Immunosequencing: applications of immune repertoire deep sequencing

Immunosequencing: applications of immune repertoire deep sequencing

Available online at www.sciencedirect.com ScienceDirect Immunosequencing: applications of immune repertoire deep sequencing Harlan Robins Advances in...

225KB Sizes 4 Downloads 28 Views

Available online at www.sciencedirect.com

ScienceDirect Immunosequencing: applications of immune repertoire deep sequencing Harlan Robins Advances in high-throughput sequencing have enabled the development of a powerful new technology for probing the adaptive immune system. Millions of B or T cell receptor sequences can be read in parallel from a single sample. The dynamics of an adaptive immune response, which is based on clonal expansion and contraction, can be monitored in real time at high sensitivity and the global properties of the adaptive immune repertoires can be studied. A large set of clinical applications for this technology are presently under study, with a few diagnostic applications for hematological malignancies already available. A review of this new field termed immunosequencing is presented. Addresses Fred Hutchinson Cancer Research Center, Seattle, WA, USA Corresponding author: Robins, Harlan ([email protected])

Current Opinion in Immunology 2013, 25:646–652 This review comes from a themed issue on Immunogenetics and transplantation Edited by Miles Davenport and Deborah K Dunn-Walters For a complete overview see the Issue and the Editorial

receptor alpha (TCRA) for T cells rearranges similarly, but with V and J only. The set of B and T cells constituting the adaptive immune system are comprised of millions of different clones defined by their specific BCR or TCR sequence rearrangement. The nucleotide sequence of the BCRs and TCRs provide a nearly unique molecular tag for each clone in the adaptive immune system. Moreover, these sequences provide a primary piece of functional information for each clone, as the receptor structures determine their antigenic binding. The highly variable CDR3 regions in both BCRs and TCRs are short, between 15 and 60 nucleotides, making them amenable to rapid interrogation by high-throughput sequencing (HTS). However, methods designed to sequence large genomic regions do not efficiently apply to these short, highly diverse regions. A new field of immunosequencing has emerged with technologies specifically tailored to sequence BCRs and TCRs, along with a set of promising applications for these technologies [1–4]. As the adaptive immune system is believed to play a role in most, if not all, human disease states, the possible applications are expansive.

Available online 16th October 2013 0952-7915/$ – see front matter, # 2013 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.coi.2013.09.017

Introduction The human adaptive immune system provides protection against an enormous variety of pathogens. This protection is mediated by receptors on the surface of B and T cells that bind to pathogenic or pathogen derived antigens. The human germline genome is limited in size, so it cannot code for a sufficient number of receptor genes to protect against the diversity of potential pathogens. The vast receptor diversity needed for protection is created dynamically by somatic rearrangement of the germline DNA at specific loci in B and T cells. Both the B cell receptor (BCR) and T cell receptor (TCR) are formed from pairing of a larger chain and smaller chain. The BCR immunoglobulin heavy chain (IGH) and the TCR beta chain (TCRB) rearrange noncontiguous variable (V), diversity (D), and joining (J) gene segments to create combinatorial diversity. At the junctions between the V– D and D–J, nucleotides are deleted and pseudo-random non-templated nucleotides are added to create massive junctional diversity. The respective small chain, immunoglobulin lambda or kappa (IGL/K) for B cells and T cell Current Opinion in Immunology 2013, 25:646–652

In this review, we present the technical challenges to high throughput sequencing of adaptive immune receptors and the present set of solutions, including both experimental and computational issues. Additionally, we will classify immune sequencing applications into categories and describe progress to date in each area and speculate on future developments.

Challenges There are two primary challenges in HTS of adaptive immune receptors. The first is the somatic rearrangement of the loci. The rearranged BCR and TCR loci are structurally different than in the germline genome. Although there are some known rules that govern the rearrangements, the resultant genes are not minor changes from a known template. Second, the rearranged sequences are highly diverse. The number of clones with different TCRB rearrangements in the blood of a healthy human is estimated to be 1–5 million with total ab T cell diversity estimated to be 20–100 million [1,5]. The B cell repertoire diversity is currently unknown, with some lower bound estimates being much larger (>10) than that of the T cell repertoire. An additional complication in sequencing BCRs is somatic hypermutation. Unlike TCRs, BCRs can evolve to increase antigen binding affinity through a process of point mutation and selection. www.sciencedirect.com

Applications of immune repertoire deep sequencing Robins 647

Figure 1

Vβ1

Vβ2

Vβn

Dβn

Jβ1.1 Jβ1.2

Jβn

Combinatorial assortment of V, D, and J genes Vβn

TdT:

Dβn

3′ V deletion

5′ J deletion

N insertion V

Jβn

N insertion N

D N

J Current Opinion in Immunology

Somatic VDJ rearrangement of the T cell receptor b chain. Caption: The adaptive immune receptor gene somatically rearrange in a process called VDJ rearrangement. This is a schematic of a human TCRb gene locus. The TCRb chain is paired with a TCRa chain to form a functional TCR on the surface of an ab T cell. The top line is a cartoon of the germline DNA in this locus with 48 V segments, 2 D segments, and 13 J segments. One each of the V, D, and J segments are rearranged adjacently. In the V–D and D–J junctions, a variable number of nucleotides are deleted and then random nucleotides are inserted by the enzyme TdT. Although the total number of nucleotides in the NDN junction is small, the resultant number of possible combinations is enormous (estimated at over 1012 for just the TCRb chain).

For illustrative purposes, this review will present the sequencing challenges and solutions as applied to the TCRB locus, except when addressing B cell specific issues such as somatic hypermutation.

sequencing falls by at least three more logs). The most common strategy to enrich for the target sequences is PCR amplification. The major impediments to quantitative accuracy for sequencing gDNA are the diversity of the V and J genes and introns [1,3,6]. In order to capture all potential combinations, a multiplex pool of PCR primers is utilized, because the constant regions, shared by all rearrangements, are too distant from the junctional region for PCR amplification. To preserve quantitative accuracy, amplification differences between any possible pair of primers in the multiplex pool must be identified and corrected. A couple of strategies have been designed to solve this problem with at least one very effective. The difficulty in quantitative accuracy for TCR cDNA is primarily a biological, rather than an assay, limitation. Each T cell can express a different number of mRNA molecules, so precisely quantitating the number of mRNA molecules does not necessarily correspond to cell count. Certainly some T (or B) cell subsets have very different numbers of receptor mRNA molecules [7]. Not only does this issue make results difficult to interpret, but it also makes testing the assay challenging. A common test employed to assess sensitivity is to spike in known clones to a complex mix at precisely measured cell counts [8]. However, T cell lines generally have much higher TCR mRNA levels, rendering such a test flawed. On the other hand, there are a few published strategies for accurately quantitating the number of cDNA molecules, even if it is not proportional to clone counts [9]. Despite these quantitative limitations, some innovative experiments have made significant breakthroughs in many disparate fields.

Target enrichment solutions A schematic of the TCRB germline locus is displayed in Figure 1 along with an example of a TCRB somatic rearrangement. Complete knowledge of a clonal rearrangement requires identification of the V and J segments utilized and the full CDR3 sequence, which includes non-templated nucleotides not present in the genome. As Figure 1 illustrates, there is an intron between the J segment and the constant region, which adds a significant difficulty for sequencing genomic DNA (gDNA). Different strategies have been developed for sequencing gDNA versus cDNA that has been reverse transcribed from mRNA. The intron shown in Figure 1 creates a fundamental difference in the structure of the sequencing target. As the adaptive immune system functions under the principle of clonal expansion, a primary goal of sequencing is to not only determine the full, rearranged sequences of each clone in a sample, but to accurately determine the quantity of each clone. For quantitative accuracy, sequencing gDNA and cDNA have their own difficulties. In order to efficiently sequence the TCRB repertoire the targets must be enriched (until the price of www.sciencedirect.com

Genomic DNA

As discussed above in challenges, enriching for rearranged adaptive immune receptor gDNA is complicated due to the J segment specific intron between J and the constant region. Although some homology exists within the 13 J’s and within the 48 V’s, there is no shared sequence of length sufficient for binding of a universal degenerate primer. In order to specifically enrich for the full set of potential TCRB rearrangements, a multiplex PCR strategy has been employed, utilizing a mixture of V and J primers inclusive of all V and J segments [1,10,11]. Competing models predict binding affinity for a primer to its target. A valid starting point is a set of primers with minimized variation in binding affinity through modulation of length and nucleotide content. Unfortunately, even with the primers optimally matched by predicted binding affinity, large empirical differences are still found in amplification efficiency. Two different strategies have been utilized to overcome the residual amplification bias within the multiplex PCR. One effective strategy is the development of a complete set of synthetic TCRB molecules, where each input Current Opinion in Immunology 2013, 25:646–652

648 Immunogenetics and transplantation

target can be precisely measured [12]. This synthetic immune system is targeted with the multiplex set of primers, and the over or under amplifying primers are titrated down or up, respectively, to reduce amplification bias. This process is repeated iteratively until the amplification bias is largely removed. The residual bias is identified by applying the final primer set to the synthetic immune system targets and removed computationally. Fortunately, potential biologically relevant variations in length and GC content between the primer binding sites have negligible effect on amplification. So, this method removes the majority of the amplification bias in a full multiplex PCR. Another method to reduce PCR bias is the use of nested PCR [6]. A multiplex set of primers are designed with universal overhangs, so that each amplified molecule has a pair of universal sequences synthesized to the flanking regions of the amplified target. In this procedure, two different PCR steps are used. First, a few rounds of PCR is used with the target specific primers to add the universal flanking regions. Second, the primers to the universal flanking regions are used to amplify a large pool of targets. The theory is that the bulk of the amplification is done with universal primers, which should not introduce amplification bias. The primary problem with this approach is that a large potential amplification bias can be introduced in the first few cycles of the PCR reaction due to under amplification from lack of priming. mRNA enrichment

Quantitatively amplifying mRNA is far easier for the rearranged adaptive immune receptors, as the introns are removed by splicing. The mRNA transcripts have constant regions 30 to the J, which are included in the transcript. After the set of mRNAs in a sample are reverse transcribed into cDNA, two different strategies have been employed to amplify the TCRB locus. One strategy uses a multiplex PCR pool of V region primers and one or two degenerate primers to the constant region [3,13–16]. This strategy is subject to the same PCR amplification bias as the gDNA methods. The second strategy uses 50 rapid amplification of cDNA ends (RACE), which only requires the gene specific primer to the constant region. The resulting cDNA pool should have limited transcript specific bias, as the same PCR primer is used for all TCRB sequences [4]. Of course, this strategy is subject to additional PCR biases such as transcript length and GC content. However, these variations are far smaller than variation due to differential priming. The 50 RACE sequencing strategy has proven successful technically. However, there are two difficulties associated with sequencing from mRNA. The first is biological interpretation. Each T (or B) cell can produce differing abundance of mRNA for their receptor genes. These amounts can change dramatically for different activation Current Opinion in Immunology 2013, 25:646–652

states of the cell. Therefore, HTS of the mRNA repertoire does not accurately reflect the abundance of the corresponding clones. The second issue is strictly practical. Compared to gDNA, mRNA is more susceptible to degradation and difficult to accurately quantitate. The stability issue can be addressed with proper sample storage and handling. The difficulty in quantitation makes control of the input amount of cDNA templates for an amplification reaction less accurate. This is not catastrophic, just technically more difficult to standardize. Somatic hypermutation

BCRs from antigen experienced B cells pose a significant additional problem for target enrichment, somatic hypermutation. B cells migrate to germinal centers where they affinity mature by point mutation, primarily in the complementary determining regions (CDR), leading to evolution and selection for stronger binding to an antigen. These somatic mutations can occur anywhere within the BCR sequence, so it is not possible to a priori design primers that bind to unmutated regions within the receptor. However, there are sequences which are far more conserved than others. A few strategies have been employed to account for SHM while sequencing BCRs from mature B cells. One is to include multiple primer sets that bind to different locations for each V or J segment, with the goal to improve the odds of missing highly mutated regions [3,17]. The resulting sequencing reads can either be combined, effectively averaging the SHM effect, or the best amplifying selected for analysis. A second strategy is to design very long primers and anchor them at highly conserved nucleotide positions that cannot be mutated without losing function for the BCR [10]. The anchor positions (the last 4–5 nucleotides at the 30 end of the primer) are the most crucial for amplification efficiency. Only a few potential amino acids are coded by a single codon, so any mutation would change the amino acid. In particular, there is a tryptophan in the V gene that is highly conserved, and can be used as part of the anchor positions. Using a very long primer (30+ nucleotides) reduces the effect of single mismatches with the target. A third strategy is to use full length cDNA and design primers outside the V and J region, which are far less likely to be subject to SHM. None of these solutions is perfect. The major difficulty is designing a gold standard to test whether a given solution is working accurately. At best, we can model SHM and create templates that fit that model for testing. However, this is limited by the accuracy of the model. Another solution is to uniquely barcode each molecule at the earliest stage of amplification [9]. In this way, the amplification bias factor is irrelevant as the output is digital, either a particular sequence was detected with a specific barcode or it was not. In rare cases, SHM might prevent amplification altogether, which will be undetectable by any method, but we expect this to be the exception. www.sciencedirect.com

Applications of immune repertoire deep sequencing Robins 649

Sequencing At present, there are a few different options for HTS technologies, each of which offer advantages and drawbacks. A common feature is that each technology requires specific DNA sequences on both ends of the target molecules to be sequenced. These sequences are added either by synthesis using PCR or by ligation. The sample preparation steps differ significantly, but an evaluation is beyond the scope of this review. We give a short list of pros and cons for each of the three most common technologies utilized for immunosequencing. One instrument, the 454 sequencer (Roche) provides long reads which is amenable to capturing full-length IgH cDNA sequences including all somatic hypermutations. The negatives are high cost per read and high rate of insertions and deletions (indels) from homopolymers. The indels from homopolymers are a large problem for sequencing TCRs and BCRs because the D segments often contain homopolymer stretches of G nucleotides. The Ion Torrent sequencing technology (Life Technologies) suffers from the same drawback of high indel rates. However, the Ion technology is very fast and inexpensive. The MiSeq and HiSeq 2500 technology (Illumina) is also fast and inexpensive. Although, Illumina’s technology has much less issues with indels, their errors are position dependent (higher as read length increases). As sequencing of TCR and BCRs begin from similar nucleotide positions, the higher error rates are consistently in the same regions. Additionally, both the Ion and Illumina technology have read lengths that are presently too short to cross the full TCR or BCR genes. However, with the rate of improvement in both of these machines, we view this as a very temporary problem. In general, the outlook is very favorable, as sequencing technology is improving very rapidly, with longer, more accurate, and less expensive reads.

template by clustering. Although, historically this was very difficult, the improved accuracy of the sequencing technology has vastly simplified this problem. A large fraction of sequences are error free, which allows for easy identification of the primary nodes in each cluster, by simply identifying the sequences with multiple copies in the data [1,18]. For the case of larger clones, there is a probability that the same error occurs multiple times, either independently with the same sequencing errors, or dependently through PCR. So, an additional rule is added for clustering, which simply asserts that if two primary nodes are different by a single nucleotide and one node is far larger than the other, the smaller node is moved into the tree anchored by the larger primary node. Then, all additional sequences are placed into the tree structure based on distance to the primary nodes. In the corrected output, each primary node is the sequence of a unique clone, and the copy number, or relative abundance, is the number of sequence reads in the tree. The different sequencing technologies determine the choice of distance metric. Additionally, the software associated with the sequencing technologies report accuracy rates at each base, and this information can be incorporated into the clustering algorithm to improve accuracy. As with every step of the sequencing process, it is vital to experimentally validate the results. Each processing algorithm makes an assumption about the likely sources of error and their form, so these assumptions must be tested. This requires a known input library. Given a known starting set of templates, they can be run through the full protocol and the output can be compared back to input sequences. The starting template can either be a synthetic set of templates, or a well studied library.

Applications Data processing Both PCR amplification and HTS generate errors in the resultant sequences. The PCR errors can propagate, with the potential for errors to compound at each PCR cycle. Fortunately, these compound errors effect an exponentially smaller fraction of the total reads at each step. Effectively, the PCR errors can be modeled as a phylogenetic tree. Since these trees do not undergo selection, the number of elements strictly decreases along each branch. And, the PCR error rate is sufficiently small that the probability of multiple errors in the same molecule at a given cycle is negligible. So, each branch of the tree represents a single nucleotide change. The second major form of error is from the sequencing. These errors are highly machine specific. The primary strategy for correcting errors is the use of redundancy. As long as any particular type of error is rare, then sequencing multiple copies from the same original template allows accurate error correction using parsimony. The key is to identify which sequence reads originated from the same www.sciencedirect.com

The set of immunosequencing applications [19] to date can be divided into three categories: clone tracking, repertoire properties, and identification of public clones. Clone tracking is based on the observation that both T and B cell repertoires are diverse, with small probability of different clones sharing nucleotide identical TCRb or IGH sequences. Therefore, the immune receptor nucleotide sequence is a nearly unique molecule tag for a clone, allowing clone tracking over time and between tissues or cell subsets. Repertoire properties are global properties of the T or B cell clone distribution in a sample. This can be a measure of diversity, oligoclonality, entropy, etc. Finally, public clones are TCR or BCR amino acid sequences that are shared between different individuals as a common response to the same antigen. For T cells, public responses are primarily specific to a shared HLA genotype of an individual. Clone tracking

The first application to become clinically available is tracking minimal residual disease (MRD) in blood Current Opinion in Immunology 2013, 25:646–652

650 Immunogenetics and transplantation

cancers [3,20,21–27,28]. Cancer is clonal, meaning for most cases, a single cell spawns the entire cancer. In the case of lymphoid malignancies, that progenitor cell is a T or B cell. So, most of these cancer types carry a unique TCR or BCR rearrangement that act as a molecular tag for the cancers. At diagnosis, a major percentage of B or T cells in blood, bone marrow, or lymph node are cancerous and contain a common TCR or BCR rearrangement. Identifying single, highly abundant TCR or BCR genes in a diagnostic blood or bone marrow sample can be used to assist in diagnosis of leukemia or lymphoma. Additionally, the clonal sequence acts as a cancer cell-specific tag for that patient’s disease. After treatment, the cancerous cells are knocked down, with the goal of complete removal. HTS of BCR or TCR sequences can detect residual cancer cells after treatment with extreme sensitivity (1 per 106 nucleated cells) and accuracy. Primary studies have been completed in acute and chronic lymphoblastic leukemia (ALL and CLL), with other cancer types currently being tested. Many applications of immunosequencing of T and B repertoires track multiple clones simultaneously. A straightforward example is adoptive T cell transfer. For this branch of cell based immunotherapy, a set of T cells (usually autologous) are selected for particular properties — often their ability to bind to a HLA:epitope complex uniquely expressed on cancer cells-then expanded and injected into patients. By sequencing an aliquot of the infusion product, the levels of the therapeutic can be monitored over time in vivo by tracking the set of infused clones in the blood of the patient [29]. Immunosequencing also enables in vivo measurement of the dynamics of the adaptive immune response in humans [30,31]. We are just beginning to apply this technology to assess vaccine response [32]. B or T cell clones can be tracked between functional subsets over time in response to a vaccine, pathogen, or therapeutic. For example, we can directly observe how many different clones reach the central memory subset in response to a vaccine. For natural pathogens, in a recent work, Herpes Simplex 2 specific T cells in skin were shown to persist for years, re-expanding upon outbreak at the site of infection [33,34]. An additional exciting application is cancer immunology. We can identify infiltrating tumor specific T lymphocytes from resected tumors and then track these clones in the blood for response to immunotherapeutics [35]. As our knowledge expands, we expect to be able to identify disease specific pathogens either by clonal expansion or directly by sequence. In the case of autoimmune disease, knowledge of these sequences will allow monitoring of flares as well as response to therapies. Studies to find these clones are presently underway for every common autoimmune disease (e.g. Crohn’s disease, lupus, multiple sclerosis, etc.). Current Opinion in Immunology 2013, 25:646–652

Repertoire properties

A diverse repertoire of T and B cell clonotypes is believed to be necessary for sufficient protection against foreign pathogens. Certainly in extreme cases of reduced diversity such as SCIDs patients and human stem cell (HSC) transplant patients, the rate of infection is vastly higher than for healthy individuals. With older technologies such as spectratyping, repertoires with abnormally large clonal expansions are readily detectable. At present PCR based tests are utilized to confirm diagnosis for some leukemia and lymphoma patients, which have single clones very highly expanded. With the much higher sensitivity of HTS, the range of applications is far larger. Although, published work has just begun to explore the large set of potential applications, we will speculate here on wider sets of applications where current research is underway. The set of measures currently employed to assess global properties of the immune repertoire fall into three categories. The first are measures of clonality, which are essentially different methods to measure the fraction of the B or T cell repertoire constituted by the largest clones (the cutoff for defining largest clones differs by study). The second property is diversity, which focuses on the less abundant clonotypes. Since, sample sizes are limited, usually by a blood draw or tissue biopsy, the full repertoire diversity is usually estimated by extrapolation from the distribution of clones in the measured sample. The third is a set of properties of the complete distribution, such as entropy. The property of the repertoire utilized is application specific. Projects are presently underway to assess one or more of the types of global adaptive immune repertoire properties in many of the settings where there is immune compromise. Broadly, these areas include the aging immune system [16,31,32,36–38], T and/or B cell depletion (e.g. before HSC transplant) [39,40,41,42], immune dysfunction diseases, immunosuppressive drugs, immunotherapy [24,27,28,43], and autoimmune disease [44]. Although the true clinical utility awaits the outcome of these studies, some interesting recent publications hint at clinical utility, particularly in the posttransplant setting. Public clones

The third area of potential clinical applications is identifying public clonotypes. The adaptive immune system forms memory B and T cells in response to most (if not all) pathogenic exposures to facilitate subsequent rapid immune responses. In many respects, examination of the memory component of the adaptive immune system could act as an ideal diagnostic. T and B cells are able to detect very small disturbances (e.g. very early tumors), and they exponentially amplify in response, and are primarily specific for each antigen. We now have the www.sciencedirect.com

Applications of immune repertoire deep sequencing Robins 651

ability through immunosequencing to identify most memory clones in an individual. If we are able to learn how to connect these clones to their target antigens, we would have a broad and specific diagnostic for pathogenic exposure, as well as non-pathogenic disturbances such as tumor formation and autoimmune reaction. Starting with a simpler problem, the field has focused on public T cell responses. A portion of memory responses are identical or very similar between different people exposed to the same pathogen. For T cells, these public responses are specific within a specific HLA context. Public T cells are most commonly identified by the use of tetramers. A tetramer with an immunodominant epitope from a particular pathogen presented by an HLA molecule is used to sort T cell clones from blood samples from a population known to have the pathogen. The TCRs are sequenced from these sorted subsets and identical or similar sequences between samples are identified. These sequences are the public T cells, with the potential to be used to diagnose pathogenic exposure in the general population. The list of public clones is growing rapidly, and already includes hundreds of sequences for common pathogens such as CMV and EBV [45,46].

References and recommended reading Papers of particular interest, published within the period of review, have been highlighted as:  of special interest  of outstanding interest 1.

Robins HS, Campregher PV, Srivastava SK, Wacher A, Turtle CJ, Kahsai O, Riddell SR, Warren EH, Carlson CS: Comprehensive assessment of T-cell receptor beta-chain diversity in alphabeta T cells. Blood 2009, 114:4099-4107.

2.

Weinstein JA, Jiang N, White RA 3rd, Fisher DS, Quake SR: Highthroughput sequencing of the zebrafish antibody repertoire. Science 2009, 324:807-810.

3.

Boyd SD, Marshall EL, Merker JD, Maniar JM, Zhang LN, Sahaf B, Jones CD, Simen BB, Hanczaruk B, Nguyen KD et al.: Measurement and clinical monitoring of human lymphocyte clonality by massively parallel VDJ pyrosequencing. Sci Transl Med 2009, 1:12ra23.

4.

Freeman JD, Warren RL, Webb JR, Nelson BH, Holt RA: Profiling the T-cell receptor beta-chain repertoire by massively parallel sequencing. Genome Res 2009, 19:1817-1824.

5.

Arstila TP, Casrouge A, Baron V, Even J, Kanellopoulos J, Kourilsky P: A direct estimate of the human alphabeta T cell receptor diversity. Science 1999, 286:958-961.

6.

Wang C, Sanders CM, Yang Q, Schroeder HW Jr, Wang E, Babrzadeh F, Gharizadeh B, Myers RM, Hudson JR Jr, Davis RW et al.: High throughput sequencing reveals a complex pattern of dynamic interrelationships among human T cell subsets. Proc Natl Acad Sci U S A 2010, 107:1518-1523.

7.

Paillard F, Sterkers G, Bismuth G, Gomard E, Vaquero C: Lymphokine mRNA and T cell multireceptor mRNA of the Ig super gene family are reciprocally modulated during human T cell activation. Eur J Immunol 1988, 18:1643-1646.

8.

Robins H, Desmarais C, matthis J, livingston RJ, Andriesen J, Reijonen H, Carlson C, Nepom GT, Yee C, Cerosaletti K: Ultrasensitive detection of rare T cell clones. J Immunol Method 2012, 375:14-19.

9.

Vollmers C, Sit RV, Weinstein JA, Dekker CL, Quake SR: Genetic measurement of memory B-cell recall using antibody repertoire sequencing. Proc Natl Acad Sci U S A 2013, 110:13463-13468.

Future developments The full T or B cell receptor is made of a large and small chain that are coded by genes on different chromosomes. In order to do functional studies and identify specific antigenic targets of TCRs or BCRs, sequencing the pair of large and small chains is important. However, there is a significant challenge to matching the large and small chain pairs from the same cell at a scale that would be high throughput (i.e. screening of thousands of cells per assay). To date, the primary strategies have isolated single cells and physically linked cDNA molecules from the alpha and beta chains for TCRs or IgH and IGL/K chains for BCRs. The most common strategy for linking is through bridge PCR, where a chimeric PCR primer is designed to extend the large chain in one direction and the small chain in the opposite direction, forming a long oligo with both chains physically linked on the same cDNA strand. The process has proved difficult, but some progress has been made recently by isolating single cells in individual wells of a large plate, and performing the bridge amplification in parallel in each well [47]. These methods are low throughput but demonstrate the proof of principle. Other groups have attempted to isolate cells in emulsion droplets and use PCR to bridge the chains inside the droplet reactor. This strategy has the potential for far higher throughput, but technical challenges have limited progress to date. Multiple other creative strategies are being employed to solve this problem, so viable solutions are expected in a short time frame. www.sciencedirect.com

10. Larimore K, McCormick MW, Robins HS, Greenberg PD: Shaping of human germline IgH repertoires revealed by deep sequencing. J Immunol 2012, 186:3221-3230. 11. Sherwood AM, Desmarais C, Livingston RJ, Andriesen J, Haussler M, Carlson CS, Robins H: Deep sequencing of the human TCR{gamma} and TCR{beta} repertoires suggests that TCR{beta} rearranges after {alpha}{beta} and {gamma}{delta} T cell commitment. Sci Transl Med 2011, 3:90ra61. 12. Carlson C, Emerson R, Sherwood A, Desmarais C, Chung M,  Parsons J, Steen M, LaMardrid-Herrmannsfeldt M, Williamson D, Livingston RJ et al.: Using synthetic templates to design an unbiased multiplex PCR assay. Nat Commun 2013 http:// dx.doi.org/10.1038/ncomms3001. An effective method to remove PCR bias in a multiplex reaction to amplify TCR genes is presented. This allows quantitative TCR and BCR repertoire analysis using genomic DNA. 13. Briney BS, Willis JR, Crowe JE Jr: Location and length distribution of somatic hypermutation-associated DNA insertions and deletions reveals regions of antibody structural plasticity. Genes Immun 2012, 13:523-529. 14. Briney BS, Willis JR, Hicar MD, Thomas JW 2nd, Crowe JE Jr: Frequency and genetic characterization of V(DD)J recombinants in the human peripheral blood antibody repertoire. Immunology 2012, 137:56-64. 15. Briney BS, Willis JR, McKinney BA, Crowe JE Jr: Highthroughput antibody sequencing reveals genetic evidence of global regulation of the naive and memory repertoires that extends across individuals. Genes Immun 2012, 13:469-473. Current Opinion in Immunology 2013, 25:646–652

652 Immunogenetics and transplantation

16. Wu YC, Kipling D, Leong HS, Martin V, Ademokun AA, DunnWalters DK: High-throughput immunoglobulin repertoire analysis distinguishes between human IgM memory and switched memory B-cell populations. Blood 2010, 116:1070-1078. 17. Boyd SD, Gaeta BA, Jackson KJ, Fire AZ, Marshall EL, Merker JD, Maniar JM, Zhang LN, Sahaf B, Jones CD et al.: Individual variation in the germline Ig gene repertoire inferred from variable region gene rearrangements. J Immunol 2010, 184:6986-6992. 18. Bolotin DA, Mamedov IZ, Britanova OV, Zvyagin IV, Shagin D, Ustyugova SV, Turchaninova MA, Lukyanov S, Lebedev YB, Chudakov DM: Next generation sequencing for TCR repertoire profiling: platform-specific features and correction algorithms. Eur J Immunol 2012, 42:3073-3083. 19. Boyd SD: Diagnostic applications of high-throughput DNA sequencing. Annu Rev Pathol 2013, 8:381-410. 20. Wu D, Sherwood A, Fromm JR, Winter SS, Dunsmore KP, Loh ML,  Greisman HA, Sabath DE, Wood BL, Robins H: High-throughput sequencing detects minimal residual disease in acute T lymphoblastic leukemia. Sci Transl Med 2012, 4:134ra163. The first clinical application of immunosequencing to detect leukemic clones and track minimal residual disease. This manuscript comparies the presently used flow cytometry to immunosequencing and shows that immunosequencing is more sensitive. 21. Logan AC, Gao H, Wang C, Sahaf B, Jones CD, Marshall EL, Buno I, Armstrong R, Fire AZ, Weinberg KI et al.: High-throughput VDJ sequencing for quantification of minimal residual disease in chronic lymphocytic leukemia and immune reconstitution assessment. Proc Natl Acad Sci U S A 2011, 108:21194-21199.

31. Wu YC, Kipling D, Dunn-Walters DK: Age-related changes in human peripheral blood IGH repertoire following vaccination. Front Immunol 2012, 3:193. 32. Ademokun A, Wu YC, Martin V, Mitra R, Sack U, Baxendale H, Kipling D, Dunn-Walters DK: Vaccination-induced changes in human B-cell repertoire and pneumococcal IgM and IgA antibody at different ages. Aging Cell 2011, 10:922-930. 33. Zhu J, Peng T, Johnston C, Phasouk K, Kask AS, Klock A, Jin L, Diem K, Koelle DM, Wald A et al.: Immune surveillance by CD8alphaalpha+ skin-resident T cells in human herpes virus infection. Nature 2013, 497:494-497. 34. Neller MA, Burrows JM, Rist MJ, Miles JJ, Burrows SR: High frequency of herpesvirus-specific clonotypes in the human T cell repertoire can remain stable over decades with minimal turnover. J Virol 2013, 87:697-700. 35. Emerson R, Sherwood A, Rieder M, Guenthoer J, Williamson D, Carlson C, Drescher C, Tewari M, Bielas J, Robins H: Highthroughput sequencing of T cell receptors reveals a homogeneous repertoire of tumor-infiltrating lymphoctyes in ovarian cancer. J Pathol 2013 http://dx.doi.org/10.1002/path.4260. 36. Le Saux S, Weyand CM, Goronzy JJ: Mechanisms of immunosenescence: lessons from models of accelerated immune aging. Ann N Y Acad Sci 2012, 1247:69-82. 37. Boyd SD, Liu Y, Wang C, Martin V, Dunn-Walters DK: Human lymphocyte repertoires in ageing. Curr Opin Immunol 2013, 25:511-515. 38. Dunn-Walters DK, Ademokun AA: B cell repertoire and ageing. Curr Opin Immunol 2010, 22:514-520.

22. Logan AC, Zhang B, Narasimhan B, Carlton V, Zheng J, Moorhead M, Krampf MR, Jones CD, Waqar AN, Faham M et al.: Minimal residual disease quantification using consensus primers and high-throughput IGH sequencing predicts posttransplant relapse in chronic lymphocytic leukemia. Leukemia 2013, 27:1659-1665 http://dx.doi.org/10.1038/leu.2013.52.

39. Kanda J, Chiou LW, Szabolcs P, Sempowski GD, Rizzieri DA, Long GD, Sullivan KM, Gasparetto C, Chute JP, Morris A et al.: Immune recovery in adult patients after myeloablative dual umbilical cord blood, matched sibling, and matched unrelated donor hematopoietic cell transplantation. Biol Blood Marrow Transplant 2012, 18 1664–1676 e1661.

23. Porter DL, Kalos M, Zheng Z, Levine B, June C: Chimeric antigen receptor therapy for B-cell malignancies. J Cancer 2011, 2:331332.

40. Giannoni F, Hardee CL, Wherley J, Gschweng E, Senadheera S, Kaufman ML, Chan R, Bahner I, Gersuk V, Wang X et al.: Allelic exclusion and peripheral reconstitution by TCR transgenic T cells arising from transduced human hematopoietic stem/ progenitor cells. Mol Ther 2013, 21:1044-1054.

24. Porter DL, Levine BL, Kalos M, Bagg A, June CH: Chimeric antigen receptor-modified T cells in chronic lymphoid leukemia. N Engl J Med 2011, 365:725-733 http://dx.doi.org/ 10.1056/NEJMoa1103849. 25. Kalos M, Levine BL, Porter DL, Katz S, Grupp SA, Bagg A, June CH: T cells with chimeric antigen receptors have potent antitumor effects and can establish memory in patients with advanced leukemia. Sci Transl Med 2011, 3:953ra73. 26. Kalos M: Biomarkers in T cell therapy clinical trials. J Transl Med 2011, 9:138. 27. Brentjens RJ, Davila ML, Riviere I, Park J, Wang X, Cowell LG, Bartido S, Stefanski J, Taylor C, Olszewska M et al.: CD19targeted T cells rapidly induce molecular remissions in adults with chemotherapy-refractory acute lymphoblastic leukemia. Sci Transl Med 2013, 5:177ra138. 28. Grupp SA, Kalos M, Barrett D, Aplenc R, Porter DL, Rheingold SR,  Teachey DT, Chew A, Hauck B, Wright JF et al.: Chimeric antigen receptor-modified T cells for acute lymphoid leukemia. N Engl J Med 2013, 368:1509-1518. CART-19 cellular immunotherapy was used to treat patients with B-ALL by targeting all CD-19 expressing B cells, including the cancerous cells. Treatment efficacy was monitored by tracking the cancerous clone using high-throughput sequencing of the BCR repertoire. 29. Grupp SA, Prak EL, Boyer J, McDonald KR, Shusterman S, Thompson E, Callahan C, Jawad AF, Levine BL, June CH et al.: Adoptive transfer of autologous T cells improves T-cell repertoire diversity and long-term B-cell function in pediatric patients with neuroblastoma. Clin Cancer Res 2012, 18:6732-6741. 30. Parameswaran P, Liu Y, Roskin KM, Jackson KK, Dixit VP, Lee JY, Artiles KL, Zompi S, Vargas MJ, Simen BB et al.: Convergent antibody signatures in human dengue. Cell Host Microbe 2013, 13:691-700. Current Opinion in Immunology 2013, 25:646–652

41. Chapman M, Warren EH 3rd, Wu CJ: Applications of nextgeneration sequencing to blood and marrow transplantation. Biol Blood Marrow Transplant 2012, 18:S151-S160. 42. van Heijst JW, Ceberio I, Lipuma LB, Samilo DW, Wasilewski GD,  Gonzales AM, Nieves JL, van den Brink MR, Perales MA, Pamer EG: Quantitative assessment of T cell repertoire recovery after hematopoietic stem cell transplantation. Nat Med 2013, 19:372-377. Measurement at the clonal level of the CD4+ and CD8+ T cell repertoire reconstitution after transplant showed rapid turnover of clones and large differences in the dynamcs of helper and killer T cell recovery. 43. Turtle CJ, Riddell SR: Genetically retargeting CD8+ lymphocyte subsets for cancer immunotherapy. Curr Opin Immunol 2011, 23:299-305. 44. Maecker HT, Lindstrom TM, Robinson WH, Utz PJ, Hale M, Boyd SD, Shen-Orr SS, Fathman CG: New tools for classification and monitoring of autoimmune diseases. Nat Rev Rheumatol 2012, 8:317-328. 45. Venturi V, Chin HY, Asher TE, Ladell K, Scheinberg P, Bornstein E, van Bockel D, Kelleher AD, Douek DC, Price DA et al.: TCR betachain sharing in human CD8+ T cell responses to cytomegalovirus and EBV. J Immunol 2008, 181:7853-7862. 46. Robins HS, Srivastava SK, Campregher PV, Turtle CJ, Andriesen J, Riddell SR, Carlson CS, Warren EH: Overlap and effective size of the human CD8+ T cell receptor repertoire. Sci Transl Med 2010, 2:47ra64. 47. DeKosky BJ, Ippolito GC, Deschner RP, Lavinder JJ, Wine Y, Rawlings BM, Varadarajan N, Giesecke C, Dorner T, Andrews SF et al.: High-throughput sequencing of the paired human immunoglobulin heavy and light chain repertoire. Nat Biotechnol 2013, 31:166-169. www.sciencedirect.com