A rapid method for semiquantitative analysis of the human Vβ-repertoire using TaqManR PCR

Ž .Journal of Immunological Methods 203 1997 181–192

A rapid method for semiquantitative analysis of the humanVb-repertoire using TaqManR PCR

Roland Lang ), Klaus Pfeffer, Hermann Wagner, Klaus HeegInstitute of Medical Microbiology, Immunology and Hygiene, Technical UniÕersity of Munich, Trogerstr. 32, 81675 Munich, Germany

Received 29 October 1996; revised 13 January 1997; accepted 10 February 1997

Abstract

Analysis of the Vb-repertoire of antigen-reactive T cell populations can be approached using either flow-cytometry orŽ .PCR-based techniques. While the former method requires a complete set of Vb-specific monoclonal antibodies mAbs and

large cell numbers for analysis, the latter is both time-consuming and labour-intensive. To circumvent the drawbacks of boththese methods we have employed the recently developed technique of TaqManR PCR to analyse the Vb-usage of human Tcell populations. TaqManR PCR is based on the 5X ™3X nuclease activity of Taq polymerase. During PCR amplification aninternal oligonucleotide probe, that is labelled with a fluorescent reporter and a quencher dye, is cleaved by Taq polymerase.After cleavage, quenching of the reporter dye is lost and reporter fluorescence can be detected with a fluorescence platereader. Using one Cb-specific fluorogenic probe and a panel of Vb-specific primers, we show that fluorescence-detectedamplification of TCR b cDNA is Vb-specific and linear within a 2–3-log range of template concentration. The sensitivity ofTaqManR PCR is comparable to conventional detection of PCR-products by agarose gel staining, while processing time isreduced. Furthermore, superantigen-induced skewing of the Vb-repertoire of human cells is readily detected with thismethod. Thus TaqManR PCR is a reliable and fast method for semiquantitative analysis of the Vb-repertoire of human T cellpopulations.

Keywords: T cell receptor; Human Vb repertoire; PCR; Fluorogenic probe

1. Introduction

T cells recognize MHC-bound peptide antigen viaŽtheir heterodimeric ab TCR Kronenberg et al.,

Abbreviations: FAM: 6-carboxy-fluorescein; FITC: fluores-ceinisothiocyanate; mAb: monoclonal antibody; PE: phyco-erythrin; RQ: ratio of reporter to quencher dye fluorescence;

Ž . Ž .DRQ: RQ sample yRQ no template ; SEA: staphylococcal en-terotoxin A; SEB: staphylococcal enterotoxin B; TAMRA: 6-carboxymethylrhodamine; TSST: toxic shock syndrome toxin

) Ž .Corresponding author. Tel.: q 49 -89-4140-4185; Fax:Ž .q49 -89-4140-4942.

.1986 . It is thought that the contact site for theantigenic peptide is formed by the CDR3 regions of

Ž .the TCR chains Katayama et al., 1995 . The CDR3region is encoded by the rearranged gene region ofthe TCR chains that are formed by the joining of

Ž . Ž . Ž .variable V , diversity D and joining J gene seg-Žments during T cell ontogeny Kronenberg et al.,

.1986; Davis and Bjorkman, 1988 . Consequently,usage of individual Vb-gene elements should bedistributed randomly within an antigen-reactive Tcell population. However, under certain conditionsthe TCR repertoire of antigen reactive T cell popula-tions is skewed towards a biased usage of specific

0022-1759r97r$17.00 Copyright q 1997 Elsevier Science B.V. All rights reserved.Ž .PII S0022-1759 97 00028-8

( )R. Lang et al.rJournal of Immunological Methods 203 1997 181–192182

Vb-elements. Those conditions include T cell stimu-Žlation with superantigens Choi et al., 1989; Fleis-

.cher et al., 1996 and even certain peptide antigensŽWucherpfennig et al., 1990; Deckhut et al., 1993;

.Dillon et al., 1994 . Thus analysis of the Vb-reper-toire of T cell populations has become an importanttool in the study of the immune response in health

Žand disease Sensi and Parmiani, 1995; Rebai et al.,.1994 .

Vb gene segments sharing more than 75% se-quence homology are grouped into subfamilies. The

Ž .recent classification by Arden et al. 1995 lists 26Vb-subfamilies in humans, while the two constant

Ž .region gene segments of the b chain Cb share ahigh homology. The Vb region repertoire of T cellpopulations can be measured either by flow-cytome-

Žtry or by RT-PCR Choi et al., 1989; Marguerie et.al., 1992 . The advantage of flow-cytometry is the

direct visualization of the protein of interest on thecell surface, allowing a direct and convenient deter-mination of the percentage of Vb-expressing cells ina given population. However, a large number of cellsmust be stained and analysed, and for the humansystem the panel of Vb-specific antibodies is notcomplete. With RT-PCR a smaller number of cells issufficient for the analyses and specific primers for allVb-segments can easily be designed because the

Žsequences of interest are known Arden et al., 1995;.Rowen et al., 1996 . Different approaches have been

proposed to quantitatively analyse the T cell reper-toire by RT-PCR, including anchored PCR followedby cloning of the product and subsequent hybridiza-

Ž .tion Rieux-Laucat et al., 1993 , semiquantitativeŽ .PCR Choi et al., 1989; Pluschke et al., 1994 and

Ž .competitive PCR Spinella and Robertson, 1994 .However, a major disadvantage of RT-PCR has beenthe labor intensive detection of the PCR product, i.e.gel electrophoresis, ethidium bromide staining anddensitometric quantitation, which are time consum-ing, likely to reduce the accuracy and increase con-tamination. Analysis of larger sample numbers orscreening of T cell populations has therefore notbeen feasible.

Recently a new technique for the detection ofPCR-amplified nucleic acids employing the 5X ™3X

nuclease activity of Taq polymerase has been re-Ž .ported Livak et al., 1995 . An internal oligonucleo-

tide probe, specific for the target sequence, with a

reporter dye attached to the 5X end and a quencherdye at the 3X end is added to the PCR reaction.Provided the target sequence is amplified, the exonu-clease activity of Taq polymerase cleaves the 3X

Ž .quencher dye Holland et al., 1991 , leaving thereporter dye fluorescence unquenched. The resultingincrease in reporter dye fluorescence can be detectedwith a plate reader. Post PCR processing time isgreatly reduced since the measurement of fluores-cence intensities is accomplished within minutes. Todate, TaqManR technology has been used only for

Žthe detection of bacterial DNA sequences Bassler et.al., 1995; Livak et al., 1995; Whitham et al., 1996 ,

but there should be no constraints in the widerapplications of this technique. The use of an internalprobe increases the specificity of PCR based detec-tion of nucleic acids without the need for additionalhybridization steps. We used TaqManR technologyin a Vb family RT-PCR with one fluorogenic Cb-specific probe to establish a fast and convenientmethod for the analysis of Vb-usage in human T cellpopulations.

2. Materials and methods

2.1. Cell culture

Ž .RPMI1640 Biochrom, Berlin, Germany , wasused as cell culture medium supplemented with 10%

Ž .FCS Seromed, Berlin, Germany , 50 mM b-mer-captoethanol and antibiotics. Blood was drawn fromhealthy volunteers, peripheral blood mononuclear

Ž .cells PBMC isolated by Ficoll density centrifuga-tion and stimulated in bulk cultures with 1 mgrml of

Ž .phythemagglutinin PHA; Wellcome, Bucks, UK or0.5 mgrml of the superantigens staphylococcal en-

Ž . Ž . Žterotoxin A SEA and B SEB Toxin Technology,.Sarasota, USA . After four days 20 Urml Inter-

Žleukin-2 IL-2; Eurocetus, Amsterdam, The Nether-.lands was added to the cultures. On day 6 cells were

either restimulated or harvested by centrifugationand counted. Samples were stained with mAbs for anumber of different Vb s and CD3 and total cellularRNA was prepared.

2.2. RNA extraction and cDNA synthesis

Total RNA was isolated from 1–5=106 cellsŽusing the RNeasy total RNA extraction kit Qiagen,

( )R. Lang et al.rJournal of Immunological Methods 203 1997 181–192 183

.Hilden, Germany according to the manufacturer’sprotocol. The amount and purity of the eluted RNAwas determined by measuring the OD at 260 and 280nm. 2 mg of RNA were reverse transcribed with 200

ŽU MMLV reverse transcriptase Gibco, Eggenstein,.Germany in a volume of 20 ml at 378C for 60 min.

cDNA synthesis was primed with random hexa-Ž .nucleotide primers Pharmacia, Freiburg, Germany

if not otherwise indicated.

2.3. Primers and fluorogenic probes

Vb-specific 5X and the common Cb 3X primerswere selected from the literature or designed accord-

Ž .ing to the published TCR b sequences see Table 1Ž .and synthesised TIB Molbiol, Berlin, Germany .

The internal fluorogenic probes Cb 20 and Cb 209were designed to match perfectly to target sequences,close to the annealing sites of the 3X reverse primersand to have a higher melting point than the PCR

Žprimers. Probes were synthesised with FAM re-

. X Žporter dye attached to the 5 and TAMRA quencher. X Ždye to the 3 end Perkin Elmer, Weiterstadt, Ger-.many .

2.4. PCR conditions

PCR was performed in a Perkin Elmer Gene Amp9600 cycler in a final volume of 100 ml. After an

Ž .initial denaturation step 948C for 5 min , Taq poly-merase was added in a volume of 20 ml, followed bythe indicated number of cycles with 958C for 25 s,588C for 30 s and 728C for 60 s. The reaction

Ž . Ž .mixture contained 10 ml PCR buffer 10= Gibco ,concentration of Mg2q was 3.0 mM, of dUTP 0.4

Ž .mM Pharmacia , dATP, dCTP and dGTP 0.2 mMŽ .Gibco , of primers 0.5 mM and of probes 0.1 mMŽ .if not otherwise indicated .

2.5. Detection and quantitation of PCR products

After PCR cycling, 80 ml of each sample wereŽtransferred to a 96-well Hiwhite plate PerkinElmer,

Table 1Oligonucleotide primers and fluorogenic probes

Name Sequence Reference

Vb 1 CACAACAgTTCCCTgACTTgCACTC Choi et al., 1989Vb 2 gACTTTCAggCCACAACTAT Rieux-Laucat et al., 1993Vb 3 AAAggAgATATTCCTgAggg Rieux-Laucat et al., 1993Vb4 CAgCCgCCCAAACCTAAC Genevee et al., 1992Vb5 gCTCTgAgATgAATgTgAACgC Rieux-Laucat et al., 1993Vb6 TCAggTgTgATCCAATTTC Genevee et al., 1992Vb 7 CCTgAATgCCCCAACAgC Choi et al., 1989Vb 8 ACTTTAACAACAACgTTCCg Rieux-Laucat et al., 1993Vb 9 TTCCCTggAgCTTggTgAC Genevee et al., 1992Vb 10 CCACggAgTCAgggAgACAC Genevee et al., 1992Vb 11 AATggAACTACACCTCATCCACTATTCCVb 12 AAAggAgAAgTCTCAgATggCTVb 13 TCCTgAAgACAggACAgAgC Rieux-Laucat et al., 1993Vb 14 gTgACTggAAAgAAgTTACAgTgACTTgVb 15 CAggAAAgAggATTATgCTggVb 16 gAAAgAgTCTAAACAggATgAgTCC Choi et al., 1989Vb 17 CACAgATAgTAAATgACTTTCAg Choi et al., 1989Vb 18 ACAggAggCAAgACTgAgATgC Rieux-Laucat et al., 1993Vb 19 gAAACggAgATgCACAAgAAgCgVb 20 gTgCCCCAgAATCTCTCAgCC Genevee et al., 1992Vb 21 AgTAgACgATTCACAgTTgCCTAAgg Rieux-Laucat et al., 1993Vb 22 CAAATCTTggggCAgAAAgTCgCb50 TgTgggAgATCTCTgCTTCT Genevee et al., 1992Cb 20 ACACAgCgACCTCgggTggg Genevee et al., 1992Cb 243 gTCCACTCgTCATTCTCCgA Rosenberg et al., 1992Cb 209 TAgAACTggACTTgACAgCggAAgTgg

Ž .Where no reference is given, primers were designed according to the Vb sequences published by Arden et al. 1995 .


.Weiterstadt, Germany and the fluorescence intensi-ties of FAM and TAMRA were measured using anLS-50B plate reader with an excitation wavelengthof 488 nm and emission at 520 or 580 nm, respec-tively. The ratio of FAM to TAMRA fluorescenceŽ .RQ was calculated. The mean RQ of at least foursamples without template DNA was subtracted fromthe RQ values of the samples to obtain DRQ values.To determine intraassay variations, PCR was per-formed in duplicate. For reasons of clarity, mean andstandard deviations of duplicate determinations are

Ž .shown and not the individual datapoints . For visu-alization of PCR products an sample of 15 ml wasloaded onto a 1.5% agarose gel. After electrophore-sis, the gel was stained for 15 min in an ethidiumbromide bath and destained in water for an addi-tional 30 min. Band intensities were estimated usingWinCam software, and using as a reference standardthe density of the 1.635 kb band of the 1 kb markerŽ .Gibco, Eggenstein, Germany .

2.6. Cloning of TCR b chain transcripts

After individual amplification with forwardprimers specific for Vb 1, Vb 2 and Vb 3, crude PCR

R Žproducts were ligated into pGEM T-vector Pro-.mega, Heidelberg, Germany overnight at 158C. E.

R Ž .coli SURE cells Stratagene, Heidelberg, Germanywere transformed by electroporation and plated onto

LB agar plates containing ampicillin. White colonieswere picked and the presence of an insert was con-firmed by electrophoresis after alkaline lysis and

Ž .digestion with PvuII Gibco, Eggenstein, Germany .

2.7. Flow-cytometry

ŽLymphocytes were stained on ice with mAb Im-.munotech, Hamburg, Germany diluted appropriately

in PBS. mAbs specific for Vb families were FITC-conjugated, anti-CD3 mAb was PE-conjugated. Afterwashing, cells were fixed in 1% paraformaldehyde inPBS. Two-colour analyses were performed using an

Ž .EPICS Elite flow-cytometer Coulter, Hialeh, USAequipped with an argon laser. Cells were gated onforward and side scatter according to standard proce-dures.

3. Results

3.1. Strategy

TCR b chain transcripts are composed of vari-able, joining, diversity and constant regions. Forwardprimers for 22 Vb regions were selected and checkedfor specificity in the EMBL Genbank. Two reverseprimers for the constant regions were designed, bothof which were perfectly matched to sequences con-

Fig. 1. Titration of the optimal probe concentration. Dilutions of plasmid Vb 2C3 were amplified using the Vb 2 forward primer and theŽ . Ž .Cb50rCb 20 right panel or Cb 243rCb 209 left panel reverse primerrprobe combination. Probe concentration was titrated as

indicated. After 30 cycles fluorescence of FAM and TAMRA was analysed at 520 and 580 nm, respectively, using an LS50B-plate reader.Calculation of DRQ values was performed as outlined in Section 2. Mean values and standard deviation of duplicate determinations, areshown.


Fig. 2. Sensitivity of TaqManR PCR vs. visualization by ethidium bromide stained agarose gels. Serial dilutions of cDNA prepared fromŽ . Ž . Ž .TSST stimulated PBL cultures left graph and Jurkat T cells right graph were subjected to PCR using the Vb 2 TSST or Vb 8 forward

primer plus the Cb50rCb 20 reverse primerrprobe combination. The probe concentration was 100 nM. After 32 cycles, fluorescence wasdetermined, followed by running an aliquot of 15 ml on an 1.8% agarose gel. Bands were visualized under UV light after staining the gelwith ethidium bromide. Quantitation of band intensity was performed using WinCam software and taking the 1635 bp band of the 1 kbDNA ladder as reference. For DRQ values, datapoints represent mean and standard deviation of duplicate determinations. For electrophore-sis only one of the two amplified samples was used.

served among Cb 1 and Cb 2. The internal fluoro-genic oligonucleotide probe was designed to annealto the conserved region of Cb 1 and Cb 2 in closeproximity to and with the same orientation as the Cb

reverse primer. During PCR amplification the re-verse primer and the internal probe can anneal to allb chain transcripts, while only the template contain-ing the sequence of the specific Vb forward primer

should be amplified in an exponential manner. Usingthis approach, Vb-specific amplification should bedetectable with the need for only one fluorogenicinternal oligonucleotide probe.

The cDNA of interest was amplified with a panelof 22 Vb-specific 5X primers and one reverseprimerrprobe combination. Since DRQ values de-

Ž .pend on template amount see Section 3.3 , raw data

Fig. 3. Specificity of TaqManR PCR. Serial dilutions of plasmids Vb 1C1, Vb 2C3 and Vb 3C2 or of Jurkat cDNA were subjected to 30cycles of amplification in the presence of probe Cb 20 at a concentration of 100 nM. The different symbols represent the differentVb-specific forward primers used. Non-specific forward primers were included only at the highest template concentration for each plasmidor the cDNA. PCR was performed in duplicate, the line in each graph is the linear regression curve.


Fig. 4. Dependance of DRQ values on template amount. Serial dilutions of plasmids Vb 2C3 and Vb 1C1 were amplified for the indicatednumber of cycles using 100 nM of two different batches of probe Cb 20. Data were fitted to a sigmoid shaped curve with a four parameter

Ž R . Ž Ž .b.logistic function fit Sigma Plot , Jandel Scientific . The non-linear regression curve was described by the equation DRQsar 1q xrc ,where a is the asymptotic plateau value, b the slope of the curve, c the inflexion point and x the amount of template.

were transformed to obtain molar amounts of Vb

transcripts. By dividing the molar amount of one Vb

by the sum of all Vb values, the relative abundanceof one Vb in a given cDNA was calculated. In thisway, a comparison can be made of the frequency ofspecific Vb subfamilies between cell populations.

3.2. Determination of the optimal probe concentra-tion

Two different reverse primers were used for theamplification of TCR b chain transcripts, both an-

Fig. 5. Optimization of Mg2q concentration for TCR b amplifica-tion. 1 attomol of plasmids Vb 2C3 and Vb 3C2 was subjected toTaqManR PCR for 26 cycles using the indicated concentrations ofMgCl . Shown are meanqstandard deviation of duplicate deter-2

minations.

nealing to the constant region and yielding productsdiffering in size by 250 bp. PCR amplification of a

Žplasmid Vb 2C3 containing a Vb 2-TCR b se-.quence showed that both combinations of reverse

primer and internal probe generated an increase inreporter dye fluorescence, yielding positive DRQvalues which rose with increasing amounts of tem-

Ž .plate Fig. 1 . Titrating the concentration of fluoro-genic probe we observed a higher positive signalwith lower probe concentrations. This is explained

Fig. 6. TaqMan PCR with plasmid mixtures as template. Plasmidy4 Ž . y6Vb 1C1 in a dilution of 1=10 open symbols or 1=10

Ž .closed symbols was mixed with the indicated dilutions of plas-mid Vb 2C3. The mixture was subjected to TaqManR PCR for 28

Ž . Ž .cycles, using the Vb 1 squares or Vb 2 circles forward primer.


by the calculation of reporter to quencher dye fluo-Ž .rescence ratio RQ values , which yields higher RQ

values for a given number of cleaved probe moleculesŽ .if the total number of uncleaved plus cleaved probe

molecules in the sample tube is lower. However,duplicate determinations showed that the increase inRQ values for low probe concentrations was com-promised by higher standard deviations of the result-

Ž .ing DRQ values Fig. 1 . The results showed that aprobe concentration of 100 nM is optimal, combin-ing high DRQ values with acceptable standard devi-ations.

Comparison of the two reverse primerrprobecombinations revealed a higher sensitivity for theCb 20 probe. The lower positive signal obtainedwith the Cb 209 probe was not due to a reducedefficiency of the respective reverse primer becausethe densitometrically determined amount of PCR

Ž .product was essentially the same data not shown .In contrast, the sensitivities of agarose gel elec-trophoresis and the fluorescence plate reader for the

Ž .Cb 20 probe Fig. 2 were comparable. Given theseresults, we used the Cb 20 probe for all furtherexperiments.

Fig. 7. Transformation of raw data and calculation of Vb percentages. 0.1 ml aliquots of cDNA prepared from SEA stimulated PBMCs wereamplified with a panel of 22 Vb specific forward primers and the Cb50rCb 20 reverse primerrprobe combination for 32 cycles. Bars

Ž .represent mean values and standard deviation of duplicate determinations. For this preparation of probe Cb 20, a the maximal DRQ wasŽ .14, b the slope of the curve 0.85, as determined by amplifying titrated plasmid dilutions.


3.3. Specificity of amplification and dependance ofDRQ Õalues on template amount

To test the specificity of the Vb 5X primers andthe internal probes used, templates containing oneknown Vb-region were amplified with the corre-sponding and a variety of other 5X Vb primers plusthe common Cb 3X primer. When the plasmidsVb 2C3, Vb 1C1 and Vb 3C2 were used as templates,only cycling with the corresponding 5X primer gener-

Ž .ated an increase in FAM-fluorescence Fig. 3 . Inaddition, cDNA of the Jurkat T cell line, which is

Ž .known to express Vb 8 Fleischer et al., 1996 , re-sulted in positive DRQ values only when the Vb 8-

X Ž .specific 5 primer was used Fig. 3 . DRQ valuesdepended on the dilution of template. Dilutions ofJurkat cDNA, for example, yielded positive DRQvalues starting at 0.004 ml of cDNA and reaching a

Ž .plateau at 10 ml of cDNA input Fig. 2 .Datapoints were fitted by a sigmoid shaped curve

Ž . ŽFig. 4 , described by the equation DRQsar 1qŽ .b.xrc , where x is the amount of template, a is theasymptotic maximum, b the slope and c the inflex-ion point. While c changes with cycle number, bwas constant for different forward primers and setsof experiments. Different batches of probe Cb 20influenced the RQ value of the controls lacking

Ž .template data not shown and the magnitude of theŽ .maximal DRQ values Fig. 4 . The values for slope

and inflexion point of the sigmoid curve fitting thedatapoints, however, remained stable.

3.4. Role of Mg 2 q concentration

Since hybridization of the probes to the templateis critical for cleavage by Taq and generation of freereporter dye during PCR, we determined the optimalconcentration of Mg2q. When using the plasmidsVb 2C3 and Vb 3C2, DRQ values rose with increas-ing Mg2q concentrations until a maximum at a final

Ž .concentration of 3.0 mM Fig. 5 . This seems to benot only the optimal concentration for hybridizationof the probe but also for the 5X and 3X primers,because the product amount as determined by gel

Želectrophoresis follows the same pattern data not.shown .

3.5. The presence of other TCR b chain sequencesdoes not alter the Vb-specific amplification of thetarget sequence

Because we used one Cb reverse primerrprobecombination for the detection of all the Vb subfami-lies, we were concerned that binding of the probeand the reverse primer to other b chains mightinfluence the efficiency of the amplification of thespecific Vb targeted with the Vb forward primer. Asshown in Fig. 6, TaqManR PCR of the Vb 2C3

Fig. 8. Correlation of Vb percentages determined with FACS and TaqMan RT-PCR. Prior to RNA extraction aliquots of cell populationswere stained with the indicated mAbs and analysed by flow-cytometry. After random primed reverse transcription, cDNAs were subjected to

R Ž . Ž .TaqMan PCR and the percentages of Vb s calculated as outlined in Fig. 7. a and c PBMC were stimulated with PHA, SEA or TSST,Ž .restimulated on day 12 and harvested on day 17. b Jurkat T cells were mixed with SEB-stimulated PBMC at the indicated percentages of

Jurkat cells.


plasmid resulted in nearly the same DRQ valuesover a range of dilutions from 1=10y3 to 1=10y6 ,when the Vb 1C1 plasmid was present in the sametube in dilutions of 1=10y4 or 1=10y6. Using theconverse configuration, increasing amounts of plas-mid Vb 2C3 only slightly decreased the fluorescent

signal generated by amplification of the Vb 1C1plasmid diluted 1=10y4 or 1=10y6.

3.6. Analysis of T cell populations

For the analysis of Vb expression in T cell popu-lations the DRQ values obtained with a panel of 22

Ž .Fig. 9. Superantigen-induced changes in Vb profiles. PBMC were cultured in the presence of PHA 1 mgrml or one of the superantigensŽ .SEA, SEB and TSST 0.5 mgrml for 6 days. After isolation of living cells by Ficoll density centrifugation, restimulation was achieved

Ž .with the same concentration of stimuli in the presence of syngeneic irradiated 1500 rad feeder PBMC. On day 12 cells were harvested,RNA extracted and cDNA prepared. The percentages of Vb s obtained with TaqManR RT-PCR for TSST, SEA, and SEB stimulations are

Ž .compared with the percentages obtained using PHA stimulated cells open bars . Mean and standard deviations of duplicate determinationsare shown.


Vb-specific forward primers and the reverseprimerrprobe combination were transformed to mo-

Ž .lar equivalents Fig. 7 . The equation used for thistransformation was derived from the curve fittingfunction described in Fig. 3. Fig. 7 depicts howpercentages of Vb s were calculated from the originaldata, generated by TaqManR RT-PCR with the 22Vb-specific forward primers and one Cb reverseprimerrprobe combination.

To evaluate the feasibility of detecting differencesin the Vb repertoire between cell populations wouldbe feasible, we induced expansions of certain Vb

families by culturing PBMC with bacterial superanti-Žgens and prepared mixtures of Jurkat T cells ex-

Fig. 10. Detection of Vb expansion in SEA- or SEB-stimulatedPBMCs from different individuals. Stimulation of PBMC andanalysis of Vb repertoire was performed as outlined in Fig. 9.PBMC from three healthy donors were stimulated with eitherPHA, SEA or SEB and analysed with TaqManR RT-PCR. Thepercentages of Vb 3, Vb 17 and Vb 22 are plotted. Data barsrepresent mean values and standard deviation of duplicate deter-minations.

.pressing exclusively a Vb 8 TCR with SEB-stimu-lated PBMC. Vb percentages measured by flow-cy-tometry after staining with mAbs to Vb 2, Vb 8 andVb 22 served as controls for the results obtained by

R Ž .TaqMan RT-PCR Fig. 8 . Comparison of the re-sults obtained by staining for TCR surface expres-sion and TaqManR RT-PCR shows that differencesin the abundance of a given Vb from sample to

Ž .sample were readily detectable Fig. 8 . Changes inVb distribution due to in vitro culture with bacterialsuperantigens SEA, SEB and TSST as analysed withTaqManR RT-PCR are depicted in Fig. 9. Thestrongest skewing in the Vb repertoire was inducedby TSST which led to an increase of Vb 2 from 7%to 35%. SEA induced outgrowth of Vb5, Vb 7, Vb 9,Vb 18 and Vb 22, while culture with SEB favouredexpansion of Vb 3, Vb5 and Vb 17. These effects

Ž .could be shown in different individuals Fig. 10 .

4. Discussion

We have shown in this study that TaqManR-basedPCR provides a rapid and reliable semiquantitativeanalysis of the Vb-repertoire of human T cell popula-tions. Amplification of cDNA was Vb-specific andlinear over a 2–3 log range of template cDNAconcentrations, sufficient for the detection of differ-ences in Vb-abundance. Using this approach, wewere able to detect specifically the skewed Vb-reper-toire induced by the superantigens TSST, SEB andSEA. Since this method is less time-consuming thanconventional PCR approaches and can easily beadopted for other gene segments of the TCR, itoffers the opportunity to analyse the TCR repertoirein response to superantigens, to conventional peptideantigens and in autoimmunity.

TaqManR PCR does not circumvent the con-straints of RT-PCR. Frequencies of T cells bearingspecific Vb s are not determined on the single celllevel but rather via measuring the abundance ofspecific Vb transcripts in the mRNA pool. Differentefficiencies of the Vb-specific primers may lead tounequal amplification of different Vb families. This,and only partial recognition of Vb families by themAbs used for flow-cytometry, may account fordiscrepancies in the percentage values obtained by


flow-cytometric or TaqManR PCR analysis. Also,using TaqManR PCR, comparisons can only be madebetween different samples and not between Vb fami-lies within one sample. In contrast to conventionalPCR, however, the procedure is less time-consum-ing. Thus, an analysis can easily be performed withinone day with multiple samples. Moreover, the intrin-sic reporter probe used in TaqManR PCR providesan internal specificity control for PCR amplification.This approach permits the semiquantitative determi-nation of Vb-abundance within a T cell population.Thus, to date, one cannot quantitatively comparedifferent Vb families in one sample, but ratherchanges in the expression of particular Vb familiesfrom sample to sample. For quantitative analysis anexact assessment of the amplification rates of theindividual primer sets would be required. Using animproved TaqManR PCR this prerequisite could beeasily achieved by measuring the fluorescence signalafter each PCR cycle, thus allowing determination ofthe individual amplification rates. From that the ini-tial template concentration could be calculated quan-titatively.

Overall, TaqManR PCR proved to be a reliableand fast method for semiquantitative analysis ofVb-expression. TaqManR PCR could also be usedfor analysis of Va-expression thus allowing the rapiddetermination of Vb and Va TCR usage in small Tcell populations. Future measurement of the amplifi-cation rates during PCR might even permit quantita-tive determinations.

Acknowledgements

The authors thank G.B. Lipford and A. Bergen-thal for careful review of the manuscript.

References

Arden, B., Clark, S.P., Kabelitz, D., Mak, T.W., 1995. HumanT-cell receptor variable gene segment families. Immunogenet-ics 42, 455.

Bassler, H.A., Flood, S.J.A., Livak, K.J., Marmaro, J., Knorr, R.,Batt, C.A., 1995. The use of a fluorogenic probe in a PCR-based assay for the detection of Listeria monocytogenes.Appl. Environ. Microbiol. 61, 3724.

Choi, Y., Kotzin, B., Herron, L., Callahan, J., Marrack, P.,Kappler, J.W., 1989. Interaction of Staphylococcus aureus

toxin ‘superantigens’ with human T cells. Proc. Natl. Acad.Sci. USA 86, 8941.

Davis, M.M., Bjorkman, P.J., 1988. T cell antigen receptors andantigen recognition. Nature 334, 395.

Deckhut, A.M., Allan, W., McMickle, A., Eichelberger, M.,Blackman, M.A., Doherty, P.C., Woodland, D.L., 1993.Prominent usage of Vb 8.3 T cells in the H-2Db restrictedresponse to an influenza A virus nucleoprotein epitope. J.Immunol. 151, 2658.

Dillon, S.R., Jameson, S.C., Fink, P.J., 1994. Vb5q T cellreceptors skew toward OVAqH-2Kb recognition. J. Im-munol. 152, 1790.

Fleischer, B., Necker, A., Leget, C., Malissen, B., Romagne, F.,1996. Reactivity of mouse T-cell hybridomas expressing hu-man Vb gene segments with staphylococcal and streptococcalsuperantigens. Infect. Immun. 64, 987.

Genevee, C., Diu, A., Nierat, J., Caignard, A., Dietrich, P.Y.,Ferradini, L., Roman-Roman, S., Triebel, F., Hercend, T.,1992. An experimentally validated panel of subfamily-specific

Ž .oligonucleotide primers V 1-w29rV 1-w24 for the study ofa b

human T cell receptor variable V gene segment usage bypolymerase chain reaction. Eur. J. Immunol. 22, 1261.

Holland, P.M., Abramson, R.D., Watson, R., Gelfand, D.H., 1991.Detection of specific polymerase chain reaction product byutilizing the 5X

™3X exonuclease activity of Thermus aquati-cus DNA polymerase. Proc. Natl. Acad. Sci. USA 88, 7276.

Katayama, C.D., Eidelman, F.J., Duncan, A., Hooshmand, F.,Hedrick, S.M., 1995. Predicted complementary determiningregions of the T cell antigen receptor determine antigen speci-ficity. EMBO J. 14, 927.

Kronenberg, M., Siu, G., Hood, L.E., Shastri, N., 1986. Themolecular genetics of the T-cell antigen receptor and T-cellantigen recognition. Ann. Rev. Immunol. 4, 529.

Livak, K.J., Flood, S.J.A., Marmaro, J., Giusti, W., Deetz, K.,1995. Oligonuleotides with fluorescent dyes at opposite endsprovide a quenched probe system useful for detecting PCRproduct and nucleic acid hybridization. PCR Methods Appl. 4,357.

Marguerie, C., Lunardi, C., So, A., 1992. PCR-based analysis ofthe TCR repertoire in human autoimmune diseases. Immunol.Today 13, 336.

Pluschke, G., Taube, H., Krawinkel, U., Pfeffer, K., Wagner, H.,Classen, M., Deusch, K., 1994. Oligoclonality and skewed Tcell receptor Vb gene segment expression in in vivo activatedhuman intestinal intraepithelial T lymphocytes. Immunobiol-ogy 192, 77.

Rebai, N., Pantaleo, G., Deamrest, J.F., Ciurli, C., Soudeyns, H.,Adelsberger, J.W., Vaccarezza, M., Walker, R.E., Sekaly,R.P., Fauci, A.S., 1994. Analysis of the T-cell receptor-chainvariable-region repertoire in monozygotic twins discordant forhuman immunodeficiency virus: evidence for perturbations ofspecific Vb segments in CD4q T Cells of the virus-positivetwins. Proc. Natl. Acad. Sci. USA 91, 1529.

Rieux-Laucat, F., Le Deist, F., Selz, F., Fischer, A., de Villarty,J.P., 1993. Normal T cell receptor Vb usage in a primaryimmunodeficiency associated with HLA class II deficiency.Eur. J. Immunol. 23, 928.


Rosenberg, W.M.C., Moss, P.A.H., Bell, J.I., 1992. Variation inhuman T Cell receptor Vb and Jb repertoire: analysis usinganchor polymerase chain reaction. Eur. J. Immunol. 22, 541.

Rowen, L., Koop, B.F., Hood, L., 1996. The complete 685-kilo-base DNA sequence of the human T cell receptor locus.Science 272, 1755.

Sensi, M., Parmiani, G., 1995. Analysis of TCR usage in humantumors: a new tool for assessing tumor-specific immune re-ponses. Immunol. Today 16, 588.

Spinella, D.G., Robertson, J.M., 1994. Analysis of human T cellrepertoires by PCR. In: Mullis, K.B., Ferre, F., Gibbs, R.A.

Ž .Eds. , The Polymerase Chain Reaction. Birhauser, Boston, p.¨100.

Whitham, P.K., Yamashiro, C.T., Livak, K.J., Batt, C.A., 1996. APCR-based assay for the detection of Escherichia coli shiga-like toxin genes in ground beef. Appl. Environ. Microbiol. 62,1347.

Wucherpfennig, K.W., Ota, K., Endo, N., Seidman, J.G., Rosen-zweig, A., Weiner, H.L., Hafler, D.A., 1990. Shared human Tcell receptor Vb usage to immunodominant regions of myelinbasic protein. Science 248, 1016.

Documents

A rapid method for semiquantitative analysis of the human Vβ-repertoire using TaqManR PCR