- Email: [email protected]
Polymerase chain reaction methods in immunology
Polymerase Chain Reaction Methods in Immunology M.J. Dallman
T
HE POLYMERASE chain reaction (PCR) has provided one of the biggest technologic advances in molecular biology of the last decade.1 Its application to the analysis of both DNA and RNA has allowed the dissection of samples either too small or in which the target sequence is present at such low concentration for previous techniques to be of use. PCR has been used extensively in transplantation as a method for both genotyping and RNA phenotyping. Discussion of PCR in this paper is confined to its use in the latter application.
REVERSE TRANSCRIPTASE-PCR: QUANTITATION OF TRANSCRIPT LEVELS
As originally described, PCR is a completely nonquantitative technique. Samples are amplified over a fixed number of cycles of denaturation, primer annealing, and extension. While this approach is perfectly satisfactory for genotyping cells, in situations in which the level of specific transcripts within different RNA pools is to be determined by reverse transcriptase-PCR (RT-PCR) RT-PCR, it is not. The reasons for this are illustrated in Fig 1. As may be observed, the exponential phase of detectable product accumulation is preceded by a phase in which product is undetectable (A) and succeeded by a phase in which product level reaches a plateau (B). Only if one measures product during the midexponential phase of amplification can a variance between two (or more) samples that contain different levels of target sequence be detected (C1, C2). Two main approaches have been described that
Fig 1. Accumulation of product in samples containing different levels of target sequence.
attempt to resolve this problem by measuring transcript levels in either a semiquantitative or quantitative fashion. These are described in detail elsewhere,2–5 but will be illustrated briefly here. Semiquantitative RT-PCR2,3 relies on running a dilution series of either the RNA or cDNA used in the PCR or by sampling product from the PCR reaction at variable cycle number (eg, every 2 to 5 cycles). In the initial analysis of samples we use control primers that amplify transcripts whose level should not vary in different RNA pools (eg, beta actin, HPRT). In this instance, we prefer to run a cDNA dilution series over a variable number of cycles to ensure that all samples have undergone equivalent reverse transcription and can be similarly amplified in PCR. Thereafter, variation in the number of cycles alone can provide data that accurately reflects twofold to fivefold differences in transcript level for the RNA of interest. We always use oligo-dT as a primer for reverse transcription such that the same sample can be examined for expression of multiple (including control) transcripts. Competitive PCR may be used to indicate absolute levels of transcript within different samples.4,5 This approach is based on the use of a competitor fragment within the PCR whose level is titrated against a fixed or variable amount of the experimental sample. The point at which the amount of PCR product from both templates is equivalent provides an indication of the level of transcript in the experimental sample (Fig 2). Different groups have used alternative competitors including genomic DNA and synthetic DNA or RNA templates. The key is that the size of product amplified from the competitor should differ in size from that of the normal transcript (although preferably by only a few tens of base pairs), and that it should contain identical primer binding sites. This method can provide beautiful results, but it is labour intensive, does not always work well enough to give absolute quatitation of data, and the avail-
From the Department of Biology, Imperial College of Science Technology and Medicine, London, UK. Work from this laboratory is supported by The Wellcome Trust, The Medical Research Council, The European Union, The national Kidney research Fund and Action Research. Address reprint requests to Margaret J. Dallman, Department of Biology, Imperial College of Science Technology and Medicine, Prince Consort Rd, London SW7 2BB, UK.
© 1998 by Elsevier Science Inc. 655 Avenue of the Americas, New York, NY 10010
0041-1345/98/$19.00 PII S0041-1345(98)00657-5
Transplantation Proceedings, 30, 2367–2369 (1998)
2367
2368
DALLMAN
Fig 2. Competitive PCR.
ability of only small amounts of material from, for instance clinical samples, may limit its use.
CYTOKINE ANALYSIS USING RT-PCR FOLLOWING TRANSPLANTATION
Both semiquantitative and competitive RT-PCR have been used to describe the expression of cytokines following transplantation in papers that are far too numerous to list here.6 –10 The work in quite a few early papers, however, did not use either quantitative or semiquantitative approaches and therefore are difficult to accurately interpret. The possibility that differential profiles of cytokine expression, based around the Th1/Th2 paradigm, might associate with graft rejection and the induction of tolerance has been a prime motivator for work in this area and has inspired many reviews on the subject.6 –10 From all of the studies in these areas there are several key findings that can be summarised: 1. The presence of cytokine transcripts within a graft may be extremely transient. 2. Proinflammatory cytokines may be expressed within grafts that are not subject to rejection, for instance syngeneic grafts, simply as a consequence of the transplant operation. 3. IL-2 is observed in both clinical and experimental transplants during the onset of the immune response to the graft, but is often absent during overt rejection episodes. 4. Graft rejection is often dominated by cytokines of the Th1 type. There are, however, exceptions to this,
exemplified by liver transplants and some xenografts where the Th2 profile is predominant. 5. Prolonged graft survival is usually associated with a decrease in Th1 cytokines (or Th2 cytokines in the case of liver and some xeno). 6. Tolerance is often associated with a rapidly downregulated expression or absence of all cytokines. 7. In clinical transplantation (where immunosuppressive cytokines clearly can influence cytokine expression), there is no profile of cytokines that has been described consistently by all workers in the field that is specifically and selectively useful in the diagnosis of rejection. 8. The onset of infection within a patient or graft can influence the cytokines expressed. While such data has been extremely useful in our understanding of cytokine activity in the context of transplantation, it can only provide associative data and does not address the causal role of cytokines in various aspects of the immune response to a graft. More direct intervention strategies have been used, based largely on the results summarised above. Such work has underscored the important role of cytokines in transplant responses, but, importantly, led us to understand the nonessential role of many single cytokines in these processes. Moreover, it has become accepted by most workers that the difference between graft rejection and tolerance cannot be explained by a simple shift between Th1 and Th2 responses. Indeed, quite the opposite appears to be the case, a dominance of either Th1 or Th2 cytokines can initiate a damaging immune response, albeit of a different nature, which can result in destruction of the graft.
DIFFERENTIAL DISPLAY RT-PCR
Differential display-PCR (DD-PCR), originally described by Liang and Pardee,11 has been a very useful technique with which to phenotype cells or tissues at the RNA level. It has also revealed the existence of hitherto unknown genes or unknown patterns of gene expression. It is based on the amplification of sequences using a set of anchor primers (59T12NA39, 59T12NC39, 59T12NG39) together with random ten-mers such that in theory any transcript, even those expressed at very low level, may be identified. Several different RNA samples can be compared in a single experiment in which the products of the DD-PCR are separated by polyacrylamide gel electrophoresis. Differentially expressed transcripts appear as a band(s) present in the products of PCR from one sample but not another. Such bands may be cut out of the gel and reamplified for further analysis and sequencing. Clearest results can be obtained from material that is very similar, but obviously not identical—samples that are too different will generate so many differentially expressed bands that the task of analysing them becomes impossible. While this sounds like a great technique, there are several drawbacks to its use, among
POLYMERASE CHAIN REACTION METHODS IN IMMUNOLOGY
which are the predominant amplification of short products from 39 untranslated regions (39UT) of the RNA and irreproducible amplification of differentially expressed bands. Clearly controls can be put in place to ensure against the latter problem. However, the former problem results in one thinking that one has found a unique gene as the sequence cannot be identified on data bases, but when more sequence is obtained it turns out not to be unique. This is because very often the complete 39UT of cDNAs is not sequenced or present on the data base. So far the use of this method has not produced many papers in the field of transplantation, but there are some results from work done on vasculopathy, liver regeneration, and chronic rejection.12–14
CONCLUSIONS
PCR continues to provide a vital technique in transplantation. The three areas described in this paper have concentrated on the use of RT-PCR, and I hope have illustrated the benefits and pitfalls of using this type of approach. Undoubtedly without the development of RT-PCR our knowledge of cytokine expression within organ grafts would have remained sparse.
2369
REFERENCES 1. Mullis K, Faloona F, Scharf S, et al: Cold Spring Harb Symp Quant Biol 51:263, 1986 2. Dallman MJ, Porter ACG In McPherson MJ, Quirke P, Taylor GR (eds): PCR: A Practical Approach. Oxford UK: IRL Press; 1991, p 215 3. Dallman MJ, Larsen CP, Morris CP: J Exp Med 174:493, 1991 4. Gilliland G, Perrin S, Bunn HF: In Innis MA, Gelfand DH, Sninsky JJ, et al (eds): PCR Protocols. A Guide to Methods and Applications. San Diego, Calif: Academic Press; 1990, p 60 5. Wang AM, Mark DF: In Innis MA, Gelfand DH, Sninsky JJ, et al (eds): PCR Protocols. A Guide to Methods and Applications. San Diego, Calif: Academic Press 1990, p 70 6. Dallman MJ: 5:788, 1993 7. Dallman MJ: Curr Opin Immunol 7:632, 1995 8. Nickerson P, Steurer W, Steiger J, et al: Curr Opin Immunol 6:757, 1994 9. Strom TB, Roy-Chadhury P, Manfro R, et al: 8:688, 1996 10. Picotti JR, Chan SY, VanBuskirk AM, et al: Transplantation 63:619, 1997 11. Liang P, Pardee AB: Science 257:967, 1992 12. Utans U, Liang P, Wyner L, et al: Proc Natl Acad Sci USA 91:6463, 1994 13. Kar S, Carr B: Biochem Biophys Res Commun 212:21, 1995 14. Chen J, Myllarniemi M, Akyurek LM, et al: Am J Pathol 149:597, 1996
T
HE POLYMERASE chain reaction (PCR) has provided one of the biggest technologic advances in molecular biology of the last decade.1 Its application to the analysis of both DNA and RNA has allowed the dissection of samples either too small or in which the target sequence is present at such low concentration for previous techniques to be of use. PCR has been used extensively in transplantation as a method for both genotyping and RNA phenotyping. Discussion of PCR in this paper is confined to its use in the latter application.
REVERSE TRANSCRIPTASE-PCR: QUANTITATION OF TRANSCRIPT LEVELS
As originally described, PCR is a completely nonquantitative technique. Samples are amplified over a fixed number of cycles of denaturation, primer annealing, and extension. While this approach is perfectly satisfactory for genotyping cells, in situations in which the level of specific transcripts within different RNA pools is to be determined by reverse transcriptase-PCR (RT-PCR) RT-PCR, it is not. The reasons for this are illustrated in Fig 1. As may be observed, the exponential phase of detectable product accumulation is preceded by a phase in which product is undetectable (A) and succeeded by a phase in which product level reaches a plateau (B). Only if one measures product during the midexponential phase of amplification can a variance between two (or more) samples that contain different levels of target sequence be detected (C1, C2). Two main approaches have been described that
Fig 1. Accumulation of product in samples containing different levels of target sequence.
attempt to resolve this problem by measuring transcript levels in either a semiquantitative or quantitative fashion. These are described in detail elsewhere,2–5 but will be illustrated briefly here. Semiquantitative RT-PCR2,3 relies on running a dilution series of either the RNA or cDNA used in the PCR or by sampling product from the PCR reaction at variable cycle number (eg, every 2 to 5 cycles). In the initial analysis of samples we use control primers that amplify transcripts whose level should not vary in different RNA pools (eg, beta actin, HPRT). In this instance, we prefer to run a cDNA dilution series over a variable number of cycles to ensure that all samples have undergone equivalent reverse transcription and can be similarly amplified in PCR. Thereafter, variation in the number of cycles alone can provide data that accurately reflects twofold to fivefold differences in transcript level for the RNA of interest. We always use oligo-dT as a primer for reverse transcription such that the same sample can be examined for expression of multiple (including control) transcripts. Competitive PCR may be used to indicate absolute levels of transcript within different samples.4,5 This approach is based on the use of a competitor fragment within the PCR whose level is titrated against a fixed or variable amount of the experimental sample. The point at which the amount of PCR product from both templates is equivalent provides an indication of the level of transcript in the experimental sample (Fig 2). Different groups have used alternative competitors including genomic DNA and synthetic DNA or RNA templates. The key is that the size of product amplified from the competitor should differ in size from that of the normal transcript (although preferably by only a few tens of base pairs), and that it should contain identical primer binding sites. This method can provide beautiful results, but it is labour intensive, does not always work well enough to give absolute quatitation of data, and the avail-
From the Department of Biology, Imperial College of Science Technology and Medicine, London, UK. Work from this laboratory is supported by The Wellcome Trust, The Medical Research Council, The European Union, The national Kidney research Fund and Action Research. Address reprint requests to Margaret J. Dallman, Department of Biology, Imperial College of Science Technology and Medicine, Prince Consort Rd, London SW7 2BB, UK.
© 1998 by Elsevier Science Inc. 655 Avenue of the Americas, New York, NY 10010
0041-1345/98/$19.00 PII S0041-1345(98)00657-5
Transplantation Proceedings, 30, 2367–2369 (1998)
2367
2368
DALLMAN
Fig 2. Competitive PCR.
ability of only small amounts of material from, for instance clinical samples, may limit its use.
CYTOKINE ANALYSIS USING RT-PCR FOLLOWING TRANSPLANTATION
Both semiquantitative and competitive RT-PCR have been used to describe the expression of cytokines following transplantation in papers that are far too numerous to list here.6 –10 The work in quite a few early papers, however, did not use either quantitative or semiquantitative approaches and therefore are difficult to accurately interpret. The possibility that differential profiles of cytokine expression, based around the Th1/Th2 paradigm, might associate with graft rejection and the induction of tolerance has been a prime motivator for work in this area and has inspired many reviews on the subject.6 –10 From all of the studies in these areas there are several key findings that can be summarised: 1. The presence of cytokine transcripts within a graft may be extremely transient. 2. Proinflammatory cytokines may be expressed within grafts that are not subject to rejection, for instance syngeneic grafts, simply as a consequence of the transplant operation. 3. IL-2 is observed in both clinical and experimental transplants during the onset of the immune response to the graft, but is often absent during overt rejection episodes. 4. Graft rejection is often dominated by cytokines of the Th1 type. There are, however, exceptions to this,
exemplified by liver transplants and some xenografts where the Th2 profile is predominant. 5. Prolonged graft survival is usually associated with a decrease in Th1 cytokines (or Th2 cytokines in the case of liver and some xeno). 6. Tolerance is often associated with a rapidly downregulated expression or absence of all cytokines. 7. In clinical transplantation (where immunosuppressive cytokines clearly can influence cytokine expression), there is no profile of cytokines that has been described consistently by all workers in the field that is specifically and selectively useful in the diagnosis of rejection. 8. The onset of infection within a patient or graft can influence the cytokines expressed. While such data has been extremely useful in our understanding of cytokine activity in the context of transplantation, it can only provide associative data and does not address the causal role of cytokines in various aspects of the immune response to a graft. More direct intervention strategies have been used, based largely on the results summarised above. Such work has underscored the important role of cytokines in transplant responses, but, importantly, led us to understand the nonessential role of many single cytokines in these processes. Moreover, it has become accepted by most workers that the difference between graft rejection and tolerance cannot be explained by a simple shift between Th1 and Th2 responses. Indeed, quite the opposite appears to be the case, a dominance of either Th1 or Th2 cytokines can initiate a damaging immune response, albeit of a different nature, which can result in destruction of the graft.
DIFFERENTIAL DISPLAY RT-PCR
Differential display-PCR (DD-PCR), originally described by Liang and Pardee,11 has been a very useful technique with which to phenotype cells or tissues at the RNA level. It has also revealed the existence of hitherto unknown genes or unknown patterns of gene expression. It is based on the amplification of sequences using a set of anchor primers (59T12NA39, 59T12NC39, 59T12NG39) together with random ten-mers such that in theory any transcript, even those expressed at very low level, may be identified. Several different RNA samples can be compared in a single experiment in which the products of the DD-PCR are separated by polyacrylamide gel electrophoresis. Differentially expressed transcripts appear as a band(s) present in the products of PCR from one sample but not another. Such bands may be cut out of the gel and reamplified for further analysis and sequencing. Clearest results can be obtained from material that is very similar, but obviously not identical—samples that are too different will generate so many differentially expressed bands that the task of analysing them becomes impossible. While this sounds like a great technique, there are several drawbacks to its use, among
POLYMERASE CHAIN REACTION METHODS IN IMMUNOLOGY
which are the predominant amplification of short products from 39 untranslated regions (39UT) of the RNA and irreproducible amplification of differentially expressed bands. Clearly controls can be put in place to ensure against the latter problem. However, the former problem results in one thinking that one has found a unique gene as the sequence cannot be identified on data bases, but when more sequence is obtained it turns out not to be unique. This is because very often the complete 39UT of cDNAs is not sequenced or present on the data base. So far the use of this method has not produced many papers in the field of transplantation, but there are some results from work done on vasculopathy, liver regeneration, and chronic rejection.12–14
CONCLUSIONS
PCR continues to provide a vital technique in transplantation. The three areas described in this paper have concentrated on the use of RT-PCR, and I hope have illustrated the benefits and pitfalls of using this type of approach. Undoubtedly without the development of RT-PCR our knowledge of cytokine expression within organ grafts would have remained sparse.
2369
REFERENCES 1. Mullis K, Faloona F, Scharf S, et al: Cold Spring Harb Symp Quant Biol 51:263, 1986 2. Dallman MJ, Porter ACG In McPherson MJ, Quirke P, Taylor GR (eds): PCR: A Practical Approach. Oxford UK: IRL Press; 1991, p 215 3. Dallman MJ, Larsen CP, Morris CP: J Exp Med 174:493, 1991 4. Gilliland G, Perrin S, Bunn HF: In Innis MA, Gelfand DH, Sninsky JJ, et al (eds): PCR Protocols. A Guide to Methods and Applications. San Diego, Calif: Academic Press; 1990, p 60 5. Wang AM, Mark DF: In Innis MA, Gelfand DH, Sninsky JJ, et al (eds): PCR Protocols. A Guide to Methods and Applications. San Diego, Calif: Academic Press 1990, p 70 6. Dallman MJ: 5:788, 1993 7. Dallman MJ: Curr Opin Immunol 7:632, 1995 8. Nickerson P, Steurer W, Steiger J, et al: Curr Opin Immunol 6:757, 1994 9. Strom TB, Roy-Chadhury P, Manfro R, et al: 8:688, 1996 10. Picotti JR, Chan SY, VanBuskirk AM, et al: Transplantation 63:619, 1997 11. Liang P, Pardee AB: Science 257:967, 1992 12. Utans U, Liang P, Wyner L, et al: Proc Natl Acad Sci USA 91:6463, 1994 13. Kar S, Carr B: Biochem Biophys Res Commun 212:21, 1995 14. Chen J, Myllarniemi M, Akyurek LM, et al: Am J Pathol 149:597, 1996