Patent Application
20120220466
Antigen receptors with diverse binding activities are the hallmark of B and T cells of the adaptive immune system in jawed vertebrates and are generated by genomic rearrangement of variable (V), diversity (D), and joining (J) gene segments separated by highly variable junction regions (Schatz (2004) Semin. Immunol. 16, 245-256). Initial calculations have been made of the combinatorial and junctional possibilities that contribute to the human immune receptor repertoire, and it is estimated that the number of possibilities may greatly exceed the total number of peripheral T or B cells in an individual (Davis and Bjorkman (1988) Nature 334, 395-402).
For example, one study in which small subsets of rearranged T cell receptor (TCR) subunit genes were extensively sequenced with a few segment-specific primers yielded extrapolations for the full TCR repertoire corresponding to 2.5×107 distinct TCRα-TCRβ pairs in the peripheral blood of an individual (Arstila et al. (1999) Science 286, 958-961). Extensive repertoire analyses for the human B cell compartment have been more limited, although small-scale studies and focused analysis of immunoglobulin (Ig) class subsets, such as IgE, have been performed (Brezinschek et al. (1995) J. Immunol. 155, 190-202, Lim et al. (2007) J. Allergy Clin. Immunol. 120, 696-706). Advanced sequencing methods have recently been used to analyze B cell receptor diversity in the relatively simple model immune system in zebrafish (Weinstein et al. (2009) Science 324, 807-810).
Against a background of continually generated novel DNA sequences, expanded clones of lymphocytes with useful antigen specificities persist over time to enable rapid responses to antigens previously detected by the immune system. Systematic means for detection of such expanded clones in human beings would provide significant opportunities for specific analysis and tracking, including measurement of clonal population sizes, anatomic distributions, and changes in response to immunological events.
In contrast to healthy immune systems, malignancies of B or T cell origin typically express a single dominant clonal Ig or TCR receptor. A variety of assays have been used to detect the presence of B cell clonality for diagnosis of lymphomas and leukemias, including analysis of Ig light chain gene restriction and Southern blotting or sizing of polymerase chain reaction (PCR) products from rearranged Ig or TCR loci (Rezuke et al. (1997) Clin. Chem. 43, 1814-1823; Arber (2000) J. Mol. Diagn. 2, 178-190). Although adequate for many applications, these strategies make limited use of the high information content inherent in rearranged immune receptor gene sequences and can give indeterminate results.
A recent study using deep sequencing of clonal IgH (Ig heavy chain) receptor genes in chronic lymphocytic leukemia revealed unexpected intraclonal heterogeneity in a subset of cases, showing that previous approaches have not captured the fundamental features of leukemic cell populations (Campbell et al. (2008) Proc. Natl. Acad. Sci. U.S.A. 105, 13081-13086). Detection of more subtle clonal populations (for example, to follow the response of lymphomas or leukemias to treatment) now relies on time- and labor-intensive multiparameter flow cytometry or custom-designed patient- and clone-specific realtime PCR assays (Sayala et al. (2007) Best Pract. Res. Clin. Haematol. 20, 499-512; Ladetto et al. (2000) Biol. Blood Marrow Transplant. 6, 241-253). Early diagnostic screening approaches may benefit from generalized and more efficient clonal detection. Indeed, a recent population-based epidemiological study showed that small amplified B cell populations can be seen in almost all individuals who go on to develop chronic lymphocytic leukemia, further underscoring the importance of assessing lymphocyte clonality in human specimens.
Detection and analysis of clonality is also of fundamental interest in characterizing and tracking normal and pathogenic immune reactions. For protective and healthy humoral immune responses, high-resolution analysis of immune receptor clonality and evolution offers the potential for definitive detection and monitoring of effective immune responses to vaccination and specific infections, whereas for some autoimmune disorders this type of analysis may facilitate diagnosis, long-term therapeutic monitoring strategies, and, eventually, specific interventions.
Methods are provided for the detection and analysis of clonality in a cell population, where the cells in the population are diverse with respect to genetic sequences at a locus of interest. Loci of interest are typically sites susceptible to somatic mutation and/or recombination, including without limitation, immune receptor genes, oncogenes, tumor suppressor genes, and the like. Cell populations for analysis are optionally complex populations having a high degree of sequence heterogeneity at the locus of interest.
Exemplary loci of interest are immunoglobulin and T cell antigen receptor genetic sequences, including without limitation each of IgH, IgLλ, IgLκ, TCRα, TCRβ, TCRγ, TCRδ, wherein genetic recombination events act to create an extensive repertoire of different sequences distinct from that of the germline. For purposes of the present invention, these loci may be referred to collectively as combinatorial antigen receptors. In biological samples, e.g. peripheral blood, lymph nodes, spleen, etc., it is rare for normal, naïve lymphocytes to share a combinatorial antigen receptor sequence. However, when stimulated by antigen, e.g. autoantigens, vaccines, infections, etc.; or when affected by hyperproliferative conditions such as cancer and other lymphoproliferative disorders; there can be a clonal expansion of cells having a single specificity. The result of such clonal expansion is the presence of multiple cells in the biological sample sharing a specific combinatorial antigen receptor sequence.
In some specific embodiments, the serological response to an antigenic stimulation, including vaccination, is determined by utilizing the methods of the invention. Following vaccination or other strong antigenic exposure, it has been found that within a defined period of time, e.g. at least about 5 days and not more than about 14 days, in some embodiments from about 6 to about 10 days, including about 7 days; there is a distinct increase in clonal B cell populations in sero-responsive individuals. The sequence of the clonal antibodies is not required for identification of the serologic responsiveness, although the methods of the invention provide for optional sequence determination if desired.
In some methods, including, without limitation, those datasets in which samples have been sequenced to different depths; which may include the determination of serologic response to vaccination, the counts of coincident sequences are normalized. Normalization can include dividing the total number of coincident sequences detected between replicates of a sample by the total number of possible pairwise comparisons between sequences in different replicates from that sample.
In the methods of the invention, parallel sequencing is applied to a DNA sample obtained from a population of cells, frequently a complex population of cells, e.g. a mammalian cell population, which may be a human lymphocyte population. The DNA sample may be genomic DNA, or cDNA obtained from cellular mRNA. Replicate samples, e.g. at least 2, 3, 4, 5, 6, 8, 10 or more replicate DNA samples are amplified, where the replicates may be derived from a single pool of cells or from multiple pools of cells.
Amplification utilizes one or more sets of amplification primers, which optionally comprise a bar-code for identification, and/or optionally a primer sequence for sequencing reactions. Where the locus of interest is a combinatorial antigen receptor; primers are often designed to amplify the hypervariable regions of the genetic loci of interest, which regions typically comprise combinatorial junctions. Other loci may be amplified with primers designed to span the locus of interest
The amplified DNA is sequenced; preferably a significant portion of one or more variable or hypervariable region(s) of interest are contained within a single “read”, e.g. at least about 32 nucleotides, at least about 50 nt., at least about 100 nt., at least about 200 nt., and may be about 500 nt. or more in length. The number of sequencing reads performed per replicate amplification reaction will vary with the specific analysis to be performed, but will generally comprise at least 10, at least 102, at least 103 or more reads per replicate. The sequences thus obtained are compared and a determination is made of coincident sequences across replicates, where coincident sequences are defined as those that share substantial sequence identity. Where the loci is a combinatorial antigen receptor, coincident sequences are those that share germline segments, e.g. V, D, J, as appropriate for the receptor being analyzed, and that share substantial identity in junctional nucleotide sequences. A variety of algorithms can be used to analyze the sequence data obtained using this method. The presence of coincident sequences across replicates is indicative of clonal expansion of a cell.
The methods of the invention provide a highly sensitive and consistent assay for determining the presence of clonal expansion even of rare cells, and further provide identification of the specific expanded sequence. The numbers and diversity of individual cells can be measured in a clinical sample in a manner that is not possible with other methods, using only very small amounts of cells. The identification of the clonal sequence provides information that can be associated with the biological relevance of the sequence, e.g. identification of antigens bound by a combinatorial antigen receptor, loss of tumor supressor activity, etc. While certain clonal expansions of lymphocytes are associated with expression of the combinatorial antigen receptor, it should be noted that other conditions that can be analyzed by the methods of the invention, such as hyperproliferative conditions, including without limitation carcinomas, leukemias and lymphomas, may not be associated with productive expression of an antigen receptor.
Conditions of interest for analysis of clonal expansion include numerous aspects of cellular proliferation and antigenic exposure, e.g. the presence of autoimmune disease; the status of transplantation; the presence of cancer, including without limitation cancers of the immune system, e.g. leukemias, lymphomas, myelomas, etc.; exposure to antigenic stimulus, e.g. exposure to cancer antigens; exposure to viral, bacterial, parasitic antigens; exposure to vaccines; exposure to allergens; exposure to foodstuffs, e.g. gluten proteins, and the like.
The information obtained from the clonality analysis may be used to monitor progression and/or treatment of hyperproliferative diseases, including detection of residual disease after treatment of patients; to monitor conditions of antigenic stimulation, including clonal expansion following vaccination, progression and/or treatment of autoimmune disease, transplantation monitoring and the like; to modify therapeutic regimens, and to further optimize the selection of therapeutic agents. With this approach, therapeutic and/or diagnostic regimens can be individualized and tailored according to the specificity data obtained at different times over the course of treatment, thereby providing a regimen that is individually appropriate. In addition, patient samples can be obtained at any point during the treatment process for analysis.
To facilitate an understanding of the invention, a number of terms are defined below.
The term “antibody,” as used herein, is intended to refer to immunoglobulin molecules comprised of four polypeptide chains, two heavy (H) chains and two light (L) chains (lambda or kappa) inter-connected by disulfide bonds. An antibody has a known specific antigen with which it binds. Each heavy chain of an antibody is comprised of a heavy chain variable region (abbreviated herein as HCVR, HV or VH) and a heavy chain constant region. The heavy chain constant region is comprised of three domains, CH1, CH2 and CH3. Each light chain is comprised of a light chain variable region (abbreviated herein as LCVR or VL or KV or LV to designate kappa or lambda light chains) and a light chain constant region. The light chain constant region is comprised of one domain, CL. The VH and VL regions can be further subdivided into regions of hypervariability, termed complementarity determining regions (CDRs), interspersed with regions that are more conserved, termed framework regions (FR). Each variable region (VH or VL) contains 3 CDRs, designated CDR1, CDR2 and CDR3. Each variable region also contains 4 framework sub-regions, designated FR1, FR2, FR3 and FR4.
As used herein, the terms “complementarity determining region” and “CDR” made with respect to the immunoglobulin loci refer to the regions that are primarily responsible for antigen-binding. There are three CDRs in a light chain variable region (CDRL1, CDRL2, and CDRL3), and three CDRs in a heavy chain variable region (CDRH1, CDRH2, and CDRH3). The particular designation in the art for the exact location of the CDRs varies depending on what definition is employed. Preferably, the IMGT designations are used (see Brochet et al. (2008) Nucleic Acids Res. 36:W503-8, herein specifically incorporated by reference), which uses the following designations for both light and heavy chains: residues 27-38 (CDR1), residues 56-65 (CDR2), and residues 105-116 (CDR3); see also Lefranc, M P, The Immunologist, 7:132-136, 1999, herein incorporated by reference.
As one example of CDR designations, the residues that make up the six CDRs have been characterized by Kabat and Chothia as follows: residues 24-34 (CDRL1), 50-56 (CDRL2) and 89-97 (CDRL3) in the light chain variable region and 31-35 (CDRH1), 50-65 (CDRH2) and 95-102 (CDRH3) in the heavy chain variable region; Kabat et al., (1991) Sequences of Proteins of Immunological Interest, 5th Ed. Public Health Service, National Institutes of Health, Bethesda, Md., herein incorporated by reference; and residues 26-32 (CDRL1), 50-52 (CDRL2) and 91-96 (CDRL3) in the light chain variable region and 26-32 (CDRH1), 53-55 (CDRH2) and 96-101 (CDRH3) in the heavy chain variable region; Chothia and Lesk (1987) J. Mol. Biol. 196: 901-917, herein incorporated by reference. Unless otherwise specified, the terms “complementarity determining region” and “CDR” as used herein, include the residues that encompass IMGT, Kabat and Chothia definitions.
As used herein, the term “framework” refers to the residues of the variable region other than the CDR residues as defined herein. There are four separate framework sub-regions that make up the framework: FR1, FR2, FR3, and FR4.
The term “T cell receptor” or “T cell antigen receptor” refers to the antigen/MHC binding heterodimeric protein product of a vertebrate, e.g. mammalian, TCR gene complex, including the human TCR α, β, γ and 8 chains. For example, the complete sequence of the human β TCR locus has been sequenced, as published by Rowen et al. (1996) Science 272(5269):1755-1762; the human α TCR locus has been sequenced and resequenced, for example see Mackelprang et al. (2006) Hum Genet. 119(3):255-66; see a general analysis of the T-cell receptor variable gene segment families in Arden Immunogenetics. 1995; 42(6):455-500; each of which is herein specifically incorporated by reference for the sequence information provided and referenced in the publication.
As used herein, “antigen” refers to any substance that, when introduced into a body, e.g., of a patient or subject, can stimulate an immune response, such as the production of an antibody or T cell receptor that recognizes the antigen. Antigens include molecules such as nucleic acids, lipids, ribonucleoprotein complexes, protein complexes, proteins, polypeptides, peptides and naturally occurring or synthetic modifications of such molecules against which an immune response involving T and/or B lymphocytes can be generated. With regard to autoimmune disease, the antigens herein are often referred to as autoantigens. With regard to allergic disease the antigens herein are often referred to as allergens. Autoantigens are any molecule produced by the organism that can be the target of an immunologic response, including peptides, polypeptides, and proteins encoded within the genome of the organism and post-translationally-generated modifications of these peptides, polypeptides, and proteins. Such molecules also include carbohydrates, lipids and other molecules produced by the organism. Antigens of interest also include vaccine antigens, which include, without limitation, pathogen antigens, cancer associated antigens, allergens, and the like.
As used herein, the term vaccine refers to a formulation comprising an antigen that is administered in a dose and regimen sufficient to produce an immune response to the antigen, usually a long-term term response, and frequently a long-term immunoglobulin response. A vaccine formulation is comprised of the antigen of interest, and frequently includes an adjuvant. Antigens include microbes, e.g. bacteria, viruses, protozoans, etc.; tumor antigens; and the like. Examples of vaccines include killed organisms, e.g. influenza, cholera, polio, rabies, hepatitis A, etc.; attenuated organisms, e.g. measles, mumps, rubella, BCG, etc.; toxins, e.g. tetanus, diphtheria, etc.; protein subunits, e.g. hepatitis B, human papillomavirus, hemagglutinin and neuraminidase of influenza, etc.; conjugates of coat proteins, e.g. H. influenzae B, etc.; DNA vaccines; and the like. Many vaccines have a schedule for primary and booster immunizations, although others are delivered as a single dose.
The time to response is very consistent for a given vaccine formulation, i.e. the time to generate detectable clonal populations will generally not vary by more than 10-20% between individuals; and will generally be consistent across vaccine formulations, i.e. the time to generate detectable clonal populations will not vary by more than 30-40% between individuals, as the response time is based on the kinetic of the underlying immune cell interactions. Factors that may alter the timing to response may include prior exposure to the antigen, dose, the presence of adjuvant, etc. Vaccines of a similar dose and adjuvant are expected to be very consistent in response time. For human use, adjuvants in the US are generally aluminum phosphate, aluminum hydroxide, or squalene. Other adjuvants include Freund's complete or incomplete adjuvant, virosomes, phosphate adjuvants, GS21, MF59, etc.
Allergens include immunogenic compounds that cause an enhanced Th2-type T cell response and IgE B cell response in a susceptible individual, also referred to as atopy, including asthma associated allergens. Allergens of interest include antigens found in food, such as strawberries, peanuts, milk proteins, egg whites, etc. Other allergens of interest include various airborne antigens, such as grass pollens, animal danders, house mite feces, etc. Molecularly cloned allergens include Dermatophagoides pteryonyssinus (Der P1); Lol pI-V from rye grass pollen; a number of insect venoms, including venom from jumper ant Myrmecia pilosula; Apis mellifera bee venom phospholipase A2 (PLA2 and antigen 5S; phospholipases from the yellow jacket Vespula maculifrons and white faced hornet Dolichovespula maculata; a large number of pollen proteins, including birch pollen, ragweed pollen, Parol (the major allergen of Parietaria officinalis) and the cross-reactive allergen Parjl (from Parietaria judaica), and other atmospheric pollens including Olea europaea, Artemisia sp., gramineae, etc. Other allergens of interest are those responsible for allergic dermatitis caused by blood sucking arthropods, e.g. Diptera, including mosquitoes (Anopheles sp., Aedes sp., Culiseta sp., Culex sp.); flies (Phlebotomus sp., Culicoides sp.) particularly black flies, deer flies and biting midges; ticks (Dermacenter sp., Ornithodoros sp., Otobius sp.); fleas, e.g. the order Siphonaptera, including the genera Xenopsylla, Pulex and Ctenocephalides felis felis. The specific allergen may be a polysaccharide, fatty acid moiety, protein, etc.
Tumor-suppressor genes, or more precisely, the proteins for which they code, either have a dampening or repressive effect on the regulation of the cell cycle or promote apoptosis, and sometimes do both. The functions of tumor-suppressor proteins may include genes that are involved cell cycle continuation, coupling of cell cycle to DNA damage, involvement in cell adhesion, DNA repair proteins, etc. Specific examples include without limitation retinoblastoma protein (pRb); p53 tumor-suppressor protein; PTEN; APC, CD95, ST5, ST7, and ST14; HNPCC, MEN1 and BRCA.
Oncogenes are tumor-inducing agents, and are often, although not necessarily, proteins involved in signal transduction and execution of mitogenic signals. Examples include, without limitation, RAS, WNT, MYC, ERK, TRK, epidermal growth factor receptor (EGFR), platelet-derived growth factor receptor (PDGFR), vascular endothelial growth factor receptor (VEGFR), HER2/neu Src-family, Syk-ZAP-70 family, BTK family of tyrosine kinases, Abl. Somatic mutations to oncogenes may include sequence alterations that cause a change in the protein structure, causing an increase in protein (enzyme) activity; a loss of regulation, an increase of protein expression, an increase of protein or mRNA stability, gene duplication, a chromosome translocation, and the like.
As used herein, the terms “subject”, “patient”, “individual” refer to any animal, usually a mammal, e.g. mouse, rat, dog, horse, monkey, and preferably a human.
Sample, as used herein, refers to a composition, often a physiological composition, e.g. a blood sample, lymph node sample, synovial fluid sample, CSF fluid, tumor biopsy sample, etc. from an individual that contains a cell population comprising genetic sequences that have, or are suspected of having, sequence diversity at a locus of interest. By diverse, it is meant that at least 2, at least 4, at least 8, at least 16, at least 32, at least 64, at least 128, at least 264, at least about 104 distinct sequences are represented at the locus of interest. Sequence diversity may include loci having at least about 105 distinct sequences; at least about 106 distinct sequences; at least about 107 distinct sequences or more. While not all such sequences may be expected to be present in a given sample, samples of interest generally provide at least a portion of the locus diversity, comprising cells representing more than 2, 4, 8, 10, 102, 103, 104, or more different sequences for a combinatorial antigen receptor of interest, where usually each cell comprises a distinct sequence.
Suitable cells for analysis include, without limitation, various hematopoietic cells, particularly including lymphocytes, tumor cells, etc. Lymphocytes expressing immunoglobulin include pre-B cells, B-cells, e.g. memory B cells, and plasma cells. Lymphocytes expressing T cell receptors include thymocytes, NK cells, pre-T cells and T cells, where many subsets of T cells are known in the art, e.g. Th1, Th2, Th17, CTL, T reg, etc.
Samples can include biopsies, or other clinical specimens containing cells. Some samples comprise cancer cells, such as carcinomas, melanomas, sarcomas, lymphomas, myelomas, leukemias, and the like.
Samples can be separated by centrifugation, elutriation, density gradient separation, apheresis, affinity selection, panning, FACS, centrifugation with Hypaque, etc. prior to analysis. For analysis of lymphocytes, often a mononuclear fraction (PBMC) comprising lymphocytes, monocytes, etc. is used.
A sample for use in the methods described herein may be one that is collected from a person with a malignancy or hyperproliferative condition, including lymphomas, leukemias, and plasmacytomas. A lymphoma is a solid neoplasm of lymphocyte origin, and is most often found in the lymphoid tissue. Thus, for example, a biopsy from a lymph node, e.g. a tonsil, containing such a lymphoma would constitute a suitable biopsy. Samples may be obtained from a patient at one or a plurality of time points in the progression of disease and/or treatment of the disease.
B lineage malignancies of interest include, without limitation, multiple myeloma; acute lymphocytic leukemia (ALL); chronic lymphocytic leukemia (CLL); diffuse large B cell lymphoma; mucosa-associated lymphatic tissue lymphoma (MALT); small cell lymphocytic lymphoma; mantle cell lymphoma (MCL); Burkitt lymphoma; mediastinal large B cell lymphoma; Waldenström macroglobulinemia; nodal marginal zone B cell lymphoma (NMZL); splenic marginal zone lymphoma (SMZL); intravascular large B-cell lymphoma; primary effusion lymphoma; lymphomatoid granulomatosis, etc. Non-malignant B cell hyperproliferative conditions include monoclonal B cell lymphocytosis (MBL).
T lineage malignancies of interest include, without limitation, precursor T-cell lymphoblastic lymphoma; T-cell prolymphocytic leukemia; T-cell granular lymphocytic leukemia; aggressive NK cell leukemia; adult T-cell lymphoma/leukemia (HTLV 1-positive); extranodal NK/T-cell lymphoma; enteropathy-type T-cell lymphoma; hepatosplenic γδ T-cell lymphoma; subcutaneous panniculitis-like T-cell lymphoma; mycosis fungoides/Sezary syndrome; anaplastic large cell lymphoma, T/null cell; peripheral T-cell lymphoma; angioimmunoblastic T-cell lymphoma; chronic lymphocytic leukemia (CLL); acute lymphocytic leukemia (ALL); prolymphocytic leukemia; hairy cell leukemia.
Inflammatory conditions are of interest for analysis by the methods of the invention, and include a number of diseases having an infectious or autoimmune component.
Neurological inflammatory conditions are of interest, e.g. Alzheimer's Disease, Parkinson's Disease, Lou Gehrig's Disease, etc. and demyelinating diseases, such as multiple sclerosis, chronic inflammatory demyelinating polyneuropathy, etc. as well as inflammatory conditions such as rheumatoid arthritis. Systemic lupus erythematosus (SLE) is an autoimmune disease characterized by polyclonal B cell activation, which results in a variety of anti-protein and non-protein autoantibodies (see Kotzin et al. (1996) Cell 85:303-306 for a review of the disease). These autoantibodies form immune complexes that deposit in multiple organ systems, causing tissue damage. An autoimmune component may be ascribed to atherosclerosis, where candidate autoantigens include Hsp60, oxidized LDL, and 2-Glycoprotein I (2GPI).
Conditions associated with immunodeficiency are also of interest for analysis, including congenital and acquired immunodeficiency syndromes.
As used herein, the term “a genetic sample’ refers to a portion of biological material containing mRNA or DNA from an individual, which portion of material is extracted, subjected to aliquoting to generate replicates, amplification and sequencing. The term an “aliquot” refers to a sub-fraction of a sample that is subjected to amplification and sequencing, i.e. a replicate. As previously discussed, the methods of the invention include at least 2 aliquots, or replicates, and may include at least 3, 4, 5, 6, 8, 10 or more aliquots.
As used herein a “read” is a single observation of DNA sequence from one amplification reaction, i.e. one aliquot from a sample. A “sequence” is derived from one or more reads, and corresponds to a single allele at a locus of interest. The set of sequences obtained from a single aliquot may be conveniently grouped for analysis
Clonal expansion, as used herein, refers to the proliferation of a cell having a specific combinatorial antigen receptor sequence, which sequence may be productively rearranged and expressed, for example where the proliferation is in response to antigenic stimulation. In other situations, e.g. with transformed or otherwise aberrantly hyperproliferative cells, the combinatorial antigen receptor sequence may not be expressed and may not be productively rearranged.
The term “clone” refers to a population of cells from an individual that have a shared allelic sequence, for example a mutation in an oncogene, a combination of germline V, D, and J regions, and junctional nucleotides, etc. Clonal combinatorial antigen receptors typically have identical germline regions and substantially identical junctional nucleotides, e.g. differing by not more than 1, not more than 2, not more than 3 nucleotides. These features are assigned by sequence comparison and alignment routines.
The term “coincident” is used herein to refer to a single sequence that is identified in two or more aliquots, or replicate samples. It may be noted that sequences appearing more than once in a single aliquot are not considered to be coincident, as the duplication can result from amplification of a single starting template.
Amplification refers to the process by which DNA templates are increased in number through multiple rounds of replication. Conveniently, polymerase chain reaction (PCR) is the method of amplification, but such is not required, and other methods, such as loop-mediated isothermal amplification (LIA); ligation detection reaction (LDR); ligase chain reaction (LCR); nucleic acid sequence based amplification (NASBA); multiple displacement amplification (MDA); C-probes in combination with rolling circle amplification; and the like may find use. See, for example, Kozlowski et al. (2008) Electrophoresis. 29(23):4627-36; Monis et al. (2006) Infect Genet Evol. 6(1):2-12; Zhang et al. (2006) Clin Chim Acta. 363(1-2):61-70; Cao (2004) Trends Biotechnol. 22(1):38-44; Schweitzer and Kingsmore (2001) Curr Opin Biotechnol. 12(1):21-7; Lisby (1999) Mol Biotechnol. 12(1):75-99. As known in the art, amplification reactions can be performed in a number of configurations, e.g. liquid phase, solid phase, emulsion, gel format, etc.
It is preferable to utilize a high fidelity polymerase in the amplification reaction to preserve sequence fidelity, typically a polymerase having an intact proof-reading function, e.g. Pfx50™ DNA Polymerase; Pfu polymerase, Vent polymerase, Phusion High-Fidelity DNA Polymerase; and the like.
Amplification by PCR is performed with at least two primers. For the methods of the invention, a set of primers is used that is sufficient to amplify all or a defined portion of the variable sequences at the locus of interest, which locus may include any or all of the aforementioned IgH and TCR loci. Exemplary IgH primers are provided in the examples.
Primer sets usually amplify at least 50% of the known rearrangements at the locus of interest, at least 75%, at least 85%, at least 90%, at least 95%, or more. Primers may further comprise nucleotides useful in subsequent sequencing, e.g. pyrosequencing. Such sequences are readily designed by commercially available software programs or companies (e.g. see Biotage). Amplification primers may optionally include a barcode sequence, to aid in the identification of clones (see Parameswaran et al. (2007) Nucleic Acids Research 35(19):e30, herein specifically incorporated by reference).
Sequencing platforms include, but are not limited to those commercialized by: 454/Roche Lifesciences including but not limited to the methods and apparatus described in Margulies et al., Nature (2005) 437:376-380 (2005); and U.S. Pat. Nos. 7,244,559; 7,335,762; 7,211,390; 7,244,567; 7,264,929; 7,323,305; Helicos BioSciences Corporation (Cambridge, Mass.) as described in U.S. application Ser. No. 11/167,046, and U.S. Pat. Nos. 7,501,245; 7,491,498; 7,276,720; and in U.S. Patent Application Publication Nos. US20090061439; US20080087826; US20060286566; US20060024711; US20060024678; US20080213770; and US20080103058; Applied Biosystems (e.g. SOLiD sequencing); Dover Systems (e.g., Polonator G.007 sequencing); Illumina as described U.S. Pat. Nos. 5,750,341; 6,306,597; and 5,969,119; and Pacific Biosciences as described in U.S. Pat. Nos. 7,462,452; 7,476,504; 7,405,281; 7,170,050; 7,462,468; 7,476,503; 7,315,019; 7,302,146; 7,313,308; and US Application Publication Nos. US20090029385; US20090068655; US20090024331; and US20080206764. All references are herein incorporated by reference. Such methods and apparatuses are provided here by way of example and are not intended to be limiting.
The term “isolated” when used in relation to a nucleic acid, as in “an isolated oligonucleotide” or “isolated polynucleotide” refers to a nucleic acid sequence that is identified and separated from at least one contaminant nucleic acid with which it is ordinarily associated (e.g. host cell proteins).
As used herein, the terms “portion” when used in reference to a nucleotide sequence (as in “a portion of a given nucleotide sequence”) refers to fragments of that sequence. The fragments may range in size from ten nucleotides to the entire nucleotide sequence minus one nucleotide (e.g., 10 nucleotides, 20, 30, 40, 50, 100, 200, etc.).
As used herein, the term “portion” when in reference to an amino acid sequence (as in “a portion of a given amino acid sequence”) refers to fragments of that sequence. The fragments may range in size from six amino acids to the entire amino acid sequence minus one amino acid (e.g., 6 amino acids, 10, 20, 30, 40, 75, 200, etc.)
As used herein, the terms “treat,” “treatment,” “treating,” and the like, refer to obtaining a desired pharmacologic and/or physiologic effect. The effect may be prophylactic in terms of completely or partially preventing a disease or symptom thereof and/or may be therapeutic in terms of a partial or complete cure for a disease and/or adverse affect attributable to the disease. “Treatment,” as used herein, covers any treatment of a disease in a mammal, particularly in a human, and includes: (a) preventing the disease from occurring in a subject which may be predisposed to the disease but has not yet been diagnosed as having it; (b) inhibiting the disease, i.e., arresting its development; and (c) relieving the disease, e.g., causing regression of the disease, e.g., to completely or partially remove symptoms of the disease.
“Suitable conditions” shall have a meaning dependent on the context in which this term is used. That is, when used in connection with an antibody, the term shall mean conditions that permit an antibody to bind to its corresponding antigen. When this term is used in connection with nucleic acid hybridization, the term shall mean conditions that permit a nucleic acid of at least 15 nucleotides in length to hybridize to a nucleic acid having a sequence complementary thereto. When used in connection with contacting an agent to a cell, this term shall mean conditions that permit an agent capable of doing so to enter a cell and perform its intended function. In one embodiment, the term “suitable conditions” as used herein means physiological conditions.
Before the present active agents and methods are described, it is to be understood that this invention is not limited to the particular methodology, products, apparatus and factors described, as such methods, apparatus and formulations may, of course, vary. It is also to be understood that the terminology used herein is for the purpose of describing particular embodiments only, and is not intended to limit the scope of the present invention which will be limited only by appended claims.
It must be noted that as used herein and in the appended claims, the singular forms “a,” “and,” and “the” include plural referents unless the context clearly dictates otherwise. Thus, for example, reference to “a drug candidate” refers to one or mixtures of such candidates, and reference to “the method” includes reference to equivalent steps and methods known to those skilled in the art, and so forth.
Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs. All publications mentioned herein are incorporated herein by reference for the purpose of describing and disclosing devices, formulations and methodologies which are described in the publication and which might be used in connection with the presently described invention.
Where a range of values is provided, it is understood that each intervening value, to the tenth of the unit of the lower limit unless the context clearly dictates otherwise, between the upper and lower limit of that range and any other stated or intervening value in that stated range is encompassed within the invention. The upper and lower limits of these smaller ranges may independently be included in the smaller ranges is also encompassed within the invention, subject to any specifically excluded limit in the stated range. Where the stated range includes one or both of the limits, ranges excluding either both of those included limits are also included in the invention.
In the following description, numerous specific details are set forth to provide a more thorough understanding of the present invention. However, it will be apparent to one of skill in the art that the present invention may be practiced without one or more of these specific details. In other instances, well-known features and procedures well known to those skilled in the art have not been described in order to avoid obscuring the invention.
Generally, conventional methods of protein synthesis, recombinant cell culture and protein isolation, and recombinant DNA techniques within the skill of the art are employed in the present invention. Such techniques are explained fully in the literature, see, e.g., Maniatis, Fritsch & Sambrook, Molecular Cloning: A Laboratory Manual (1982); Sambrook, Russell and Sambrook, Molecular Cloning: A Laboratory Manual (2001); Harlow, Lane and Harlow, Using Antibodies: A Laboratory Manual: Portable Protocol NO. I, Cold Spring Harbor Laboratory (1998); and Harlow and Lane, Antibodies: A Laboratory Manual, Cold Spring Harbor Laboratory; (1988) Protein Methods (Bollag et al., John Wiley & Sons 1996); Nonviral Vectors for Gene Therapy (Wagner et al. eds., Academic Press 1999); Viral Vectors (Kaplift & Loewy eds., Academic Press 1995); Immunology Methods Manual (I. Lefkovits ed., Academic Press 1997); and Cell and Tissue Culture: Laboratory Procedures in Biotechnology (Doyle & Griffiths, John Wiley & Sons 1998). Reagents, cloning vectors, and kits for genetic manipulation referred to in this disclosure are available from commercial vendors such as BioRad, Stratagene, Invitrogen, Sigma-Aldrich, and ClonTech.
Methods are provided for the detection and analysis of clonality in a cell population, where the cells in the population are diverse with respect to genetic sequences at a locus of interest. Exemplary loci of interest are immunoglobulin and T cell antigen receptor genetic sequences, including without limitation each of IgH, IgLλ, IgLκ, TCRα, TCRβ, TCRγ, TCRδ, wherein genetic recombination events act to create an extensive repertoire of different sequences.
In the methods of the invention, a sample comprising a complex cell population is obtained from an individual, particularly an individual that is, or that will be, subject to clonal expansion of a cell population. Samples of interest may include samples obtained from patients having a hyperproliferative condition, for example involving lymphocytes, e.g. lymphomas, leukemias, myelomas and benign lymphoproliferative conditions. Such individuals may be tested at one or a plurality of time points, including, without limitation, at the time of diagnosis, prior to, during and/or after cytoreductive treatment; and at various timepoints to monitor disease progression. Samples of interest also include individuals before and/or after specific antigenic stimulation, e.g. following vaccination, where vaccines include cancer antigens, pathogen antigens, allergens, autoantigens for tolerization. Antigenic stimulation may also be monitored before and/or after infection, e.g. to monitor epidemic or pandemic situations. Samples of interest also include individuals before and/or after transplantation of an allogeneic tissue. Samples of interest also include individuals suspected of having an autoimmune or inflammatory disease. Such individuals may be tested at one or a plurality of time points, including, without limitation, at the time of diagnosis, prior to, during and/or after cytoreductive treatment; and at various timepoints to monitor disease progression.
Patient samples include a variety of bodily fluids in which cells are present, e.g. blood and derivatives thereof, CSF; synovial fluid, tumor biopsy samples, spleen, lymph nodes, bone marrow, cord blood, etc. Samples may be obtained at one or more suitable time points, depending on the needs of the specific analysis. As described above, patient samples of interest may comprise complex cell populations, and the desired cell population may further be subjected to selection or sorting by various methods.
DNA is obtained from the cell sample. In some embodiments, genomic DNA is utilized, which is readily extracted from cells using conventional methods known in the art. It will further be understood by one of skill in the art that a lymphocyte may comprise two distinct rearranged alleles at a locus of interest, although generally only one allele is productively rearranged. In other embodiments, mRNA is obtained from the cells and converted to cDNA for amplification purposes using conventional methods. It will be understood by those of skill in the art that determination of clonality based on mRNA samples will require that aliquots be obtained prior to extraction of RNA from the cells, as one cell may comprise multiple copies of mRNA corresponding to a sequence of interest.
An important feature of the methods of the invention is the analysis of replicate samples, where replicates are frequently aliquots of cells or DNA from a single sample. Replicates may also be two cell samples obtained from an individual. Replicates are obtained prior to amplification of the DNA, as the amplification reaction has the potential to generate coincidence by duplication of the initial template sequence.
The DNA replicates are amplified by any convenient methods and sequenced, as previously described. Generally at least about 10, 102, 103 or more reads are obtained for each replicate sample, where a preferred read includes the variable sequence at the locus of interest. The sequences thus obtained are compared and a determination is made of coincident sequences across replicates, where coincident sequences are defined as those that share germline segments, e.g. V, D, J, as appropriate for the receptor being analyzed, and that share junctional nucleotide sequences. A variety of algorithms can be used to analyze the sequence data obtained using this method. The presence of coincident sequences across replicates is indicative of clonal expansion of a cell.
Clonally expanded cell populations are detectable by the presence of coincident sequences in distinct amplicon pools, e.g. in at least 2 replicates, in at least 3 replicates, usually in at least 4 replicates, in at least five replicates, or more.
The minimum expected number of binary sequence coincidences is achieved under the condition of equal representation of each sequence type (any preferential representation of one or more sequence types for a given value of total repertoire number would only increase the coincidence frequency). For example, in analysis of immunoglobulin loci, where (“IgHR”, the total number of distinct IgH sequences present in the peripheral blood of an individual), the minimal expected coincidence number (for example, for time point 1) is given by the following formula: (½)*Σi=1 to 6(Si*Σ(j=1 to 6; j≠i){1−[1−(1/IgHR)]Sj}) where S1 . . . S6 are the numbers of distinct sequences determined for the six independent amplicon pools prepared from that time point. Starting with an arbitrary sequence in one amplicon pool, (1/IgHR) is the probability that any single determined sequence in another pool would match that sequence, 1−(1/IgHR) is the probability that any single determined sequence in another pool would not match that sequence, and [1−(1/IgHR)]Sj is the probability that no determined sequence from an amplicon pool with Sj determined sequences would match that sequence. The value [1(1/IgHR)]Sj, summed for every element in the sequence set, yields a value that is twice the expected number of coincidences (because this sum counts each binary coincidence once for each participating sequence and hence twice in total). The formula above represents this sum. Similar calculations can be used for greater than binary sequence coincidence.
In some preferred embodiments, the methods of the invention are used in determining the efficacy of a therapy for treatment of a hyperproliferative or autoimmune disease, or for efficacy of vaccination, either at an individual level, or in the analysis of a group of patients, e.g. in a clinical trial format. Such embodiments typically involve the comparison of two time points for a patient or group of patients. The patient status is expected to differ between the two time points as the result of a therapeutic agent, therapeutic regimen, or other intervention to a patient undergoing treatment.
Examples of formats for such embodiments may include, without limitation, testing the effect of a therapy or vaccination at two or more time points, where a first time point is a diagnosed but untreated patient; and a second or additional time point(s) is a patient treated with a candidate therapeutic agent or regimen. An additional time point may include a patient vaccinated with a candidate agent or regimen, and challenged with the antigen, e.g. by community exposure to an infectious agent, ex vivo antigen challenge, skin test, etc.
In another format, a first time point is a diagnosed patient in disease remission, e.g. as ascertained by current clinical criteria, as a result of a candidate therapeutic agent or regimen. A second or additional time point(s) is a patient treated with a candidate therapeutic agent or regime.
In such clinical trial formats, each set of time points may correspond to a single patient, to a patient group, e.g. a cohort group, or to a mixture of individual and group data. Additional control data may also be included in such clinical trial formats, e.g. a placebo group, a disease-free group, and the like, as are known in the art. Formats of interest include crossover studies, randomized, double-blind, placebo-controlled, parallel group trial is also capable of testing drug efficacy, and the like. See, for example, Clinical Trials: A Methodologic Perspective Second Edition, S. Piantadosi, Wiley-Interscience; 2005, ISBN-13: 978-0471727811; and Design and Analysis of Clinical Trials: Concepts and Methodologies, S. Chow and J. Liu, Wiley-Interscience; 2003; ISBN-13: 978-0471249856, each herein specifically incorporated by reference. Specific clinical trials of interest include analysis of therapeutic agents for the treatment of hyperproliferative conditions involving lymphocytes, analysis of immunosuppressive therapies, including antigen-specific immunotherapies; analysis of vaccine responses, and the like
In some embodiments, a blinded crossover clinical trial format is utilized. In another embodiments a randomized, double-blind, placebo-controlled, parallel group trial is used to test drug efficacy.
In other embodiments, a clinical trial format is utilized to test the efficacy of a vaccine, for example by determining the percent of vaccinated individuals that are serological responders following vaccination, e.g. following a single dose of vaccine, following a booster, etc. The methods of the invention allow determination of which individuals are responder within a short, defined time period, for example within less than about 10 days, less than about 9 days, less than about 8 days, including within 7 days following immunization. The ability to assess the efficacy of the vaccine in such a short time frame provides substantial advantages over the prior art, which can require 21 days or more to determine serological responsiveness. In one embodiment, a cohort of individuals, e.g. a mammal or avian, including without limitation humans, dogs, cats, horses, cows, sheep, pigs, chickens, ducks, and the like, are vaccinated with a test vaccine formulation, usually in combination with a randomized control group immunized with the formulation in the absence of antigen. At a defined period of from about 7 to about 10 days, a sample, e.g. a blood sample, is drawn from the individuals, and analyzed for the presence of clonal B cell populations. The presence of such clonal populations is indicative that the individual is serologically response to the vaccine antigen. As discussed above, the data may be normalized by dividing the total number of coincident sequences detected between replicates of a sample by the total number of possible pairwise comparisons between sequences in different replicates from that sample.
Also provided are databases of sequence analyses relating to a treatment or condition of interest. Such databases will typically comprise analysis profiles of various individuals following a clinical protocol of interest etc., where such profiles are further described below.
The profiles and databases thereof may be provided in a variety of media to facilitate their use. “Media” refers to a manufacture that contains the expression profile information of the present invention. The databases of the present invention can be recorded on computer readable media, e.g. any medium that can be read and accessed directly by a computer. Such media include, but are not limited to: magnetic storage media, such as floppy discs, hard disc storage medium, and magnetic tape; optical storage media such as CD-ROM; electrical storage media such as RAM and ROM; and hybrids of these categories such as magnetic/optical storage media. One of skill in the art can readily appreciate how any of the presently known computer readable mediums can be used to create a manufacture comprising a recording of the present database information. “Recorded” refers to a process for storing information on computer readable medium, using any such methods as known in the art. Any convenient data storage structure may be chosen, based on the means used to access the stored information. A variety of data processor programs and formats can be used for storage, e.g. word processing text file, database format, etc.
As used herein, “a computer-based system” refers to the hardware means, software means, and data storage means used to analyze the information of the present invention. The minimum hardware of the computer-based systems of the present invention comprises a central processing unit (CPU), input means, output means, and data storage means. A skilled artisan can readily appreciate that any one of the currently available computer-based system are suitable for use in the present invention. The data storage means may comprise any manufacture comprising a recording of the present information as described above, or a memory access means that can access such a manufacture.
A variety of structural formats for the input and output means can be used to input and output the information in the computer-based systems of the present invention. Such presentation provides a skilled artisan with a ranking of similarities and identifies the degree of similarity contained in the test expression profile.
The methods of the invention find use in a centralized setting, e.g. the analysis of patient samples in a clinical laboratory. Such assays may conveniently utilize one or more of the primer sets provided herein. Such assays may include an analysis of the clonality present in a sample, e.g. the presence of clonal changes in cancer cells, the presence of residual disease in a cancer patient, the presence of clonal lymphocytes specific for an antigen of interest, including vaccine antigens, and the like.
Also provided are reagents and kits thereof for practicing one or more of the above-described methods. The subject reagents and kits thereof may vary greatly. Reagents of interest include reagents specifically designed for use in production of the above described analysis. Kits may include amplification primers, including without limitation one or more of the sets of primers identified herein, reagents amplification and sequence, and such containers as are required for sample collection.
The kits may further include a software package for statistical analysis of the sequences. In addition to the above components, the subject kits will further include instructions for practicing the subject methods. These instructions may be present in the subject kits in a variety of forms, one or more of which may be present in the kit. One form in which these instructions may be present is as printed information on a suitable medium or substrate, e.g., a piece or pieces of paper on which the information is printed, in the packaging of the kit, in a package insert, etc. Yet another means would be a computer readable medium, e.g., diskette, CD, etc., on which the information has been recorded. Yet another means that may be present is a website address which may be used via the internet to access the information at a removed site. Any convenient means may be present in the kits.
The following examples are put forth so as to provide those of ordinary skill in the art with a complete disclosure and description of how to make and use the present invention, and are not intended to limit the scope of the invention or to represent that the experiments below are all or the only experiments performed. Efforts have been made to ensure accuracy with respect to numbers used (e.g., amounts, temperature, and the like), but some experimental errors and deviations may be present. Unless indicated otherwise, parts are parts by weight, molecular weight is weight average molecular weight, temperature is in degrees Centigrade, and pressure is at or near atmospheric.
Using a bar-coding strategy to allow pooling of multiple libraries of rearranged IgH V-D-J gene loci from many human blood samples, high-throughput pyrosequencing was performed to characterize the B cell populations in a series of human clinical specimens. Deep sequencing of immune receptor gene populations offered specific and detailed molecular characterization as well as high sensitivity for detecting sequences of interest and to transform understanding of the human immune system while aiding in diagnosis and tracking of lymphoid malignancies.
Bar-Coded High-Throughput Pyrosequencing of Rearranged IgH Loci.
We amplified rearranged IgH loci in human blood samples with BIOMED-2 nucleic acid primers adapted for high-throughput DNA pyrosequencing. A unique 6-, 7-, or 10-nucleotide sequence “bar code” in the primers used for a particular sample allowed pooling and bulk sequencing of many libraries together and subsequent sorting of sequences from each sample (
An overview of the IgH amplicon sequences in the data sets from experiments 1 and 2 is shown in
Evaluation of Clonal Malignancies.
Human cancers are clonal proliferations of cells that have sustained mutational damage, leading to dysregulated proliferation, survival, and response to the extracellular environment. Molecular clonality testing of IgH receptor and TCR y loci, accomplished with a PCR and capillary electrophoresis, is a helpful adjunct to morphological and immunophenotypic evaluation of suspected B or T cell malignancies.
Blood or bone marrow samples from some patients give indeterminate or oligoclonal patterns of reactivity for a variety of reasons: Few lymphocytes may be present, there may be genuine oligoclonal lymphocyte populations, or clonal lymphocytes may have separately detected rearrangements from two chromosomes. We compared the results from DNA sequencing of the products of independent PCR replicates for such samples. One such difficult case is represented by the bone marrow and liver specimens from patient 5 in Table 1. The patient had undergone liver transplantation and subsequently developed a large B cell lymphoma in the liver as a manifestation of post-transplant lymphoproliferative disorder, a condition in which immunosuppression leads to B or T cell lymphomas that are typically associated with Epstein-Barr virus infection (
Another diagnostically challenging case, the chronic lymphocytic leukemia of patient 4, showed an oligoclonal pattern by standard PCR and capillary electrophoresis analysis. A consistent pattern was seen with deep sequencing of the sample. Finally, the two distinct V-D-J rearrangements in a lymph node from patient 3 indicated that there were two separate clonal B cell populations in the specimen, a conclusion supported by morphological and immunophenotypic evidence of two different B cell lymphomas (follicular lymphoma and small lymphocytic lymphoma) in the tissue.
Minimal Residual Disease Testing by Sequencing.
To evaluate the sensitivity of deep sequencing for detection of a clonal lymphoid population in a background of polyclonal cells, we performed serial 10-fold dilutions of a known clonal chronic lymphocytic leukemia blood sample into normal peripheral blood. The percentage of clonal sequences detected at each dilution is shown in
We next evaluated clinical specimens from patients with chronic lymphocytic leukemia who had undergone total lymphoid irradiation and anti-thymocyte globulin therapy followed by human leukocyte antigen-identical allogeneic peripheral blood progenitor cell transplantation and compared the results of deep sequencing analysis to results from patient- and clone-specific real-time PCR assays (Table 2). In these experiments, the patients with chronic lymphocytic leukemia were different from the patients tested in our initial experiments described in Table 1, and the minimal residual disease (MRD) sequencing was performed in a separate instrument run. Realtime PCR assay results were reported as confidently positive if at least 100 copies per microgram of template DNA were detected. Table 2 demonstrates that all specimens showed agreement between the high throughput sequencing data and real-time PCR assay, although for the lowest confidently positive real-time PCR result for chronic lymphocytic leukemia patient A the clone was detected in only one of the two high throughput sequencing sample replicates.
†Second replicate.
Peripheral Blood B Cell Repertoire in Healthy Subjects.
To identify potentially expanded B cell clones within healthy peripheral blood, we looked for independent occurrences of “coincident” IgH sequences (identical V, D, and J segments and identical V-D and D-J junction sequences) in independent pools from the same individual. Such coincidences could have resulted from clonally related cells; indeed, clonal relations are likely for a majority of these coincidences, given both the diversity of the potential repertoire of IgH rearrangements and the absence of rearrangements found in this individual from comparable sequence samples from different individuals. We note that any population with a limited IgH rearrangement repertoire would be expected to show large numbers of such coincidences. Instead, we observed only small numbers of coincident sequences in our data. From six independent amplification pools derived from the blood of a single individual at one time point, we observed only 19 potential coincidences from a total of 10,921 distinct IgH rearrangements sequenced. Seven independent amplification pools from a second time point (14 months later) gave comparable results (25 potential coincidences from a total of 7450 distinct rearrangements sequenced) (Table 3).
It is noteworthy that we see only slightly fewer coincidences when comparing aliquots between the two time points (0.76 coincidences per sample comparison versus 1.22 for comparisons within the same time point). Although the difference is statistically significant (P<0.05, Fisher's exact one-tailed test), the modest ratio between intratemporal and intertemporal coincidence levels indicates a considerable degree of persistence in the clonal populations in this individual.
The numbers of coincident sequences observed when comparing sequence data from any two aliquots provide strong evidence for substantial diversity in the IgH repertoire. Minimal estimates obtained with approaches similar to the “birthday problem” in probability theory yield a lower bound of ˜2 million different IgH rearrangements in these samples. The analysis leading to this lower bound estimate does not yield an upper bound on repertoire; in particular, it is not possible from these data to rule out a category of IgH rearrangements that are very diverse but present in single- or low-copy number in ˜2×109 B cells in peripheral blood. Thus, the true complexity of the blood IgH repertoire could certainly be much greater than 2×106. In addition to the total complexity of the IgH pool, it is of interest to evaluate the degree to which clonal cell populations above a certain size are present in normal peripheral blood. No sequence was identified in more than 2 of the 13 sequence sets from independent amplicon pools (Table 3).
Using a similar analysis to that described above, we can derive an upper bound for the most abundant IgH rearrangements. For the healthy individual examined in these experiments, this analysis yields a maximum contribution to the sequence pool of 1 of 1000 for any individual clone (P<0.01) in this individual. Within these experimental estimates of the lower bound of the IgH repertoire size, and the upper bound of the largest clone size, a variety of combinations of clonally expanded populations of different sizes could give rise to our observed data. Estimation of the upper limit of the IgH repertoire would require much more extensive sequencing to evaluate the extent of single-copy or very small clonal expansions of B cells and would require characterization of a significant fraction of the blood volume of a healthy donor, which presents ethical concerns. It should be noted that this analysis of the blood does not exclude the possibility that other tissues may contain B cells that are clonally related to circulating cells and does not address the exchange of B cells between the blood and other hematolymphoid compartments of the body.
Diversity of Clonal B Cell Expansions in Healthy Subjects of Various Ages.
We extended our analysis of healthy human patients to an additional 23 subjects ranging in age from 19 to 79 years by sequencing sixfold replicate samples of peripheral blood IgHs from each individual. We detected considerable interindividual variation in the number of expanded lymphocyte clones and expanded clone sizes (Table 4). Using an analysis similar to that performed for the healthy donor in Table 3, we calculated the minimum IgH repertoire size and the largest clone size for these additional subjects. Our data confirm that at least 15 of the 23 additional normal human samples had IgH pools of >1,000,000 different rearrangements. Although the additional eight individuals may have comparable diversity, the lower bound estimates were somewhat lower, relative to the other 15 subjects, because of the greater numbers of weakly amplified clones detected and the lower total yield of sequences from these samples. For a majority of the healthy samples, no sequence appeared in more than two of six sequenced DNA aliquots; for these individuals, this places an upper limit of 0.1% to 0.3% of the measured B cell repertoire that could be dedicated to any single clone, similar to the results from the individual in Table 3. Two of the apparently healthy blood donors in our sample set had expanded B cell clones that were large enough to be detected in all six sequencing replicates. The size of these larger clones can be estimated by the expanded clonal sequence's proportion of total sequences obtained from these patients: For the 54-year-old patient, this value was 0.15%, whereas for the 68-year-old patient the value was 1.5% of the total sequences.
These data demonstrate that detection of clonal populations that make up >0.1% of the total B cell population is readily possible with the small blood samples used for this work (<0.1 ml of blood was sufficient for the multiple replicates from these specimens). Further, these results suggest that searches for persistent premalignant or pathological clonal populations at the 0.1% level might be facilitated in certain cases by the limited set of amplified candidates in the normal repertoire.
Deep sequencing data sets of this kind enables explicit detection of preferentially rearranged or selected combinations of V, D, or J segments in IgHs in specific populations. Using the healthy control specimens in our current data sets, we have seen evidence of preferential pairwise segment associations for at least three combinations (D2-2 with J6, D3-22 with J3, and D3-3 with J6) across the group of individuals. Overrepresentation of these D-J combinations (that is, a frequency of the D-J combination that is greater than the products of the D and J frequencies) was observed in 122 of 138, 113 of 138, and 119 of 138 sequenced aliquots, respectively. With a false discovery rate of <10−7 (no examples of overrepresentation in this number of aliquots were found in 107 randomly shuffled data sets), these were the most consistent nonrandom associations seen with the data set. We interpret these results as reflecting nonrandom character in rearrangement or selection in this specific population of individuals (Stanford's blood donor pool in a fixed time frame). One could expect different specific nonrandom characters in other populations with distinct histories of community immune response and genetic compositions.
Modern DNA sequencing methods open a new window of investigation into the complex gene rearrangements necessary for human lymphocyte function. Our results using multiplexed bar-coded IgH sequencing of multiple replicate samples of blood from 24 healthy subjects represent the most extensive characterization to date of human B cell populations. For a majority of the healthy individuals, our results were sufficient to place a lower limit of 1,000,000 on the number of distinct IgH rearrangements in circulating lymphocytes and an upper bound of 0.1% to 0.3% of total B cells on the representation of any single clone within the repertoire. A small number of individual amplified clones with greater representation were observed in healthy individuals in our sample set, with the largest clonal populations (seen in patients aged 54 and 68 years) accounting for 0.15% to 1.5% of total sequences of the observed sequence space from circulating B cells. These larger expanded clones may be the result of physiological responses to environmental antigens or pathogens; alternatively, these could represent the precursors to lymphoid malignancies, such as chronic lymphocytic leukemia, which have a strong association with advanced patient age. Recent and older literature describing monoclonal B cell lymphocytosis (MBL) using multiparameter flow cytometry assays to detect B cells with aberrant surface protein expression has indicated that between 5% and 12% of adults have these atypical B cell populations, and essentially all patients who develop chronic lymphocytic leukemia can be shown to have had preceding MBL. An important caveat is that most patients who show MBL do not go on to develop chronic lymphocytic leukemia.
High-throughput immune receptor sequencing provides an unprecedented degree of sensitivity and specificity in tracking monoclonal B cell expansions and enables detection of clonal B cell populations that do not show aberrant cell surface marker expression. Deep sequencing of IgH rearrangements simplifies the assessment of overt populations of suspected malignant B cells in clinical samples and shows success in MRD testing after treatment of leukemia patients. A substantial advantage of the MRD detection approach used here is that all patient samples can be analyzed with a single uniform assay rather than having to tailor individual real-time PCR assays to each patient's clonal malignant sequence and to validate these assays individually as unique clinical tests, an expensive and laborious process likely to limit the accessibility of MRD testing. Having a sequence-based assay that can detect variants from the original malignant clonal sequences present at diagnosis is an advantage in screening for disease relapse. Recent microarray-based data from studies of acute lymphoblastic leukemias suggest that genomic copy number changes may occur relatively frequently at immune receptor loci between initial diagnostic specimens and relapse specimens. For the most sensitive detection of residual disease and clonal variants in a variety of B cell neoplasms, particularly those such as follicular lymphoma that have ongoing hypermutation of rearranged IgH gene loci, one may use several different primer sets (for example, making use of all three framework regions of the IgH V genes) to avoid false-negative results that arise from mutations at primer-binding sites.
The deep sequencing approach of the present invention to lymphocyte population analysis provides insights into autoimmune and infectious diseases, medical manipulations of the immune system such as vaccination, and harmful outcomes of current therapies such as graft versus host disease after stem cell transplantation. Immune receptor sequencing in medical scenarios that involve lymphoid malignancies or immune-mediated diseases are broadly useful for gathering diagnostic, prognostic, and disease-monitoring information.
Specimens
Specimens of human peripheral blood and tissues were obtained under Institutional Review Board approved protocols at our institution (Stanford University). Samples for testing for minimal residual disease in chronic lymphocytic leukemia patients were initial diagnostic specimens from lymph node or bone marrow, and blood or bone marrow specimens taken at various time-points after chemotherapy and allogeneic stem cell transplantation. Anonymized healthy control samples from adults of various ages were obtained from blood donors. Subjects gave informed consent for blood donation and were determined to be healthy via evaluation of their suitability to act as blood donors, including screening for malignant or infectious disease history, pregnancy, current infections, travel history, and recent vaccination with live attenuated viral vaccines. Healthy donor samples were also tested and found to be negative for serologic and/or nucleic acid-based evidence of infection by hepatitis B virus, hepatitis C virus, human immunodeficiency virus types 1 and 2, West Nile virus, Treponema pallidum, and Trypanosome cruzi. Additional screening to rule out donors with allergic disorders was conducted by measuring total plasma IgE. Donors were included as healthy controls if their total IgE levels were below 25 IU/mL.
DNA Template Preparation
Peripheral blood mononuclear cells were isolated by centrifugation of diluted blood layered over Hypaque 1077 (Sigma-Aldrich, St. Louis, Mo.). The peripheral blood of a healthy adult typically contains between 200-500 B cells per microliter. Twenty-micron sections of formalin-fixed and paraffin-embedded tissue samples were extracted with xylenes, washed with ethanol, and subjected to proteinase K digestion prior to DNA purification. Column purification (Qiagen, Valencia, Calif.) or magnetic bead—based isolation (Magnapure, Roche Diagnostics Corporation, Indianapolis, Ind.) was used to purify the DNA templates.
PCR Primer Design
The BIOMED-2 consortium has developed a clinically validated set of DNA primers for immune receptor amplification, including 7 sequences that anneal to framework region 2 (FR2) of IgH V gene segment family members and a common IgH J sequence. The initial evaluation of patient samples in this study was done by capillary electrophoresis of BIOMED-2 amplicons. For sequencing experiments, these primers were augmented with additional sequence elements at the 5′ ends to permit emulsion PCR, amplicon capture, and pyrosequencing. A 6-, 7- or 10-nucleotide unique sequence “barcode” was also added to identify the sample from which particular amplicon products are derived. Barcodes were designed to differ from each other at 2 or more nucleotide positions and to not contain polynucleotide repeats. High-fidelity “ultramer” synthesis chemistry was used for all primers (Integrated DNA Technologies, Coralville, Iowa).
PCR Amplifications and Sequencing Sample Preparation
PCR amplifications were performed using 100 or 200 ng of template genomic DNA, 10 pg of each primer, and 0.5 μL of AmpliTaq Gold enzyme (Applied Biosystems, Foster City, Calif.) per 50-μL reaction. Initial PCR amplification used the following program: (95° C. for 10 min); 35 cycles of (95° C. for 30 s, 58° C. for 45 s, 72° C. for 45 s); (72° C. for 10 min). To minimize the incidence of heteroduplexes in the final sample, 10 μL of the PCR products were amplified for 2 additional cycles in fresh PCR mix. The length of the PCR products obtained using the FR2 primer set in these experiments was 250-300 base pairs. Amplicons were pooled in equal amounts and purified by 1.5% agarose gel electrophoresis and gel extraction, with dissolution of the gel slice at room temperature in lysis buffer prior to column purification (Qiagen, Valencia, Calif.). For each specimen used for Experiments 1 and 2, a single PCR amplification was performed, with the exception of healthy donor 1, where 6 replicate amplifications were performed from the sample obtained at time point 1, and 7 replicate amplifications were performed from the sample collected at time point 2. For the minimal residual disease specimens described in Table 2, 2 independent PCR amplifications were performed. For the additional healthy control subjects described in Table 4, 6 independent PCR amplifications were performed and pooled for sequencing.
High-Throughput Pyrosequencing
The total DNA concentration in amplicon library pools was quantified with the PicoGreen fluorescence assay (Invitrogen, Carlsbad, Calif.). Sequencing data presented in this paper are derived from 4 independent experiments performed on the 454 GS-FLX instrument (454 Life Sciences, a Roche Company, Branford, Conn.). Two of the runs were performed using Standard chemistry and the remaining two runs were performed using Titanium chemistry, with long-range amplicon pyrosequencing beginning from the “B” primer in the manufacturer's protocol. The two Titanium sequencing runs also contained sequences for experiments and samples apart from those discussed in this study.
Sequence Data Analysis
Sequences from each input specimen were sorted based on recognition of a perfect match of the sample barcode, as well as a perfect match to the first 3 bases of the IgH J common primer. Sequences without perfect match barcodes were not considered further. Alignment of rearranged IgH sequences to germ line V, D and J segments and determination of V-D and D-J junctions were performed using the IgBLAST algorithm (National Center for Biotechnology Information). Sequences that contained single base pair insertions or deletions in the V or J gene segments were filtered from the data set, based on the known error properties of pyrosequencing. One 454 Standard FLX-derived data set (Experiment 1) in which two additional PCR cycles were performed on the pooled sample library before sequencing showed some evidence of trace cross-contamination of highly abundant sequences from malignant specimens in other samples, but this artifact was absent from a complete replicate experiment in which PCR of the pooled libraries was not performed (Experiment 2). Some apparently artifactual non-Ig sequences were present in the sequenced samples and were filtered prior to analysis.
Clonality Calculations
“Coincident sequences” are defined as those with identity in V, D, and J segment usage, and in V-D and D-J junctional bases. Clonally expanded cell populations should be detectable by the presence of coincident sequences in distinct amplicon pools. We drew no conclusion from sequences repeatedly observed within a single amplicon pool, as such “intra-pool” identities can conceivably result from amplification of a single initial molecule during PCR. Our initial calculations of diversity in a healthy B cell repertoire were predominantly derived from 13 replicate samples from a healthy individual: Six independent amplicon pools from an initial time point and seven independent amplicon pools from a time point taken 14 months later. Inter-pool coincident sequences were rare in our data, accounting for a total of 19 sequences in the six pools from time point 1 and 25 sequences from the seven pools from time point 2 (these were from a total of 10,921 and 7450 distinct sequences determined for the two time points, respectively). We did not identify any sequences that were present in 3 or more amplicon pools in this initial subject. A lower bound for the blood B cell IgH repertoire (“IgHR”, the total number of distinct IgH sequences present in the peripheral blood of an individual) can be calculated from these data based on the minimum expected number of coincidences that would have been generated by various values of IgHR. The minimum expected number of binary sequence coincidences is achieved under the condition of equal representation of each sequence type (any preferential representation of one or more sequence types for a given value of IgHR would only increase the coincidence frequency).
The minimal expected coincidence number (for example, for time point 1) is given by the following formula: (½)*Σi=1 to 6(S i*Σ(j=1 to 6; j≠i){1-[1−(1/IgHR)]Sj}) where S1 . . . S6 are the numbers of distinct sequences determined for the six independent amplicon pools prepared from that time point. Starting with an arbitrary sequence in one amplicon pool, (1/IgHR) is the probability that any single determined sequence in another pool would match that sequence, 1−(1/IgHR) is the probability that any single determined sequence in another pool would not match that sequence, and [1−(1/IgHR)]Sj is the probability that no determined sequence from an amplicon pool with Sj determined sequences would match that sequence. The value [1(1/IgHR)]Sj, summed for every element in the sequence set, yields a value that is twice the expected number of coincidences (because this sum counts each binary coincidence once for each participating sequence and hence twice in total). The formula above represents this sum. A related probabilistic calculation is that of the largest clonal expansion that could be present in the blood without being detected in more than 2 amplicon pools.
In our analysis of the 13 independent amplicon pools from the two distinct time points for blood samples from healthy donor 1, the lack of any sequence detected in three or more pools gives an upper bound on the maximum clone size. As a sample calculation, if 10% of sequences derived from an individual were from a single clone, then an arbitrary group of 1000 independent sequences could only avoid this sequence if each independent sequence were from the other 90% of available rearrangements. For 1000 sequences, this probability is (0.9)1000 or 1.7×10−46; under these circumstances it is virtually certain that at least three (and in fact all six) of the amplicon pools would contain the 10% clone. Taking into account the slightly different numbers of sequences from the amplified clones, our data yield a conclusion that a clone making up >1/1000 of the sequenced repertoire (or approximately 2 million cells) would have a >99% chance of being recovered in at least three different amplicon pools at the first time point (see Clonality and Diversity Calculations for detailed computation). These calculations were repeated for the additional 23 healthy subjects described in Table 4; for the two subjects in which larger clones were detected in all 6 replicate sequencing samples, an estimation of clone size was obtained by dividing the number of sequence reads from the amplified clone by the total number of sequence reads for that individual.
Analysis of V, D, and J Segment Combination Frequencies
Nonrandom representation of pairs of V, D or J segments in the healthy control IgH sequence data sets was assessed by comparing the frequency of pairwise combinations of segments to the product of the individual segment frequencies. Combinations of segments that showed consistent under- or over-representation in large numbers of independent samples were considered as candidates for nonrandom association. The three most significant over-represented combinations in our data set were validated using a false discovery test with 10 million randomly shuffled versions of the experimental dataset.
PCR-Based CLL Minimal Residual Disease Tests
Quantitative real-time PCR monitoring of minimal residual disease was performed using patient allele-specific oligonucleotides as previously described. When consensus probes were unsuitable, probes specific for the clonal third complementarity region were designed. Minimal residual disease assays were performed on the ABI 7900 (Applied Biosystems).
Sequencing primers have the general design:
5′ [sequencing instrument primer sequence][barcode][gene-specific primer sequence] 3′
For clarity, the primer tables listed below contain only nucleotide sequence of the gene-specific region, which will be understood by one of skill in the art to be linked to an appropriate barcode (for example as described in the table above); and linked to a sequencing instrument primer sequence. For example, an exemplary 454 sequencing instrument primers would be the
However, other sequencing instrument primers may be used as appropriate for the platform.
Detailed Description of Clonality and Diversity Calculations
i. Introduction to this section: In what follows, we describe three calculations related to diversity and clonality in the immune response. The repertoire of immune rearrangements consists of a remarkable diversity of potential receptors that can be encoded. For the immunoglobulin heavy chain, the product of 27 D regions*6 J regions*>50 V regions*two junctions [which can have any sequence and variable length, hence millions of possible sequences] yields a virtually limitless set of possibilities. We use this information to categorize each sequence. This assignment also yields a grouping of individual sequences in terms of their origins. Several points should be made in advance about the calculations to be described. 1. Each yields a lower bound as to the diversity in the populations of cells present in the B cell repertoire. Upper bounds cannot simply be estimated, as we cannot rule out the possibility of very large numbers of rare clones (e.g., single-cell clones) that would contribute rather modestly to the total number of cells but substantially to the diversity of specificities. The lower bounds derived from this analysis are nonetheless remarkable and are of considerable interest in understanding the immune repertoire and immune responses. 2. For several of the calculations, we begin with a very restrictive (and likely unrealistic) assumption that a subset of individual rearrangements are represented in equal numbers in a population; we calculate frequencies based on this assumptions, and then show that the result provides an lower bound for a “real world” situation in which individual rearrangements are represented at different frequencies in the population. 3. There is a PCR-based redundancy in the amplicon pools. That is, there are many cases in which a single sequence has been captured in the PCR reaction that gave rise to a given amplicon pool, with several amplified products from the same original template molecule sequenced. This is highly evident in looking at the complex normal samples in that the number of coincidences between samples (identical sequence assignments that occur in more than one aliquot) is much fewer than the number of coincidences within each sample. We also see a high number of coincident sequences (data not shown) in duplicate runs of sequence from the same amplification reaction in different 454 sequencing runs. Thus much of the sequence redundancy in the individual aliquots is due to simple “oversampling” of the sequence pool. ii.
For the purposes of complexity estimates, we define a number of terms that refer to groups of sequences either in the biological repertoire or in the experimental dataset.
Estimates of upper bound for amplified clonal populations: For any given clone (which we'll arbitrarily designate “Clone1”) in a given individual (who we'll arbitrarily designate “Individual1”), a key value is the number of times that clone is represented in the total blood of the individual. The number of total rearranged rIgH-DNA segments in Individual1 with the Clone1 rearrangement is designated Individual1.incidence(Clone1). We can also define a total number of rearranged IgH DNA segments as Individual1.incidence(AllClones). The relative incidence of Clone1 in the population is then given by Individual1.incidence(Clone1)/Individual1.incidence(AllClones) We will abbreviate this with the value f1, or long-form Individual1.Fraction(Clone1). So far, f1 is a property of the individual and the clone and not an experimental value. We can reasonably assume that for each rIgH-DNA included in an Aliquot, there is a probability f1 that that rIgH-DNA corresponds to Clone1. The segment then needs to jump through two hoops before we sequence it. First it has to be captured using the V- and J-segment specific primers during the initial amplification reaction, so that it starts amplifying, second it needs to be present in the pool of sequences that are actually utilized for emulsion PCR and sequencing in the 454 instrument protocol. It is conceivable for any given sequence that either of these processes is highly inefficient (or unusually efficient), so that Clone1 would be under-represented or over-represented in the pool of determined sequences. We use the adjusted value fa to indicate the value f adjusted to account for sequence-specific differences in capture and/or amplification. Thus fa is the fraction of sequences (in a large sample set) that would be available and sequence-ready following a requirement for initial capture and amplification. Each Aliquot's data set derives from a sample subset with a certain number of cells, each represented a variable number of times in the sequence data from that Aliquot. Of the clones present in the original sample and not represented in the data obtained from a specific Aliquot's dataset, a fraction will have been lost due to not being in the Aliquot (due to a finite size of the aliquot), a fraction will be absent due to not having been captured in the initial PCR amplification reaction, and a fraction will be present in the amplified mixture of sequences from the aliquot but not been fortunate enough to attach to a bead that was actually sequenced. For an exemplary Aliquot (“Aliquot1”) and an exemplary clone (“Clone1”), fa1 is the probability that any individual independently-derived read that is present in Aliquot1's dataset would have come from Clone1. The number of opportunities to choose an instance of Clone1 in the sample will be the number of independently-derived sequences present in the Aliquot1 dataset (which we'll call “S1”). For a blood sample containing an extensive diversity of B cells (i.e. cases in which coincidences between independent aliquots of the same blood sample are a small fraction of the total sequence diversity, a condition met by all of the healthy samples in our dataset), S1 is very close to the number of unique Clone identities (which we'll call “U1”) represented in the aliquot's DNA sequence read dataset. For samples with greater redundancy (e.g. for future analysis of samples from individuals undergoing concerted immune responses), there is some potential under-estimation of S1 due to the possibility that more than one sequence from an individual clone was independently captured and is represented in a given Aliquot's dataset. That this effect is extremely small in our normal samples is evident in that the numbers of coincidences between samples is very small relative to the total number of unique sequences in each sample. The correction value in the case where there were a more substantial number of coincidences would be S1=U1/(1−frco) where frco is the fraction of sequenced rIgH-DNAs that are co-detected in an equivalent but independent aliquot of the same DNA. Although a precise value for frco is not easily calculated from the data, a rough estimate and hard upper bound can be experimentally determined as the fraction of sequence Reads that are shared between independent and equivalent aliquots. Given the observed coincidence frequencies, we again stress that the adjustment for our normal samples (on S1 values that are in the hundreds and thousands and coincidence values generally in the single digits) is negligible. At this point, we can describe the probability of an arrangement of positive and negative results for detection of a particular clonal sequence within n different aliquots based on S1, S2 . . . Sn and fa. As an example, the probability that S1 and S2 would be positive for Clone1 but S3, S4, and S5 would be negative would be (1−(1−f)S1)*(1−(1−f)S1)*((1−f)S3)*((1−f)S4)*((1−f)S5). This is just one way that we can get a situation where there are two positive and three negative aliquots. To get a more complete picture, we can take all subsets P of the set 1 . . . n which contain up to two members and calculate
Plugging the values of Si in for the six independent replicates of the first time point on the normal blood sample, we get an estimate of how frequently a clone Clone1 of frequency f would have appeared in at least three of the individual aliquots. Reassuringly, this value is near 100% for a value of fa that is near 1, and 0% if fa is vanishingly small. At fa=1/1040, this value crosses 99%. This says we'd have had a >99% chance of having 3 or more aliquots positive for any clone whose occurrence in the detected B cell repertoire was >1/1040. This would correspond to any clone consisting of more than approximately 2,000,000 B cells in a total blood B cell population of 2×109 cells.
Estimating a lower bound for Ig-rearrangement diversity in each sequenced aliquot Next we calculate the minimum numbers of individual sequences that were captured and amplified in the sequenced libraries. Here we start with the assumption of a set of sequences that are present in the amplified pool, each with a distinct probability of being recruited in the emulsion PCR for 454 sequencing. Some of these sequences may be (and in this case certainly are) represented at higher levels than others in the 454 sequencing pool. For this calculation, we are interested in what number of sequences are present in this pool at levels comparable to those that are captured for sequencing in the incidence=1 class but which were “unlucky” in not getting picked. The distribution of frequencies for individual classes of sequence can be thought of as a sum of Poisson distributions for different probabilities of inclusion. We can then use P[1] and P[2] to get a lower bound for P[0], using the fact that any variation in values of the probability of capture fa between different groups of clones will actually increase the frequency of P[0] relative to what would be predicted from the simple Poisson model. To make our lower bound estimate of the P[0] class, we use the facts
P[0]=exp(−fa*Si)
P[1]=(fa*Si)*exp(−fa*Si)
P[2]=(fa*Si)̂2*exp(−fa*Si)/2
So P[0]=P[1]*(P[1]/P[2]*2)
It should be stressed that this leads us to a lower limit on P[0], since any variation, particularly at the low end in fa will yield a class of clones with lower inclusion frequency that could be quite large. When these calculations are performed on the real data from the canonical healthy human sample, we obtain the following numbers:
Table of Calculated Sequence Incidences in PCR Amplicon Pools
Samples S01, S02, S24, S25, S28, and S29 are from time point 1 while S03, S04, S26, S27, S30, S31, and S32 are from time point 2. Note that the total estimated available complexities of the pools used for 454 sequencing are in the range of 2490-4097 for time point 1 of this healthy human sample. The number of B cells added to these amplifications was estimated at 2750. This argues strongly against a large number of B cells with poorly captured, non-amplifiable, or non-sequencable rearrangements. Note also that there is an uncertainty in these calculations as to what fraction of B cells have one versus two amplifiable rearrangements. We observe in-frame stop codons in approximately 20% of the sequences from our dataset (data not shown), suggesting that non-productive rearrangements are likely to be only a modest fraction of the total sequence space. From this analysis, it is clear that a substantial fraction of functional rearrangements can be amplified in this protocol. v. Estimating a lower bound for Ig-rearrangement diversity in an individual following sequencing of several aliquots. Next we calculate a minimal diversity for the sequence space being explored in an individual for whom we have sequences from several aliquots. Start in this case with the assumption that there are a large number of micro-clones each with a frequency fa in the population. For any instance of a unique sequence in a set, the probability of that sequence showing up in any given instance in another set is fa. The probability of showing up in another specific aliquot (j) is 1−(1−fa)Sj. So the number of expected coincidences from this sequence is Σj≠i(1−(1−fa)Sj). The total number of expected coincidences from all of the sequences in Aliquot (i) is then Si*Σj≠i(1−(1−fa)Sj). The total number of coincidences between sequences in different aliquots is then expected to be Σi(Si*Σj≠i(1−(1−fa)Sj))/2 (divided by two since we've counted each coincidence twice). Note that we operate on real observed sequences P[1], P[2], . . . , not on the P[0] class above since we have no way to estimate coincidences that involve the P[0] class.
If the distribution of frequencies fa were to be non-uniform (i.e. not the same for every clone), the complexity could potentially be larger than the minimum estimates here, not smaller. The argument for this is reasonably intuitive: a non-uniform distribution of frequencies will serve to increase the frequency of coincidences for a given base of complexity. Thus the coincidences may all frequently come from a class of clones that are over-represented relative to the bulk of (relatively rare clones). We note that the above complexity estimates were based on a situation where there were binary coincidences (sequences that were present in two aliquots) but no higher order coincidences (sequences present in more than two aliquots). Where higher order coincidences are present, the reflect evidence for amplified clones (see above) but not necessarily the complexity of the IgHDNA population as a whole. Fortunately it is possible to provide an upper bound on complexity in which only true binary coincidences are considered. The following calculation allows this. The number of expected true binary coincidences from a given sequence is a Σj≠i[(1−(1−fa)Sj)*Πk≠(i or j)(1−fa)Sk]. The total number of expected coincidences from all of the sequences in Aliquot (i) is then Si*Σj≠i[(1−(1−fa)Sj)*Πk≠(i or j)(1−fa)Sk]. The total number of coincidences between sequences in different aliquots is then expected to be Σi{Si*Σj≠i[(1−(1−fa)Sj)*Πk≠(i or j)(1−fa)Sk]}/2.
Specimens.
De-identified specimens of genomic DNA from human peripheral blood mononuclear cells were obtained under Institutional Review Board approved protocols at our institution (Stanford University). Subjects were recruited at Duke University and provided informed consent, and the Duke University Institutional Review Board approved the protocols for all studies. Subjects were given trivalent inactivated seasonal influenza vaccine. Blood was drawn from immunized subjects on day 0 before vaccination and on days 7 and 21 after challenge.
Serological measurements. Subjects were classified as influenza vaccine ‘seroconverters’ or ‘non-seroconverters’ based on measurements of pre-vaccination and day 21 post-vaccination plasma antibody titers in hemagglutination inhibition assays (Coffey et al. Influenza virus. Curr Protoc Immunol Chapter 19, Unit 19 11 (2001)). ‘Seroconverters’ were those whose titer increased 4-fold above pre-vaccination baseline, or increased from undetectable at baseline to at least 1:40. Influenza stocks were grown in embryonated eggs and were titered for hemagglutination units on turkey red blood cells. To perform HAI assays, serial dilutions in PBS of plasma or transfected cell supernatants were placed into 96-well plates and were mixed with an equal volume of washed turkey red blood cells (0.5%) and incubated at room temperature for 30 min before hemagglutination was read directly from the wells.
DNA Template Preparation. Peripheral blood mononuclear cells were isolated by centrifugation of diluted whole blood over Hypaque 1077 (Sigma-Aldrich). Column purification (Qiagen, Valencia, Calif.) was used to isolate genomic DNA template.
PCR amplifications and sequencing sample preparation. PCR amplifications were performed using 100 ng of template genomic DNA for each of 6 replicate PCR amplifications for each sample. 10 pg of each primer, and 0.5 μL of AmpliTaq Gold enzyme (Applied Biosystems, Foster City, Calif.) per 50 μL reaction were used. Primers are shown in Table 6. Initial PCR amplification used the following program: (95° C. for 10 minutes); 35 cycles of (95° C. for 30 seconds, 58° C. for 45 seconds, 72° C. for 90 seconds); (72° C. for 10 minutes). 10 μL of the products were amplified for 2 additional cycles in fresh PCR mix to minimize heteroduplexes in the final product (Boyd, S. D., et al. Sci Transl Med 1, 12ra23). Amplicons from the various replicate PCR reactions for all samples were pooled in equal amounts and purified by 1.5% agarose gel electrophoresis and gel extraction, with dissolution of the gel slice at room temperature in lysis buffer prior to column purification (Qiagen, Valencia, Calif.).
High-throughput pyrosequencing. Amplicon library pools were quantitated by real-time PCR (Roche, Connecticut) or PicoGreen fluorescence assay (Invitrogen, Carlsbad, Calif.), (Parameswaran et al. (2007) Nucleic Acids Res 35, e130). Sequencing was performed on the 454 instrument using Titanium chemistry, with long-range amplicon pyrosequencing beginning from the “B” primer in the manufacturer's protocol (Roche, Connecticut).
Sequence data analysis. Sequences from each input specimen were sorted based on recognition of a perfect match of the sample barcode and the IgHJ common primer, while individual replicate libraries from each sample were identified by a perfect match to the V primer barcode and IgHV segment primers. Alignment of rearranged IgH sequences to germline V, D and J segments, and determination of V-D junctions and D-J junctions was performed using the IgBLAST algorithm (NCBI) and the iHMMune-align algorithm (Gaeta et al. (2007) Bioinformatics 23, 1580-1587). Sequences containing single base-pair insertions or deletions in the V or J gene segments were filtered from the dataset, based on the known error properties of pyrosequencing (Huse et al. (2007) Genome Biol 8, R143; Margulies et al. (2005) Nature 437, 376-380; Johnson et al. (2006) Genome Res 16, 1505-1516). Artifactual non-IgH sequences in the data were filtered out prior to further analysis.
Detection of ‘Coincident Sequences’ Providing Evidence for Amplified B Cell Clones.
“Coincident sequences,” (those with identity in V, D, and J segment usage, and in V-D and D-J junctional bases) were detected as previously described (van Dongen et al. (2003) Leukemia 17, 2257-2317) by being identified in IgH amplicons sequence libraries from independent PCR replicates from an individual, and provide evidence of clonally expanded B cell populations. Sequences repeatedly observed within a single amplicon pool were not taken in isolation as evidence of an expanded B cell clone, as such “intra-pool” multi-copy sequences could be the result of amplification of a single initial molecule during PCR. In comparing the total number of ‘coincident sequences’ between samples or individuals, the total number of copies of a sequence within a replicate PCR library were included.
Normalization of ‘Coincident Sequence’ Counts.
To normalize the counts of ‘coincident sequences’ detected in samples that had been sequenced to different depths (i.e., where more total sequences had been obtained from the replicate amplification pools of one sample compared to another), the total number of coincident sequences detected between the PCR replicate libraries of a sample was divided by the total number of possible pairwise comparisons between sequences in different PCR replicate library data sets from that sample.
Prediction of Seroconversion Status by Measurement of Normalized Coincident Sequence Counts Following Vaccination.
Evidence of expanded B cell clones in the blood following vaccination is provided by detection of coincident IgH VDJ sequences. Normalized counts of these coincident sequences using the approach described in the Materials and Methods reveal a diagnostic characteristic at day 7 post-vaccination of increased normalized coincident sequence counts in individuals who demonstrate vaccine-specific seroconversion (adequate increase of neutralizing anti-viral titers) at day 21 post-vaccination. There was a significant difference in the day 7 coincident sequence counts between the two groups (p=0.03, Mann-Whitney test) prior to normalization.
Change of hypermutation level of IgH V segments in coincident sequences at day 7 post-vaccination correlates with seroconversion. B cell clones in the blood (as assessed by detection of coincident sequences) following vaccination show consistently higher levels of hypermutation of IgH V segments compared to pre-vaccination samples in subjects who demonstrate vaccine-specific seroconversion (adequate increase of neutralizing anti-viral titers) at day 21 post-vaccination. Hypermutation levels are counted by comparing V segments in rearranged V(D)J to germline V segment sequences. The sequences set forth in Table 6 generally include a sequencing instrument primer, as indicated, linked to a barcode and a gene specific primer.
| Number | Date | Country | |
|---|---|---|---|
| 61459666 | Dec 2010 | US | |
| 61476182 | Apr 2011 | US |
