WO2018162696A1 - Common genetic variations at the tcra-tcrd locus control thymic function in humans - Google Patents

Abstract

The invention relates to a method of evaluating thymic function in a subject comprising the detection of a genetic polymorphism in the T-cell receptor alpha-T cell receptor delta (TCRA-TCRD) locus associated with the level of T lymphocyte production by the thymus. The invention relates also to the use of said method and genetic polymorphism for the diagnostic, prognostic, treatment or monitoring of conditions or clinical situations where thymopoiesis is impacted, or that are impacted by thymopoiesis efficiency and/or quality.

Description

COMMON GENETIC VARIATIONS AT THE TCRA-TCRD LOCUS CONTROL THYMIC FUNCTION IN HUMANS

The invention pertains to the field of precision medicine using genetic biomarkers. The invention relates to a method of ev aluating thymic function in a subject comprising the detection of a genetic polymorphism in the T-cell receptor alpha-T cell receptor delta (TCRA-TCRD) locus associated with the level of T lymphocyte production by the thymus. The invention relates also to the use of said method and genetic polymorphism for the diagnostic, prognostic, treatment or monitoring of conditions or clinical situations where thymopoiesis is impacted, or that are impacted by thymopoiesis efficiency and/or quality, such as for example aging, allograft transplantation, acquired immunodeficiencies such as HIV/AIDS, vaccination, infectious diseases, cancer, autoimmune diseases and immunotherapy.

The thymus is the primary lymphoid organ where T lymphocytes are generated in the adaptive immune system of all vertebrates, through spatio-temporal interactions between thymocytes and specialized microenvironments (Shah et al., J Immunol, 2014, 192, 4017-4023: Calderon el al.. Cell, 2012, 149, 159-172). It is an organ sensitiv e to insults received throughout life upon inflammation and infections, reflected in its functional decline with age (Douek et al.. Nature, 1998, 396, 690-695; Palmer, Front. Immunol., 2013, 4, 3 16). It is however an extremely plastic tissue endowed with endogenous regenerative capacities following an acute damage during chemotherapies or irradiation (Boehm. J. B. Svvann, Nat. Rev . Immunol., 2013, 13, 83 1 -838; Lopes. EM BO Moi. Med., 201 7, 9, 835-851; Wertheimer et al., Sci Immunol. 201 8. 3 ). Thymus is a v ital organ for homeostatic maintenance of the peripheral immune system. In healthy individuals, continuous production of naive self-tolerant T cells by the thymus ensures potent immune responses towards newly encountered antigens from pathogens or tumors and contributes to maintenance of the naive T-cell repertoire.

Although thymic function has been extensively studied for its capacity to shape the adaptiv e immune repertoire through positiv e and negativ e selection during thymocyte differentiation ( Von Boehmer et al., Immunol. Rev., 2003, 191 . 62-78; Mathis et al, Annu. Rev . Immunol., 2009, 27, 287-3 1 ), much less is known about the env ironmental or genetic determinants affecting the quantitative aspects o f thymopoiesis in normal individuals, especially in humans.

Thymic function is high in newborns and during infancy, and diminishes with age (C. L. Mackall, R. E. Gress, Immunol. Rev., 1997, 160, 91 -102). Dysregulated thymopoiesis is associated with an increased risk of opportunistic infections, autoimmunity, cancer and inefficient vaccination in the elderly (D.D. Taub, D. L. Longo, Immunol. Rev., 2005, 205, 72-93). However, levels of thymic function vary significantly among individuals, and some adults have persistent, although reduced, thymopoiesis until at least their fifth decade of life (C. L. Mackall, R. E. Gress, Immunol. Rev., 1997, 160, 91-102; Mitchell WA, Lang PO, Aspinall R, Clin. Exp. Immunol, 2010, 161 , 497- 503). To date, the biological mechanisms underlying these natural variations have not been identified.

In this context, assessing the environmental and genetic determinants of variation in thymic function across healthy adults is of primary importance to delineate targets for new thymic regenerative or boosting therapies (T. Boehm, J. B. Swann, Nat. Rev. Immunol., 2013, 13, 831-838) in particular in conditions where thymopoiesis is impacted, such as aging (S. Ferrando -Martinez et al, Age (Dordr), 2013, 35, 251 -259), human immunodeficiency virus (HIV) infection (M. L. Dion et al , Immunity, 2004, 21 , 757-768) or allogeneic hematopoietic stem cell transplantation (ailo-HSCT) (W. Krenger, B. R. Blazar, G. A. Hollander, Blood, 201 1 , 1 17, 6768-6776). Thymopoiesis is exquisitely dependent on the thymic microenvironment which is broadly demarcated into an outer cortex and inner medulla, each defined by different subsets of thymic epithelial cells (TEC) (Abramson et al, Annu Rev Immunol, 2017, 35, 85-1 18). A bilateral crosstalk between thymocytes and thymic stromal cells directs sequential intrathymic T- cell development and maintains thymic stromal niches activity (Shah et al, J Immunol, 20 1 4, 1 92, 40 1 7-4023; Kurd et al., Immunol Rev, 20 1 6, 27 1 , 1 14- 1 26 ). Thymocyte progenitors receive signals from cortical TEC (cTEC) for their commitment to the T-cell lineage via the engagement of the NOTCH 1 receptor with Delta-like 4 (DLL4) ligand ( Hozumi et al.. J Exp Med, 2008, 205. 2507-25 1 3 ) a major FOXN l target in the thymic epithelium (Calderon et al.. Cell, 20 1 2, 1 49, 1 59- 1 72; Zuklys et al, Nat Immunol, 20 1 6, 1 7, 1 206- 1 2 1 5 ). The medulla, via medullary TEC (mTEC) and dendritic cells, has a critical role in establishing self-tolerance by negative selection and induction of regulatory T cells (Tregs), especially but not exclusively via mTEC expressing the autoimmune regulator (AIRE) gene (Abramson et al., Annu Rev Immunol, 2017, 35, 85- 1 18). The thymic function decline is an early hallmark of aging, linked with an increased risk in opportunistic infections, autoimmunity, cancer and incHicacy of vaccination in the elderly. Naive T cells are heterogeneous including so-called recent thymic emigrants (RTE), a subset which undergoes further post-thymic maturation after positive selection ( Fink et al.. Nature reviews. Immunology 201 1 , 1 1 , 544-549). It is for instance more prone than mature naive T cells to differentiate into Tregs (Paiva et al., Proc Natl Acad Sci U S A, 2013, 1 10, 6494-6499) and possess specific metabolic characteristics (Cunningham et al., J Immunol 201 7, 1 98, 4575-4580). Some phcnotypic markers have been proposed to stain RTE, such as CD31 (PECAM-1) in CD4 T cells. However, CD31 expression can be maintained during cytokine-driven proliferation of CD4 T cells, making expression staining uncertain to interpret in terms of thymopoiesis (Kimmig et al., J Exp Med, 2002, 1 95, 789-794; Azevedo et al.. Blood, 2009, 113, 2999-3007 ). RTE are enriched in T-celi receptor excision circles (TRECs) produced during thymic TCR somatic recombination (Junge et al., Eur J Immunol, 2007, 37, 3270-3280) (Figure 1 A and I B ). TRECs persist within mature T cells as episomal DNA (de Villartay et al.. Nature, 1 988, 335. 1 70- 1 74) cannot replicate and are diluted out by peripheral cell divisions. Their quantification in peripheral blood provides a non-invasive surrogate marker of thymopoiesis, especially relevant in steady-state homeostatic conditions of the T-cell compartment.

The level of thymic function of a given individual can be evaluated on peripheral blood by a direct, non-invasive quantification of signal-joint (sj) and beta (β) T-cell

Receptor (TCR) Excision Circles (TRECs) that does not require in vitro cell culturing ( Figure 1 A, IB). TRECs are small circular DNAs generated during TCR somatic recombination that persist within T cells as episomal DNA (J. P. de Villartay et al , Nature, 1988, 335, 1 70- 1 74 ). Signal joint TRECs (sj TRECs) are generated during the recombination of the alpha chain of the TCR, in all double-positive (DP) CD4 CD8 thymocytes, before positiv e and negative selection and lineage commitment (de Villartay et al.. Nature, 1988, 335, 1 70- 1 74). They result from the deletion and recirculanzation of the T-cell receptor delta (TCRD or TRD) locus embedded within the T-cell receptor alpha ( TCRA or TRA ) locus ( Figure I B). A vast majority of recently exported naive αβ T cells circulating in periphery, do carry one sjTREC, indicator of a successful TCRA rearrangement. As sjTRECs do not replicate, their quantification in peripheral blood by R T -PCR provides a reliable assessment of the proportion of recent thymic emigrants (RTE) within a given population - and thus of the level of thymic function. The sjTRECs quantification assay is used in clinical laboratories as a diagnostic test for recovery of the naive T-cell repertoire during H IV treatment, after allo-HSCT and in the screening of severe combined immunodeficiencies in newborns (M. L. Dion et al , Immunity, 2004, 2 1 , 757-768; D. C. Douek et al, Nature, 1998, 396, 690-695; E. Clave et al, Blood, 2009, 1 13, 6477-6484: A Kwan et al , JAMA, 2014, 3 1 2, 729-738 ). Similar PCR-based assays are available to measure β TRFK^'s, generated during the TCRB recombination at the double-negative (DN) 2/3 thymocyte differentiation stage transition (T. Kreslavsky et al, Immunity, 201 2. 37, 840-853 ) (Figure 1.4). pTRECs are much less frequent than sjTRECs in the periphery and frequently below the detection threshold in quantitative PGR. Given the dilution of βTRECs at each cell division between pTRECs and sjTRECs generation, the iog2 -transformation of the sjTRECs/pTRECs ratio gives an estimate of the number of intrathymic div isions occurring between CD4-CD8- double negative 3 (DN3) and DP stages ( Figure 1A) (M. L. Dion et al , Immunity, 2004, 2 1 , 757-768 ).

To quantify natural variation in thymopoiesis under homeostatic conditions, the inventors have quantified sjTRECs and PTRECs in peripheral blood of 1 ,000 age- and sex-stratified healthy adults of the Milieu Interieur cohort (S. Thomas et al., Clin. Immunol. 1 57, 201 5. 277-293 ). Among 40 non-genetic variables, age and sex were the only factors substantially impacting thymic function. TRECs values decreased with age (5% per year, P =2.x 10^{- 16}') and were higher in women compared to men (66% increase, P =2x10^-16). Genome-wide association studies (GWAS) (Sex-stratified and two-stage) revealed common variants within the TCRA- TCRD locus between DDI and DD3 gene segments associated with TRECs generation in men and women. The variants included twenty-two SNPs in linkage disequilibrium at a genome-wide significant threshold ( P < 1 .5 x 10^-7) with four SNPs (rs8013419, rs 1087301 8, rs12147006, rs2204985 ) in linkage disequilibrium at a genome-wide statistical significance ( P <2x 10^-8 ) in metaanalysis. Strikingly, transplantation of human hematopoietic stem cells with the rs2204985 GG genotype into immunodeficient mice led to thymopoeisis with higher levels of TRECs, of thymocytes counts and a higher TCR repertoire diversity. This systems immunology approach identifies a genetic control of thymopoiesis in healthy adults, which has a broad impact in precision medicine. The detection of these variants provides a simple, cheap and hightroughput method to evaluate thymic function, i.e. a immune competence index, in a subject by determining a genotype associated with thymic function level in a subject. The method and variants are useful for the diagnostic, prognostic or monitoring of conditions or clinical situations where thymopoiesis is impacted, or that are impacted by thymopoiesis efficiency and/or quality, such as for example aging, allograft transplantation, acquired immunodeficiencies such as H IV/A IDS, vaccination, infectious diseases, cancer, autoimmune diseases , and immunotherapy.

Therefore, the invention relates to an in vitro method of evaluating thymic function in an indiv idual, comprising:

- determining, in a biological sample containing nucleic acid obtained from the individual, the presence or absence of at least one allele of at least one polymorphic marker in the T-cell receptor alpha-T cell receptor delta (TCRA-TCRD) locus that is associated with increased level of production of T lymphocytes by the thymus, wherein the presence of said allele is indicative of increased thymic function for the individual.

In the present invention, the allele that is associated with increased level of T lymphocyte production by the thymus (i.e. , increased thymic function) is named "effect allele". Unless specified otherwise, the aiiele(s) that are not associated with increased level of T lymphocyte production by the thymus are named other allele(s). Thus determination of the presence of the effect allele is indicative of increased thymic function for the indiv idual. Determination of the absence of the effect allele is indicative that the individual does not have the increased thymic function conferred by the effect allele. The level of production of T lymphocytes by the thymus is proportional to the number of effect alleles (0, 1 or 2) present in the individual (Figure 1 1 A and 1 I B). The highest level is found in individuals homozygous for the effect allele, the lowest level is found in individual homozygous for the other allele and an intermediate level is found in heterozygous individuals.

Definitions

- "Thymic function", as described herein, refers to thymopoiesis, which is the production of (naive) T lymphocytes by the thymus. The level of thymic function of a given indiv idual at a given time can be evaluated on peripheral blood by quantification of signal-joint (sj), beta (β) T-cell Receptor (TCR) Excision Circles (TRECs) and or number of intrathymic divisions (log: of the ratio of sjTREC number over β TREC number (log₂ sjTREC/pTREC), according to standard methods based on the principles illustrated in Figure 1 A and IB. These standard methods which are well-known in the art are disclosed in the examples of the present application.

- A "polymorphic marker", sometimes referred to as a "marker", as described herein, refers to a genomic polymorphic site. Each polymorphic marker has at least two sequence variations characteristic of particular alleles at the polymorphic site. Thus, genetic association to a polymorphic marker implies that there is association to at least one specific allele of that particular polymorphic marker. The marker can comprise any allele of any variant type found in the genome, including SNPs, mini- or microsatellites, translocations and copy number variations (insertions, deletions, duplications). Polymorphic markers with population frequency higher than 5- 10% are in general most useful.

- A "polymorphic site" referred to a nucleotide position at which more than one sequence is possible in a population.

- An "allele" refers to the nucleotide sequence of a given locus (position) on a chromosome. A polymorphic marker allele thus refers to the composition (i.e., sequence) of the marker on a chromosome. Genomic DNA from an indiv idual contains two alleles (i.e. allele-specific sequences) for any given polymorphic marker, representative of each copy of the marker on each chromosome.

- A "Single Nucleotide Polymorphism" or "SNP" is a DNA sequence variation occurring when a single nucleotide at a speci fic location in the genome differs between members of a species or between paired chromosomes in an individual. Most SNP polymorphisms have two alleles. Each individual is in this instance either homozygous for one allele of the polymorphism (i.e. both chromosomal copies of the individual have the same nucleotide at the SNP location), or the individual is heterozygous (i.e. the two sister chromosomes of the individual contain different nucleotides). The SNP nomenclature as reported herein refers to the official Reference SNP (rs) ID identification tag as assigned to each unique S N P by the National Center for Biotechnological In formation (NCBI )

- A "variant", as described herein, refers to a segment of DNA that differs from the reference DNA. A "marker" or a "polymorphic marker", as defined herein, is a variant. Alleles that differ from the reference are referred to as "variant" alleles.

- "Biological sample" refers to a biological material comprising nucleic acid that is obtained from an i ndivi dual . The biological material that may be derived from any biological source is removed from the individual by standard methods which are well- known to a person having ordinary skill in the art. The biological sample is a lso named "sample" or "nucleic ac id sam ple".

- A "Biomarker" refers to a distinctive biological or biologically derived indicator of a process, event or condition. A biomarker includes a genetic marker, a protein marker and other molecular marker.

- A "woman" as described herein, refers to a female human of any age such as for example a baby, adult or elderly.

- The terms "a", "an", and "the" include plural referents, unless the context clearly indicates otherwise. For example "a marker" as used herein is understood to represent one or more markers. As such, the term "a" (or "an"), "one or more" or "at least one" can be used interchangeably herein.

The TCRA-TCRD locus is situated on human chromosome 14 and corresponds to the nucleotide sequence from positions 2 1 ,62 1 ,904 to 22,552, 132 of NCBI reference sequence NC 000014.9. In some embodiments of the above method, the polymorphic marker is in the TCRD locus and corresponds to the nucleotide sequence from to positions 22,422,546 to 22,466.577 of NCBI reference sequence NC 000014.9 or positions 22,891 ,537 to 22.935,569 of NCBI reference sequence NC 000014.8. In a more preferred embodiment. the polymorphic marker is in the region from the 5 ' end of the D delta 2 (D62) gene segment to the 3 ' of the D delta 3 ( D03 ) gene segment corresponding to positions 22,439.007 to 22.449.125 of NCBI reference sequence NC 000014.9 (Figure 3B).

In other embodiments of the above method, the polymorphic marker is characterized by the following features:

a) an effect allele frequency higher than 20, 25, 30, 35, 40, 45 or 50 %; preferably higher than 40, 45, 47 or 50 %;

b) an association between the effect allele and increased level of T lymphocyte production by the thymus (i.e. , increased thymic function ) characterized by a P-v alue of less than 5x 10°, 10^-3, 10^-4, 10^-5, 10^-6 or 10^-7: preferably of less than 7x10^-8, 5x10^-8, 2x 10^-8 or 1 .5x 1 0^-8; and

c) a linkage disequilibrium with at least one polymorphic marker selected from the group consisting of: rs381 1236, rs23() l 199, rs2301200, rs3216790, rs62762262, rs2331618, rs11091 30, rs2072616 , rs801248 1 , rs6572448. rs2141988. rs916052, rs8021297, rs7492759, rs2204984, rs71 1 1 5550, rs201497432. rs8013419, rs 1087301 8, rs12147006, rs2204985, and rs l 1 84471 5 (SEQ ID NO: 1 to 22: Table V) characterized by a numerical value of the linkage disequilibrium correlation measure r of greater than 0.2; preferably of greater than 0.3, 0.4, 0.5. 0.6 or 0.7; still more preferably of greater than 0.8 or 0.9; even more preferably of greater than 0.95 or 0.97.

The frequency values of the effect allele as presented in this application are found at least in the French population which is representative of west European Caucasian population.

The association between the effect allele and increased level of T lymphocyte production by the thymus is found at least for the number of signal -joint T-cell Receptor

Excision Circles (sjTRECs) in the indiv idual. The polymorphic marker is preferably in linkage disequilibrium with at least rs1087301 8.

In other embodiments of the abov e method, the level of T lymphocyte production by the thymus is increased by at least 10% (0.1 fold), 25% (0.25 fold), 50% (0.5 fold), 75% (0.75 fold), 100% (1 fold) , 125% ( 1 .25 fold). 150% ( 1 .5 fold), 175% ( 1 .75 fold), 200% (2 fold), 225% (2.25 fold) or 250% (2.5 fold) in individuals carrying at least one effect allele of at least one polymorphic marker. The lev el of T lymphocyte production by the thymus is preferably determined by measuring the number of signal-joint T-cell Receptor Excision Circles (sjTRECs) in the indiv idual. In other embodiments of the abov e method, the polymorphic marker is a single- nucleotide-polymorphism (SNP).

The SNP is preferably selected from the group consisting of rs38 1 1236, rs2301 199, rs2301200, rs32 16790, rs62762262, rs2331618, rs l 1 09130, rs2072616, rs8012481 , rs6572448, rs2 141988, rs916052, rs8021297, rs7492759, rs2204984, rs7 1 115550. rs201497432, rs8013419, rs l 087301 8, rs12147006, rs2204985. rs1 184471 5, and other polymorphic markers in linkage disequilibrium therewith.

The C allele of rs381 1236, the G allele of rs2301 199. the C allele of rs2301200, the CG allele of rs3216790, the TAGTC allele of rs62762262, the A allele rs233 161 8, the G allele of rsl 109130, the G allele of rs2072616, the G allele of rsSO 1248 1 , the T allele of rs6572448. the A allele of rs2141988, the A allele of rs916052, the T allele of 8021297, the G allele of rs 7492759, the A allele of rs2204984, the A allele of rs7 1 115550, the T allele of rs201497432, the A allele of rs8013419, the G allele of rs1087301 8, the A allele of rs12147006, the G allele of rs2204985 and the C allele of rsl 1844715 are indicative of an increased thymic function in an indiv idual. Thus, in certain embodiments, determination of the presence of at least one of the above-listed allele is indicativ e of increased thymic function for the individual. Determination of the absence of any of these alleles is indicative that the indiv idual does not have the increased thymic function conferred by the allele. More preferably, the SNP is selected from the group consisting of rs381 1236, rs2141988, rs2204984, rs8013419, rs1087301 8, rs12147006, rs2204985, and rs11844715. Still more preferably, the SNP is selected from the group consisting of rs8013419, rs l 087301 8, rs12147006 and rs2204985. These 4 SNPs are in the region from the 5 'end of the D delta 2 (D62) gene segment to the 3 ' end of the D delta 3 (D53) gene segment. Even more preferably, the SNP is rs2204985.

More preferably, the other polymorphic markers are characterized by numerical values of the linkage disequilibrium correlation measure r of greater than 0.2: still preferably of greater than 0.3, 0.4, 0.5, 0.6, 0.7 or 0.8; still more preferably of greater than 0.7 or 0.8.

More preferably, the other polymorphic markers are preferably in linkage disequilibrium with at least rs l 087301 8.

The 22 SNPs listed above (SEQ ID NO: 1 to 22 ) are characterized by the following features in a whole population (women (Table II); men and women (Table III)):

a) an effect allele frequency higher than about 49 % (> 48.5 %) at least in the

French population;

b) an association between the effect allele and increased blood level of sjTRECs characterized by a P-valuc of less than 1 .5.x 10^-7 : and

c) a linkage disequilibrium with rs1087301 8 characterized by a r² value of greater than 0.7.

These 22 SNPs are associated with a number of signal-joint T-cell Receptor Excision Circles (sjTR ECs) that is increased by up to 30 % in individuals heterozygous for the effect allele and up to 80 % in indiv iduals homozygous for the effect allele. Highest increase is observed with rs2204985, rs8013419, rs10873018, and rs12147006. In other embodiments of the above method, the individual is a human individual.

The individual can be of any age (baby, child, adult, elderly).

In other embodiments of the above method, the individual is a female, preferably a woman. The biological sample comprises genomic DNA. Such biological sample can be obtained from any source that contains genomic DNA, including tissue or body fluid. Non-limiting examples of body fluids include blood (whole-blood), cerebral spinal fluid (CSF), amniotic fluid, urine and mucosal secretions. Tissue sample can be with n o - limitati ons from skin, mucosa including buccal or conjunctival mucosa and gastrointestinal tract, muscle, hair, nail, tooth, placenta, or other organs. Sample includes swab.

In other embodiments of the above method, the biological sample is blood, in particular whole-blood. The blood is prelerably peripheral blood. For example, the sample is dried blood spot or dried pellet of unseparated peripheral blood lymphocytes.

In other embodiments of the above method, the biological sample is derived from blood, in particular the biological sample comprises peripheral blood mononuclear cells (PBMC).

Analyzing a nucleic acid sample may include the step of isolating genomic nucleic acid from the sample using standard methods used for the isolation of nucleic acids from biological samples.

Any method that provides the allelic identity at particular polymorphic sites is useful in the method of the invention. Suitable methods are disclosed in the art and include, for instance, whole genome sequencing methods, w hole genome analysis using SNP chips, cloning for polymorphisms, non-radioactive PCR-single strand conformation polymorphism analysis, denaturing high pressure liquid chromatography (DHPLC), DNA hybridization, computational analysis, single-stranded conformational polymorphism (SSCP), restriction fragment length polymorphism (RFLP), automated fluorescent sequencing, clamped denaturing gel electrophoresis (CDGE), denaturing gradient gel electrophoresis (DGGE), mobility shift analysis, restriction enzyme analysis, heteroduplex analysis, chemical mismatch cleavage (CMC), RNase protection assays, use of polypeptides that recognize nucleotide mismatches, such as E. coli mutS protein, allele-speci fic PCR, and direct manual and automated sequencing. In other embodiments, the method of the invention comprises at least one procedure selected from the group consisting of:

a) amplification of nucleic acid from the biological sample;

b) hybridization assay using a nucleic acid probe and nucleic acid from the biological sample;

c) hybridization assay using a nucleic acid probe and nucleic acid obtained by amplification of the biological sample, and

d) sequencing, in particular high-throughput sequencing.

Amplification is preferably performed by Polymerase Chain Reaction (PGR) techniques. Pair of oligonucleotide primers that hybridize to opposite strands of a genomic segment comprising at least one polymorphic marker are used for amplification. Preferably, each oligonucleotide primer pair is designed to include an ailele-specific oligonucleotide to selectively amplify a fragment of the genome of the individual that includes at least one polymorphic marker of the invention. Standard techniques for genotyping can be used to detect particular marker alleles, such as fluorescence-based techniques utilizing PCR, LCR, Nested PCR and other techniques for nucleic acid amplification.

Hybridization is preferably sequence-specific hybridization, i.e., hybridization with a nucleic acid probe that specifically hybridizes to a nucleic acid which contains a specific allele at a polymorphic site (ailele-speci fic oligonucleotide or ailele-specific oligonucleotide probe). Arrays of oligonucleotide probes can be used to identify several genetic markers including one or more polymorphic markers according to the invention.

Oligonucleotide primers and probes, including ailele-specific oligonucleotide probes and primers are usually of 10 to 30 or 10 to 50 nucleotides. The oligonucleotide can be DNA, RNA, PNA or mixed, and may comprise locked nucleic acids ( LNA). The oligonucleotide is advantageously labeled with a suitable label such as for example fluorescent label, radioisotope or magnetic label. Such oligonucleotide primers and probes, including ailele-specific oligonucleotide probes and primers can be prepared using standard methods. Specific commercial methodologies available for SNP genotyping include, but are not limited to, TaqMan genotyping assays and SNPlex platforms (Applied Biosystems), gel electrophoresis (Applied Biosystems), mass spectrometry (i.e., MassARRAYsystem from Sequenom ), minisequencing methods, real-time PGR, Bio- Plex system (BioRad), CEQ and SNPstream systems (Beck man ), array hybridization technology (i.e., Affymetrix GeneChip; Perlcgen): Bead Array Technologies (i.e., Illumina GoldenGate and Infinium assays), array tag technology (i.e., Parallele), and endonuclcase-based fluorescence hybridization technology ( InvadenThird Wave).

In other embodiments, the method of the inv ention comprises the determination of at least one allele of at least two different polymorphic markers.

The method of the inv ention which allows to determine a genotype associated with thymic function lev el in an individual is useful for the diagnostic, prognostic, and/or monitoring of conditions or clinical situations where thymopoiesis is impacted or that are impacted by thymopoiesis efficiency and/or quality. In particular, the method of the inv ention is useful for the diagnostic, prognostic, and/or monitoring of aging, of conditions or clinical situations of lymphopenia where immune regeneration is required, and of conditions or clinical situations where production of T-cell naive response is required.

The method of the inv ention is performed on a biological sample that may be from the patient or from the donor (in case allo-hematopoietic stem cell transplantation

(HSCT)).

According to the inv ention, a patient refers to an indiv idual, preferably a human, affected by a disease where thymopoiesis is impacted or that is impacted by thymopoiesis efficiency and/or quality, as defined abov e. Conditions or clinical situations where immune regeneration is required include acquired immunodeficiencies, allo-hematopoietic stem cell transplantation (HSCT) and cellular therapies, gene therapy, immunosuppressive treatments such as in solid-organ transplantation, immunotherapy and immunoregenerative therapies. In cellular and gene therapies, expanded or manipulated hematopoietic stem cells are infused to give rise to lymphoid progenitors and T-cells as an adoptive or replacement immunotherapy. Acquired immunodeficiencies include in particular Human Immunodeficiency Virus infection and acquired immune deficiency syndrome (HIV/AIDS ). Immunoregenerative therapies of thymic function include in particular sex steroid ablation. IL-7 and or II. -22 therapy.

Conditions or clinical situations where production of T-cell naive response is required include vaccination such as v accination against pathogens or tumors (anticancer vaccine), immunotherapy, infectious diseases including opportunistic infections, such as immunodeficiencies, and cancer. Immunotherapy includes checkpoint inhibitor therapies for treatment of cancer.

Clinical situations that are impacted by thymopoiesis efficiency and/or quality include autoimmunity. Autoimmune disorders include those with a high sex bias such as Systemic Lupus Erythematosus (SLE), Rheumatoid arthritis (RA) and Type 1 Diabetes.

During autoimmune disorders, the quality of T cells exiting the thymus is altered, due to defects in the so-called T-cell selection process, therefore generating autoreactive T-cells. It is therefore reasonable to anticipate that a higher production of autoreactive T cells resulting from the SNP polymorphism according to the inv ention may be associated with a higher susceptibility to the development or persistence of autoimmune diseases.

In conditions or clinical situations where thymopoiesis is impacted or that are impacted by thymopoiesis efficiency and/or quality as disclosed herein, the presence of a particular allele at a polymorphic site is indicative of a different degree of susceptibility and/or severity of the disease. Such allele is useful as prognostic marker. For example, in conditions or clinical situations where immune regeneration or production of T-cell naive response is required as described herein, the presence of the effect allele is indicative of a decreased susceptibility to the disease and/or a decreased severity of the disease for the indiv idual compared to individuals not having the effect allele. For autoimmune disorders, the presence of the effect allele is indicative of an increased susceptibility to the disease and/or an increased severity of the disease for the individual compared to indiv iduals not having the effect allele. Furthermore, in conditions or clinical situations where thymopoiesis is impacted or that are impacted by thymopoiesis efficiency and/or quality as disclosed herein, the presence of a particular allele at a polymorphic site is also indicative of a different response to a particular treatment. This means that an individual such as a patient carrying at least one effect allele of at least one polymorphic allele according to the invention would response better to, or worse to, a specific therapeutic, drug, and/or other therapy used to treat the disease. This has consequences not only in terms of efficacy but also in terms of safety of therapy because an excessive response in the indiv idual may produce undesirable side-effects. Therefore, the identity of a marker allele would help in deciding what treatment should be used for the patient. For example, for a newly diagnosed patient, the presence of an effect allele o f a p o l y m o rp h i c a l l e l e of the present invention may be assessed. I f the patient is positive for the marker allele, then the physician recommends one particular therapy, w hile if the patient is negative for the at least one allele of a marker, then a different course of therapy may be recommended. Thus, the patient's carrier status could be used to help determine whether a particular treatment modality should be administered. The p o l y m o rp h i c markers of the invention, as described herein, may be used to assess response to these therapeutic options, or to predict the progress of therapy using any one of these treatment options. Thus, genetic profiling can be used to select the appropriate treatment strategy based on the genetic status of the individual, or it may be used to predict the outcome of the particular treatment option, and thus be useful in the strategic selection of treatment options or a combination of available treatment options. For example, it may be an important marker predicting the response of immunotherapies, for example those utilized in the treatment of cancers, and guiding the choice of the dose to employ or the duration of the therapy. It may also influence the choice of adjuvant to be joined to the antigenic stimulation in the context of vaccination.

In other embodiments, the method of the invention further comprises the determination of at least one other marker, such as for example polymorphic marker(s) different from those of the invention and or biomarker(s). The other marker(s) may be determined concomitantly to the polymorphic marker(s) of the invention, or before or after the polymorphic marker(s) of the invention.

Other markers include markers of thymic function, HLA hapiotype, in particular in the context of allogeneic HSCT, and drug-related or disease-related biomarkers related to conditions or clinical situations where thymopoiesis is impacted or that are impacted by thymopoiesis efficiency and/or quality, as described herein.

In preferred embodiments, the polymorphic marker of the invention is combined with HLA allele(s), in particular alleles of the HLA class I genes, such as HLA-A,-B and -C alleles, and HLA class 11 genes, such as H L A- DP, -DQ, -DR, so as to determine the H LA hapiotype of the indiv idual. Determination of the H LA hapiotype of an individual is determined by standard methods such as by standard 1 1 LA genotyping methods. Determination of H LA allele(s) in addition to the polymorphic marker of the invention is useful in allo-HSCT to improv e donor choice algorithms in the search of unrelated donors, including cord blood or H LA haploidentical related donors. In other preferred embodiments, the polymorphic marker of the invention is combined with another biomarker of thymic function such as sjTRECs, PTRECs, and or intrathymic divisions number, preferably sjTRECs number.

The method of the invention is also useful in drug screening and drug development, in particular to increase safety and effectiv eness of clinical trials. In particular, indiv iduals carrying at least one effect allele of at least one polymorphic allele according to the inv ention may be more likely to respond fav orably to therapeutic agent or drug. Therefore, the strati fication of patients according to the genotype status of the polymorphic marker(s) according to the inv ention (presence or absence of marker effect allele) in a clinical trial can improv e the safety of clinical trials, but can also enhance the chance that a clinical trial will demonstrate statistically significant efficacy.

Another object of the inv ention is a kit for performing the method of the inv ention, comprising reagents necessary for selectiv ely detecting at least one allele of at least one polymorphic marker of the present inv ention in the genome of the individual, in particular, at least one SNP chosen from SEQ I D NO: 1 to 22, preferably chosen from SEQ ID NO: 1 , 1 1 , 15 and 18 to 22, more preferably chosen from SEQ I D NO: 18 to 21 , even more preferably SEQ ID NO: 21. The kit usually comprise means for amplification of the nucleic acids of the invention (i.e., nucleic acid segment comprising one or more polymorphic markers of the invention), means for analyzing nucleic acid sequence of the nucleic acids, and/or means for allele-specific detection of the nucleic acids or amplified fragments thereof. In certain embodiments, the kit comprises primers for amplification of nucleic acids of the invention, and/or hybridization probes for sequence specific hybridization to said nucleic acids, in particular allele-specific oligonucleotide probe and or primers. In some preferred embodiments, the kit comprises allele-specific oligonucleotide probe and/or primers for the specific detection and/or amplification of one or both alleles of one or more SNPs chosen from SEQ ID NO: 1 to 22, preferably chosen from SEQ ID NO: 1 , 1 1 , 15 and 1 8 to 22, more preferably chosen from SEQ ID NO: 1 8 to 21 , even more preferably SEQ ID NO: 2 1 . The kit can comprise necessary buffers and enzymes. In certain embodiments, the kit comprises reagents for detecting at least two different polymorphic markers according to the invention, as described herein.

In certain embodiments, the kit further comprises reagents for detecting at least another marker as described herein, preferably HLA allele(s), in particular alleles of the HLA class I genes, such as HLA-A,-B and -C alleles, and HLA class II genes, such as HLA-DP, -DQ, -DR., so as to determine the H LA haplotype of the individual.

Another object of the invention is the use of said polymorphic marker, in vitro, for evaluating thymic function level in a subject.

In certain embodiments, the polymorphic marker is used for the diagnostic, prognostic or monitoring of conditions or clinical situations where thymopoiesis is impacted or that are impacted by thymopoiesis efficiency and/or quality as described herein. In some preferred embodiments, the polymorphic marker is used for predicting the response to therapy of said conditions or the outcome of said clinical situations.

In certain embodiments, the polymorphic marker of the invention is used with another marker as described herein. Another object of the invention is the use of said polymorphic marker for drug screening and/or drug development as described herein.

In preferred embodiments, the polymorphic marker of the invention is used w ith another marker as described herein. Another object of the invention is a method of treating a disease where thymopoiesis is impacted or that is impacted by thymopoiesis efficiency and/or quality in a patient, comprising:

- determining, in a biological sample containing nucleic acid obtained from the patient, the presence or absence of at least one allele of at least one polymorphic marker according to the inv ention (desired allele), and

- administering a treatment to the patient, when the patient has the desired allele.

According to the present invention, a "desired allele" refers to an allele of at least one polymorphic marker according to the inv ention that is beneficial for a specific application, in particular a therapeutic application, including the treatment of a patient and allo-hematopoietic stem cell transplantation (HSCT) from a donor to a recipient.

Depending upon the particular disease or condition where thymopoiesis is impacted or that is impacted by thymopoiesis efficiency and or quality, the desired allele is either the effect allele (i.e. allele associated w ith increased lev el of T lymphocyte production by the thymus) or the other allele (allele not associated w ith increased lev el of T lymphocyte production by the thymus).

Another object of the inv ention is a method of selecting hematopoietic cells of interest, in particular hematopoietic stem cells or T lymphoid progenitor cells of interest, comprising:

- determining, in nucleic acid of the hematopoietic cells, the presence or absence of at least one allele of at least one polymorphic marker according to the inv ention

(desired allele), and

- selecting the hematopoietic cells hav ing said desired allele.

Depending upon the particular application of the hematopoietic cells, the (desired) allele that is selected in the hematopoietic cells is the effect allele (i.e. allele associated with increased level of T lymphocyte production by the thymus) or the other allele (allele not associated with increased level of T lymphocyte production by the thymus).

For example, the method may be used for selecting hematopoietic cells for allograft transplantation. For this application, hematopoietic stem cells having the effect allele are selected.

In some embodiments of the method of selection, said polymorphic marker is chosen from SEQ ID NO: 1 to 22, preferably SEQ ID NO: 1 , 1 1 , 15 and 18 to 22, more preferably SEQ ID NO: 18 to 21 , even more preferably SEQ ID NO: 21. The method of selection is usually performed on a nucleic acid sample comprising genomic DNA from the cells, using standard methods that prov ide the allelic identity at particular polymorphic sites as described above.

The method of selection of the invention may be performed using hematopoietic cells including stem cells obtained from a variety of sources, using conventional methods known and available in the art. For example, hematopoietic cells may be recovered from bone marrow, mobilized peripheral blood mononuclear cells (PBMCs), umbilical cord blood, embryonic stem (ES) cells or induced pluripotent stem (IPS) cells. The hematopoietic cells are advantageously human hematopoietic cells. In some embodiments, the cells are CD34+ hematopoietic cells, preferably human CD34+ hematopoietic cells.

In some embodiments, the method of selection further comprises the determination of another marker as defined above, in particular H LA allele( s) in the context of allogeneic HSCT, to improve donor choice algorithms in the search of unrelated donors, including cord blood or HLA haploidenticai related donors. Another object of the invention is a genetically engineered hematopoietic cell as defined above, in particular a genetically engineered hematopoietic stem cell or T lymphoid progenitor cell, in which at least one allele of at least one polymorphic marker according to the invention has been replaced with the other allele (desired allele), in particular the effect allele. In some embodiments of the genetically engineered hematopoietic cell, said polymorphic marker is chosen from SEQ ID NO: 1 to 22, preferably SEQ ID NO: 1 , 1 1 , 1 5 and 1 8 to 22, more preferably SEQ ID NO: 18 to 21 , even more preferably SEQ ID NO: 2 1 . The cells of the invention may be genetically-engineered using any known gene- editing system such as TALEN, Zinc-Finger meganucleases, CRISPR Cas and others. The gene editing system is engineered for replacing specifically at least one allele with the other allele (desired allele), in particular the effect allele, according to the invention in the T-cell receptor alpha-T cell receptor delta locus of the hematopoietic stem cells or T lymphoid progenitors, using standard method that are well-known in the art. The hematopoietic stem cells or T lymphoid progenitors are obtained from an indiv idual

(donor) using conventional methods known and available in the art, as mentioned above. The gene editing system is introduced in the hematopoietic stem cells or T lymphoid using standard nucleic acid and/or protein delivery agents or systems. The genetically engineered hematopoietic cells may be administered to the donor (autologous graft transplantation/identical donor and recipient) or to another individual (allograft or xenograft transplantation/recipient different from the donor).

Another object of the invention is a pharmaceutical composition comprising an effective amount of hematopoietic cells having the desired allele of a polymorphic marker according to the invention, in particular hematopoietic stem cells or T lymphoid progenitor cells, either genetically engineered or obtained by the selection method according to the inv ention, and a pharmaceutically acceptable carrier, vehicle, and or excipient. The hematopoietic cells, either genetically engineered or obtained by the selection method according to the inv ention, may be from the patient or from a donor.

In some embodiments, the hematopoietic cells are human hematopoietic cells.

In some embodiments, the hematopoietic cells have the effect allele.

In some embodiments, the hematopoietic cells are genetically engineered hematopoietic cells, in particular genetically engineered hematopoietic stem cells or T lymphoid progenitor cells.

In some preferred embodiments, the hematopoietic stem cells are human hematopoietic cells from a donor individual having the effect allele for ailo-HSCT to a recipient indiv idual (patient). The human hematopoietic cells from the donor, preferably human hematopoietic stem cells or T lymphoid progenitor cells, are either genetically engineered or obtained by the selection method according to the invention, preferably obtained by the selection method according to the invention. Another object of the inv ention is a pharmaceutical composition comprising an effective amount of gene-editing system engineered for replacing specifically one allele of at least one polymorphic marker according to the inv ention w ith the other allele (desired allele), in the T-celi receptor aipha-T cell receptor delta locus of at least one hematopoietic stem cell or T lymphoid progenitor of a patient, and a pharmaceutically acceptable carrier, vehicle, and/or excipient. The gene-editing system can be based on any-known gene-editing system as mentioned above.

A pharmaceutical composition according to the invention comprises a therapeutically effective amount of active agent (genetically engineered hematopoietic cells or gene editing system), which is a dose sufficient for reversing, allev iating or inhibiting the progress of the disorder or condition to which such term, applies, or reversing, alleviating or inhibiting the progress of one or more symptoms of the disorder or condition to which such term applies.

The effective dose is determined and adjusted depending on factors such as the composition used, the route of administration, the physical characteristics of the indiv idual under consideration such as sex, age and weight, concurrent medication, and other factors, that those skilled in the medical art will recognize.

A "pharmaceutically acceptable carrier, v ehicle, and/or excipient" refers to compounds, materials, compositions, and/or dosage forms that do not produce an adverse, allergic or other unwanted reaction when administered to a mammal, especially a human, as appropriate. The pharmaceutical v ehicles, carriers, and/or excipients are those appropriate to the planned route of administration, which are well known in the art. A pharmaceutically acceptable carrier, vehicle and/or excipient includes with no limitations, non-toxic solid, semi-solid or liquid filler, diluent, encapsulating material or formulation of any type. Another object of the invention is a pharmaceutical composition comprising an effective amount of hematopoietic cells having the desired allele and/or gene-editing system for introducing the desired allele in hematopoietic cells as defined above, for use in the treatment of a condition where thymopoiesis is impacted or that is impacted by thymopoiesis efficiency and/or quality, as defined above.

Another object of the inv ention is a method of treating a disease where thymopoiesis is impacted or that is impacted by thymopoiesis in a patient, comprising administering an effective amount of hematopoietic cells hav ing the desired allele, gene- editing system for introducing the desired allele in hematopoietic cells, or pharmaceutical composition thereof to the patient.

Depending upon the type of disease or condition of the patient, the desired allele is either the effect allele (i.e. allele associated with increased level of T lymphocyte production by the thymus) or the other allele (allele not associated with increased lev el of T lymphocyte production by the thymus). In some embodiments, the above therapeutic method or use is for gene therapy, wherein hematopoietic cells from a patient are genetically engineered ex vivo or in vivo using a gene-editing system as defined abov e. Follow ing ex vivo modification using the gene-editing system, the genetically engineered hematopoietic cells are then reintroduced into the patient using standard methods. In some other embodiments, the above therapeutic method or use is for cell therapy, wherein allogenic/xenogenic hematopoietic stem cells or T lymphoid progenitors hav ing the desired allele, either selected using the method of selection of the invention or genetically engineered to replace one allele of at least one polymorphic marker according to the inv ention with the other (desired) as defined abov e, are administered to the patient.

In some embodiments, the abov e therapeutic method or use for gene or cell therapy comprises :

- determining, in a biological sample containing nucleic acid obtained from the patient, the presence or absence of at least one (desired) allele of at least one polymorphic marker according to the invention, and - administering an effective amount of hematopoietic cells hav ing the desired allele, gene-editing system for introducing the desired allele in hematopoietic cells, or pharmaceutical composition thereof to the patient, when the patient does not have the (desired) allele. The practice of the present invention will employ, unless otherwise indicated, conventional techniques, which are within the skill of the art. Such techniques are explained fully in the literature.

For a better understanding of the invention and to show how the same may be carried into effect, there will now be described by way of example a specific mode contemplated by the Inventors with reference to the accompanying drawings in which:

- Figure 1. Thymic function associates with na^'ive T-cell immune phenotypes. (A) T cells differentiate in the thymus from double negative 1 (DN1) to single positiv e (SP) stage, β T-cell receptor excision circles (βΤΡνΕ-Cs) are episomai DNA generated during the TCRB recombination. Signal joint TRECs (sjTRECs) deriv e from the deletion of the TCRD locus during TCRA locus recombination (figure 1 B). The log2 of the sjTRECs/pTRECs ratio is used as an estimate of intrathymic proliferation between double negative 3 (DN3) and DP stages. (B) Mapping at the TCRA-TCRD locus of primers (sjTREC F and sjTREC R ) and probe (sjTREC P) used for the sjTRECs assay. Sequences of primers and probe are listed in Table I. (C) Effect sizes of significant associations (adj. p-value < 0.05) between sjTRECs levels and the number of relevant circulating immune cells measured by flow cytometry in the 1 .000 healthy indiv iduals. Effect sizes were estimated in a multiple regression that includes age, sex, CMV infection status and smoking status, together with batch effects as random effects. The confidence intervals are false coverage-adjusted and calculated using the profile likelihood and based on the likelihood ratio test. (D) Relationships between sjTRECs levels and the Log-transformed number of na^'ive CD4 and CD8 T cells, naive regulatory T cells (Treg), and in variant natural killer T cells (MKT). Regression lines were fitted using linear regression.

- Figure 2. Age and gender strongly impact thymic function in healthy donors. (A) sjTRECs levels as a function of age. Regression lines were fitted using linear regression. Light grey colour indicates men and darker grey colour indicates women. (B) β T R ECs levels as a function of age in donors with detectable amounts. Regression lines were fitted using linear regression. Light grey indicates men and darker grey colour indicates women. (C) Scatter plots showing numbers of intrathymic div isions as a function of age in donors with detectable PTREC. Regression lines were fitted using linear regression. Grey indicates men and black indicates women. (D) sjTRECs levels as a function of sex. Light grey colour indicates men and darker grey colour indicates women.

- Figure 3. Genome-wide association study (GWAS) of thymic function in healthy women and men. (A) Manhattan plot for sjTRECs in the 500 men and 500 women of the Milieu Interieur cohort. The GWAS revealed a genome-wide significant association at chromosome 14 in women (rs 1 087301 8,

Gray line indicates the threshold for genome-wide significance (P=5x 10^-8 ). (B) Detailed view of the genetic association signal found in the TCRA-TCRD locus. The 22 most strongly associated SNPs with sjTREC levels are indicated in black. Primers (sjTREC F/R ) and probe (sjTREC P) used to quantify sjTRECs are shown in grey. (C) Physical position of the four most strongly associated variants (rs8013419, rs 10873018, rs 12 147006 and rs2204985), relative to active transcription activity measured by the 1 13 K27 Ac histone acetylation mark and transcription factor binding sites ( identified by ChlP-seq (ENCODE Project Consortium, Nature, 2012, 489, 57-74)).

- Figure 4. Genome-wide association study reveals an impact of TCRA-TCRD genetic variation on thymic function. ( A) Manhattan plot for genetic association with sjTREC levels in the 969 donors of the Milieu Interieur cohort. Light and dark grey lines indicate the threshold for suggestive association (P =5x 10^-5 ) and genome-wide significance (P=5.0x 10^-8 ), respectively. ( B) Detailed view of the TCRA- TCRD locus. Primers (sjTREC F/R) and probe (sjTREC P) used to quantify sjTRECs are shown. (C) Fine mapping of the genetic association between the TCRA-TCRD locus and sjTREC levels. Meta-analysis /'-values were obtained by combining array-based, probe-based and imputed genotypes of the Milieu Interieur and MARTHA cohorts. Grey line indicates the threshold for genome-w ide significance (P= 5.0x 10^-8 ). Variants that are significantly associated at the genome-wide level are indicated in black. ( D ) Physical position of the four most strongly associated variants, relative to active transcription activ ity measured by the H3K27Ac historic acetylation mark and transcription factor binding sites (identified by ChlP-seq) (E. P. Consortium, Nature, 2012, 489, 57-74). Position of the DD3 gene segment (D53) is indicated. - Figure 5. Age, sex and rs2204985 impact thymic function in MARTHA cohort. (A) sjTREC levels as a function of age in the independent MA RTHA cohort (n=612). Regression lines were fitted using linear regression. Grey indicates men and black indicates women. Indicated adj. P is the FDR adjusted /'-v alue for the large sample chi-square likelihood ratio test of an age effect obtained using a mixed model for the response variable logio sjTRECs, including sex and TREC processing plate as fixed effect covariates, and additional batch variables as random effects. (B) sjTREC levels as a function of sex in the independent MARTHA cohort (n=612). Light grey color indicates men (n=160) and dark grey indicates women (n=452). Regression lines were fitted using linear regression. Indicated adj. P is the FDR adjusted /'-v alues for the large sample chi-square likelihood ratio test of a sex effect obtained using a mixed model for the response variable logio sjTRECs including sex and TREC processing plate as fixed effects, and additional batch variables as random effects.

- Figure 6. Effect of the human TCRA-TCRD genetic variation on thymic function in immunodeficient mice. (A) H IS mice were generated in Balb/c Rag2^-/-Il2rg- ^/-Sirpa^NOD ( BRGS ) hosts with rs2204985 A A (in pale grey), GA (in dark grey) or GG genotype (in black ) CD34 CD38- human fetal liver hematopoietic stem cells. sjTREC levels were measured in reconstituted thymi and spleen from 8-29 weeks old mice. Quantification of (B) sjTREC/ 150 000 cells and (C) total sjTREC per thymus (left: AA, n= 1 9; GA n=58; GG, n= 15) or spleen (right: AA, n=12; GA n=38; GG, n= 5). ( D) Number of CD3^pos human thymocytes according to rs2204985 genotype (AA, n=5; GA n=3 1 : GG, n= 13 ). (E ) Influence of the SNP rs2204985 polymorphism (AA, n=5; GA n=3 1 ; GG, n= 1 3 ) on human thymocytes developmental stages in consecutive thymocyte subsets (DN, double negative; ISP, immature single positiv e: DP, double positiv e: SP, single positive), /'-v alues are indicated on the top of the boxpiot (Kruskai-Waliis test). - Figure 7. Role of the mouse recipient sex on the rs2204985 polymorphism effect. Comparison of the SNP rs2204985 polymorphism effect on thymic sjTREC production in male (AA, n=10; AG n=29; GG, n= 4) and female (AA, n=9; AG n=29; GG, n= 1 1) immunodeficient mice. Values were log 10-transformed and /'-values are indicated (Kruskal-Wallis test).

Figure 8. Effect of the human TC RA-TCRD genetic v ariation on thymic I CR repertoire in immunodeficient mice. Human TCRA- TCRD was sequenced using genomic DNA from 8 (3 males, 5 females) and1 2 (4 males, 8 females) immunodeficient mice thy mi grafted with A A and GG human fetal livers, respectively (Table IV ). (A)

Whisker Boxplot of the % of each V (left panel) and J (right panel) gene usage among all the TCR alpha and TCR delta productive rearrangements according to the donor genotype (AA in Grey, GG in Black ). ( B) Ratio of median percentage of V (left panel) or J (right panel) gene usage in GG grafted mice/ median percentage of V or J gene usage in A A grafted mice. Gene segments used specifically by TCR delta are indicated in black, by I CR alpha in dark grey and shared by TCR. alpha and delta in light grey. (C) Percentages of specific TCR delta J genes {TCRDJOl to 04) among total TCR alpha and delta J genes used in productive rearrangements according to A A (grey) and GG (black ) genotypes. (D) Percentages of D V ( left panel ) DD (central panel ) and DJ (right panel) genes usages among TCRD productive rearrangements. Genes are ordered according to their genomic location (see Fig. 4B) and p values are obtained using the non-parametric Mann-Whilncy test.

Figure 9. Combined effects of sex, age and TCRA-TCRD genetic variation on thymic function. (A) Box plots showing sjTRECs levels as a function of sex and rs2204985 SNP alleles in the Milieu Interieur cohort (n=1000). (B) Scatter plots showing sjTRECs levels as a function of age, sex and genotype (SNP rs2204985). Regression lines were fitted using simple linear regression. Black indicates A A genotype, dark grey colour indicates AG genotype, and ligt grey colour indicates GG genotype. (C) Representation of number of thymic age versus chronological age as a function of gender and SNP rs2204985 modalities. Figure 10. sjTRECs levels as a function of sex and SNP genotype in the Milieu Interieur cohort (n=1000) with SNP rs l 087301 8 (A), rs2204985 (B), rs2 141988

(C), rsl 1844715 (D), rs2204984 (E), rs12147006 (F), rs38 1 1236 (G) and rs8013419 (H).

Figure 11. Combined effects of sex, age and TCRA-TCRD genetic variation on thymic function. sjTREC levels as a function of age and rs2204985 genotypes in (A) the Milieu Interieur cohort (n=969) and (B) the replication MARTHA cohort (n=612). Regression lines were fitted using linear regression. Genetic associations were tested ( indicated /'-value ) with a mixed model of logio sjTREC levels, including as predictors rs2204985 genotypes, and covariates selected using a data-driven variable selection scheme, and correcting for population stratification using ancestry as a random effect. rs2204985 genotypes were obtained by additional by-design genotyping in both the Milieu Interieur and MARTHA cohorts. Light grey indicates A A genotype, dark grey indicates GA genotype, and black indicates GG genotype. (C) Proportions of variance of sjTREC levels explained by age, sex and TCRA-TCRD genetic variation. The surface area indicates the total variance explained by the multiple regression model, in Milieu Interieur (left) and MARTHA (right) cohorts, and the area and colour of sub -rectangles indicate proportions attributed to specific predictors (as measured by the R² of the regression model ). (D) Difference between (chronological) age and thymic age as a function of sex and rs2204985 variant. Thymic age is predicted from our regression model, and AA men are assumed as the baseline of thymic function.

In the following description numerous specific details are set forth in order to provide a thorough understanding. It will be apparent however, to one skilled in the art, that the present invention may be practiced without limitation to these specific details. In other instances, well known methods and structures have not been described, so as not to unnecessarily obscure the description. EXAMPLE 1: Materials and Methods

- Experimental design

1 ,000 healthy donors were recruited by BioTrial (Rennes, France), stratified by gender (500 men, 500 women) and age (200 individuals from each decade of life, between 20 and 69 years of age). Donors were selected based on stringent inclusion and exclusion criteria, detailed elsewhere (S. Thomas et a!., Clin. Immunol. 157, 2015, 277- 293). Briefly, recruited individuals had no evidence of any severe/chronic/recurrent pathological conditions. Main exclusion criteria were: seropositivity for pathological chronic infections (human immunodeficiency virus, HIV; hepatitis B virus, HBV; hepatitis C virus, HCV), travel to foreign countries in the last three months, recent vaccine administration, as well as alcohol abuse. To avoid the influence of hormonal fluctuations in women during the peri-menopausal phase, only pre- or post-menopausal women were included. To avoid the presence of population structure, the study was restricted to French citizens with Metropolitan French origin for three generations (i.e. , with parents and grandparents born in continental France). Metabolic syndrome was defined in the MI donors based on six criteria: increased abdominal circumference (>94cm European men, >80cm European women), elevated systolic blood pressure (>130mmHg), elevated diastolic blood pressure (>85mmHg), elevated triglyceride levels (>1.7mM), diminished levels of high density lipoprotein (HDL<lmM men, <1.3mM women) and glucose concentration (>6.1mM). To generate a composite metabolic score, 1 point was assigned for each of the assessed variables, taking blood pressure elevation as a single value (i.e., elevated systolic and/or diastolic blood pressure = 1 point). Whole blood samples were collected from the 1 ,000 fasting healthy donors on lithium heparin tubes, from September 201 2 to August 201 3, in BioTrial, Rennes. France. The clinical study was approved by the Comite de Protection des Personnes - Quest 6 on June 13th, 201 2, and by the French Agence Nationale de Securite du Medicament (ANSM) on June 22nd, 201 2. The study is sponsored by the Institut Pasteur ( Pasteur ID-RCB Number: 201 2-A00238-35 ), and was conducted as a single center study without any investigational product. The protocol is registered under CiinicaiTriais.gov (study# NCT01699893). DNA extraction from human whole Mood

Blood was collected in 5ml sodium EDTA tube and was kept at room temperature (18-25°) until processing. DNA extraction was performed using the Nucleoli BACC3 kit (#RPN8512, GE-Healthcare). Upon arrival at the processing site, blood was transferred into a 50ml polypropylene tube. 20ml of sterile Reagent A Ix (lysis buffer) were added to the blood sample in aseptic conditions and mixed by rotation for 4 minutes at room temperature. After red blood cell lysis, the tube was centrifuged 1300g for 5m in and the supernatant was discarded. The cell pellet was resuspended with 40μ1 of PBS IX, transferred to a 0.5ml 2D-cap tube and stored at -80°C before processing. After thaw ing, 1ml of sodium Reagent B was added directly to the cell pellet for resuspension before transfer to a 15ml screw capped propropylene centrifuge tube. 25()μ1 of sodium perchlorate solution were then added for deproteinisation and the tube was mixed by inverting the tube at least 7 times. 1 ml of chloroform was then added to the tube, which was mixed by inversion. Without remixing the phases, 1 50μ1 of Nucleoli resin were added and the tube, which was then centrifuged at 1300g for 3 minutes. Without disturbing the Nucleoli resin layer, the upper phase was transferred to a clean 1 5ml tube and 2 volumes of cold absolute ethanol were added. The tube was then mixed by inversion until the DNA precipitate appeared. Using a heat-sealed Pasteur pipette, the precipitated DNA was hooked out and placed into a clean 1 .5ml microcentrifugation tube. 1ml of cold 70% ethanol was added, the DNA was washed and the supernatant was discarded after centri i ligation at 4000g for 5min. After DNA pellet air dry for 10m in, 400μΙ of deionized water were added and the tube kept overnight at 4°C to complete the resuspension before DNA quantification.

- T-cell receptor excision circles (TRECs) assays

sjTRECs and pTRECs are episomal circular DNAs generated during TCR a and β chain recombination, respectively (Figures 1 A and IB). The protocol from Clave et al,, Blood, 2009, 1 13, 6477-6484) was implemented to quantify sjTRECs and βTRECs levels. The protocol is based on a quantitative PGR of genomic DNA extracted from whole blood, using the Biomark HI) system ( Fluidigm France, Paris, France). 1 to 2 μg of genomic DNA was preamplified for 3 min at 95 °C and then 1 8 cycles of 95 °C 1 5s, 60°C 30s and 68°C 30s, in a 50 μΐ reaction that contained the primers listed in Table I, 200 μΜ of each dNTP, 2.5 niM MgS0₄ and 1 .25 unit of Platinum Taq DNA pol High Fidelity (ThermoFisher Scientific. Courtaboeuf France) in1 x buffer. Columns of 48.48 Dynamic array IFCs were loaded with 5 μΐ containing 2.25 μΐ of a 1/400 dilution of preamplified DNA, 2.5 μ ΐ of 2x Takyon Low Ro.x Probe MM (Eurogentec) and 0.25 μΐ of sample Loading Reagent and raws with an equal mixture of 2x Assay loading Reagent and 2x Assay Biomark that contains only the 2 primers and the probe specific for each assay. The sum of the 10 pTRECs was multiplied by 1 .3 in order to take into account the 3 Οβ-Ιβ that were not quantified (Ι)β2-.Ιβ2.5, 2.6 and 2.7). Sj and pTRECs were normalized to 1 50 000 cells (around 1 μg of DNA) using the Albumin gene quantification. As iogio-transformed levels of PTRECs (logio PTRECs) showed a bimodal distribution, it was analyzed as either a binary variable (logio pTRECs=0, undetectable; versus logio PTRECs>0, detectable), or a quantitative variable in donors with detectable PTRECs only (logio PTRECs>0). Number of intrathymic division was obtained using the Log2 of sjTRECs/pTRECs ratio, and could thus be calculated in donors with detectable PTRECs only. When indicated, a shorter probe (sjTREC-LNA ) was used for sjTRECs quantification, which contains only the 26 first nucleotides of the standard probe and 4 Locked Nucleic Acids (LNA) in order to keep the same Tin as the original probe (Table I). From the 1000 Milieu Interieur donors, only 979 DNA were available for TREC analysis and 10 were further excluded because of low Albumin quantification (less than 50 000 copies).

- Association between TRECs measurements and immunophenotypes

Immunophenotyping was conducted on whole blood from ail donors, and details on technical procedures and complete results are available in M. Hasan et al (201 5 ). Ten 8-color flow cytometry panels were deveiopped (M. Hasan et al , Clin. Immunol. 201 5, 1 57, 261 -276), allowing for the measurement of 168 traits, including 76 cell counts, 89 Mean Fluorescence Intensity (MFI) and 3 ratios. To evaluate association between TRECs levels and immunophenotypes, a mixed model was fitted for each pair of immunophenotype and thymic variable using the log-transformed immunophenotype as response variable, and logio sjTRECs, logio PTRECs (in donors with detectable amounts), and the number of intrathymic div isions as treatment variable. Age, sex, CMV serostatus, and smoking were included as fixed effects covariates since they have been shown to largely impact immunophenotypes. It is well known that age affects also thymic function, and there is some prev ious evidence showing that also sex has an impact on it (W.A. Mitchell et al, Clin. Exp. Immunol., 2010, 161, 497-503), something that is confirmed in this study. Therefore, age and sex were considered common causes of the immunophenotypes and the thymic phenotypes, making them con founders. Since it is known that immunophenotype measurements can vary across season and be sensitive to batch effects, day of whole-blood sampling was included as a random effect. Whole- blood sampling could also affect the TREC v alues, thus potentially being a confounder. The hypothesis tests were chi-square likelihood ratio tests relying on the sample size of the MI cohort. All of the tests were considered as one family with the false discovery rate (FDR) as error rate. A total of 531 tests were made for this family. The associations that had an FDR less than 0.05 were considered significant. Confidence intervals were constructed for these significant associations using the profile likelihood with likelihood ratio test based cutoffs. The simultaneous confidence level for these intervals was chosen to be 0.05 and was calculated using the false coverage rate adjustment (Y. Benjamin! & D. Yekutieli, Journal of the American Statistical Association, 2005, 100, 71-81).

- Impact of non-genetic factors on thymic function

To identify non-genetic factors associated with thymic function, 40 physiological and demographic variables were chosen from the detailed data set available in the Milieu Interieur cohort (S. Thomas et al, Clin. Immunol. 157, 2015, 277-293), based on their relevance to immune traits. A mixed model was run for each pair of thymus function variable and non-genetic variable, using logio-transformed sjTRECs, login-transformed PTRECs and the number of intrathymic divisions as response and the non-genetic variables as treatment. As mentioned previously, it is known that age and sex affects thymic function (something we confirm in this study). Therefore these variables were included as controls (except when said variable was the treatment). The batch variables, i.e., the day of whole-blood sampling and the day of TREC processing, were included as random effects. Also plates used for TREC processing were included as a fixed effect. For PTRECs, the box used for processing was also included. Testing the effect of age and sex on the probability of having detectable amounts of PTRECs (using the binary variable described above) was done using logistic regression together with a Wald test. All tests were considered for association between the non-genetic treatment variables and the TREC response variables as one family of tests. The FDR was used as error rate and a significance cut-off of 0.05. A total of 120 models were fitted, and tests were performed. - DNA genotyping and imputation

The 1 ,000 subjects were genotyped at 719,665 SNPs by the HumanOmniExpress- 24 BeadChip (!llumina, California). To increase coverage of rare and potentially functional variation, 966 of the 1,000 donors were also genotyped at 245,766 exonic SNPs by the HumanExome-12 BeadChip ( lllumina, California). A total of 945,213 unique SNPs were thus genotyped. SNP quality-control filters yielded a total of 661 ,332 and 87,960 SNPs for the H umanOm ni Ex press and HumanExome BeadChips. respectively. The two datasets were then merged. Average concordance rate for the 16,753 SNPs shared between the two genotyping platforms was 99.9925%. The final dataset included 732.341 QC-filtered genotyped SNPs. Genotype imputation was performed by IMPUTE v.2, considering 1-Mb windows and a buffer region of I Mb. After quality-control filters, a total of 1 1 ,395,554 high-quality SNPs were obtained, which were further filtered for minor allele frequencies >5° o, yielding a final set of 5,699,237 SNPs for association analyses.

- Genotype-wide association analyses

Univariate genome-wide association study (GWAS) was conducted for each trait

(iog 10-transformed sjTRECs; iog 10-transformed β^'1^'RECs in donors with detectable v alues; number of intrathymic divisions), using the linear mixed model implemented in Genome Wide Efficient Mixed-Model Analysis for association studies (GEMMA; X. Zhou, M. Stephens, Nat. Methods. 2014, 11, 407-409). Genetic relatedncss matrix estimation was done with GEMMA, using a leav e-one-chromosomc approach. Age was included as a covariate in ail GWAS. Covariates were selected from batch and non- genetic variables using a variable selection scheme based around the elastic net (X. Zhou & M. Stephens, Nat. Methods, 2014, 11, 407-409) and stability selection (N. Meinshausen & P. Buhlmann, Journal of the Royal Statistical Society: Series B (Statistical Methodology), 2010, 72, 41 7-473 ). - Interaction model Including the polymorphism (Figures 3, 9, 10)

The interaction model used was a mixed model having logio sjTRECs as response variable and sex, age, plate used for T REC processing, rs2204985 genotypes, and the interaction between the rs2204985 genotypes and sex, as fixed effects, and ancestry (encoded by the genetic relatedness matrix), day of TREC processing and day of whole- blood sampling as random effects. Confidence intervals and tests based on this model was done using large-sample normal distribution approximations.

- Association between immunophenotypes and the rs2204985 variants (Figures 4, 5 and 11)

The effect of the rs2204985 variant on the immunophenotypes was tested using mixed models, with log-transformed immunophenotypes as response variables, the rs2204985 variant as treatment variable, and age, sex, CMV serostatus and smoking as fixed effects covariates, and blood sampling day as random effect. The P-values were adjusted to control the false discovery rate at 5 % within this family of tests. - Replication cohort

The replication cohort included 612 patients from the MARTHA cohort (Thrombophilia center of La Timone hospital, APHM, Marseille, France (M orange et ah, Blood, 201 1 , 117, 3692-3694). Donors are all of European descent, and were included between January 1994 and October 2005 for having suffered a single venous thrombosis event, without detectable cause. The study was approved by institutional ethic committee ("Departement Same de la Direction Generate de la Recherche et de Γ Innovation^"; Projects DC: 2008-880 & 09.576), and written informed consent was obtained from each subject. MARTHA biobank is hosted by the HEMOVASC bioresource center (CRB APHM). SjTREC levels of all donors were measured in DNA extracted from blood, using the approach described above. Genotypes for candidate variants were obtained from the lllumina Human610-Quad SNP array (Morange et ah, Blood, 201 1 , 1 17, 3692- 3694) or probe-based genotyping, as described below. Replication was tested with mixed models including age, sex and TREC processing plate as fixed effect covariates, and batch effects as random effects, using the lme4 R package. Mixed effect models including the GRM were fitted using the Imekin function in the coxme R package. Meta- analysis of the Milieu Interieur and MARTHA cohorts was conducted with the rna.mi function in the metafor R package

(Viechtbauer, W\, Journal of Statistical Software, 201 0, 36, 48-) with the random effects approach based on the DerSimonian and Laird estimator. - Probe-based genotyping of candidate variants

From the 22 SNPs associated with sjTRECs, 8 SNPs (rs381 1236, rs2141988, rs2204984, rs8013419, rs 1 087301 8, rs 1 2 147006, rs2204985, and rs 1 1844715 ) initially imputed on the GW AS analysis were selected on the basis of their informativity. Genotyping was done on the Biomark™ HD. Briefly, 60 to 1 20 ng of genomic DNA were preampiified with 0.2 X of the 8 TaqMan Genotyping Assay (ThermoFisher Scientific) and 1 x Preamp Master Mix ( Fluidigm) for 14 cycles (95°C 15s, 60°C 1 min). The FLEXsix™ Genotyping IFCs (Fluidigm) were loaded with l/50th dilution of the preampiified product. 2X Takyon™ Low Rox Probe MM (Eurogentec) and 40 X TaqMan™ Genotyping Assay (ThermoFisher Scienti fic) according to manufacturer's instructions.

Reconstitution of Human Immune System (HIS) mice

Mouse strain

H IS mice were generated in Balb/c Rag2^-/Ti2rg^-/"Sirpa^NOD ( BRGS ) recipients using human fetal liver hematopoietic stem cells as previously described (Lopez-Lastra et al, Blood Advances, 201 7, 1, 601 -614). Briefly, newborn mice (3 to 5 clays of age) received sublethal irradiation (3 Gy) and were injected intrahcpatically with the equivalent of 2 X 1 ()⁵ CD34 CD38^" human fetal liver cells. A total of 92 H IS mice in 1 5 independent experiments (4- 10 mice per experiment) were analyzed at 8-29 weeks of age. Thymocytes and spienocytes were mechanically dissociated using a Ceil Strainer ( Ι ΟΟμιπ nylon Falcon^®). Ceils (5 X 10⁵) were frozen as dry pellet. DNA was prepared using the Proteinase K method ( 54°C for 1 min, 95 °C for 10 min ).

Human and mouse sex determination

Sexing of human donors was made by single amplification of the ZFX/ZFY genes in 25 μΐ PGR using 200 nM of primers hSex2-F (AAGTGCCCTCTTGCACATA; SEQ ID NO: 44) and hSe.x2-R (CTCGACTTAAACTTCTTCCC; SEQ ID NO: 45), 200 μΜ each dNTPs, 1.5 niM MgS04 and 1 unit of HiFi Taq Platinum (Thermo fisher). Cycling conditions were 94°C for 5 min and 40 cycles of 94°C for 30 sec, 56°C for 30 sec and 72°C for 2 min. PGR product was subsequently loaded on a 0.8% agarose gel giving a 1 329 pb band for ZFX and a 906 pb band for ZFY. Sexing of mouse recipient was made single amplification of the SRY/IL3 genes in 25 μΐ PCR using 200 tiM of primers mSRY (5 ' -TGGGACTGGTGACAATTGTC-3 '; SEQ ID NO: 46) and mSRY (5 '- GAGTACAGGTGTGC AGCTCT-3 ' ; SEQ ID NO: 47) and 200 nM of primers mIL3 (5 '- GGGACTCCAAGCTTCAATCA-3 '; SEQ ID NO: 48) and m IL3 (5 '- T G G A G G A G G A A G A A A A G C A A - 3 ' : SEQ ID NO: 49), 200 iiM each dNTPs, 1 .5 mM MgS04 and 1 unit of HiFi Taq Platinum (Thermofisher). Cycling conditions were 94°C for 3 min and 35 cycles of : 94°C for 1 5 sec, 57°C for 15 sec and 72°C for 30 sec, and. 72°C for 5 min. PCR product was subsequently loaded on a 1.5% agarose gel giving a 402 pb band for SRY and a 544 pb band for 11.3.

Quantification of T -cell Receptor (ICR) Excision Circles (TRECs) in HIS mice TREC quantification was adapted from Clave et ai (Taub et al, Immunol Rev

2005, 205, 72-93 ). Real time quantification was made using ViiA7 (Applied Biosystems by Life Technologies, Austin, TX, USA) in .W-well plates loaded with 20 μΐ containing 5 μΐ of DNA (0.5 to 1 μg of genomic DNA), 1 0 μΐ of 2x Takyon Low Rox Probe MM (Eurogentec) and 5 μΐ of specific primer-probe mix (Table I). In addition, the quantification of the alternative rearrangement in the TCR0 locus (called 6Rec-Ja58) of TRECs in HIS mice was made using V'iiA7 in 384-woli plates loaded with 20 μΐ containing 5 μΐ of DNA (0.5 to 1 μg of genomic DNA), 10 μΐ of 2.x Takyon Low Rox Probe MM (Eurogentec) and 5 μΐ of specific primer-probe mix (Table I). sjTRECs were normalized to 150 000 cells using the Albumin gene quantification. DNA genotyping in HIS mice

The samples were genotyped to the SNPs rs2204985 and rs 1087301 8 with 5 μΐ containing 2 μΐ of DNA (10 to 20ng of genomic DNA), 0.25 μΐ of 2.x Takyon Low Rox

Probe MM (Eurogentec) and 2.5 μΐ of 40X TaqMan Genotyping Assay (ThermoFisher Scientific) according to manufacturer's instructions. Immunophenotyping of human thymocytes in HIS mice

For the quantification of human thymocyte subsets from the humanized mouse thymus, 10⁶ thymocytes were labelled with CD45 PERCP-Cy5.5, CDS PE, CD4 APC H7, CD3 V500 AmCyan and CD la FITC monoclonal antibodies (Ail from BD Biosciences), then read on a FACSCanto I I llow-cytometer. Data were analyzed using FACSDiva software (BD Biosciences)

Next Generation sequencing of the TCHAD locus recombination

Sequencing of TCRAD locus in H IS mouse thy mi was performed on genomic

DNA at Adaptativc Biotechnologies (Seattle, USA) using the immunoSEQ assay at the survey level. Analysis of diversity was made using the ImmunoSEQ software tool and

Morisita's index using the R tcR package (W. Viechtbauer, Conducting Meta-Analyscs in R with the metatbr Package. 2010 36, 48 (2010)).

- Thymic age estimation

To illustrate how thymic function is impacted by non-genetic and genetic factors, it was sought to estimate the effect of an individual's rs2204985 genotype and sex in terms of "thymic age". First, a simple linear regression model was fitted with logio sjTRECs as response and age, sex, and rs2204985 genotypes as predictor variables. This regression model defines expected values of logio sjTRECs as a function of age, sex and rs2204985 genotypes. Thymic age was then defined as the expected age when donors are A A homozygous men, which were assumed as the baseline of thymic function. Equating expected logio sjTRECs of A A homozygous men and thymic age, with expected logio sjTRECs of another category of donors - for instance, GG homozygous women - we could estimate the difference between the age of GG homozygous women and expected thymic age. This difference is equal to the sum of the regression parameters (i.e., beta) of sex and rs2204985 genotypes divided by the parameter of age. According to Slutsky^'s theorem, this is therefore the maximum likelihood estimate of the age and thymic age difference for a particular donor group. Confidence intervals for this estimate where calculated by simulation of the parameters from the linear regression model. - Proportion of variance estimation

The contribution of rs2204985 genotypes, age and sex to the explained variance of logio sjTREC values was estimated by fitting linear regression models. The proportion of variance explained by a particular predictor was estimated by averaging the sum of squares for that particular variable over different orderings in the regression model. The estimation was done using the relaimpo R package.

- Statistical analysis and plots

All plots were done using the R programming language ( R Development Core Team ). Mixed effects models that do not include the GRM where fitted using lmc4. Mixed effect models including the GRM were fitted using the Imekin function in the cox me package. Graphs where primarily done using ggpiot2 and related packages.

EXAMPLE 2: Results

Validation of TRECs as surrogate markers of thymic function in the Milieu Interieur cohort

To assess natural variations in thymic function under homeostatic conditions, sjTRECs and (iTRECs were quantified in the Milieu Interieur cohort, which includes 500 men and 500 women of western European ancestry, stratified across five decades of age from 20 to 69 years-old (S. Thomas et al , Clin. Immunol. 157, 201 5. 277-293). sjTRECs and |¾TRECs were standardized and validated by high-throughput assays on DNA samples extracted from whole blood of the 1 ,000 subjects. sjTRECs counts normalized per 150 000 cells were used in subsequent analyses and correlated (r =0,99) with sjTRECs calculated as absolute numbers per til of blood which are not affected by T-cell peripheral divisions.

Log 10-trans formed levels of sjTRECs (logio sjTRECs) showed a normal distribution, with a mean of 2.4 +/- 0.03 (min to max range 0.2 to 4. 1 ). In contrast, Log 10-transformed levels of [VTRECs (logio (VFRECs) showed a bimodal distribution, with 368 donors under the threshold of detection of the assay (logio β^"Π¾Ε('=Ό). In donors with detectable fVl^'R ECs in whole-blood, the logio PTREC approximately followed a normal distribution with a mean of 1 .8 ± 0.05 (min to max range 0.02 to 3. 1 ). The number of intrathymic divisions was also normally distributed across the healthy donors, with a mean of 2.6 ± 0.16 (min to max range -4.3 to 9.5).

Immune cell populations or parameters associated with logio sjTRECs and PTRECs, or with the number of intrathymic divisions were searched by taking advantage of the extensive flow cytometry based immunophenotyping performed on the 1 .000 Milieu Interieur healthy donors (M. Hasan et al, Clin. Immunol. 2015, 157, 261-276). Multiple regression analyses controlling for potential confounding and batch effects were performed, and threshold for significance was set at 5x10- . Strong associations were found between logio sjTRECs and naive CD8 ^' and CD4 ^" T-celi counts, as well as with counts of other lymphocytes generated within the thymus such as naive regulatory T (Tregs) and invariant natural killer T (NKT) cells (Figure 1C). Naive CD8 T-celi counts doubled with a 10-fold increase in sjTREC levels (CI: [75%, 133%], adj. p- value = 8x 10- ⁴⁷), while naive CD4^' T-cell, NKT cell and naive Treg counts showed 60% (CI: [38%, 85%], adj. P =1.6x10^-20), 50% (CI: [10%, 92%], adj. P =1 .5x l 0^-3), and 40% (CI: [22%, 61%)], adj. P =5.7.x 1 0^-12 ) increases respectively, per 10- fold increase in sjTREC levels (Figure ID). A significant association of naive CDS T-celi counts with logio PTRECs levels was observed (adj. p-value=2x l 0^-2). By contrast, no significant association was found with numbers of intrathymic divisions. These results demonstrate that the sjTREC quantification assay provides a global assessement of thymic production. Non-heritable factors and TRECs levels in the Milieu Interieur cohort

Several non-heritable factors have been proposed to impact thymic function during aging, in particular endocrine factors such as the sex steroid and growth hormones, the body mass index (BMI), and the metabolic syndrome in the context of "inflammaging" (D.D. Taub, D. I Longo, Immunol. Rev ., 2005, 205, 72-93: H. Yang et al , Blood, 2009, 1 14, 3803-38 1 2:Y. I I. Youm et al , Proc. Natl. Acad. Sci. USA, 2016, 1 13, 1026-1031). Subjects included in the Milieu Interieur cohort were surveyed for hundreds of variables relating to nutrition, sleep, smoking, vaccination and medical histories (S. Thomas et al , Clin. Immunol. 1 57, 201 5, 277-293 ). 40 variables that represent potential sources of variation of thymic function were selected and used in a mixed model analysis, controlling for potential confounders and batch effects, to identify those significantly contributing to thymic function variance across the 1 ,000 healthy donors. Correcting for age and sex, no significant impact of cytomegalovirus (CMV) and Influenza A serostatus, metabolic score index and healthy inflammation on thymopoiesis was found in the studied cohort. In contrast, age had a strong effect on all three parameters of thymic function (Figure 2A-C). sjTREC values showed a decrease of 5% per year (CI: [4%, 6%], adj. p-value<2x l 0^{- 16}, Figure 2A), yet sjTRECs remained detectable in >99.9% of the 60-69 year-old donors. At the other stages of thymopoiesis, a decrease of 3% per year was seen in the odds of hav ing detectable amounts of [VTRECs in periphery (CI: [2%, 5%], adj. p-value=9,3x l 0^-8), a decrease of 2% per year in pTRECs in donors with detectable levels (CI: [1%, 4%], adj. p-value=l x l 0^-4, Figure 2B) and a decrease of 0.3 intrathymic divisions every 10 years (CI: [0, 0.5], adj. p-value 3x 1 0 \ Figure 2C).

Strikingly, a strong effect of sex on sjTRECs was also identified. sjTRECs were 66% higher in women in all age ranges, as compared to men (CI: [38%, 100%], adj. p- value<2x 10^-16) (Figure 2D). A small interaction was observed between age and sex, reflecting an increasing impact of sex over age ranges (P=0.03). By contrast, any impact of sex on JVfRECs or number of intrathymic divisions was not observed.

A common genetic variation at the TCRA-TCRD locus associates with sjTRECs levels

A major unanswered question is the contribution of genetic factors to the variations of thymic function across individuals. Herein, advantage was taken of the data arising from the genome-wide genotyping of the Milieu Interieur healthy donors at 945,2 13 SNPs, enriched in rare e.xonic SNPs (sec Methods). After quality control, genotype imputation was performed and yielded a total of 5,699,237 highly accurate common SNPs. Using a linear mixed model that adjusts for genetic relalcdness and specific covariates (X. Zhou, M. Stephens, Nat. Methods, 2014, 1 1 , 407-409), association of SNPs with login sjTRECs, logio PTRECs, and the number of intrathymic divisions was tested. No genome-wide significant association signal (P<5x \ Q- ) was observed for login PTRECs and the number of intrathymic div isions.

For which concerns logio sjTRECs, giv en the strong effect of sex, a genome-w ide association study (GWAS ) was conducted in men and women separately. Strikingly, a unique genome-wide significant signal was found in women (PfemaksrT . lx 10- ;

corresponding to a cluster of 3 genotyped and 19 imputed SNPs in strong linkage disequilibrium (r²>0.8) on chromosome 14 (Figure 3A, Table II).

Furthermore, seven independent genomic regions on chromosomes 2, 4, 5, 10, 1 1 , 14 and 17 showed suggestive evidence for association in the whole population (men and women together; P<5.0x 1 0^-5 Figure 4 A and Table III). To test for replication of these results, sjTREC levels were measured in the independent MARTHA cohort, which includes 612 unrelated patients of European descent affected by deep venous thrombosis (Morange et al , Blood, 201 1 , 117, 3692-3694). The results validated the association of decreased sjTREC levels with increasing age (4.05% per year, CI: [3.55%, 4.56%]; P< 1 0^-16) ( Figure 5A), and their higher levels in women, relative to men (86 %, CI: [60%, 1 16%]; P= 1 .4x l 0⁴⁵) (Figure 5B). Among 14 SNPs tagging the 7 suggestive loci (Table III), only variants on chromosome 14 showed statistical evidence for replication in the MARTHA cohort (Table III).

All these SNPs are located within a 25 kb region inside the TCRA- TCRD locus, and excised during I CR a chain recombination ( Figures 3B and 4B-C).

To fine map the signal, the 8 more informative SNPs of the cluster in the 1 .000 Milieu lnterieur donors (rs38 1 1 236, rs2 141988, rs2204984, rs8() 1 341 9, rs l 087301 8, rs12147006, rs2204985, and rs l 1 84471 5 ) were genotyped (Table II) and these data were combined with array-based or imputed genotype data from the MARTHA cohort. This led to the identification of 22 SNPs in linkage disequilibrium at a genome-wide significant threshold (P'< 1 .5x 10^-7 ) with 4 SNPs in linkage disequilibrium (rs8013419, rs l 087301 8, rs12147006, rs2204985) at a genome-wide statistical significance (P'< 2x 10- ) in meta-analysis (Table II I ) located in the intergenic DD2 and DD3 segments ( Figure 4D). Among them, rs2204985 (located 472 bases upstream of DD3) was considered the most likely candidate variant (effect allele A with frequency of 0.49; P= 1 . 1 x 10^-8 (Table I I ): meta-analysis P- 1 .9x 10^-8 (Table III)) based on the ENCODE consortium reference dataset (Consortium, Nature, 2012, 489, 57-74), because (i) it provided, with three other SNPs (rs8013419, rs12147006 and rs1087301 8 ), the strongest association signal, and (ii) it is located in a chromatin-open region, between the 1)62 and D63 gene segments, which is targeted by the transcription factors RUNX3, ELF1 , FOXM1 and RNA polymerase 11 (Figures 3C and 4D) (Thurman et al, Nature, 2012, 489, 75-82). This variant is commonly distributed in human healthy populations with slight differences in allele frequencies, the A allele being more frequent in Asian populations and the G allele in

South America (http://www.ncbi.nim.nih.gov/variation/toois/1000genomes/). Therefore, the biological significance of the rs2204985 variant was further studied.

Human thymopoiesis in immunodeficient mice is controlled by TCRA-TCRD genetic variation at early developmental stages.

Immune deficient mice engrafted with human hematopoietic stem cells (HSC) sustain a conventional T-cell development close to human thymopoiesis (Huntington et al , Eur J Immunol 201 1 , 41 , 2883-2893; Halkias et al , Immunol Cell Biol 2015, 93, 716-726). To directly evaluate in vivo the impact of rs2204985 genotype on thymopoiesis, immunodeficient BRGS (Balb/c Rag2^-/-Il2rg^-/-Sirpa^NOD) mice ( Lopez- Las tra et al., Blood Advances 201 7, 1 , 601 -614) were reconstituted with human CD34 hematopoietic progenitors harv ested from foetal livers (FL) of di fferent genotypes (Figure 6A). Higher sjTRECs levels were observed in thymi and spleen of mice reconstituted with CD34 ⁺ progenitors of the GG genotype at rs2204985, as compared to mice reconstituted with AA or GA genotypes (Figure 6B). The total sjTRECs counts per organ (thymus or spleen) ( Figure 6C) as well as total thymocyte numbers ( Figure 6D) were also dependent on SNP polymorphism. This effect was independent of the sex of human FL cells and was observed in both male and female recipient mice studied separately ( Figure 7 ). In a parametric bootstrap test, mouse or donor sex had no con founder effect whereas taking into account the graft group, the impact of genotype on sjTRECs remained significant (/'=().049). In ail, these data support the hypothesis of a T- cell intrinsic effect of the identified genetic v ariant, with rs2204985 genotype being associated with higher thymocyte counts.

Having validated this mouse model, it was sought to get a closer look to the influence of SNP polymorphism on thymocyte sequential differentiation stages and TCRA-TCRD rearrangements. First, although 5Rec-J661 rearrangement is the prominent primary TCRD deleting element measured in sjTRECs (Figure IB), additional rearrangements may occur, especially between 5Rec and J 658 gene segments in about 23% of all 5Rec rearrangements (Verschuren et al , J Immunol 1997, 1 58, 1 208- 12 1 6). The rs2204985 polymorphism had the same effect on the alternative 6Rec-J558 rearrangement than on sjTRECs, therefore excluding an effect of the SNP polymorphism on the .10 segment usage during primary TCRA rearrangements.

Then, thymocyte developmental stages was studied by flow cytometry on human CD45 gated cells. An increase in total thymocyte counts was observed at all stages, starting as early as the CD3-CD4-CD8- double negative (DN) population ( Figure 6E). This led us to analyse in humanized mice thym i the TCRA-TCRD repertoire by next- generation sequencing (NGS) according to the A A vs. GG rs2204985 genotype in terms of I CR diversity and genes segments usage. Twenty thymi ( 1 2 GG vs. 8 AA) were analysed indiv idually (Table IV). The numbers of total and productive rearrangements did not differ in function of the genotype. No preferential overlap of TCRA and TCRD sequences calculated by the Morosita-Horn similarity index was found in association with the SNP genotype or between mice grafted with the same FL CD34+ cells excluding a methodological bias. Repertoire diversity was assessed with different indexes such as productive clonality. Shannon equitability or inverse Simpson^'s D dominance indexes. These approaches showed more diversity and less clonality in mice grafted with the GG genotype (Table IV). There was no difference in TCRA V and TCRAJ gene segments usage according to the SNP polymorphism (Figure 8 A, 8B). Conversely, the TCRDV and TCRDJ usage was quite different according to the genetic variation, with a preferential usage of gene segments close to the SNP region ( DJ and DV2, DV3) in rs2204985 A A individuals ( Figure 8B), translating in a higher frequency of T-cells carrying a productive TCRD rearrangement based on DJ vs. AJ usage frequencies ( Figure 8C). A fine analysis of TCRD V and TCRDJ usages showed that DVl, DD2 and DJl segments were used preferentially in GG whereas DV2, DD3 and DJ3 were used preferentially in A A indiv iduals ( Figure 8D).

In all, these data indicate that the SNP rs2204985 polymorphism impacts locally

TCRD rearrangements.

Modeling the variance of thymic function in healthy adults

An important issue is the impact of the genetic variation in the physiologic process of thymopoiesis.

When the rs2204985 variant was included in a first mixed interaction model analysis for the Milieu lnterieur cohorte, controlling for age, sex, population stratification and batch variables, an 80% increase in sjTRECs was observed in GG homozygous women, as compared to AA homozygous women (CI: [46%, 126%], P=5x10 , Figure 9A), with no significant impact found in men (/'=().44) using this mixed interaction model analysis. Strikingly, females carrying the GG genotype showed a 140% increase in sjTREC levels compared to males carrying the GG genotype (CI: [91 %, 200%] , P = 8x 10- ¹⁴). In contrast, there was no significant impact of rs2204985 variant allele in men using this mixed interaction model analysis. Similar results were also observed with the other more informative SNPs (rs 1 087301 8, rs2141988, rs1 1 8447 15, rs2204984, rs12147006, rs38 1 1 236, s8013419; Figure 10 )

Then, the effect of the rs2204985 variant on the immunophenotypes was tested using other mixed models, with iogio-transformed immunophenotypes as response variables, the rs2204985 variant as treatment variable, and age, sex, CMV serostatus and smoking as fixed effects covariates, and blood sampling day as random effect.

Using these other mixed models, a 44% increase of sjTRECs was observed in rs2204985 GG homozygotes. relative to A A homozygotes in the Milieu Interieur cohort (men and women) (marginal CI: [29%, 62%]) ( Figure 11 A).

Similarly, in the MARTHA cohort (men and women), a 41% increase of sjTRECs was observed in rs2204985 GG homozygotes, relative to AA homozygotes (marginal CI:

[19%, 69%]) ( Figure 1 I B). The rs2204985 variant also associated with increased numbers of circulating naive CD4 ^' and CD8 T cells (Ρ= 5x10^-2 and 3x l (P, respectively), as well as increased numbers of CD4- CD8- T cells (Ρ=8χ 10^-3 ) in the Milieu Interieur cohort in women.

To better understand the combined effects of age, sex and genetics on thymopoiesis, combined data were used in a mixed model that included rs2204985, age and sex, controlling for population stratification and batch variables. The variance decomposition of the obtained model was comparable between both the Milieu Interieur and MARTHA cohorts: age, sex and TCRA- TCRD genetic variation respectively explained 36.7%, 4.9%, and 1.3% of the variance of sjTREC levels in the Milieu Interieur cohort, and 25.6%, 8.5%, and 1.3% of the variance in the MA RTHA cohort ( Figure 1 1 C ).

The effect of age on sjTRECs was independent from that of rs2204985 genotypes (Figures 9B and 11A (Milieu Interieur); Figure 11B (MARTHA); slope of -0.01 for GG, -0.02 for AG and AA, ns in Milieu Interieur ( Figure 9B); CI: [0.94, 0.95], [0.94, 0.96] and [0.94, 0.96] respectively for AA, GA and GG in Milieu Interieur (Figure 1 1 4); CI: [0.95, 0.97], [0.95, 0.96] and [0.95, 0.98] respectively for AA, GA and GG in MARTHA ( Figure 11B)). As such, "thymic age" could be directly modeled approximated by sjTRECs values as a function of the rs2204985 SNP, and it was estimated that carrying the GG genotype is equiv alent to a 19-year difference in logio sjTRECs for women (CI: [ 1 5.3, 22.1]), and a 7-year difference in logio sjTRECs for men (CI: [4.8, 10.0]), relative to A A homozygous men ( Figures 1 I D and 9C).

It is believed that these efforts will help providing foundations for precision medicine strategies, in particular stratification of patients based on their thymic function potential.

In sum, this systemic approach in a large cohort of healthy adults identifies age, gender, and host genetics as important determinants of thymic function.

Regarding non-heritable factors impacting thymopoiesis, the only parameters defined in physiological conditions were age and sex, in contrast to the many pathological conditions affecting thymopoieisis, such as acute inflammation in sepsis (Venet et al, Nat Rev Nephrol, 201 8, 14, 1 2 1 - 1 37 ). obesity (Yang et al, Blood 2009, 1 14, 3803-3812) or endocrine dysfunction (Youm et al, Proc Natl Acad Sci U S A 2016, 1 1 3, 1 026- 1 03 1 : Ventevogel et al., Curr Opin Immunol, 2013, 25, 5 16-522 ).

The negative impact of age on thymic function has been reported since decades.

Our study brings additional insight into this effect, ev idencing a linear decrease at all stages of thymic differentiation and from 20 to 69 years of life. Additionally, a measurable thymic function until age 69 was reported in a large majority of the healthy donors, which reinforces the hypothesis of a thymus being functional even late in life.

Besides age, this study reports higher thymic function in women relativ e to men.

Previous studies reported higher thymic mass - as measured by computed tomography in young (20-30 yrs old) women as compared to young men (J. B. Ackman et al , Radiology, 2013, 268, 245-253 ). Herein, this observation is extended to thym ic function, and to all decades of life from 20 to 69 years-old. Interestingly, it was observed that the impact of gender on sjTRECs increases gradually with age, which suggests that both effects are mediated by common or synergistic biological mechanisms. In mice, androgens have a direct detrimental effect on stromal thymic epithelial cells (TECs; Rode et al, Proc Natl Acad Sci U S A. 2012, 109. 3463-3468), and male cortical TECs express low levels of genes implicated in thymocyte expansion and positive selection (M. Dumont-Lagace, C. St-Pierre, C. Perreault, Sci. Rep. 2015, 5, 1 2895 ). It is believed that the gender differences observed in this study could similarly reflect sex differences in TEC function, resulting in a more efficient bilateral crosstalk between thymocytes and thymic stroma and higher thymopoiesis in women, and synergizing with age-related changes in thymic stroma (Palmer, Front Immunol 2013, 4, 316; D. K. Shah, J. C. Zuniga-Pflucker, J. Immunol., 1 92, 2014, 401 7-4023; M. hi a et al, Immunity, 2008, 29, 45 1 -463 ). This may also account for the preferential association in women of the SNP rs2204985 with naive T-cell counts in the Milieu Interieur cohort. In ail, the direct evidence of sex as a covariate for thymopoiesis reinforces the need of stratifying immunological studies by sex (Markle et al.. Trends Immunol, 2014, 35, 97- 104).

Other cohort-based studies identified genetic loci controlling immune phenotypes. It has been reported that RTE numbers are strongly regulated, with a strong heritability but no defined genetic association (M. Roederer et al.. Cell, 201 5, 1 61 , 387-403 ). Also, naive CD27+ CD4 T-cell counts have a strong influence from heritable factors in healthy twins (Brodin et al., Ceil 160, 2015, 37-47) as well as in the Milieu Interieur cohort (Patin et al., Nat Immunol. 20 1 7. in press). Interestingly, in this last study, naive CD8+ and CD4+ T cell age-related decreases were estimated at 1 .6% and 3.6% per year respectively, which can be compared to the 5% decrease in sjTRECs per year estimated here in the same cohort. This reflects the long half-life of some naive T-cell populations in homeostatic conditions (Thome et al., Sci Immunol 2016, 1 ,). Collectiv ely, these studies show a stronger genetic association with naive rather than differentiated T-cells in the adaptive compartment. This engaged us to focus our search on a genetic contribution in T-cell generation using the TREC approach as the closest readout of TCR rearrangements.

The most striking result of this study is the demonstration that sjTRECs levels are controlled by genetic variations at the TCRA- TCRD locus, within the genomic region that is excised during TCR alpha recombination and sjTRECs generation, which offers novel insights into the TCR locus function. The TCRA-TCRD locus is organized in a single genetic locus contributing to 2 different TCR specificities, TCR /cS and TCRa(:k at 2 differential developmental stages, therefore requiring a complex program that regulates chromatin accessibility of TCRA and TCRD gene segments to the recombination machinery (Carico et al, Adv Immunol. 201 5, 128, 307-361). The 4 SNPs identified are located in a short segment spanning 4kb within the DD2 and DD3 intergenic region, in a close 5 ' position to the TCR5 enhancer (Εδ) (Figure 4). The best candidate variant. rs2204985, locates in an open-chromatin region that is a target for numerous transcription factors such as RUNX3 and ELF l , and for the RNA polymerase 11 (Consortium, Nature 2012, 489, 57-74).. This region regulates the expression of the TCRD enhancer Εδ close by (R. E. Thurman et al, Nature, 201 2, 489, 75-82 ). rs2204985 is also close to a CCCTC-binding factor (CTCF) binding element, mediating chromatin looping, and modulating the access of the recombination machinery to the chromatin (Z. Carico, M. S. Krangel, Adv . Immunol., 201 5, 1 28, 307-361 ; Chen et al., Nat Immunol 201 5, 1 6, 1085- 1093 ). As such, it is believed that the rs2204985 variant is likely to interfere directly, within the thymocytes, with the recombination events. The precise molecular mechanisms underlying this genetic polymorphism will require further studies. However, the data collected in immunodeficient mice experiments allow to draw some hypothesis. TCR5 rearrangement is the first to occur at the earliest CD34+CD38-CDla- DN stage (Dik et al., J Exp Med, 2005, 201 , 1 71 5- 1 723 ) and is highly ordered in humans due to RUNX1 interaction with DD2, DD2-DD3 rearrangements occurring before DD2-DJI rearrangements (Cieslak et al., J Exp Med, 2014, 2 1 1 , 1 82 1 - 1 832 ).

rearrangements measured in sjTRECs are first detected in immature single positive (ISP) cells and reach peak levels in SP thymocytes ( Dik et al.. J. Exp. Med., 2005. 201 , 1 71 5- 1 723 ). It was observed in this study an influence of SNP rs2204985. or a close genetic element in linkage disequilibrium, on DN thymocyte numbers before sjTREC generation ( Figure 6) and differences in DV and DJ usages according to the genotype (Figure 8), supporting a direct role of the genetic variation at an early stage of thymocyte differentiation. The higher sjTRECs levels and TCR diversity in mice engrafted with the rs22()4985 GG genotype could be the consequence of a higher rate of T-cell generation starting from this early DN stage. The higher usage of DV1, as well as sjTRECs levels, in GG compared to AA rs2204985 genotypes, ruled out a possible effect of the genetic variation on the reciprocal usage of DV1 and 5REC. The Εδ element is a major regulator of TCRD accessibility in DN thymocytes (Monroe et al., Immunity 1999, 10, 503-513), functioning over a limited chromosomal distance (Bassing et al., Proc Natl Acad Sci U S A, 2003, 100, 2598-2603). It has been suggested that Εδ may require additional upstream elements to promote TCRD accessibility ( Monroe et al.. Immunity 1999, 10, 503-5 13 ). Especially, matrix attachment regions (MAR), DNA regions that anchor chromatin loops to the nuclear matrix, have been identified in the human TCRD locus, a potent MAR being located about 20 kb upstream Εδ (Zhong et al., Proc Natl Acad Sci U S A 1999. 96, 1 1970- 1 1975 ). Taking this into account, it was hypothesized that the SNP rs2204985, or a genetic element nearby, could influence chromatin conformation and TCRD accessibility directly or through the binding of transcription factors and participate in the regulation of the TCRD recombination center (Zhao et al., J Exp Med, 2016. 213, 192 1 - 1936). It remains also to be explained how the I CR genetic polymorphism could be linked to survival and/or thymocyte proliferation at the DN stage. Indeed, the lack of difference in TCRA V and TCRAJ usages according to the genotype is not in favor for a differential lifespan in DP thymocytes (Guo et al., Nat. Immunol.. 2002, 3, 469-476). Of note, physiological DNA double-strand breaks generated in developing lymphocytes activate a broad transcriptional program (Bredemeyer et al.. Nature, 2008, 456, 819-823 ) some of them promoting lymphocyte survival as for instance the activation of p38MAPK in DN thymocytes (Pedraza-Aiva G. Pedraza-Alva et al., Activation of p38 MAP kinase by DNA double-strand breaks in V(D).I recombination induces a G2/M cell cycle checkpoint. EMBO J 25, 763-773 (2006., EM BO J, 2006, 25, 763-773 ). Also, transcription factors binding to rs22()4985 might affect DN survi val/prol i feration. such as FOXM 1 which is required for cellular proliferation in normal cells (Bella et al., Semin Cancer Biol, 2014, 29, 32-39). It is puzzling to show evidence for a genetic control of T- cell generation in a locus deleted in all mature peripheral T cells, TCR.y5 through TCRD rearrangements as well as TCRo.f} T cells through sjTRECs generation. This indicates a control at a critical step in T-cell development, which might be otherwise unnecessary or even harm ful if still functional in the periphery. In support to the importance of this genetic region, is its involvement during oncogene activation in T-cell acute lymphoblastic leukemia (Lc Noir et al.. Blood, 2012, 120, 3298-3309). Finally, the intensity of the effect of the genetic variant in women, accounting for up to 19 years differences in thymic function, strongly highlights its potential impact in the numerous clinical situations where the level of thymic function matters.

Finally, it will be interesting to assess the link between the level of thymic function triggered by SNP rs2204985 allelic variations in women, and the shaping of a tolerant T cell repertoire, and its connection with the strong female bias observed in autoimmune diseases (K. Rubtsova, P. Marrack, A. V. Rubtsov , J. Clin. Invest. 125, 201 5, 2 1 87-2 193 ).

Another issue raised by these findings is their medical applications.

This study contributes to a better understanding of aging of the immune system, a major public health concern (W. H. Organization, "World report on ageing and health," (World Health Organization, 201 5 ). This study shows clear differences between aging of thymic output and senescence of the immune system mostly driven by environmental factors, especially CMV infection , leading to exhaustion of differentiated T cells (Patin et al., Nat Immunol. 201 7. in press; N ikolich-Zugich et al., Curr Op in Immunol 201 7, 48, 23-30). About 50% of the variance of sjTRECs generation in this study remains undefined and may include other unknown environmental or genetic parameters. Nonetheless, differences in healthy thymic aging are shown depending on the TCRA- TCRD genetic variation in both independent cohorts of Caucasian origin. It would be interesting to evaluate such impact in other ethnic groups given the allele frequency differences in SNP rs2204985 across populations. The impact of the genetic v ariation in disease is probably not univocal and will depend on the pathological context. It could be discussed on the basis of our knowledge of the thymic crosstalk, a key feature of thymic function (Abramson et al., Annu Rev Immunol, 201 7, 35, 85- 1 1 8: Shah et al., J Immunol, 2014, 192, 401 7-4023 ).

We may anticipate that in case of a functional thymic environment it would be beneficial to get a high potential of T-cell production. This would be the case for instance in an uncomplicated alio-HSCT setting or in the recovery of lymphopenic conditions in young patients. At the opposite, it would be detrimental to fuel the system if the thymic environment is damaged as for instance in older individuals, in graft- versus host disease in allo-HSCT (W. Krenger, B. R. Bla/ar, G. A. Hollander, Blood. 201 1 . 1 1 7, 6768-6776; Clave et al.. Blood, 2009, 1 13, 6477-6484) or in autoimmune conditions where women are known to have an overall higher susceptibility (N. Dragin et al., J Clin Invest, 2016, 1 26, 1525-1537). In such cases, the consequence could be the generation of T cells defective in their selection process with an autoreactive potential which could be pathogenic. This will deserv e further dedicated studies integrating SNP rs2204985 genotyping into clinical algorithms. In particular, it is believed that the genotyping of SNP rs2204985 can hav e important implications in allo-HSCT, where its integration into clinical algorithms could help selecting donors with best potential for triggering an optimal recov ery of thymic function and a related optimal outcome of the transplantation. It will also be of interest to ev aluate the rs2204985 v ariant as a biomarker predicting disease severity or immune reconstitution (ev olution under therapy) in patients with H IV chronic infection (M. L. Dion et al , Immunity. 2004, 2 1 , 757-768; Douek et al.. Nature, 1998, 396, 690-695) a condition where II. -7 may restore CD4 lymphopenia (Thiebaut et al., Clin Infect Dis, 2016, 62, 1 178-1 185) and response to vaccination in adults and in the elderly which has been associated with RTE numbers in elderly humans (Schulz et al., J Immunol, 2015, 195, 4699-471 1).

In conclusion, by providing reference values of thymic function in a large healthy population and the definition of its genetic control, this study offers an unprecedented resource that may be used as a comparator for disease studies and as a platform for precision medicine and regenerative strategies, in particular reinforcing the need of stratifying immunological studies by gender (J. G. Markie, E. N. Fish, Trends Immunol., 2014, 35. 97- 104). Together, results from this study will help establish a path towards personalized care.

Claims

1 . An in vitro method of evaluating thymic function in an indiv idual, comprising: - determining, in a biological sample containing nucleic acid obtained from the indiv idual, the presence or absence of at least one allele of at least one polymorphic marker in the T-eell receptor alpha-T cell receptor delta ( TCRA- TCRD) locus that is associated with increased level of production of T lymphocytes by the thymus, wherein the presence of said allele is indicative of increased thymic function for the individual.

2. The method according to claim 1 , wherein the polymorphic marker is characterized by the following features:

a) an allele frequency higher than 40 %;

b) an association between the allele and increased level of T lymphocyte production by the thymus characterized by a P-value of less than 10^-7; and

c) a linkage disequilibrium w ith at least one polymorphic marker selected from the group consisting of: rs381 1236, rs2301 1 99, rs2301200, rs32 16790, rs62762262, rs2331618, rs l 1 091 30. rs2072616, rs801248 1 , rs6572448, rs2141988, rs916052, rs8021297, rs7492759, rs2204984. rs71 1 1 5550, rs201497432, rs8013419, rs l 087301 8, rs 12 147006, rs2204985. and rs l 1 8447 1 5, characterized by a numerical value of the linkage disequilibrium correlation measure r of greater than 0.7.

3. The method according to claim 1 or 2. wherein the polymorphic marker is a single-nucleotide-polymorphism (SNP).

4. The method according to claim 3, wherein the allele of the SNP is selected from the group consisting of: the C allele of rs38 1 1 236, the G allele of rs2301199. the C allele of rs2301200, the CG allele of rs32 1 6790, the TAGTC allele of rs62762262, the A allele rs2331618, the G allele of rs l 1 09130, the G allele of rs2072616, the G allele of rs8012481 , the T allele of rs6572448, the A allele of rs2141988, the A allele of rs916052, the T allele of rs802 1297, the G allele of rs7492759, the A allele of rs2204984, the A allele of rs71 115550, the T allele of rs201497432, the A allele of rs8013419, the G allele of rS1087301 8, the A allele of rs12147006, the G allele of rs2204985 and the C allele of rs l 1 84471 5.

5. The method according to claim 4, wherein the allele of the SNP is selected from the group consisting of: the A allele of rs8013419, the G allele of rsl 0873018, the

A allele of rs12147006 and the G allele of rs2204985.

6. The method according to claim 5, wherein the allele of the SNP is the G allele of rs2204985.

7. The method according to any one of claims 1 to 6, w herein the indiv idual is a woman of any age.

8. The method according to any one of claims 1 to 7, wherein the biological sample is whole-blood.

9. The method according to any one of claims 1 to 8, which comprises at least one procedure selected from the group consisting of:

a) amplification of nucleic acid from the biological sample;

b) hybridization assay using a nucleic acid probe and nucleic acid from the biological sample;

c) hybridization assay using a nucleic acid probe and nucleic acid obtained by amplification of the biological sample, and

d) sequencing, in particular high-throughput sequencing.

10. The method according to any one of claims 1 to 9, which comprises the determination of at least one other marker in the biological sample from the individual, selected from the group consisting of: HLA haplotype, signal-joint T-cell Receptor Excision Circles (sjTRECs) number, and drug-related or disease-related biomarkers related to conditions or clinical situations where thymopoiesis is impacted or that are impacted by thymopoiesis efficiency and quality.

1 1 . A kit for performing the method of evaluating thymic function according to any one of claims 1 to 1 0, which comprises:

- reagents necessary for selectively detecting at least one allele of at least one polymorphic marker as defined in any one of claims 1 to 6, and

- reagents necessary for detecting at least one other marker as defined in claim 1 0.

12. Use of the method according to any one of claims 1 to 10, for the in vitro diagnostic, prognostic, and/or monitoring of conditions or clinical situations where thymopoiesis is impacted or that are impacted by thymopoiesis efficiency and/or quality.

13. The use of the method according to claim 12, w herein said conditions or clinical situations are selected from the group consisting of: acquired immunodeficiencies, allo-hematopoietic stem cell transplantation and cellular therapies, gene therapy, immunosuppressive treatments, immunotherapy, immunoregenerativ c therapies, vaccination, infectious diseases, cancer and autoimmune disorders.

14. Use of the polymorphic marker as defined in any one of claims 1 to 6 for drug screening and/or drug dev elopment.

15. A method of treating a disease where thymopoiesis is impacted or that is impacted by thymopoiesis thymopoiesis efficiency and/or quality in a patient, comprising:

- determining, in a biological sample containing nucleic acid obtained from the patient, the presence or absence of at least one desired allele of at least one polymorphic marker in the T-cell receptor alpha-T cell receptor delta (TCRA-TCRD) locus that is associated with increased level of production of T lymphocytes by the thymus, using the method according to any one of claims 1 to 1 0, and

- administering a treatment to the patient, when the patient has the desired allele.

16. A method of selecting hematopoietic cells of interest, in particular hematopoietic stem cells or T lymphoid progenitor cells of interest, comprising: - determining, in nucleic acid of the hematopoietic cells, the presence or absence of at least one desired allele of at least one polymorphic marker in the T-cell receptor aipha-T cell receptor delta (TCRA-TCRD) locus that is associated with increased level of production of T lymphocytes by the thymus, using the method according to any one of claims 1 to 7 and 9 to 1 0, and

- selecting the hematopoietic cells having the desired allele.

17. The method according to claim 16, wherein the hematopoietic cells having the desired allele are human hematopoietic cells from a donor indiv idual hav ing the effect allele for allo-hematopoietic stem cell transplantation.

18. A genetically engineered hematopoietic cell, in particular a genetically engineered hematopoietic stem cell or T lymphoid progenitor cell, in which at least one allele of at least one polymorphic marker in the T-celi receptor aipha-T cell receptor delta (TCRA-TCRD) locus that is associated with increased level of production of T lymphocytes by the thymus as defined in any one of claims 1 to 6 has been replaced with the desired allele.

19. The genetically engineered hematopoietic cell according to claim 18, which is a genetically engineered human hematopoietic cells having the effect allele.

20. A pharmaceutical composition comprising an effective amount of hematopoietic cells hav ing the desired allele of a polymorphic marker as defined in any one of claims 1 to 6, in particular hematopoietic stem cells or T lymphoid progenitor cells, either genetically engineered according to claim 1 8 or 1 9 or obtained by the selection method according to claim 16 or 17, and a pharmaceutically acceptable carrier, vehicle, and/or excipient.

21. A pharmaceutical composition comprising an effective amount of gene- editing system engineered for replacing specifically one allele of at least one polymorphic marker as defined in any one of claims 1 to 6 with the desired allele, in the T-cell receptor alpha-T cell receptor delta locus of at least one hematopoietic stem cell or T lymphoid progenitor of a patient, and a pharmaceutically acceptable carrier, vehicle, and or excipient.

22. A pharmaceutical composition according to claim 20 or 21 , for use in the treatment of a condition where thymopoiesis is impacted or that is impacted by thymopoiesis efficiency and/or quality, as defined in claim 1 2 or 13.

23. A method of treating a disease where thymopoiesis is impacted or that is impacted by thymopoiesis efficiency and or quality in a patient, comprising administering an effective amount of the pharmaceutical composition according to claim 20 or 2 1 to the patient.

24. The method according to claim 23, which comprises :

- determining, in a biological sample containing nucleic acid obtained from the patient, the presence or absence of at least one desired allele of at least one polymorphic marker in the T-cell receptor alpha-T cell receptor delta (TCRA-TCRD) locus that is associated with increased lev el of production of T lymphocytes by the thymus, using the method according to any one of claims 1 to 1 0, and

- administering an effectiv e amount of the pharmaceutical composition, when the patient does not hav e the desired allele.