High-resolution analysis of the human T-cell receptor repertoire

Abstract

Unbiased dissection of T-cell receptor (TCR) repertoire diversity at the nucleotide level could provide important insights into human immunity. Here we show that TCR ligation-anchored-magnetically captured PCR (TCR-LA-MC PCR) identifies TCR α- and β-chain diversity without sequence-associated or quantitative restrictions in healthy and diseased conditions. TCR-LA-MC PCR identifies convergent recombination events, classifies different stages of cutaneous T-cell lymphoma in vivo and demonstrates TCR reactivation after in vitro cytomegalovirus stimulation. TCR-LA-MC PCR allows ultra-deep data access to both physiological TCR diversity and mechanisms influencing clonality in all clinical settings with restricted or distorted TCR repertoires.

ARTICLE

Received 8 Jul 2014 | Accepted 15 Jul 2015 | Published 1 Sep 2015

DOI: 10.1038/ncomms9081 OPEN

High-resolution analysis of the human T-cell receptor repertoire

Eliana Ruggiero1,*, Jan P. Nicolay2,3,*, Raffaele Fronza1, Anne Arens1, Anna Paruzynski1, Ali Nowrouzi1, Gke renden1, Christina Lulay1, Sven Schneider4, Sergij Goerdt3, Hanno Glimm1, Peter H. Krammer2, Manfred Schmidt1,** & Christof von Kalle1,**

Unbiased dissection of T-cell receptor (TCR) repertoire diversity at the nucleotide level could provide important insights into human immunity. Here we show that TCR ligation-anchored-magnetically captured PCR (TCR-LA-MC PCR) identies TCR a- and b-chain diversity without sequence-associated or quantitative restrictions in healthy and diseased conditions.

TCR-LA-MC PCR identies convergent recombination events, classies different stages of cutaneous T-cell lymphoma in vivo and demonstrates TCR reactivation after in vitro cytomegalovirus stimulation. TCR-LA-MC PCR allows ultra-deep data access to both physiological TCR diversity and mechanisms inuencing clonality in all clinical settings with restricted or distorted TCR repertoires.

1 Department of Translational Oncology, National Center for Tumor Diseases and German Cancer Research Center, Im Neuenheimer Feld 280, 69120 Heidelberg, Germany. 2 Division of Immunogenetics, German Cancer Research Center, Im Neuenheimer Feld 280, 69120 Heidelberg, Germany. 3 Department of Dermatology, Venereology and Allergology, University Medical Center, Theodor-Kutzer-Ufer 1-3, 68167 Mannheim, Germany. 4 Institute for Clinical Chemistry, University Medical Center, Ruprecht-Karls-University of Heidelberg, Theodor-Kutzer-Ufer 1-3, 68167 Mannheim, Germany. * These authors contributed equally to this work. ** These authors jointly supervised this work. Correspondence and requests for materials should be addressed to C.v.K. (email: mailto:[email protected]

Web End [email protected] ).

NATURE COMMUNICATIONS | 6:8081 | DOI: 10.1038/ncomms9081 | http://www.nature.com/naturecommunications

Web End =www.nature.com/naturecommunications 1

& 2015 Macmillan Publishers Limited. All rights reserved.

ARTICLE NATURE COMMUNICATIONS | DOI: 10.1038/ncomms9081

The naive TCR repertoire is assembled in the thymus and formed by somatic rearrangements of non-contiguous genes belonging to the variable (V), diversity (D; only for

the b- and d-chain) and joining (J) families (combinatorial diversity). These are adjacent to the constant (C) gene in the heterodimeric ab or gd TCR. Further random insertion and deletion of nucleotides at the rearrangement positions create junctional diversity of the highly variable complementarity-determining region 3 (CDR3 region). The CDR3 region is unique for every T-cell clone and encodes the receptor portion that makes the majority of TCR contacts with antigenic peptides bound by the major histocompatibility complex. The frequency of a specic CDR3 sequence indicates the abundance of its T-cell clone. Comprehensive and unbiased TCR deep sequencing could provide a representative and quantitative repertoire analysis of specic cellular immunity. Outside of rapid amplication of cDNA ends (RACE) PCR15, current DNA-based TCR-sequencing technologies require extensive primer multiplexing69 and are almost exclusively focused on the b-chain repertoire. To obtain the highest possible resolution at high specicity, we have made use of previously established PCR-based methods: ligation-anchored (LA) PCR10,11 and non-restrictive linear amplication-mediated PCR1214, and developed TCR-LA-MC PCR as a new sequencing technology for the immunogenetic characterization of normal and reduced clonality in the a- and b-chain TCR repertoire. Here we show the identication of the events leading to the generation of the TCR diversity in healthy donors (HDs), discrimination of different stages of cutaneous T-cell lymphoma in vivo and monitoring of the reactivation of specic TCR clones after in vitro stimulation with peptide pools of two immunodominant cytomegalovirus (CMV) antigens, providing high-resolution characterization of TCR clonality repertoires.

ResultsT-cell receptor direct sequencing. TCR transcripts are composed of the known C gene and an adjacent unknown anking region representing the rearranged V(D)J genes. A comparable situation, an unknown sequence adjacent to a known one, is also given in the context of viral vector integration site analyses in gene therapy, where the vector edges represent the known sequence. TCR-LA-MC PCR starts with TCR-specic cDNA synthesis using a C gene-specic biotinylated primer. The resulting cDNA strands are immobilized by streptavidin and a single-stranded linker cassette is ligated to their 30 end to allow subsequent exponential amplication and deep sequencing (Fig. 1).

Using peripheral blood mononuclear cells (PBMC) and sorted T-cell fractions, we rst compared available TCR-sequencing methods (Supplementary Table 1ae). We further assessed the sensitivity and applicability of our approach. The results showed that even 10 ng of cDNA still provides a reliable representation of the existing TCR repertoire diversity (Supplementary Fig. 1ad). Furthermore, TCR-LA-MC PCR allows T-cell identication with a resolution capacity of at least 1: 10,000 and down to the single-cell level (Supplementary Table 2a,b) and readily identies even rare cells such as invariant natural killer T (iNKT) cells and mucosal associated invariant T-cells15 (Supplementary Table 2c).

Dissection of the TCR repertoire in healthy individuals. To visualize the TCR diversity in HDs with TCR-LA-MC PCR, we rst analysed the TCR repertoire in six HDs. The 454 sequencing platform was used for HDs 1, 2 and 3 and the MiSeq platform for HDs 4, 5 and 6 (Supplementary Table 3). TCR-LA-MC PCR results were consistent with even highly variable sequence numbers, ranging from hundreds to hundreds of thousands (Supplementary Fig. 2ad). TCR-LA-MC PCR sequencing

allowed a detailed, unbiased representation of functional TCR gene families, pseudogenes and open reading frame regions16 (Supplementary Table 4). Preferential usage of particular a- and b-chain genes was similar between all HDs (Supplementary Fig. 3ad). A highly efcient retrieval of both a- and b-chain rearrangements of all donors was obtained, revealing the existence of nonrandom, over- as well as underrepresented, VJ pairings (Supplementary Fig. 4a,b). Preferred CDR3 lengths were observed in both chains (Supplementary Fig. 5a,b). Interestingly, we found that a similar G (guanine) preponderance, known for the G-rich composition of the D gene17,18 in b-chain rearrangements, is also present in the centre of a-chain rearrangements, whose recombination does not involve a D gene (Supplementary Fig. 5c,d).

The presence of convergence is a strong indicator of positive selection pressure on particular TCR specicities. Convergent recombination describes the phenomenon that identical TCR specicities at the amino acid level can be obtained through different DNA codes or alternative molecular routes of recombination1921. We observed that up to 15.8% of the a-chain and up to 17.4% of the b-chain sequences in our data set encoded antigen specicities that were generated by convergent recombination (Supplementary Table 5). The importance of functional selection becomes exceedingly obvious in identical CDR3 amino acid clonotypes shared between different individuals, termed public clones19,22. Our results on human leukocyte antigen (HLA)-independent individuals

C

C

RNA

C

C

C

C

TCR cDNA

C

C

C

C

ssLC

Solid phase

C

ssLC

C

ssLC

C I LC I

C

ssLC

Solid phase

C

ssLC

C

ssLC

C II LC II

Constant (C) gene cDNA synthesis

RNA digestion

Magnetic capture + ssDNA linker ligation

First PCR

Second PCR

Sequencing adaptors addition

C

ssLC

C

ssLC

C

ssLC

Primer A

BC

C

LC

Primer B

Primer A

BC

C

LC

Primer B

Primer A

BC

C

LC

Primer B

Figure 1 | TCR-LA-MC PCR methodology. For TCR analysis, RNA was used as the starting material and gene-specic cDNA synthesis was performed using a biotinylated primer located in the constant (C) gene of the TCR chains. The resulting full-length cDNA fragments were magnetically captured on streptavidin beads and ligated to a known single-stranded linker cassette (ssLC). After exponential PCR amplication, sequencing adaptors were added and the PCR products were deep sequenced. C, constant gene; BC, sample-specic barcode.

2 NATURE COMMUNICATIONS | 6:8081 | DOI: 10.1038/ncomms9081 | http://www.nature.com/naturecommunications

Web End =www.nature.com/naturecommunications

& 2015 Macmillan Publishers Limited. All rights reserved.

NATURE COMMUNICATIONS | DOI: 10.1038/ncomms9081 ARTICLE

a b

CMV + donor 1 CMV + donor 2

d0

d9

d15

d0

d9

d15

CMV-enriched fraction CMV-enriched fraction Clones (%)

Relative seq count (%)

Clones (%) Relative seq count (%)

Relative seq count (%)

Clones (%) Relative seq count (%)

Clones (%) Relative seq count (%)

Clones (%) Relative seq count (%)

Clones (%) Clones (%)

CMV-enriched fraction

Sequence

Sequence

Sequence

-chain

-chain-chain

CADYYGQNFVF

CAVAFGNQFYF

CILEDQTGANNLFF

CALSEAHGSSNTGKLIF

CVVSESGGGNKLTF

CAANNNDMRF

0%

0.010.1%

0.11%

15%

510%

1020%

2030%

3040%

>40%

CAFMKPPLIQGAQKLVF

CAVISGGSYIPTF

CAFMKPPVIQGAQKLVF

CAVNNADSSYKLIF

CAFEAGNQFYF

CAVNIIGYALNF

CALTGAGSYQLTF

CAVRQGSSNTGKLIF

CAYRTPLRGSRLTF

CLVARGDSSYKLIF

CAVGGPQFSTISGGGNKLTF

CAEIYSGNTPLVF

CAYRTPLRDDKIIF

CGTVGNDMRF

d0

d9

d15

d0

d9

d15

Relative seq count (%)

Relative seq count (%)

Clones (%)

Relative seq count (%)

Clones (%)

Relative seq count (%)

Clones (%)

Relative seq count (%)

Clones (%)

Relative seq count (%)

Clones (%)

CMV-enriched fraction

Sequence

-chain

CASSFQGYTEAFF

CASSLSAGVPLEQYF

CATSESGTRGEKLFF

CASSRDTWSEQYF

CASTEGRDTQYF

CASSLEAENTQYF

CASNRLAGTSYNEQFF

CASSLEGEKEQFF

CASSSFGPSSYNEQFF

CASSLAEGSTEAFF

CSASRTAGSGNQPQHF

CASSEGTGANTEAFF

CASSGTSGSNNNEQFF

CASSLEQLLYNEQFF

CASTLSGSAYNEQFF

CASSPGSGGKNIQYF

CASSLGGSQDNEQFF

CSVEEGMMYEQYF

CASSIVNEQFF

CASRGGPGQPYNEQFF

Figure 2 | Immunomonitoring after CMV stimulation. TCR-LA-MC PCR was performed on RNA samples for two CMV-seropositive donors ((a) donor 1; (b) donor 2; upper panel, a-chain; lower panel, b-chain) at different time points after CMV stimulation. At day 15, an additional enrichment of the

IFNg-secreting fraction was performed. The 10 most predominant CMV-reactive TCR sequences identied at day 15 and their relative contribution to the total repertoire (sequence count) are shown. Retrieval frequency is indicated with a colour code ranging from light pink (0.010.1%) to black (440%).

We observed enrichment of dened TCR specicities over time. The number of T-cell clones (shown as %) contributing to the generation of those specicities also increased, indicating that a portion of the T-cell population is committed to the generation of specic TCR molecules. The CMV specicity of the analysed sequences was further conrmed by analysis of the enriched fraction. Two replicates were performed for day 0, 9 and 15 of each donor and sequencing results were combined. , sequence was identied in the CMV enriched fraction; aa, amino acid; CDR3, complementarity-determining region 3; d, day; IFN, interferon; Seq, sequence.

(Supplementary Table 6) showed that most inter-individually shared TCR amino acid sequences were encoded by different nucleotide sequences. In HDs 13 we identied 626 and 272 public amino acid sequences for the a- and the b-chains, respectively. We estimated the public clone frequency for our HDs 16, notably HLA-independent individuals, with not more than 5.8% a- and 2.2% b-chain nucleotide clonotypes in each donor (Supplementary Fig. 6ad). Furthermore, we screened our complete HD TCR sequence data set to identify known public clones and found TCR sequences reactive against inuenza, CMV and EpsteinBarr virus8,2327 (Supplementary Table 7).

TCR repertoire in an in vitro model of CMV infection. CMV infection persists for life in the lymphatic tissue in 6080% of the population. CMV is mainly controlled by CD4 and CD8 T cells28 directed against viral phosphoprotein 65 (pp65) and immediate early protein-1 (IE-1) (refs 2831).

To model vaccination and to study specically activated TCR clones over time, we stimulated T-cells from two healthy CMV-seropositive donors in vitro with a pool of 15-mer CMV

antigenic peptides spanning the full-length pp65 and IE-1 protein sequences. After stimulation, cells from both donors showed an increasing amount of IFNg-secretion in the CD8 fraction for donor 1 and in the CD4 fraction for donor 2 (Supplementary Table 8). TCR-LA-MC PCR deep sequencing showed a polyclonal TCR repertoire at day 0 and a substantially increased contribution of CMV-reactive sequences to the TCR repertoire at day 9 after the rst stimulation (Fig. 2). The intensity of convergence strongly increased after stimulation, pointing to a selection of specicities in both TCR chains. Between 48 and 169 CMV-reactive CDR3 amino acid sequences were found. These sequences showed an increasing or stable contribution over time to the total repertoire. Reactive TCR clonotypes present at day 15 could be readily backtracked on day 0 and day 9. Notably, between 0.1 and 0.4% of the T-cell clones at day 0 turned out to be reactive (Supplementary Fig. 7ad). These changes entailed a decrease in the diversity of the TCR repertoire (Supplementary Fig. 7e,f) and the preferential usage of individual V and J genes (Supplementary Fig. 7g,h). Statistical calculation showed that even the presence of three identical amino acid is unlikely by chance alone even in 6-mer sequences (P value o0.05%; see

NATURE COMMUNICATIONS | 6:8081 | DOI: 10.1038/ncomms9081 | http://www.nature.com/naturecommunications

Web End =www.nature.com/naturecommunications 3

& 2015 Macmillan Publishers Limited. All rights reserved.

ARTICLE NATURE COMMUNICATIONS | DOI: 10.1038/ncomms9081

Healthy donors Szary patients

100%

Retrieval frequency

DI (Shannon) CDR3

Szary cell count per l

Skin involvement

50%

0%

HD1

HD2

HD3

P1

P2

P3

P4

P5

P6

P7

P8

P9

P10

DI (Simpson) CDR3

87

3.2

346

3.6

1,664

3.5

1,568

3.8

552

3.3

411

3.4

1.4

0.5

1.5

0.5

2.6

0.7

1.6

0.6

17.9

1.6

23.4

1.8

1.3

0.5

1.2

0.3

2.0

0.8

5.1

1.2

45

1.9

8.4

1.7

49

2.1

223

2.8

62

2.2

162

2.7

1.4

0.6

2.3

1.0

1.9

0.8

11.5

1.6

/

/

/

/

/

/

2,195

Severe

324

Severe

15

Mild

4,882

Severe

75

Severe

730

Mild

171

Mild

380

Mild

752

Severe

817

Severe

Figure 3 | Contributions of CDR3 clonotypes in Szary patients. The gure shows the contributions of the CDR3 aa sequences to the ab TCR repertoire in HDs and in 10 Szary patients. Each bar represents an individual CDR3 aa clonotype, with red and violet indicating the rst and tenth most predominant sequences, respectively. Grey bars indicate the contribution of the remaining sequences identied in the analysed sample. TCR-LA-MC PCR results show that the stage of the disease correlated well with the observed TCR clonality. These results were conrmed by calculating SI and SA diversity indices (DI). The higher is the SA or the SI, the more diverse is the population. Two replicates were performed for each TCR chain of each patient and sequencing results were combined. aa, amino acid; CDR3, complementarity-determining region 3; HD, healthy donor; P, patient; SA, Shannon index; SI, Simpson index.

Methods), demonstrating functional selection after CMV stimulation of specic TCR clones. We also identied published public clones specic for CMV infection in our two CMV-seropositive donors (Supplementary Table 9).

TCR repertoire characterization in Szary syndrome. Repertoire information is even more relevant in pathological distortions of immunity. The pathogenesis of Szary syndrome, an aggressive form of cutaneous T-cell lymphoma characterized by a signicant number of malignant CD4 T cells circulating in the peripheral blood32, is still unknown and a lack of reliable clinical markers complicates clear diagnosis of the disease. We reasoned that TCR diversity analysis might readily dene the extent of the pathologically distorted T-cell repertoire in this disease33 in a more simple and objective manner. TCR-LA-MC PCR was performed on PBMC from 10 Szary patients at different disease stages, allowing unequivocal quantication of the degree of repertoire distortion in these patients (Fig. 3; Supplementary Table 10). Contributions of CDR3 clonotypes showed an oligoclonally distorted a- and b-chain repertoire with the emergence of a single predominant clone in patients 1, 2, 4, 5, 9 and 10, who were characterized by severe skin involvement. Patients 3, 6, 7 and 8, with mild skin involvement, showed a less-restricted clonal repertoire. Shannon (SA) and Simpson (SI) diversity indices of CDR3 amino acid for patients with a mild skin involvement indicated a less-reduced TCR repertoire, but still substantially different from a healthy situation. The clonal CDR3 sequence was different in every patient sample analysed. In several cases, the clonal V and J genes were found to be identical (Supplementary Table 11) and, interestingly, they are not commonly used in healthy individuals TCR.

DiscussionTCR-LA-MC PCR provides a detailed sequencing analysis of the clonal TCR repertoire. Cells, including bulk populations as well as sorted fractions (Supplementary Fig. 8af), blood and tissues34 from humans as well as from in vivo (Supplementary Table 12a,b,

Supplementary Fig. 9aj) and in vitro models can be analysed. Detailed, non-invasive monitoring of immune reactions can be obtained, enabling the discrimination of uniform and skewed TCR repertoires and its classication, even when only limiting amounts of material are available. TCR-LA-MC PCR is suitable for the identication of specic TCR clonotypes that initiate, maintain and progress infectious as well as acquired diseases.

Methods

Samples. To show the reproducibility of the results concerning the status of the immune repertoire constitution in different immunological conditions, we have performed TCR repertoire analysis on at least two biological human replicates. Samples from 6 healthy donors, 2 CMV-seropositive donors and 10 Sezary patients have been included in our analyses. We obtained buffy coats (B40 ml) from healthy unrelated donors and CMV healthy seropositive individuals from the blood bank in Heidelberg and peripheral blood from 10 Szary patients through the University Medical Center Mannheim. Each sample was obtained after informed consent of the donors. All the experiments with patient material are covered by the Medical Ethics Committee II of the Medical Faculty Mannheim of the Ruprecht-Karls-Universitat Heidelberg. For the Szary patients samples, a blinded analysis was performed: disease stage of the patients was assessed according to our TCR repertoire analysis results, then afterwards these results were compared with the clinical data/status of the patients.

After Ficoll Histopaque gradient, we cryopreserved the isolated PBMC.

Cell line. Jurkat E6.1 cells were used for the spike-in experiments. Cells were obtained from the American Type Culture Collection (ATCC) company (ATCC, TIB-152) and tested with a multiplex cell contamination test (performed in Heidelberg by the Multiplexion company), which was conducted for the following contaminations: mycoplasma, squirrel monkey retrovirus, human-, Macaca cynomolgus-, mouse-, rat-, Chinese hamster-, Syrian hamster-, feline-, canine-, rabbit-, guinea pig- and Drosophila cells. DNA quality was determined by an internal DNA quality control. Positive and negative controls were included to monitor PCR performance. Results have shown purity of the cell line and absence of contamination.

Mice. Spleen was taken from a 10- to 14-week-old female C57Bl/6J mouse (Jackson Laboratory, Bar Harbor, ME) raised and housed at a specic pathogen-free animal facility (DKFZ, Heidelberg) according to all applicable laws and regulations following approval by the responsible animal care (Regierungsprasidium Karlsruhe) and ethical committee.

4 NATURE COMMUNICATIONS | 6:8081 | DOI: 10.1038/ncomms9081 | http://www.nature.com/naturecommunications

Web End =www.nature.com/naturecommunications

& 2015 Macmillan Publishers Limited. All rights reserved.

NATURE COMMUNICATIONS | DOI: 10.1038/ncomms9081 ARTICLE

Single-cell sorting. We stained PBMC with uorescein isothiocyanate (FITC)-labelled antibody to CD3 (Beckton Dickinson, Clone HIT3a, Catalogue Number 555339, 1:100) for 30 min at room temperature. We sorted single cells with a BD FACS Aria cell sorter in a 96-well plate with a V-bottom containing 6.4 ml of water and 1.6 ml of reverse transcription buffer (Life Technologies).

RNA isolation and cDNA synthesis. We isolated RNA from the PBMC of HDs as well as CMV-seropositive donors with the RNeasy Mini (or Micro) kit (Qiagen). For iNKT cells the Arcturus PicoPure RNA Isolation kit (Life Technologies) was used. For the Szary patients we used the RiboPure blood kit (Life Technologies) to isolate RNA from whole blood. For RNA isolation from murine spleen, the organ was rst homogenized using a tissue homogenizer (Qiagen) and then processed with the RNeasy Mini kit (Qiagen). Afterwards, we performed DNA digestion with the Turbo DNA-free kit (Life Technologies). We carried out cDNA synthesis following the manufacturers instructions using the Superscript II enzyme (Life Technologies) and biotinylated gene-specic primers for the constant gene of the TCR (Supplementary Data 1).

After cDNA synthesis, we carried out RNase H (Life Technologies) treatment and puried the resulting product with the QIAquick PCR purication kit (Qiagen).

cDNA synthesis on single cells. After sorting single T-cells, we heated the 96-well plate at 65 C for 2 min followed by 5 min on ice and performed cDNA synthesis using the Superscript II protocol.

TCR ligation-anchored-magnetically captured PCR. Biotinylated fragments originated during the cDNA synthesis were captured on streptavidin magnetic beads (Life Technologies). For each sample, 200 mg magnetic beads were exposed for 60 s to a magnetic eld, washed twice with PBS/0.1% BSA, once with 3 M LiCl and nally resuspended in 6 M LiCl (puried cDNA volume:beads volume 1:1).

Samples were incubated on a shaker at 300 r.p.m. After an overnight enrichment of the biotinylated cDNA molecules, samples were washed with water to remove the unbound unspecic fragments and then resuspended in 10 ml of the ligation mastermix. We used 0.1 ng1 mg (1 mg for healthy donors and CMV samples, 200 ng for Szary samples, everything for the enriched CMV fraction, the spike-in experiment and for the sorted fractions, 1 mg0.1 ng for the limiting dilution experiment) of the captured cDNA molecules as a template for the ligation procedure. We ligated a single-stranded DNA linker phosphorylated at its 50 end and carrying a dideoxycytidine at its 30 end (single-stranded linker cassette (ssLC)) to the unknown genomic end of the amplicons. Ligation was carried out using the CircLigase enzyme (Epicentre) at 60 C for 1 h in a total volume of 10 ml.

At the end of the reaction, samples were washed with water, resuspended in 10 ml water and 1 ml was further processed. To amplify the CDR3 region of the TCR molecules, we performed two exponential PCR (50 ml total volume) using nested primers for the constant gene of the TCR chains and for the ssLC (Supplementary Data 1). Each primer was used at a concentration of 8.4 pmol. We used the following PCR programme: initial denaturation at 95 C for 120 s followed by10 cycles (95 C for 45 s, 55 C for 45 s and 72 C for 20 s) and an additional denaturation at 90 C for 120 s followed by 25 cycles (90 C for 45 s, 55 C for 45 s and 72 C for 20 s). A biotinylated primer for the constant gene of the TCR molecule was used in the rst exponential PCR. Between rst and second exponential PCR an additional magnetic capture was performed to increase the specicity of the reaction by removing all the unwanted fragments. After washing as described above, 20 ml beads in 6 M LiCl were mixed with 20 ml of PCR product (1:1 ratio) and the samples were incubated for at least 1 h on a shaker at 300 r.p.m. At the end of the incubation, samples were washed with water to remove the unspecic fragments that have not been bound to the beads, resuspended in 10 ml0.1 N NaOH and incubated 10 min on a shaker at 300 r.p.m. Denaturation with NaOH was performed to obtain the non-biotinylated strands. A volume of 2 ml of the denatured product was used for the second exponential PCR. We used the same PCR programme of the rst exponential PCR with the addition of a nal extension step at 72 C for 120 s. Afterwards, we puried the PCR amplicons using AmPure beads (Beckman Coulter) and prepared them for sequencing. Repetitive TCR-LAMC PCR analyses were performed for healthy donors, CMV donors, Szary patients and for the spike-in experiment.

High-throughput sequencing of TCR-LA-MC PCR amplicons. We used the 454 GS Flx Titanium Platform for long amplicon sequencing. To add specic GS Titanium adaptors (A and B) to both ends of the PCR amplicons, we performed Fusion primer PCR. To allow simultaneous analysis of multiple samples, we incorporated barcodes of 6 or 10 bp into the fusion primer A-TCR. We amplied 40 ng of puried PCR products with the following programme: initial denaturation for 120 s at 95 C, followed by 12 cycles (95 C for 45 s, 60 C for 45 s and 72 C for 60 s) and nal elongation for 300 s at 72 C. We puried PCR amplicons with AmPure beads (Beckman Coulter) and sequenced them. With recent improvements in the Illumina sequencing technology, the sequencing of long fragments (4250 bp from each side) has become possible allowing detection of the CDR3 region without the need to fragment the sequences. Thus, the MiSeq platform was used for the sequencing of iNKT cells, T-cell subsets, HDs 4, 5 and 6 and limiting dilution on Szary samples. As described for sequencing with the 454 platform, a

Fusion primer PCR was performed with 50 ng of puried amplicon product and special fusion primers harbouring Illumina MiSeq sequencing adaptors. A double barcoded strategy was used allowing the tagging of individual samples with a barcoded (12 bp) linker cassette in addition to the barcoded fusion primer. Fusion primers used are listed in Supplementary Data 1. Each primer was used at a concentration of 5 pmol.

Comparison RACE PCR and TCR-LA-MC PCR. Different methodologies have been tested for the sequencing of the TCR repertoire. We have performed several variants of a multiplex PCR35 approach both on DNA and RNA (MP1, Multiplex PCR; MP2, Multiplex PCR followed by a nested Multiplex PCR; MP3, Multiplex PCR followed by a nested Multiplex PCR (reactions performed with the Pfx polymerase); MP4, Multiplex PCR with biotinylated J primers followed by magnetic capture and by a nested Multiplex PCR; MP5, Multiplex PCR followed by Microcon30 treatment and by a nested Multiplex PCR; MP6, Linear amplication with biotinylated J primers followed by a Multiplex PCR with biotinylated primers, magnetic capture and a nested Multiplex PCR), a commercially available kit (TCR Express Human TCR Vb Repertoire Clonality Detecting System (BioMed

Immunotech)) on RNA and different variants2,5 of RACE PCR on RNA. In the case of the different variants of the multiplex PCR and for the amplicons derived from the TCR Express Kit, we added by ligation two double-stranded fragments harbouring known sequences to the PCR amplicons. Sequencing adaptors were included by Fusion Primer PCR. RACE PCR2 amplicons were prepared for sequencing as described for the TCR-LA-MC PCR.

As the focus of the paper is the use of PCR-based approaches devoid of multiplexing primers for the TCR sequencing of expressed molecules, after an initial test we have concentrated our work on TCR-LA-MC PCR and a recently published adaptation of the conventional RACE PCR5.

Three experiments were performed to compare RACE PCR5 and TCR-LA-MC PCR efciency as described in Supplementary Table 1be. RNA has been isolated from PBMC of two healthy donors starting from 3 107 cells for donor 1 (labelled

as PBMC1) and 5 107 for donor 2 (labelled as PBMC2). A unit of 18 mg of RNA

was retrieved for PBMC1 and 12 mg for PBMC2. A unit of 150 ng RNA was used for each sample condition. In the rst experiment, TCR-LA-MC PCR has been performed either by using Superscipt II (Life Technologies) as described in the manuscript or SmartScribe (Clontech) reverse transcriptase. SmartScribe reverse transcriptase (Clontech) has been used for cDNA synthesis following the manufacturers instruction. In the second experiment, the inuence of magnetic capture introduction in the RACE PCR protocol has been evaluated. In the third experiment, primers used in the RACE PCR protocol have been introduced in the TCR-LA-MC PCR procedure.

The respective PCR (50 ml total volume) was prepared with 2 Pfx

amplication buffer, 0.3 mM dNTP, 10 pmol of each primer and 1 unit of Platinum Pfx DNA polymerase (Life Technologies). First PCR amplication was carried out as follows: initial denaturation at 94 C for 2 min, 18 cycles of 94 C for 30 s, 55 C for 30 s, 68 C for 45 s and a nal elongation at 68 C for 5 min. For the second exponential PCR 27 cycles were used for both methodologies. Sequencing adaptors were directly added in the second PCR and amplicons were puried using the AmPure beads (Beckman Coulter). Samples were sequenced with MiSeq paired-end 250 bp and the MiTCR software36 was used for the analysis of the sequencing results.

RACE PCR protocol was performed as described in Bolotin et al.5 First-strand cDNA synthesis was performed using the SmartScribe enzyme in the presence of a constant gene-specic primer and of a template-switching primer added during the reaction. For the rst and second exponential PCR, 1 ml undiluted rst-strand cDNA and 1 ml rst PCR product were used as input material, respectively.

A volume of 15 pmol of each primer was used in the rst exponential PCR. Second exponential PCR primers have been generated by fusing the adaptor sequence required for the Illumina platform with the MmeStep1 and MmeSmart20 primers5. A volume of 7.5 pmol of MiSMmeStep1 and 0.75 pmol of MiSMmeSmart20 were used. TCR-specic primers were prepared accordingly as described in the main text for the TCR-LA-MC PCR and used at a concentration of 7.5 pmol.

When magnetic capture was introduced in the RACE PCR protocol to capture TCR-specic cDNA fragments, a biotinylated BC1R C gene-specic primer was used during the reverse transcription and afterwards TCR-specic cDNA molecules were captured with 200 mg pre-washed streptavidin magnetic beads (cDNA volume:beads volume 1:1). Samples were incubated overnight on a

shaker at 300 r.p.m. The day after, samples were washed and resuspended in 10 ml water. A volume of 1 ml was then used for the rst exponential PCR. When magnetic capture was performed both after cDNA and rst exponential PCR or only after the rst expo, a biotinylated C gene-specic BC2R primer was used in the PCR. Then 20 ml of the biotinlyted PCR products were captured with 200 mg pre-washed streptavidin magnetic beads and incubated for at least 1 h on a shaker at 300 r.p.m. Samples were washed with water to remove the unspecic fragments that have not been bound to the beads, resuspended in 10 ml 0.1 N NaOH and incubated 10 min on a shaker at 300 r.p.m. Then 1 ml of the denatured product was used for the second exponential PCR.

TCR-LA-MC PCR was performed as described in the manuscript with the following differences: (1) cDNA synthesis was perfomed using either Superscript II or Smartscribe reverse transcriptase; (2) between rst and second exponential PCR

NATURE COMMUNICATIONS | 6:8081 | DOI: 10.1038/ncomms9081 | http://www.nature.com/naturecommunications

Web End =www.nature.com/naturecommunications 5

& 2015 Macmillan Publishers Limited. All rights reserved.

ARTICLE NATURE COMMUNICATIONS | DOI: 10.1038/ncomms9081

no magnetic capture was performed; instead 1 ml of rst PCR product was directly used as input for the second exponential PCR; and (3) adaptor sequences were directly added in the second exponential PCR.

In addition, when RACE PCR primers were used in the TCR-LA-MC PCR protocol, BC1R biotinylated primer was used for the cDNA synthesis and BC2R biotinylated primer was used in the rst exponential PCR. Furthermore,a different linker cassette was ligated during the single-stranded ligation (PHO-LC2-DDC) allowing then the usage of the RACE primers M1 in the rst exponential PCR and MiSMmeStep1 as well as MiSMmeSmart20 in the second exponential PCR.

Antigens. We used a pool of 15-mer peptides with 11 amino acid overlap spanning the full CMV pp65 and IE-1 proteins (Peptivator, Miltenyi Biotec).

Cell stimulations with antigenic peptides. We thawed cryopreserved PBMC and incubated them overnight in RPMI (Gibco) supplemented with 5% human serum (Sigma-Aldrich), 100 U ml 1 penicillin/streptomycin (Lonza) and 2 mM

L-glutamine (Lonza).

To measure the number of CMV-specic T cells in the isolated PBMC, we stimulated cells with a nal concentration of 1 mg each peptide per ml. After 6 h of incubation at 37 C and 5% CO2, we diluted the cells to a concentration of 2 million per ml and added 30 U ml 1 human recombinant IL-2. We refreshed the cell medium and IL-2 every 23 days and added IL-7 (10 ng ml 1) once a week. After 9 days of culture, we performed restimulation of the CMV-specic cells. We loaded autologous irradiated (50 Gy) PBMC for 4 h with the peptide pools and used them as feeder cells (stimulator:effector ratio of 1:3). We kept the cells in culture for additional 6 days. At days 0, 9 and 15, we used an aliquot of cells for IFNg FACS staining and an additional aliquot for TCR repertoire analysis.

Flow cytometric analysis. We performed phenotypic characterization of the stimulated PBMC by using directly conjugated antibodies (Beckton Dickinson). After 2 h of incubation of the PBMC with the peptide pools, we added Brefeldin A (BFA) at a concentration of 7.5 mg ml 1 and incubated the cells for additional 4 h.

We washed the cells with PBS supplemented with 0.1% BSA (w/v) containing1 mg ml 1 BFA (PBS/BSA/BFA) and stained them with an Alexa 700-labelled antibody to CD4 (Beckton Dickinson, Clone RPA-T4, Catalogue Number 557922, 1:50) and an allophycocyanin (APC)-labelled antibody to CD8 (Beckton Dickinson, Clone RPA-T8, Catalogue Number 555369, 1:50) in a total volume of 100 ml for 30 min at 4 C. We then washed the cells in PBS/BSA/BFA, xed them with PBS containing 4% (w/v) paraformaldehyde for 20 min at room temperature and permeabilized them with PBS with 1% BSA (w/v) and 0.1% saponin (w/v) (Sigma-Aldrich) (perm buffer) at room temperature for 10 min. Afterwards, we washed the cells and incubated them for intracellular staining with a FITC-labelled antibody to IFNg (Beckton Dickinson, Clone B27, Catalogue Number 554700, 1:50) in 100 ml of perm buffer for 30 min at room temperature. We acquired uorescent events with a LSRII ow cytometer (Beckton Dickinson) and performed lymphocyte gating to determine the T-cell population. Sorting of T-cell subsets was performed using directly conjugated antibodies (Beckton Dickinson) for membrane staining of CD3 (Beckton Dickinson, Clone HIT3a, Catalogue Number 555339, 1:100), CD4 (Beckton Dickinson, Clone RPA-T4, Catalogue Number 557922, 1:50),CD8 (Beckton Dickinson, Clone RPA-T8, Catalogue Number 555369, 1:50) and CD45RO (Beckton Dickinson, Clone UCHL1, Catalogue Number 555493, 1:100) molecules. The stained cells were then sorted using the BD FACS Aria (Beckton Dickinson).

Enrichment of IFNc-secreting cells. To isolate IFNg-secreting cells,we stimulated the cells with CMV pp65 and IE-1 pools for 4 h on day 15 as previously described. We then collected the cells, washed them with PBS and performed staining and enrichment with the IFNg capture assay (Miltenyi Biotec).

We rst added an IFNg-catch reagent to label the cells, incubated them at 37 C for 45 min and then stained them with a phycoerythrin (PE)-labelled antibody to the IFNg. We enriched for IFNg-secreting cells by using anti-PE microbeads and separated them using the MACS system (Miltenyi Biotec). We stored the isolated cells for TCR repertoire analysis.

Isolation of iNKT cells. iNKT cells represent a very rare population of T cells, being present at a frequency ranging between 0.01 and 1% (ref. 15). First, we performed magnetic separation using anti-iNKT microbeads (Miltenyi Biotec) followed by staining of the labelled cells. A total of 50 million frozen PBMC was used as starting material. After centrifugation, the cells were resuspended in 400 ml of cold buffer solution (PBS, pH 7.2, 0.5% BSA, 2 mM EDTA) and 100 ml anti-iNKT microbeads were added. Samples were incubated for 15 min in the fridge. At the end of the incubation time, 50 ml of Anti-iNKT-PE antibody (Miltenyi

Biotec, Clone 6B11, Catalogue Number 130098128) were used for the staining and the cells were kept in the fridge for 5 min. After a washing step, magnetic separation using a MACS separator (Miltenyi Biotec) and MACS columns (Miltenyi Biotec) was carried out. To enrich the purity of the separated fraction, an additional separation was performed over a second column and cells were stained

with Fluorogold (Life Technologies) to label dead cells. Second, FACS sorting was performed, retrieving 10,215 cells.

Bioinformatic analysis of TCR amplicons. After removal of the introduced linker cassette, we aligned the trimmed sequences via BLAT37 (-stepSize 5 -minIdentity 0

-minScore 0) to a reference set of TCR genes. We obtained these gene data sets for

mouse and human from the IMGT/GeneDB database38. If we identied the J gene as well as the V gene for a given sequence in the amplicon via BLAT, we further analysed its CDR3 junction. We selected the terminal conserved cysteine in the V gene and the phenylalanine of the mainly conserved FGXG motif in the J gene as junction boundaries in accordance with multiple amino acid sequence alignments of corresponding alleles obtained from IMGT.

For the b-chain we identied the D gene within the junction nucleotides via pairwise sequence alignment39,40. We reported a D gene name if for the relatively short D sequences an alignment of 10 or more nucleotides with Z90% sequence identity or of 3 or more nucleotides with 100% sequence identity between the reference sequence and a part of the junction could be veried. We identied and reported additional nucleotides (N nucleotides) inserted between the D and J genes, V and D genes or V and J genes (a-chain). The source code of our own TCR sequence analysis bioinformatical tool may be available on request.

For some experiments, analysis of the TCR-sequencing results was performed by using the MiTCR software36.

Error estimation. Being aware of the errors deriving from PCR amplication and sequencing, we performed an estimation of the potential error rate by comparing the sequenced V gene portion of the CDR3 region with its germline sequence. We performed a sequence alignment using BLAST and carried out the calculation of the error rate by considering sequences harbouring one or more mismatches. The results showed that the error rate was B2% per nucleotide.

Statistics to identify identical amino acids in a sequence. Under the assumption that the amino acids are equiprobable (p 1/20) and independent we

calculated the probability of nding subsequences of m amino acids in sequences of length k using the following formula

Pk;m k m 1

pm: 1

derived from the elementary rules of probability.

Heat-map. We used log-odd scores as non trivial representations of the abundance of VJ pairs. Log-odds can be written as lV,J log2(pV,J/pVpJ). pV,J is the frequency

that gene V is fused with gene J and pV and pJ are the background uniform probability with which the V and J genes occur in the samples. We expressed the resulting lV,J score in bit due to the logarithm base.

CMV prole selection. From the total sequences retrieved, we selected the amino acid sequences consistent with an expansion prole. We dened as expansion a prole where the contribution of a sequence s between day 0 and day 9 is increasing and between day 9 and day 15 is at least stable. The latter condition was due to a saturation phenomenon that could affect the frequencies of strongly activated species. We computed the distribution of the negative differences between day 15 and day 9 and then considered all those species where this difference lays under the 25th percentile as stable.

Diversity indices. To estimate the diversity in the TCR repertoire, we calculated two different diversity indices: Shannon index (SA)41 and Simpson index (SI)42. Both of them take in consideration the two components that constitute the concept of diversity, the richness of a population and its homogeneity. The richnessof a population is dened by its total number of species (for example, CDR3 amino acid sequences) and is estimated with the Shannon index as follows:

SA Xi ni=N

lg ni=N

: 2 Homogeneity measures the distribution of the species (for example, CDR3 amino-acid sequences) in the population and is calculated with the Simpson index as follows:

SI 1 = Xi ni=N

2: 3

where i is an index that is chosen between 1 and the number of species s, ni is the number of sequencing reads in species i and N is the total number of reads.

References

1. Douek, D. C. et al. A novel approach to the analysis of specicity, clonality, and frequency of HIV-specic T cell responses reveals a potential mechanism for control of viral escape. J. Immunol. 168, 30993104 (2002).

2. Freeman, J. D., Warren, R. L., Webb, J. R., Nelson, B. H. & Holt, R. A. Proling the T-cell receptor beta-chain repertoire by massively parallel sequencing. Genome Res. 19, 18171824 (2009).

6 NATURE COMMUNICATIONS | 6:8081 | DOI: 10.1038/ncomms9081 | http://www.nature.com/naturecommunications

Web End =www.nature.com/naturecommunications

& 2015 Macmillan Publishers Limited. All rights reserved.

NATURE COMMUNICATIONS | DOI: 10.1038/ncomms9081 ARTICLE

3. Warren, R. L. et al. Exhaustive T-cell repertoire sequencing of human peripheral blood samples reveals signatures of antigen selection and a directly measured repertoire size of at least 1 million clonotypes. Genome Res. 21, 790797 (2011).

4. Quigley, M. F., Almeida, J. R., Price, D. A. & Douek, D. C. Unbiased molecular analysis of T cell receptor expression using template-switch anchored RT-PCR. Curr. Protoc. Immunol. Chapter 10, Unit10: 33. doi: 10.1002/0471142735.im1033s94 (2011).

5. Bolotin, D. A. et al. Next generation sequencing for TCR repertoire proling: platform-specic features and correction algorithms. Eur. J. Immunol. 42, 30733083 (2012).

6. Wang, C. et al. High throughput sequencing reveals a complex pattern of dynamic interrelationships among human T cell subsets. Proc. Natl Acad. Sci. USA 107, 15181523 (2010).

7. Robins, H. S. et al. Comprehensive assessment of T-cell receptor beta-chain diversity in alphabeta T cells. Blood 114, 40994107 (2009).

8. Robins, H. S. et al. Overlap and effective size of the human CD8 T cell

receptor repertoire. Sci. Transl. Med. 2, 47ra64 (2010).9. Klarenbeek, P. L. et al. Human T-cell memory consists mainly of unexpanded clones. Immunol. Lett. 133, 4248 (2010).

10. Troutt, A. B., McHeyzer-Williams, M. G., Pulendran, B. & Nossal, G. J. Ligation-anchored PCR: a simple amplication technique with single-sided specicity. Proc. Natl Acad. Sci. USA 89, 98239825 (1992).

11. Argaet, V. P. et al. Dominant selection of an invariant T cell antigen receptor in response to persistent infection by Epstein-Barr virus. J. Exp. Med. 180, 23352340 (1994).

12. Schmidt, M. et al. High-resolution insertion-site analysis by linear amplication-mediated PCR (LAM-PCR). Nat. Methods 4, 10511057 (2007).

13. Paruzynski, A. et al. Genome-wide high-throughput integrome analyses by nrLAM-PCR and next-generation sequencing. Nat. Protoc. 5, 13791395 (2010).

14. Gabriel, R. et al. Comprehensive genomic access to vector integration in clinical gene therapy. Nat. Med. 15, 14311436 (2009).

15. Greenaway, H. Y. et al. NKT and MAIT invariant TCRa sequences can be produced efciently by VJ gene recombination. Immunobiology 218, 213224 (2013).

16. Yous Monod, M., Giudicelli, V., Chaume, D. & Lefranc, M. P. IMGT/JunctionAnalysis: the rst tool for the analysis of the immunoglobulin and T cell receptor complex V-J and V-D-J JUNCTIONs. Bioinformatics 20(Suppl 1): i379i385 (2004).

17. Rosenberg, W. M., Moss, P. A. & Bell, J. I. Variation in human T cell receptor V beta and J beta repertoire: analysis using anchor polymerase chain reaction. Eur. J. Immunol. 22, 541549 (1992).

18. Gauss, G. H. & Lieber, M. R. Mechanistic constraints on diversity in human V(D)J recombination. Mol. Cell. Biol. 16, 258269 (1996).

19. Venturi, V., Price, D. A., Douek, D. C. & Davenport, M. P. The molecular basis for public T-cell responses? Nat. Rev. Immunol. 8, 231238 (2008).

20. Quigley, M. F. et al. Convergent recombination shapes the clonotypic landscape of the naive T-cell repertoire. Proc. Natl Acad. Sci. USA 107, 1941419419 (2010).

21. Venturi, V. et al. Sharing of T cell receptors in antigen-specic responses is driven by convergent recombination. Proc. Natl Acad. Sci. USA 103, 1869118696 (2006).

22. Li, H., Ye, C., Ji, G. & Han, J. Determinants of public T cell responses. Cell Res. 22, 3342 (2012).

23. Price, D. A. et al. Avidity for antigen shapes clonal dominance in CD8 T cell

populations specic for persistent DNA viruses. J. Exp. Med. 202, 13491361 (2005).24. Lehner, P. J. et al. Human HLA-A0201-restricted cytotoxic T lymphocyte recognition of inuenza A is dominated by T cells bearing the V beta 17 gene segment. J. Exp. Med. 181, 7991 (1995).

25. Trautmann, L. et al. Selection of T cell clones expressing high-afnity public TCRs within Human cytomegalovirus-specic CD8 T cell responses.J. Immunol. 175, 61236132 (2005).26. Miles, J. J. et al. Genetic and structural basis for selection of a ubiquitous T cell receptor deployed in Epstein-Barr virus infection. PLoS Pathog. 6, e1001198 (2010).

27. Moss, P. A. et al. Extensive conservation of alpha and beta chains of the human T-cell antigen receptor recognizing HLA-A2 and inuenza A matrix peptide. Proc. Natl Acad. Sci. USA 88, 89878990 (1991).

28. Sinclair, E. et al. CMV antigen-specic CD4 and CD8 T cell IFNgamma

expression and proliferation responses in healthy CMV-seropositive individuals. Viral Immunol. 17, 445454 (2004).29. Slezak, S. L. et al. CMV pp65 and IE-1 T cell epitopes recognized by healthy subjects. J. Transl. Med. 5, 17 (2007).

30. Kern, F. et al. Cytomegalovirus (CMV) phosphoprotein 65 makes a large contribution to shaping the T cell repertoire in CMV-exposed individuals.J. Infect. Dis. 185, 17091716 (2002).31. Wills, M. R. et al. The human cytotoxic T-lymphocyte (CTL) response to cytomegalovirus is dominated by structural protein pp65: frequency, specicity, and T-cell receptor usage of pp65-specic CTL. J. Virol. 70, 75697579 (1996).

32. Olsen, E. et al. Revisions to the staging and classication of mycosis fungoides and Sezary syndrome: a proposal of the International Society for Cutaneous Lymphomas (ISCL) and the cutaneous lymphoma task force of the European Organization of Research and Treatment of Cancer (EORTC). Blood 110, 17131722 (2007).

33. Yawalkar, N. et al. Profound loss of T-cell receptor repertoire complexity in cutaneous T-cell lymphoma. Blood 102, 40594066 (2003).

34. Martins, V. C. et al. Thymus-autonomous T cell development in the absence of progenitor import. J. Exp. Med. 209, 14091417 (2012).

35. Willenbrock, K. et al. Analysis of T-cell subpopulations in T-cell non-Hodgkins lymphoma of angioimmunoblastic lymphadenopathywith dysproteinemia type by single target gene amplication of T cell receptor- beta gene rearrangements. Am. J. Pathol. 158, 18511857 (2001).

36. Bolotin, D. A. et al. MiTCR: software for T-cell receptor sequencing data analysis. Nat. Methods 10, 813814 (2013).

37. Kent, W. J. BLATthe BLAST-like alignment tool. Genome Res. 12, 656664 (2002).

38. Lefranc, M. P. et al. IMGT, the international ImMunoGeneTics information system. Nucleic Acids Res. 33, D593D597 (2005).

39. Smith, T. F. & Waterman, M. S. Identication of common molecular subsequences. J. Mol. Biol. 147, 195197 (1981).

40. Gotoh, O. An improved algorithm for matching biological sequences. J. Mol. Biol. 162, 705708 (1982).

41. Shannon, C. E. The mathematical theory of communication. 1963. MD Comput. 14, 306317 (1997).

42. Simpson, E. H. Measurement of diversity. Nature 163, 688 (1949).

Acknowledgements

We gratefully acknowledge the clinical staff taking care of Szary patients. We thankS. Fessler, A. Mengering and S. Wenzel for technical assistance, U. Abel, A. Deichmann,T. Holland-Letz and A. Slynko for their help with the statistical analyses,G.P. Balasubramanian and B. Brors for the support with the bioinformatical analyses. Funding was provided by the Initiative and Networking Fund of the Helmholtz Association within the Helmholtz Alliance on Immunotherapy of Cancer and by the European Commissions 7th Framework Program (http://ec.europa.eu/research

Web End =http://ec.europa.eu/research), contract HEALTH-F2-2012-305380-DIABIL-2 and ERC-2012-ADG20120314 TRiPod. We thank the DKFZ-Heidelberg Center for Personalized Oncology (DKFZ-HIPO) for technical support and funding through HIPO_H007.

Author contributions

E.R., H.G., M.S. and C.v.K. developed the concept of TCR-LA-MC PCR, designed experiments and interpreted data; E.R., J.P.N., A.P., A.N., G., C.L. and S.S. performed experiments; A.A. and R.F. established the bioinformatic analysis of TCR-LA-MC PCR amplicon sequences; J.P.N. and S.G. diagnosed and treated the patients and evaluated the patient data; E.R., J.P.N., P.H.K., M.S. and C.v.K. prepared and wrote the manuscript.

Additional information

Accession codes: The TCR-sequencing data generated in this paper have been deposited in the SRA database under the accession code SRP059581.

Supplementary Information accompanies this paper at http://www.nature.com/naturecommunications

Web End =http://www.nature.com/ http://www.nature.com/naturecommunications

Web End =naturecommunications

Competing nancial interests: The authors declare no competing nancial interests.

Reprints and permission information is available online at http://npg.nature.com/reprintsandpermissions/

Web End =http://npg.nature.com/ http://npg.nature.com/reprintsandpermissions/

Web End =reprintsandpermissions/

How to cite this article: Ruggiero, E. et al. High-resolution analysis of the human T-cell receptor repertoire. Nat. Commun. 6:8081 doi: 10.1038/ncomms9081 (2015).

This work is licensed under a Creative Commons Attribution 4.0 International License. The images or other third party material in this article are included in the articles Creative Commons license, unless indicated otherwise in the credit line; if the material is not included under the Creative Commons license, users will need to obtain permission from the license holder to reproduce the material. To view a copy of this license, visit http://creativecommons.org/licenses/by/4.0/

Web End =http://creativecommons.org/licenses/by/4.0/

NATURE COMMUNICATIONS | 6:8081 | DOI: 10.1038/ncomms9081 | http://www.nature.com/naturecommunications

Web End =www.nature.com/naturecommunications 7

& 2015 Macmillan Publishers Limited. All rights reserved.