one can perform immunohistochemistry, RNA, and DNA amplification at a single-cell level (the so-called "triple-labeling" [9,14,15]). This chapter presents a detailed protocol currently used in our laboratory.
Before one attempts to conduct in situ PCRs on tissues or cells, we strongly recommend that all investigators first optimize the PCRs in solution to get the primers, probes, and the incubation temperatures all working right before throwing the additional complication of cellular matrices into the mix. Investigators universally report that amplification reactions are more problematic in tissues and cells, and virtually no one gets in situ amplification correctly the first time they try if they are working alone. Unfortunately, troubleshooting these early reactions is usually very difficult because there are so many variables involved.
Solution-based reactions are relatively easy, and one can have success rather quickly—often on the first try. Then the reactions can be quickly optimized so that the results will be even better, and optimized parameters almost always transfer more successfully to the in situ protocols. Solution-based reactions also can serve as controls for in situ PCR.
However, before DNA or RNA can be amplified in a solution, it must be extracted from the cellular matrix in some manner. There are several well-established extraction procedures. Readers are advised to follow the precise extraction methods from any of the well-established molecular biology methods (i.e., Short Protocol in Molecular Biology or Molecular Cloning [16-19]). In addition, several manufacturers of molecular biology products sell extraction kits that can be used for the specific purposes for which they are intended.
PCR is now one of the most widely used procedures in almost every aspect of biological sciences. Readers who are still new to PCR should consult any of the numerous articles, animation and simulation available on almost all the search engines (i.e., Google, Yahoo, etc.). Our favorite is from the Cold Spring Harbor Laboratory (http://www.dnalc.org/).
2.1. Reverse Transcription: Making Complementary DNA From RNA
Of course, DNA serves as the primary "library" of genetic information for any organism, and it resides primarily in the nucleus of a cell (at least with eukaryotes). RNA serves as the working transcript of genetic information, and it is from RNA templates that proteins are eventually synthesized in the ribo-somes. RNA molecules tend to be much smaller than DNA molecules and they move freely about the cell, and having shorter lifetimes (RNA is sometimes described as being like a photo copy of an individual page of an organism's book of life, which can be discarded after they have done their job). RNA molecules almost always are single stranded and somewhat less stable physically and chemically than double-stranded DNA. Also, they are quickly degraded by ubiquitous enzymes called "RNases"—these continuously get rid of used RNA and recycle it in vivo but can also quickly chew up an investigator's target RNA, especially after a cell has died and before it is fixed.
In all life forms, whether prokaryotic or eukaryotic, the direction of transcription is almost always from DNA to RNA to protein—the permanent record to the working copy. Very rarely does transcription occur in the reverse direction. However, some RNA viruses—namely retroviruses—contain an enzyme called reverse transcriptase that can do just that, make DNA copies (called complementary [c]DNA) from RNA originals. The enzymes from various types of retroviruses have been isolated and cloned and are available in many different variations, and their activity can be exploited to make cDNA from RNA in vitro. This ability can be put to great advantage with in situ PCR because it allows an investigator to indirectly amplify RNA within cells so that it can be readily detected even when the target is in very low abundance. This amplification is achieved by first converting single-stranded RNA into double-stranded cDNA with the RT enzyme, then PCR is conducted on the cDNA copy to amplify the sought-after signal (so far, PCR works only with DNA, not with RNA). The result is that an investigator can determine whether a specific gene is actively expressed within a cell by determining whether RNA copies of the gene are present within the cellular structure (there is an especially elegant manner to amplify RNA in a single step, discussed separately in Subheading 4.5.).
All RT enzymes exhibit at least four specific enzymatic activities, namely reverse transcription, a RNA-dependent DNA polymerase activity, ribonuclease H (RNase H) activity and integration capability (we do not use the last one for our purpose). The reverse transcription maneuver allows the enzyme to use any RNA as a template to be copied, provided that the action is initiated by an annealed primer at the beginning of the target sequence. The copy being made is always DNA (cDNA). The ribonuclease H activity serves to peel away the RNA template from the newly created RNA-DNA hybrid, and then the enzyme cuts down the original RNA molecule into tiny bits. The single-stranded cDNA
molecule that remains is manipulated by the enzyme once again; this time, another domain of RT enzymes acts like DNA polymerase and weaves a complementary strand along single-stranded cDNA to make a double stranded cDNA molecule.
The different versions of the RT enzyme each have slightly different characteristics as the result of the different origins of the enzymes. Each was derived from a specific retrovirus that has evolved a particular capsid protein to provide a structural milieu within which the enzyme can do its work. However, in the artificial environment of an in vitro experiment, there are no capsid proteins available, and hence no optimized milieu for the enzyme. Therefore, the investigator must take special care in using RT enzymes; otherwise, various untoward biochemical characteristics of the enzyme can be summoned forth.
In particular, RT enzymes in vitro have a tendency to give up on RT too early and start chopping up the RNA template before transcription is complete. This destroys the RNA signal before a complete cDNA copy is synthesized. An investigator minimizes this effect by optimizing reagents, particularly manganese ion and a specific detergent that substitutes for some of the characteristics of the capsid proteins. Generally, vendors of RT enzymes supply an optimized buffer solution containing proper concentrations of the manganese ion and detergent (be certain to always use the optimized buffer with each RT enzyme).
Last but not least, there are several newer RT enzymes available with special characteristics. One, called rTth, is available from Perkin Elmer and can serve not only as the RT enzyme but also as the thermostable DNA polymerase for PCR, allowing RT and PCR reactions to be conducted in the same buffer in one thermal cycling regime. Few other RTs with similar characteristics have been isolated from prokaryotes, which survive at very high temperatures. Many RTs have other qualities, including proofreading ability, longer life-span, and the ability to reverse transcribe longer messenger (m)RNAs. Investigators ought to search for the right enzymes, which is best suited for their tasks. One way to look for various reagents is to log onto http://www.sciquest.com to search for various reagents.
Currently, we are using a ready to go kit from Amersham Biosciences. There are several kits available in the market and a researcher can use their own judgments in utilizing one of these kits that is suitable for their particular project.
One of the most important keys to performing PCR successfully is designing a proper primer pair and then optimizing the annealing temperature so that amplification can proceed smoothly. Fortunately, numerous resources and computerized tools are available to assist in these matters, and an investigator is well advised to pay particular attention to these details early in the development of an experimental protocol, for it will save much agony later on.
Primers are synthetic oligonucleotides, typically between 18 and 22 bases in length. Some might consider an 18-mer primer a little short with too much chance that random annealing occurring. Others would argue that a 22-mer primer is extravagantly long. It is our experience that primers in this range work effectively.
Among the four primary nucleotides (i.e., G, A, T, and C) in DNA and primers, two are "stickier"—G and C—than the others are, as they form six hydrogen bonds between them (A and T only form four). This varying character can affect the performance of the primer substantially. Therefore, it is desirable to have G and C nucleotides at the 3' (downstream) end of the primer because this will facilitate annealing. However, one does not want a triple GGG or CCC at the 3'-end because this combination is too sticky, nor should one have AAA or TTT. Rather, the ideal sequence is to have two GC nucleotides followed by an AT type at the 3'-end, like GCT or GGA. The overall GC-content of the primer should be between 45 and 50%.
The two primers should be designed so that they have approximately the same annealing temperature. They also should be designed so that they do not form intrastrand or interstrand base pairs, which may result in hairpin formation. Single strands of DNA can twist and form loops in such a way that part of the primer can anneal to a target, with another part annealing to a spot on the target tens or hundreds of bases away from the first. One must be careful to select sequences that have negligible complementary regions, even in short fragments, in the larger segment of the DNA intended as the target. Primers that anneal to multiple regions of the target will neither identify the targeted sequence, nor will they extend the target properly.
Furthermore, primers should never be complementary to one another, particularly around their 3'-ends. If they complement in this region, they often anneal to each other and thus form "primer dimmers." These get extended by the polymerase into double-stranded DNA the length of two primers combined, minus the length of the complementary region. This is always undesirable, because it consumes the primers and greatly lowers the efficiency of amplification. Fortunately, there are computer programs available that can analyze various factors, as well as many other more subtle ones, thus being well worth the investment.
If one wishes to amplify RNA targets, the design of primers becomes somewhat more complex. Four strategies for RNA amplification are possible. The first three techniques begin with destroying the entire genomic DNA with an RNase-free DNase treatment to eliminate all DNA copies of a gene that could lead to false-positive results regarding an RNA target. The DNase is then inactivated, thus allowing one to convert all mRNA to cDNA with an RT reaction by using one of two types of primers.
The first type of primer is called an oligo d(T) primer. All mRNA molecules are single-stranded and have a poly (A) tail, meaning there is a long series of AAAAAAAAA... at 3'-end of the RNA molecule. An oligo d(T) primer is simply a long series of TTTTTTTs that will anneal to the poly(A) tail. When a reaction is performed with the reverse transcriptase enzyme (RT), the enzyme will extend the primer, making a complementary copy of the mRNA. The RT enzyme synthesizes DNA from the RNA template—substituting thymine for uracil. This reverse transcribed DNA is known as "cDNA." If the reverse transcription reaction is properly carried out, all the mRNA will be converted to the more stable cDNA, which is then available for amplification through conventional PCR techniques.
The second type of primer is called random primers. These are sets of very short oligos of random sequence, generally "hexamers," only six base pairs long. They anneal to complementary strands of mRNA, and the RT enzyme extends them in a manner similar to that described for oligo d(T) primers. Because any hexamer represents a common sequence as the result of its short length and because so many types of hexamers are included in a random primer set, essentially all of the mRNA converted to cDNA by this method (but not necessarily the early parts of every sequence).
Next, one can use a specific primer to reverse-transcribe only the gene-of-interest from the mRNA rather than all mRNA, as with the oligo d(T) and random primers. The cDNA copy can usually be created using the downstream (antisense) primer for the subsequent PCR reaction, resulting in unbounded transcription of the downstream target.
For the last type of RT reaction, please bear in mind that cDNA—which represent copies of mRNA—is fundamentally different from genomic DNA because it represents only the expressed sequences or exons of DNA. Therefore, the cDNA will be missing all of the introns and the controlling regions of DNA that are found in the genomic copies, which in fact compose 70 to 90% of most eukaryotic genomes.
An investigator can exploit this fundamental difference to design special, RNA-specific primers that spans introns in the genomic DNA, eliminating the need for the oligo d(T) primers and the wholesale conversion of mRNA to cDNA (or the alternative downstream specific-primers). Rather, one designs primers that will anneal only to targeted mRNA sequences by designing the primers to span introns in the genomic DNA. The primers will only then adhere to the mRNA templates and the cDNA copies of mRNA, and not to any genomic DNA copy of the same gene.
If one combines these special primers with a special polymerase enzyme that has both RT and DNA polymerase activity, such as the rTth enzyme described earlier, then one can amplify mRNA sequences directly without going through any specific RT step. This simplifies the whole PCR procedure by eliminating the need for a harsh DNase treatment, as well as a buffer change between the RT and polymerase enzyme steps. Better yet, this procedure allows for the amplification of multiple mRNAs or two types of nucleotide signals simultaneously—both the mRNAs and the genomic DNAs—because there is no need to destroy all the endogenous DNA: primers for each nucleotide-type can also be included without interfering with the activity of the other. However, one must know a considerable amount about the sequence of the gene in question to design these RNA-specific primers.
Recent publications have shown that the amplification of genes up to 50,000 base pairs is possible. However, this "long PCR" is not frequently used for in situ work because the primary purpose of the amplification in most circumstances is to detect specific genes, not clone them.
For most in situ PCR work, relatively short amplicons are used. Our laboratory has had great success with amplicons in the 150- to 500-bp range, and that is what we usually target. The amplicons should not be so small as to be prone to diffusion away from the original locus of the target, nor should they be so long as to lower the efficiency of the amplification in the difficult environment of a cellular matrix.
2.2.4. Sources for Sequence Data and Computerized Design of Primers
There are several useful sources. First is the scientific literature, particularly if one's project follows earlier research on a similar matter. But be aware: errors in the transcription of tedious DNA sequences seem to be common in the published literature.
A much more useful and up-to-date source for sequence information is GenBank, an extensive, freely available database operated out of the Los Alamos National Laboratory in New Mexico, and accessible on the World Wide Web through the National Library of Medicine. GenBank has sequence data available for a wide variety of genes from various species, though its collection is most extensive with human, primate, and rodent species. The GenBank can be contacted through the Internet or via modem, with sequences downloaded digitally with little or no transcription error (http://www. ncbi.nlm.nih.gov/).
These data are most useful if used in combination with software that is specially designed to process these data and select primer sequences, after various desired characteristics for the primers (or hybridization probes) have been input. One can get free Internet programs to design primers. Also, numerous PCR primer pairs are available as stock items from commercial biotechnology com panies. More information on primers, and other products can be obtained via the Internet (i.e., http://www.sciquest.com)
3. Preparation of Tissues
To use peripheral blood leukocytes, first isolate cells on a Ficoll-Hypaque density gradient. Tissue-culture cells or other single-cell suspensions also can be used.
There are several types of slides that are designed to support in situ PCR after they are attached on the glass slides. The cells are grown on these types of slides. If certain primary cell cultures require attachment factors or growth media, these could be used in conjunction with this cell culture system. After appropriate confluency (usually >60%), cells are gently washed with 1X phosphate-buffered solution (PBS), heat fixed, and then fixed with 2% paraformal-dehyde overnight.
Routinely fixed paraffin tissue sections can be used for amplification purpose quite successfully. This permits the evaluation of individual cells in the tissue for the presence of a specific RNA or DNA sequence. For this purpose, tissue sections are placed on routine histological slides. Tissue sections should be sliced to a 5- to 6-^m thickness. However, if one is using tissues that contain particularly large cells, such as ovarian follicles, then thicker sections may be appropriate. Before in situ PCR, the slides need to be deparaffinized.
4. In Situ PCR (Basic Preparation, All Protocols)
For all sample types, the following steps comprise the basic preparatory work which must be done before any amplification-hybridization procedure. The overview is depicted in Fig. 1.
4.1. Creating Micro-Well for In Situ PCR
Before the actual amplification process can be initiated, we have to create a vapor-tight sealing chamber (MJ Research; www.mjr.com/ or www.biorad.com) which can withstand high temperatures required for PCR. Once an artificial well is created, one can place the PCR solution so the putative genes of interest can be amplified in situ. For this purpose, we can use "Frame-Seal Incubation Chambers." These frames have double-sided adhesive surfaces: one on the bottom that sticks to the surface of slides, and one on the top that can be sealed by a plastic cover after it has been placed in the correct amount of PCR solution on
the tissue surface. To create an artificial well, first the adhesive frame is attached to the slide, enclosing the specimen area. Next, the reaction cocktail is added and the well is sealed with the flexible plastic cover slip. This vapor-tight chamber can withstand temperatures up to 99°C. After the completion of the amplification process, the entire artificially created well can be removed by simply pilling-off the adhesive-frame. In addition, if one requires growing cells or small fragments of tissues inside the artificially created well, then the entire adhesive frame and the slide can be sterilized either by ultraviolet treatment or autoclaving the whole slide with frame before seeding the cell cultures (see Fig. 2 for details).
Place the slides with the adhered tissues or cells on a heat-block at 105°C for 0.5 to 2 min to stabilize the cells or tissue on the glass surface of the slide. This step is absolutely critical. One may need to experiment with different periods in order to optimize the heat treatment for specific tissues. Our laboratory routinely uses 90 s for DNA target sequences, and approx 30 s for the RNA sequences (for RT in situ PCR).
4.3. Fixation and Washes
Place the slides in a solution of 2% paraformaldehyde in PBS (pH 7.4) for 2 h at room temperature. Then, wash the slides once with 3X PBS and 1X PBS for 10 min, agitating periodically with an up and down motion. Repeat once with fresh 1X PBS. At this point, slides with adhered tissues can be stored at -80°C until use. If biotinylated probes or peroxidase-based color developments are to be used, the samples should further be treated with a 3% solution of hydrogen peroxide in PBS in order to inactivate any endogenous peroxidase activity. Incubate the slides or store at room temperature for 10 to 20 min. If other probes are to be used, proceed directly to the proteinase K digestion, which is perhaps the most critical step in the protocol.
This step is the most critical step in the protocol. Treat samples with 6 pg/mL proteinase K in PBS for 5 to 10 min at room temperature (see ref. 20 for a detailed protocol).
4.5. RT Variation: In Situ RNA Amplification
One has two choices in order to detect an RNA signal. The first and more elegant method is to simply use primer pairs that flank spliced sequences of mRNA, as these particular sequences will be found only in RNA and be split into sections in the DNA (see Fig. 3). Thus, by using these RNA-specific primers, one can skip the following DNase step and proceed directly to reverse transcription. The second, more brutal, yet often necessary approach is to treat the cells or tissue with a DNase solution subsequent to the proteinase K digestion. This step destroys the entire endogenous DNA in the cells so that only RNA survives to provide the signals for amplification.
All reagents for RT in situ amplification should be prepared with RNase-free water (i.e., diethylpyrocarbonate [DEPC]-treated water). In addition, the silanized glass slides and all glassware should be RNase-free, which we insure by autoclaving the glassware in a small autoclave before use in the RT-amplifi-cation procedure.
One has to take a great care in performing the RT reaction in situ, and there are many caveats that one needs to know before jumping into the action of RT-in situ PCR. Usually, a researcher is well-trained in RNA isolation and analysis methods by someone or by technical manuals, which are so abundant. Experimental procedures often are not questioned and quickly become an absolute truth. Furthermore, it is difficult to sort out the "correct" procedure from the literature to document the "facts" from the company line and kits that all claim to be the best in the market. Therefore, we would like to mention certain ground rules that might help the reader in their quest of preserving the RNAs in the cells. One should use DEPC treatment to make solutions RNase-free. However, DEPC may not be the only choice and may not be the only
Original DNA Copy Of a Spiked Gene
Primary RNA Transcript
111111 f frrrrrr
End Modi (It ¡it toil s (polyadenylatioa & cupping) & Primer Design ♦ ♦
Future imicalins site of an I 8-mer prtmc tprimer cannot hthridfec lo limplknl lnnKri|il)
m m ihr y side of primer should he designed lo anneal «¡lh nnly J - S l»:m"> nTthr upstream mon
Was this article helpful?