FIELD OF THE INVENTION

This invention refers to the field of human medicine. More specifically, the present invention is focused on a method and kit for the in vitro diagnosis and/or prognosis of the tolerant state of a patient to be submitted to liver transplantation, which comprises assessing the level of systemic and/or intra-hepatic iron stores in a biological sample obtained from the patient under investigation, and comparing it either with the level of iron stores of a reference sample, or with a pre-determined threshold.

STATE OF THE ART

The long-term survival of transplanted grafts critically depends on the life-long administration of immunosuppressive drugs to prevent graft rejection. These drugs are very effective at preventing graft rejection, but they are also associated with severe side effects, such as nephrotoxicity, an augmented risk of opportunistic infections and tumors, and metabolic complications such as diabetes, hyperlipidemia and arterial hypertension. Due to the side effects of immunosuppressive drugs, the induction of tolerance, defined as a state in which the graft maintains a normal function in the absence of chronic immunosuppression, is one of the main goals of research in transplant immunology. Tolerance induction is possible in a great number of experimental models of transplant in rodents. Nevertheless, the application of these experimental treatments in the clinic has been a failure to a large extent. One of the reasons why clinical application in humans of experimental treatments of tolerance induction has not been successful relates to the lack of an accurate tool to non- invasively diagnose tolerance in human transplant recipients. Recent publications point out the urgent need for this tool (N. Najafian et al 2006 and Newell et al. 2006). On the other hand, maintenance of a normal allograft function despite complete discontinuation of all immunosuppressive drugs is occasionally reported in clinical organ transplantation, particularly following liver transplantation. Patients spontaneously accepting grafts are conventionally considered as "operationally" tolerant, and provide a proof of concept that immunological tolerance can actually be attained in humans. Liver transplantation is the only clinical setting in which tolerance spontaneously occurs in a substantial proportion of patients. Indeed, complete immunosuppression withdrawal can be achieved in around 21% of patients (Lerut, J. et al 2006). Unfortunately, there are currently no means to identify these patients before immunosuppression withdrawal is attempted. For this reason, complete discontinuation of immunosuppressive drugs is rarely attempted in liver transplantation, and thus many patients continue to be unnecessarily immuno suppressed, with the health and economic problems that this involves. Prior attempts to identify tolerance in transplantation, mainly in kidney and liver recipients, have employed either antigen-specific functional assays or antigen nonspecific tests. In the functional assays recipient T lymphocytes are challenged with donor antigens either in vitro or in vivo (J. Cai et al 2004), (J. Cai et al 2004) and (E. Jankowska-Gan E et al 2002), (P. Sagoo et al. 2010). These assays are very valuable from a mechanistic point of view, since they are the only tests capable of revealing which pathways are responsible for the specificity of the tolerance state. Unfortunately, these assays are also difficult to perform, highly variable from laboratory to laboratory (difficult to standardize), and require the availability of carefully cryopreserved donor cells. For these reasons, functional assays are not optimal for widespread clinical application, and are currently employed only in selected, highly specialized laboratories, and basically for research purposes.

The antigen-non specific immune monitoring tests constitute a variety of methodologies aiming at the phenotypic characterization of the recipient immune system, without the use of donor antigen challenges. Among these tests, the study of T cell receptor CDR3 length distribution patterns (TcLandscape), peripheral blood cell immunophenotyping by employing flow cytometry, and gene expression profiling have been employed to identify biomarkers characteristic of tolerance in humans. The TcLandscape technique has been employed in peripheral blood to discriminate between tolerant kidney recipients and recipients experiencing chronic rejection (S. Brouard et al. 2005). However, this technique is expensive, is currently only available at one laboratory (Inserm 643 and TcLand Expression in Nantes, France), and has never been validated in liver transplantation. The use of peripheral blood immunophenotyping has been used with peripheral blood samples from both liver and kidney tolerant transplant recipients. At least four studies addressing this methodology are known to inventors. In the first one, from the University of Pittsburgh in USA (G.V. Mazariegos et al 2003), it is said that the ratio between pDC and mDC dendritic cell subsets could discriminate between tolerant and non-tolerant recipients in pediatric liver transplantation. In the second study, from Kyoto (Y. Li et al 2004), it is said that an increased ratio between delta-1 and delta-2 gammadelta T cells in peripheral blood is more prevalent in tolerant than in non-tolerant liver recipients. In the third study, which was coordinated by the inventors (Martinez-Llordella et al 2007), an increased number of CD4+CD25+ T cells and an increased ratio of delta-1 to delta-2 gammadelta T cells were noted in peripheral blood of tolerant liver recipients as compared with non-tolerant recipients. The value of the ratio between delta-1 and delta-2 gammadelta was however questioned in a subsequent study from the same group (Puig-Pey et al. Transplant Int 2010). Furthermore none of these tests offers the accuracy required for a widespread clinical application. The use of gene expression profiling techniques to identify biomarkers of tolerance has been employed both in kidney and in liver transplantation (S.Brouard et al. PNAS 2007; M. Martinez- Llordella et al. J Clin Invest 2008; K.Newell et al . J Clin Invest 2010; P.Sagoo et al. , J Clin Invest 2010). These techniques are easier to standardize than the tests described before. Furthermore, the referenced studies have shown that the use of the identified transcriptional biomarkers is an extremely accurate means to differentiate between tolerant recipients off immunosuppressive drugs and non-tolerant recipients who require maintenance immunosuppression. The main limitation of the studies published in the literature so far is that they have not attempted to prospectively validate their results. In other words, they have not been able to demonstrate whether these biomarkers can identify tolerant recipients before immunosuppression is discontinued. In the absence of this demonstration it is not possible to be certain that the differences observed in gene expression are not actually caused by the effect of pharmacological immunosuppression in the group of non-tolerant recipients. Furthermore, none of the previously reported studies have attempted to investigate whether differences in gene expression also exist at the level of the graft itself. While the chronic use of immunosuppressive drugs is currently the only means to ensure long-term survival of transplanted allografts, these drugs are expensive and are associated with severe side effects (nephrotoxicity, tumor and infection development, diabetes, cardiovascular complications, etc.) that lead to substantial morbidity and mortality. Hence, any strategy capable of significantly reducing the use of immunosuppressive drugs in transplantation may have a large impact on the health and quality of life of transplant recipients.

In conclusion, the provision of a validated method able to predict tolerance in liver transplant patients, and thus capable of indicating that the administration of immunosuppressive drugs to said patients can be dispensed with, remains being a challenge.

DESCRIPTION OF THE INVENTION

Therefore, the present invention aims to solve the above cited problem by providing an in vitro method to identify tolerant liver transplant recipients by assessing the level of systemic and/or intra-hepatic iron stores in a biological sample obtained from the patient under investigation, and comparing it either with the level of iron stores of a reference sample, or with a pre-determined threshold. This is based on the fact that the levels of systemic and/or intra-hepatic iron stores (the total amount of iron present in the body, either as free iron or bound to proteins such as transferring or ferritin) are significantly higher in tolerant liver transplant recipients as compared with non- tolerant liver transplant recipients. In a preferred embodiment of the invention, the assessment of the level of systemic and/or intra-hepatic iron stores is carried out by means of the evaluation, in the liver biopsy of the patient under investigation, of the expression profile of a specific group of genes directly involved in iron metabolism, which are reliable biomarkers able to predict tolerance in liver transplant patients. Thus, with the objective of identifying those genes showing a statistically significant difference in expression level profiles between liver recipients who can discontinue immunosuppressive therapy (tolerant) and those who require maintenance immunosuppressive drugs (non-tolerant), biopsy liver tissue samples were collected from a group of stable liver transplant recipients, under maintenance immunosuppression therapy, who were enrolled in a prospective clinical trial of immunosuppressive drug withdrawal. The expression profile may be determined by any technology known by a man skilled in the art. In particular, each gene expression level may be measured at the genomic and/or nucleic and/or protein level. In a preferred embodiment, the expression profile is determined by measuring the amount of nucleic acid transcripts of each gene. In another embodiment, the expression profile is determined by measuring the amount of protein produced by each of the genes.

The amount of nucleic acid transcripts can be measured by any technology known by a man skilled in the art. In particular, the measure may be carried out directly on an extracted messenger RNA (mRNA) sample, or on retrotranscribed complementary DNA (cDNA) prepared from extracted mRNA by technologies well-known in the art. From the mRNA or cDNA sample, the amount of nucleic acid transcripts may be measured using any technology known by a man skilled in the art, including nucleic microarrays, quantitative PCR, and hybridization with a labelled probe. In a particular embodiment, which should not be considered as limiting the scope of the invention, the determination of the expression profile of these biopsies were conducted by using Illumina Beadchip whole-genome expression microarrays, which identified genes with p-value<0.01 and false discovery rate (FDR) < 25% (Table 1). Table 1. List of genes differentially expressed between tolerant and non-tolerant liver biopsy samples

NCBI reference

Gene sequence FDR p-value

(RefSeq)

TFRC No Annotation 0 1.19E-005

LOC644037 XR 017337 0 2, 13E-005

LOC729266 XM 001721977 10,0141821 1 4,50E-006

ST70T1 NR 002330 10,0141821 1 9, 12E-006

MY019 NM 001033580 10,0141821 1 1.16E-005

TP53I3 NM 004881 10,0141821 1 0,000143629

HAMP NM 021 175 10,0141821 1 0,001 152774

MCOLN1 NM 020533 15,02127316 3,52E-005 OTUD7A NM 130901 15,02127316 4 21 E-005

EXT2 NM 000401 15,02127316 7 92E-005

KLHL28 NM 017658 15,02127316 0 000219206

UHMK1 NM 175866 15,02127316 0 000252044

FIGF NM 004469 15,02127316 0 000482129

SLC1A7 NM 006671 15,02127316 0 000917616

ADORA3 NM 020683 15,02127316 0 001 104349

SLC5A12 NM 178498 15,02127316 0 001594655

TAF15 NM 139215 15,02127316 0 001995782

TPPP3 NM 016140 16,89893231 2 00E-005

TAGLN NM 001001522 16,89893231 2 71 E-005

NFKBIL2 NM 013432 16,89893231 3 36E-005

OR2C3 NM 198074 16,89893231 0 000221256

UNG NM 08091 1 16,89893231 0 000297046

GHSR NM 004122 16,89893231 0 000320149

KRTAP5-10 NM 001012710 16,89893231 0 00041 1491

UNC13A NM 001080421 16,89893231 0 000434094

G3BP1 NM 198395 23,05590765 0 000139785

ANKRD5 NM 022096 23,05590765 0 000181261

RBM23 NM 018107 23,05590765 0 000181688

RAG2 NM 000536 23,05590765 0 000191286

TUBA8 NM 018943 23,05590765 0 000201266

DGKK NM 001013742 23,05590765 0 000210462

C1orf61 NM 006365 23,05590765 0 000286859

ADSSL1 NM 199165 23,05590765 0 0004024

FBXL4 NM 012160 23,05590765 0 000406323

VAC14 NM 018052 23,05590765 0 000426174

LOC643668 XR 039201 23,05590765 0 000427565

RNASE13 NM 001012264 23,05590765 0 000451496

SAGE1 NM 018666 23,05590765 0 000474797

RTP2 NM 001004312 23,05590765 0 000516765

SYNE2 NM 182910 23,05590765 0 00052836

TSPAN2 NM 005725 23,05590765 0 000583719

SCRG1 NM 007281 23,05590765 0 000676665

ACSL1 NM 001995 23,05590765 0 000734359

STRN4 NM 013403 23,05590765 0 000767142

TUBA4A NM 006000 23,05590765 0 000924509

RIBC1 NM 144968 23,05590765 0 000965443

MCHR1 NM 005297 23,05590765 0 00099149

MUTED NM 201280 23,05590765 0 001028448

TANK NM 004180 23,05590765 0 001031594

DPP4 NM 001935 23,05590765 0 001 174649

CHD3 NM 001005271 23,05590765 0 00129579

KLK15 NM 017509 23,05590765 0 00133

NFIX NM 002501 23,05590765 0 001358456

FAM3B NM 206964 23,05590765 0 001386029

DOC2A NM 003586 23,05590765 0 001388434

NQ02 NM 000904 23,05590765 0 001470921

KIAA1143 NM 020696 23,05590765 0 001478818

PLCD3 NM 133373 23,05590765 0 001551069

PLXNA4 NM 181775 23,05590765 0 001589304

ADAMTS3 NM 014243 23,05590765 0 001589498

ABAT NM 001 127448 23,05590765 0 001742645

POF1B NM 024921 23,05590765 0 001746171

CYP2W1 NM 017781 23,05590765 0 001815383

HSPA1A NM 005345 23,05590765 0 002150872 FAM162A NM 014367 23,05590765 0,002262613

KIAA1274 NM 014431 23,05590765 0,002458413

CP NM 000096 23,05590765 0,002664451

OR5A2 NM 001001954 23,05590765 0,002806537

C9orf127 NM 001042589 23,05590765 0,003069775

AMPD2 NM 203404 23,05590765 0,004156098

KRT19 NM 002276 23,05590765 0,006219448

ATP1B1 NM 001677 23,05590765 0,006336133

C20orf71 NM 178466 24,92892142 0,001022188

LTBP4 NM 001042544 24,92892142 0,00131758

CNTNAP1 NM 003632 24,92892142 0,002377442

FNDC3A NM 014923 24,92892142 0,00241448

ZNF665 NM 024733 24,92892142 0,0033221 1

On the basis of the microarray results and a number of studies conducted experimental animal models of immunological tolerance, a set of 104 genes (listed

Table 2) were then selected for validation employing quantitative real time PCR.

Table 2. List of genes analysed by real-time PCR

Gene NCBI Gene ID Selection

Name

criteria

18S 100008588 Eukaryotic 18S rRNA HK

TP53I3 9540 tumor protein p53 inducible protein 3 M

HAMP 57817 hepcidin antimicrobial peptide M

SAGE1 55511 sarcoma antigen 1 M

DPP4 1803 dipeptidyl-peptidase 4 M

MY019 80179 myosin XIX M

MCOLN1 57192 mucolipin 1 M

ACSL1 2180 acyl-CoA synthetase long-chain family member 1 M

UNG 7374 uracil-DNA glycosylase M

TFRC 7037 transferrin receptor (p90, CD71) M

TUBA4A 7277 tubulin, alpha 4a M

COG5 10466 component of oligomeric golgi complex 5 M

F AMI 62 A 26355 family with sequence similarity 162, member A PK

FKBP1A 2280 FK506 binding protein 1A, 12kDa PK

ABAT 18 4-aminobutyrate aminotransferase M

CP 1356 ceruloplasmin (ferroxidase) PK

HLA-E 3133 major histocompatibility complex, class I, E PK

CXCR7 57007 chemokine (C-X-C motif) receptor 7 PK

SRGN 5552 serglycin PK

PRF1 5551 perforin 1 (pore forming protein) PK

TLR8 51311 toll-like receptor 8 PK ST ATI 6772 signal transducer and activator of transcription 1 PK

IL18BP 10068 interleukin 18 binding protein PK

PSMB9 5698 proteasome (prosome, macropain) subunit 9 PK

HFE 3077 hemochromatosis PK

IRF1 3659 interferon regulatory factor 1 PK

CXCL9 4283 chemokine (C-X-C motif) ligand 9 PK

UBD 10537 ubiquitin D PK

CD8A 925 CD8a molecule PK

IL32 9235 interleukin 32 PK

CXCL10 3627 chemokine (C-X-C motif) ligand 10 PK

CCL3 6348 chemokine (C-C motif) ligand 3 PK

CD3D 915 CD3d molecule, delta (CD3-TCR complex) PK

IL6 3569 interleukin 6 (interferon, beta 2) PK

ILIA 3552 interleukin 1, alpha PK

IL1B 3553 interleukin 1, beta PK

TFR2 7036 transferrin receptor 2 PK

HFE2 148738 hemochromatosis type 2 (juvenile) PK

BMP4 652 bone morphogenetic protein 4 PK

SMAD4 4089 SMAD family member 4 PK

FTH1 2495 ferritin, heavy polypeptide 1 PK

PDCD1 5133 programmed cell death 1 PK

HLA-G 3135 major histocompatibility complex, class I, G PK

FOXP3 50943 forkhead box P3 PK

IL10 3586 interleukin 10 PK

TGFB1 7040 transforming growth factor, beta 1 PK

IL2RB 3560 interleukin 2 receptor, beta PK

KLRF1 51348 killer cell lectin-like receptor subfamily F, 1 PK

SLAMF7 57823 SLAM family member 7 PK

KLRD1 3824 killer cell lectin-like receptor subfamily D, 1 PK

CX3CR1 1524 chemokine (C-X3-C motif) receptor 1 PK

LING02 158038 leucine rich repeat and Ig domain containing 2 PK

BNC2 54796 basonuclin 2 PK

NCRl 9437 natural cytotoxicity triggering receptor 1 PK

COL13A1 1305 collagen, type XIII, alpha 1 PK

IGFBP7 3490 insulin-like growth factor binding protein 7 PK

SH2D1B 117157 SH2 domain containing IB PK

NCAM1 4684 neural cell adhesion molecule 1 PK

KLRK1 22914 killer cell lectin-like receptor subfamily K, 1 PK

KLRC1 3821 killer cell lectin-like receptor subfamily C, 1 PK MICA 4276 MHC class I polypeptide-related sequence A PK

MICB 4277 MHC class I polypeptide-related sequence B PK

TLR4 7099 toll-like receptor 4 PK

GZMB 3002 granzyme B (granzyme 2) PK

APIS! 8905 adaptor -related protein complex 1, sigma 2 PK

SMARCD3 6604 SWI/SNF related, matrix associated PK

CD 37 951 CD37 molecule PK

FCER2 2208 Fc fragment of IgE, low affinity II, (CD23) PK

MS4A1 931 membrane-spanning 4-domains PK

CXCR3 2833 chemokine (C-X-C motif) receptor 3 PK

CXCL11 6373 chemokine (C-X-C motif) ligand 11 PK

IFNG 3458 interferon, gamma PK

CD274 29126 CD274 molecule PK

PDCD1LG2 80380 programmed cell death 1 ligand 2 PK

C3 718 complement component 3 PK

TBX21 30009 T-box 21 PK

GAT A3 2625 GATA binding protein 3 PK

FAS 355 Fas (TNF receptor superfamily, member 6) PK

FASLG 356 Fas ligand (TNF superfamily, member 6) PK

RORC 6097 RAR-related orphan receptor C PK

HMOX1 3162 heme oxygenase (decycling) 1 PK

TNFA1P3 7128 tumor necrosis factor, alpha-induced protein 3 PK

BCL2 596 B-cell CLL/lymphoma 2 PK

SOCS1 8651 suppressor of cytokine signaling 1 PK

TNF 7124 tumor necrosis factor (TNF superfamily, 2) PK

NOS2 4843 nitric oxide synthase 2, inducible PK

IL12B 3593 interleukin 12B (natural killer cell stimulatory 2) PK

IL18 3606 interleukin 18 (interferon-gamma- inducing factor) PK

IRF3 3661 interferon regulatory factor 3 PK

CCL21 6366 chemokine (C-C motif) ligand 21 PK

HPRTl 3251 hypoxanthine phosphoribosyltransferase 1 HK

GAPDH 2597 glyceraldehyde-3 -phosphate dehydrogenase HK

DTWD2 285605 DTW domain containing 2 M

POF1B 79983 premature ovarian failure, IB M

MYD88 4615 myeloid differentiation primary response gene (88) PK

DAB2 1601 disabled homolog 2 M

TIPARP 25976 TCDD-inducible poly(ADP-ribose) polymerase M

RBM23 55147 RNA binding motif protein 23 M

TTC3 7267 tetratricopeptide repeat domain 3 M MIF 4282 macrophage migration inhibitory factor M

PEBPl 5037 phosphatidylethanolamine binding protein 1 M

SLC5A12 159963 solute carrier family 5 member 12 M

FABP4 2167 fatty acid binding protein 4 M

PCDH24 54825 protocadherin 24 M

VNN3 55350 vanin 3 M

ADORA3 140 adenosine A3 receptor M

TAF15 8148 TATA box binding protein (TBP)-associated factor M

(M=significant in microarray, PK=previous knowledge, HK=houskeeping control)

NCBI accession date: 28 ^th July 2010

The results of the experiments conducted by real-time PCR revealed that the genes listed in Table 3 shows a statistically significant difference in expression between biopsies taken from liver transplant p atients who can safely ab andon immunosuppressive drugs (tolerant) and patients who undergo rejection when immunosuppressive drugs are discontinued (non-tolerant). As shown in Table 3 the genes TFRC and MIF are down-regulated, and the genes CDHR2, HMOX1, HAMP, IFNG, PEBPl, SLC5A12, ADORA3 and DAB2 are up-regulated, in tolerant liver transplant recipients as compared with non-tolerant liver transplant recipients. Identical results can be obtained if biopsies taken from liver transplant patients are compared with a reference RNA sample (which can be a pool of RNAs obtained from healthy non-transplanted liver tissue, a reference RNA such as the commercially available Human Liver Total RNA from Ambion, or an absolute reference consisting in a sample containing a previously quantified number of RNA molecules).

Table 3

p- value

Gene Student t Wilcox fold change

TFRC 0,000035 0,000026 -2,505329

CDHR2 0,006059 0,004665 1 ,747146

HMOX1 0,0071 95 0,005044 1 ,399586

MIF 0,008793 0,003526 -1 ,547565

HAMP 0,012583 0,077430 2, 1 73470

IFNG 0,013215 0,020816 1 ,319508

PEBP1 0,023371 0,01 1 855 1 , 1 32884

SLC5A12 0,032834 0,022216 2,345670

ADORA3 0,039721 0,041 890 1 ,464086

DAB2 0,046314 0,0491 69 1 , 1 93336 It is important to note that the genes comprised in Table 3 share a functional pathway because they are involved in the regulation of iron metabolism. In fact, the biopsies of tolerant patients who can successfully discontinue the immunosuppressive medication showed a greater accumulation of iron, as shown in Figure 1A. Furthermore, these differences in intra-hepatic iron content were independent from any clinical parameter such as time since transplantation or type of immunosuppressive therapy employed at baseline (Figure IB). It is known that the genes TFRC, HAMP, IFNG and HMOX1 are directly involved in the control of cellular iron metabolism. In particular, in a situation of systemic iron deficiency TFRC expression is typically increased while HAMP expression is decreased. In our experiments, the involvement of the genes found to be differentially expressed between tolerant and non-tolerant patients in the regulation of iron metabolism was further illustrated by the observation that TFRC, HAMP, CDHR2, MIF, SLC5A12, ADORA3, HMOX1, IFNG and DAB2 significantly correlated with the deposition of intra-hepatic iron (measured by the modified method of Scheuer or the total iron score method; see Figure 2). Therefore it can be stated that the expression profile of the genes comprised in Table 3, is a clear indication of the presence of significantly higher levels of systemic and/or intra-hepatic iron stores in tolerant patients as compared with non-tolerant patients (Figure 2), and consequently a reliable biomarker for the to identification of tolerant liver transplant recipients.

More specifically, these results indicate that the expression level of the genes involved in said regulation of iron metabolism should be particularly relevant for the design of a method according to the present invention. Hence, any expression profile of genes belonging to the regulation of iron metabolism at the intrahepatic level should be considered as an equivalent expression profile, or pattern, in the described and claimed invention. Accordingly, the expression of any gene belonging to said regulation of iron metabolism in liver should be considered as a functional equivalent of the genes described in the present invention. Thus, a preferred embodiment of the present invention refers to a method for the in vitro diagnosis and/or prognosis of the tolerant state of a patient subjected to a liver transplantation that comprises: a. Obtaining a biological sample from the liver allograft of the patient under investigation;

b. Measuring the expression levels of at least one of the following genes or combination or equivalent thereof: TFRC, CDHR2, HMOX1, MIF, HAMP, IFNG, PEBP1, SLC5A12, ADORA3 and DAB 2; c. Assessing the tolerance or non-tolerance of the patient under investigation to the transplanted liver allograft by comparing the intra- graft expression level of at least one of the genes of the step b), in a sample taken a liver biopsy, with the expression level of the same genes taken from a reference RNA sample.

The reference sample is a predetermined expression profile, obtained from a biological sample of the liver tissue of a healthy non-transplanted subject. It can be a pool of RNAs, a reference RNA such as the commercially available Human Liver Total RNA from Ambion, or an absolute reference consisting in a sample containing a previously quantified number of RNA molecules).

As shown in Table 4 below, measurement of the expression level of each of the genes comprised in Table 3 is useful for the identification of patients who can safely discontinue all immunosuppressive medication without undergoing rejection (tolerance). Therefore, this Table 4 shows the capacity of the individual genes listed therein to statistically differentiate the patients who will tolerate the transplanted liver in the absence of immunosuppressive therapy, from those recipients who will reject when immunosuppressive medications are discontinued. Table 4

GENES AUC SN SP ER PPV NPV

TFRC 0, 76 51, 72 90,48 25,35 78,95 73,08

CDHR2 0, 70 58,62 83,33 26, 76 70,83 74,47

HMOX1 0, 70 55,17 83,33 28,17 69,57 72,92

PEBP1 0,68 44,83 90,48 28,17 76,47 70,37

MIF 0,68 93,1 52,38 30,99 57,45 91,67

SLC5A12 0,66 44,83 85, 71 30,99 68,42 69,23

DAB2 0,65 58,62 73,81 32,39 60, 71 72,09

IFNG 0,64 31,03 90,48 33,8 69,23 65,52

HAMP 0,63 86,21 52,38 33,8 55,56 84,62

ADORA3 0,65 20,69 95,24 35,21 75 63,49

AUC: area under the curve

SN: sensitivity

SP: specificity

ER: error rate

PPV: positive predictive value

NPV: negative predictive value However, although the genes cited in Tables 3 or 4 have an individual predictive capacity, different clusters were made departing from some combinations of said genes, with the aim of identifying a predictive method as accurate as possible. Moreover, the genes listed in Tables 3 or 4 were also grouped with other genes which did not show a predictive value per se (as taken independently), for example: LC5A12, VNN3, SOCS1, TTC3, RBM23, SH2D1B, NCR1, TFRC, TUBA4A, TAF15, TIPARP, MOXl, MCOLNl, EBPl, DHR2, and AB2. Therefore, in a preferred embodiment, the present invention further comprises measuring the expression levels of at least one of the following genes: LC5A12, VNN3, SOCS1, TTC3, RBM23, SH2D1B, NCR1, TFRC, TUBA4A, TAF15, TIPARP, MOXl, MCOLNl, EBPl, DHR2, and AB2 in combination with at least one of the genes listed in Table 3 or 4.

In order to identify the combination/s of gene expression bio markers with the best performance in the diagnosis of the outcome of immunosuppression drug withdrawal in liver transplantation, we conducted an exhaustive search for predictive models employing the linear discriminate analysis and logistic regression algorithms implemented in the misclassification penalized posterior (MiPP) software. First, we conducted a 10-fold cross-validation step on a group of liver samples (18 tolerant and 31 non-tolerant) collected from patients enrolled in Hospital Clinic Barcelona. Next, random splitting cross-validation of the diagnosis models was conducted on the whole data set (which included the 56 samples from Barcelona and 21 additional samples from Rome and Leuven) by repeatedly partitioning it into training set (2/3) and independent test set (1/3) for external model validation. In addition, for each model identified in the training set the optimal probability cut-off of tolerance was computed employing ROC (Receiver Operating Curves) analysis. To demonstrate that the performance of the models was not center-dependent, we then computed SN, SP, NPV, PPV and overall error rates for the samples collected from Barcelona and those obtained from Rome and Leuven. Importantly, all gene expression measurements were performed on samples obtained before immunosuppression medications were discontinued. Our results therefore indicate that the identified genetic markers are capable of predicting the success of immunosuppression drug withdrawal.

As cited above, this type of analysis takes into account not only those genes found to be differently expressed genes (Table 3) but also genes that, although they are not statistically differentially expressed, as taken independently, they do contribute to optimize the diagnosis in combination with the genes of Table 3. Table 5 shows the groups of samples employed for the design and evaluation of the predictive models based on the expression in liver biopsy of the genes measured by real-time PCR. Importantly, while the samples collected from Barcelona recipients were employed for both microarray and qPCR experiments, none of the samples obtained from Rome and Leuven were employed in the microarray experiments.

Table 5

TOL (n) Non-TOL (n) origin

Training set 18 31 Barcelona

Test set 10 11 Rome/Leuven Thus, Table 6 shows the combinations of genes whose expression best classifies patients into the non-tolerant or tolerant categories according to the results of the qPCR expression measurements. A classification error of less than 15% in the learning group and less than 15% in the validation group was arbitrary selected to select the most accurate and clinically useful models.

Table 6

Barcelona Rome + Leuven

Genes n SN SP ER SN SP ER

SLG5A12+ VA/A/3+ TFRG+ S0CS1+ IW/F+ TTG3+RB

M23+PEBP1+SH2D1B+NCR1+DAB2+ADORA3 12 100 90,32 6,12 90 90,91 9,52

TFRC+PEBP1+MIF+CDHR2+HAMP+TUBA4A+TT

G3+HM0X1 + VNN3+NCR 1 +AD0RA3+ TAF15+/FA/

G+S0CS1+TIPARP 15 94,44 93,55 6,12 80 100 9,52

HM0X1 + CDHR2+MIF+PEBP1+ TFRC+ SLC5A 12+

S0CS1+HAMP+VNN3+IFNG 10 94,44 90,32 8,16 80 100 9,52

TFRC+PEBP1+MIF+CDHR2+SLC5A12+HAMP+S

0CS1+IFNG+HM0X1 9 94,44 90,32 8,16 80 100 9,52

TFRC+IFNG+CDHR2+ADORA3+HAMP+MIF+PEB

P1+ VNN3+ S0CS1 +HM0X1 +DAB2 11 88,89 90,32 10,2 80 100 9,52

TFRC+DAB2+MIF+PEBP1+IFNG+HAMP+SLC5A1

2+S0CS1+ VNN3+ADORA3+ CDHR2+MC0LN1+H

M0X1 13 77,78 96,77 10,2 80 100 9,52

TFRC+IFNG+HM0X1+MC0LN1+MIF+HAMP+AD

ORA3+CDHR2+PEBP1+SOCS1 10 88,89 90,32 10,2 80 100 9,52

PEBP1+TFRC+HM0X1+IFNG+MC0LN1+S0CS1

+MIF+CDHR2+HAMP+ADORA3 10 88,89 90,32 10,2 80 100 9,52

TFRC+PEBP1+IFNG+CDHR2+ADORA3+VNN3+H

M0X1 +DAB2+ S0CS1 +MIF+HAMP 11 88,89 90,32 10,2 80 100 9,52

CDHR2+ADORA3+IFNG+ TFRC+ VNN3+HM0X1 +

PEBP1+MIF+SLC5A12+HAMP+SOCS1+MCOLN1 12 77,78 96,77 10,2 80 100 9,52

SLC5A 12+ TFRC+IFNG+MIF+DAB2+HM0X1 + CD

HR2+SOCS1+HAMP+PEBP1+VNN3+ADORA3+M

C0LN1 13 77,78 96,77 10,2 80 100 9,52

TFRC+ S0CS1 +HM0X1 +PEBP1+ VNN3+ CDHR2+

HAMP+IFNG+DAB2+MCOLN1+ADORA3+MIF 12 83,33 93,55 10,2 80 100 9,52

TFRC+PEBP1+ VNN3+ S0CS1 +MIF+HM0X1 +DAB

2+HAMP+IFNG+CDHR2+ADORA3+MCOLN1 12 83,33 93,55 10,2 80 100 9,52

SLC5A12+MIF+CDHR2+TFRC+IFNG+ADORA3+

HAMP+VNN3+S0CS1+MC0LN1+PEBP1+HM0X1 12 77,78 96,77 10,2 80 100 9,52

TFRC+IFNG+CDHR2+ADORA3+PEBP1+VNN3+

MIF+HMOX1+MCOLN1+SOCS1+SLC5A12+DAB2

,+HAMP 13 77,78 96,77 10,2 80 100 9,52

TFRC+VNN3+HAMP+CDHR2+SLC5A12+HMOX1

+SOCS PEBP MIF 9 94,44 83,87 12,24 80 100 9,52

DAB2+ TFRC+ MIF+ CDHR2+PEBP1+ VNN3+ TTC3

+HM0X1+S0CS1 9 83,33 90,32 12,24 80 100 9,52 TFRC+PEBP1+MIF+CDHR2+VNN3+IFNG+MCOL

N1+S0CS1 8 88,89 83,87 14,29 80 100 9,52

TFRC+PEBP1+MIF+S0CS1+CDHR2 5 94,44 80,65 14,29 80 100 9,52

ADORA3+CDHR2+MIF+PEBP1+TAF15+TFRC 6 94,44 82,14 11 ,94 80 100 13,04

CDHR2+MIF+PEBP1+SLC5A12+SOCS1+TAF15+

TFRC 7 83,33 92,85 11 ,94 70 100 10,87

ADORA3+CDHR2+HAMP+MIF+PEBP1+SOCS1,

TAF15+TFRC 8 94,44 82,14 11 ,94 80 100 13,04

CDHR2+HAMP+IFNG+MCOLN1+MIF+PEBP1+SO

CS1+TFRC+VNN3 9 88,88 89,28 10,44 80 100 10,87

SN: sensitivity

SP: specificity

ER: error rate

PPV: positive predictive value

NPV: negative predictive value

One embodiment of present invention refers to use of at least one of the following genes or combinations thereof: TFRC, CDHR2, HMOXl, MIF, HAMP, IFNG, PEBPl, SLC5A12, ADORA3 and DAB2, in a method for the in vitro diagnosis and/or prognosis of the tolerant state of a patient subjected to a liver transplantation. In a preferred embodiment the method of the invention is carried out using a combination of at least one of the above cited genes with at least one of the following genes: LC5A12, VNN3, SOCSl, TTC3, RBM23, SH2D1B, NCR1, TFRC, TUBA4A, TAF15, TIPARP, MOX1, MCOLN1, EBP1, DHR2, and AB2. In a particularly preferred embodiment the method of the invention is carried using one of the following gene combinations: LC5A12, VNN3, TFRC, SOCSl, MIF, TTC3, RBM23, PEBPl, SH2D1B, NCR1, DAB2 and ADORA3; TFRC, PEBPl, MIF, CDHR2, HAMP, TUBA4A, TTC3, HMOXl, VNN3, NCR1, ADORA3, TAF15, IFNG, SOCSl and TIPARP; MOX1, CDHR2, MIF, PEBPl, TFRC, SLC5A12, SOCSl, HAMP, VNN3 and IFNG; TFRC, PEBPl, MIF, CDHR2, SLC5A12, HAMP, SOCSl, IFNG and HMOXl; TFRC, IFNG, CDHR2, ADORA3, HAMP, MIF, PEBPl, VNN3, SOCSl, HMOXl and DAB2; TFRC, DAB2, MIF, PEBPl, IFNG, HAMP, SLC5A12, SOCSl, VNN3, ADORA3, CDHR2, MCOLN1 and HMOXl; TFRC, IFNG, HMOXl, MCOLN1, MIF, HAMP, ADORA3, CDHR2, PEBPl and SOCSl; EBP1, TFRC, HMOXl, IFNG, MCOLN1, SOCSl, MIF, CDHR2, HAMP and ADORA3; TFRC, PEBPl, IFNG, CDHR2, ADORA3, VNN3, HMOXl, DAB2, SOCSl, MIF and HAMP; DHR2, ADORA3, IFNG, TFRC, VNN3, HMOXl, PEBPl, MIF, SLC5A12, HAMP, SOCSl and MCOLNl; LC5A12, TFRC, IFNG, MIF, DAB2, HMOXl, CDHR2, SOCSl, HAMP, PEBPl, VNN3, ADORA3 and MCOLNl; TFRC, SOCSl, HMOXl, PEBPl, VNN3, CDHR2, HAMP, IFNG, DAB2, MCOLNl, ADORA3 and MIF; TFRC, PEBPl, VNN3, SOCSl, MIF, HMOXl, DAB2, HAMP, IFNG, CDHR2, ADORA3 and MCOLNl; LC5A12, MIF, CDHR2, TFRC, IFNG, ADORA3, HAMP, VNN3, SOCSl, MCOLNl, PEBPl and HMOXl; TFRC, IFNG, CDHR2, ADORA3, PEBPl, VNN3, MIF, HMOXl, MCOLNl, SOCSl, SLC5A12, DAB2 and HAMP; TFRC, VNN3, HAMP, CDHR2, SLC5A12, HMOXl, SOCSl, PEBPl and MIF; AB2, TFRC, MIF, CDHR2, PEBPl, VNN3, TTC3, HMOXl and SOCSl; TFRC, PEBPl, MIF, CDHR2, VNN3, IFNG, MCOLNl and SOC SI; TFRC, PEBPl, MIF, SOCSl and CDHR2; ADORA3, CDHR2, MIF, PEBPl, TAF15 and TFRC; CDHR2, MIF , PEBPl, SLC5A12, SOCSl, TAF15 and TFRC; ADORA3, CDHR2, HAMP, MIF, PEBPl, SOCSl, TAF 15 and TFRC; CDHR2, HAMP, IFNG, MCOLNl, MIF, PEBPl, SOCSl, TFRC and VNN3.

Some clinical variables may influence the development of operational tolerance following liver transplantation in humans. In particular, the recipients who have been transplanted for a longer period of time or who are older have a higher likelihood of being capable of successfully discontinuing immunosuppressive medications. We conducted additional logistic regression analyses to exclude the potentially confounding effect of these 2 clinical variables on the gene expression measurements. Out of the genes depicted in Tables 3 or 4, the following genes were found to be statistically significant after excluding the effect of recipient age and time since transplantation (Table 7).

Table 7

Genes p-value

TFRC 0,0033

PEBPl 0,0125

HMOXl 0,0177

IFNG 0,0198

CDHR2 0,0234

DAB2 0,0572 In order to develop a genetic predictor independent from clinical variables, we employed logistic regression and the MiPP software on the genes shown in Table 7. The following model was found to be the best predictor in both the learning and the validation groups of patients (Table 8).

Table 8

Thus, an additional embodiment of the present invention refers to a method for the in vitro diagnosis and/or prognosis of the tolerant state of a patient subjected to a liver transplantation by measuring gene expression of the following combination of genes: TFRC, IFNG and CDHR2.

In a particular embodiment of the method according to the invention, said method may further comprise determining at least one additional parameter useful for the diagnosis and/or prognosis. Such "parameters useful for the diagnosis" are parameters that cannot be used alone for a diagnosis but that have been described as displaying significantly different values between tolerant subjects and subjects who clearly need immunosuppressive treatment and may thus also be used to refine and/or confirm the diagnosis according to the above described method according to the invention. Therefore, a further embodiment of the invention is a method such as described above and which further comprises the determination of the age of the patient and/or the post-transplantation time.

Another embodiment of the present invention refers to a kit, for performing the method of the invention for the in vitro diagnosis and/or prognosis of the tolerant state of a patient subjected to a liver transplantation, comprising (i) means for measuring the gene expression levels of the corresponding genes, and (ii) instructions for correlating said gene expression levels above or below the expression level of the same genes taken from a reference R A sample. Said reference samples can be a pool of R As obtained from healthy non-transplanted liver tissue, a reference RNA such as the commercially available Human Liver Total R A from Ambion, or an absolute reference consisting in a sample containing a previously quantified number of RNA molecules). In a preferred embodiment the means comprise a microarray or a gene chip which comprises nucleic acid probes, said nucleic acid probes comprising sequences that specifically hybridize to the transcripts of the corresponding set of genes, along with reagents for performing a microarray analysis. In another preferred embodiment of the invention the kit comprises oligonucleotide primers (i.e. HS02559818sl for gene TFRC or HsOI 125168ml for gene VNN3), for performing a quantitative reverse transcription polymerase chain reaction, said primers comprising sequences that specifically hybridize to the complementary DNA derived from the transcripts of the corresponding set of genes. Moreover the kit of the invention may comprise a solid support wherein nucleic acid probes which comprises sequences that specifically hybridize to the transcripts of the corresponding set of genes, are displayed thereon.

In another embodiment, the means comprises a microarray or a protein chip which comprises specific binding moieties such as monoclonal antibodies or fragments thereof.

In one embodiment the kit of present invention measures the expression of at least one of the following genes or combinations thereof: TFRC, CDHR2, HMOX1, MIF, HAMP, IFNG, PEBPl, SLC5A12, ADORA3 and DAB2, for the in vitro diagnosis and/or prognosis of the tolerant state of a patient subjected to a liver transplantation. In a preferred embodiment the kit of the invention measures the gene expression of a combination of at least one of the above cited genes with at least one of the following genes: LC5A12, VNN3, SOCS1, TTC3, RBM23, SH2D1B, NCR1, TFRC, TUBA4A, TAF15, TIPARP, MOX1, MCOLN1, EBP1, DHR2, and AB2. In a particularly preferred embodiment the kit of the invention measures gene expression of the following gene combinations: LC5A12, VNN3, TFRC, SOCS1, MIF, TTC3, RBM23, PEBPl, SH2D1B, NCR1, DAB2 and ADORA3; TFRC, PEBPl, MIF, CDHR2, HAMP, TUBA4A, TTC3, HMOX1, VNN3, NCR1, ADORA3, TAF15, IFNG, SOCS1 and TIPARP; MOX1, CDHR2, MIF, PEBPl, TFRC, SLC5A12, SOCSl, HAMP, VNN3 and IFNG; TFRC, PEBPl, MIF, CDHR2, SLC5A12, HAMP, SOCSl, IFNG and HMOXl; TFRC, IFNG, CDHR2, ADORA3, HAMP, MIF, PEBPl, VNN3, SOCSl, HMOXl and DAB2; TFRC, DAB2, MIF, PEBPl, IFNG, HAMP, SLC5A12, SOCSl, VNN3, ADORA3, CDHR2, MCOLNl and HMOXl; TFRC, IFNG, HMOXl, MCOLNl, MIF, HAMP, ADORA3, CDHR2, PEBPl and SOCSl; EBP1, TFRC, HMOXl, IFNG, MCOLNl, SOCSl, MIF, CDHR2, HAMP and ADORA3; TFRC, PEBPl, IFNG, CDHR2, ADORA3, VNN3, HMOXl, DAB2, SOCSl, MIF and HAMP; DHR2, ADORA3, IFNG, TFRC, VNN3, HMOXl, PEBPl, MIF, SLC5A12, HAMP, SOCSl and MCOLNl; LC5A12, TFRC, IFNG, MIF, DAB2, HMOXl, CDHR2, SOCSl, HAMP, PEBPl, VNN3, ADORA3 and MCOLNl; TFRC, SOCSl, HMOXl, PEBPl, VNN3, CDHR2, HAMP, IFNG, DAB2, MCOLNl, ADORA3 and MIF; TFRC, PEBPl, VNN3, SOCSl, MIF, HMOXl, DAB2, HAMP, IFNG, CDHR2, ADORA3 and MCOLNl; LC5A12, MIF, CDHR2, TFRC, IFNG, ADORA3, HAMP, VNN3, SOCSl, MCOLNl, PEBPl and HMOXl; TFRC, IFNG, CDHR2, ADORA3, PEBPl, VNN3, MIF, HMOXl, MCOLNl, SOCSl, SLC5A12, DAB2 and HAMP; TFRC, VNN3, HAMP, CDHR2, SLC5A12, HMOXl, SOCSl, PEBPl and MIF; AB2, TFRC, MIF, CDHR2, PEBPl, VNN3, TTC3, HMOXl and SOCSl; TFRC, PEBPl, MIF, CDHR2, VNN3, IFNG, MCOLNl and SOC Sl; TFRC, PEBPl, MIF, SOCSl and CDHR2 ; ADORA3, CDHR2, MIF, PEBPl, TAF15 and TFRC; CDHR2, MIF , PEBPl, SLC5A12, SOCSl, TAF15 and TFRC; ADORA3, CDHR2, HAMP, MIF, PEBPl, SOCSl, TAF15 and TFRC; CDHR2, HAMP, IFNG, MCOLNl, MIF, PEBPl, SOCSl, TFRC and VNN3.

One of the preferred embodiments of the present invention refers to a kit for the in vitro diagnosis and/or prognosis of the tolerant state of a patient subjected to a liver transplantation which measures gene expression of the following combination of genes: TFRC, IFNG and CDHR2.

Kits, according to present invention may further comprise reagents for performing a microarray analysis and/or solid supports wherein nucleic acid probes which comprises sequences that specifically hybridize to the transcripts of the corresponding set of genes, are displayed thereon. Another embodiment of the present invention refers to a kit for selecting or modifying an immunotherapy treatment protocol by assessing the tolerant state of the liver recipient by using the above disclosed method or kit. The last embodiment of the present invention refers to a method for adapting the immunosuppressive treatment of a liver grafted patient, said method comprising the use of above disclosed method and kits.

The state of the art comprises (Benitez C et al, Abstract #517, American Transplant Congress, San Diego, CA, May 3 - May 5, 2010) the measure in peripheral blood samples of the expression of a group of genes (KLRF1, PTGDR, NCALD, CD 160, IL2RB, PTCH1, ERBB2, KLRB1, NKG7, KLRD1, FEZ1, GNPTAB, SLAMF7, CLIC3, CX3CR1, WDR67, MAN1A1, CD9, FLJ14213, FEM1C, CD244, PSMD14, CTBP2, ZNF295, ZNF267, RGS3, PDE4B, ALG8, GEMIN7) different from that presented in the present invention, as a method to identify tolerant liver recipients. A comparative assay was carried out (see Example 10) in order to determine whether the genes which form part of the present invention have a higher discriminative power as compared with the previously disclosed genes in peripheral blood. It was concluded the measurement of the expression of the genes comprised in the present invention in liver tissue samples, appears as least as accurate than the measurement of the genes comprised in the state of the art in peripheral blood, in order to identify the liver recipients who can successfully leave the immunosuppressive medication because they are tolerant to the transplantation. Moreover, the present invention offers additional evidences supporting the role of iron metabolism in the acquisition of operational tolerance to liver allografts:

• The liver iron content (intra-hepatic iron stores), measured in a semiquantitative manner after Perls' staining employing either the Scheuer modified method or the total iron score method, is significantly higher in tolerant than in non-tolerant liver recipients (see Figure 1).

• Serum levels of hepcidin, the most important hormone in the regulation of systemic iron homeostasis, which in fact is encoded by HAMP, are significantly higher in tolerant than in non-tolerant liver recipients (see Figure 3 and Example 9). This is consistent with the observation that iron stores are higher in tolerant than in non-tolerant liver recipients, given that the physiological response in a situation of iron deficiency is to decrease the production of hepcidin.

Serum levels of ferritin, one of the most accurate markers of total body iron storage, are significantly higher in tolerant than in non-tolerant liver recipients (see Figure 4A and B and Example 11);

The liver tissue phospho-Stat3 protein levels are significantly higher in tolerant than in non-tolerant recipients (see Figure 5 and Example 12). This is consistent with the observation that hepcidin phosphorylates Stat3, and therefore that in the absence of hepcidin (for instance as a consequence of iron deficiency) the levels of phospho-Stat3 are decreased. Taken together, these results indicate that the regulation of iron metabolism at the systemic and intrahepatic level plays a role in the control of allo-immune responses in liver transplantation by the intra-graft regulation of the activation of the transcription factor Stat3. Similarly, the quantification of intra-hepatic iron levels, serum hepcidin and serum ferritin should also be considered as functional equivalents of the genes described in the present invention.

Therefore, another embodiment of the present invention refers to a method for the in vitro diagnosis and/or prognosis of the tolerant state of a patient subjected to a liver transplantation that comprises: obtaining a biological sample from the liver allograft of the patient under investigation, measuring in said sample the level of intra-hepatic iron stores, and assessing the tolerance or non-tolerance of the patient under investigation to a liver transplantation by comparing the level of his intra-hepatic iron stores with level of intra-hepatic iron stores taken from a reference sample, knowing that, as cited above, the level of intra-hepatic iron stores is significantly higher in tolerant liver transplant recipients as compared with non-tolerant liver transplant recipients. The assessment of the level of intra-hepatic iron stores can be carried out by any means known in the state of the art, for example (non-exhaustive list): by direct staining of liver biopsy slides with iron-specific stains (e.g. Perls Prussian blue), by quantification of iron from liver tissue biopsies employing atomic absorption spectrophotometry, or by magnetic resonance imaging of the whole liver. In this case, the reference value is a threshold pre-defined on the basis of the differences observed between tolerant and non-tolerant liver transplant patients as shown in Figure 1.

Another embodiment of the present invention refers to a method for the in vitro diagnosis and/or prognosis of the tolerant state of a patient subjected to a liver transplantation that comprises: obtaining a biological sample from the serum of the patient under investigation, measuring in said sample the level of the protein ferritin, and assessing the tolerance or non-tolerance of the patient under investigation to a liver transplantation by comparing his level of ferritin with the level of the same protein taken from a reference sample, knowing that, as cited above, the serum level of ferritin are significantly higher in tolerant liver recipients than in non-tolerant liver recipients. The assessment of the level of the protein ferritin can be carried out by any means known in the state of the art, for example (non-exhaustive list): ELISA and radioimmunoassay. In this case, the reference value is a threshold pre-defined on the basis of the differences observed between tolerant and non-tolerant liver transplant patients as shown in Figure 4.

Another embodiment of the present invention refers to a method for the in vitro diagnosis and/or prognosis of the tolerant state of a patient subjected to a liver transplantation that comprises: obtaining a biological tissue sample from the liver allograft of the patient under investigation, measuring in said sample the protein level of phospho-Stat3, and assessing the tolerance or non-tolerance of the patient under investigation to a liver transplantation by comparing his protein level of phospho- Stat3, with the protein level of the same protein taken from a reference sample, knowing that, as cited above, the liver tissue level of phospho-Stat3 is significantly higher in tolerant liver recipients than in non-tolerant liver recipients. The assessment of the level of liver tissue phospho-Stat3 can be carried out by any means known in the state of the art, for example (non-exhaustive list): immunohistochemistry, immunofluorencence and Western-Blot. In this case, the reference value is a pre- defined threshold or a reference sample, knowing that, as cited above, the level of intra-hepatic phospho-Stat3 is significantly higher in tolerant liver transplant recipients as compared with non-tolerant liver transplant recipients as shown in Figure 5.

Another embodiment of the present invention refers to a method for the in vitro diagnosis and/or prognosis of the tolerant state of a patient subjected to a liver transplantation that comprises: obtaining a biological sample from the patient under investigation, measuring in said sample the protein level of hepcidin, and assessing the tolerance or non-tolerance of the patient under investigation to a liver transplantation by comparing his protein level of hepcidin, with the protein level of the same protein taken from a reference sample, knowing that, as cited above, the serum level of hepcidin is significantly higher in tolerant liver recipients than in non-tolerant liver recipients. The assessment of the level of hepcidin can be carried out by any means known in the state of the art, for example (non-exhaustive list): mass spectrometry and ELISA. In this case, the reference sample consists in a hepcidin analogue with known concentration, knowing that the serum level of hepcidin is significantly higher in tolerant liver recipients than in non-tolerant liver recipients as shown in Figure 3. Brief description of the figures

Figure 1.

(A) The liver iron content (measured in a semi-quantitative manner after Perls' staining employing either the Scheuer modified method or the total iron score method) is significantly higher in the livers of tolerant patients who may successfully leave the immunosuppressive therapy (TOL) than in those non-tolerant patients where this is not possible (Non-TOL).

(B) The presence or absence of intra-hepatic iron staining as assessed by the Scheuer method is independent from the time elapsed since transplantation (left panel) and from the type of immunosuppressive drugs employed at baseline (right panel), and can be employed to discriminate between tolerant (black bars) and non-tolerant (white bars) patients. Figure 2. This figure shows the influence of individual gene expression measurements on the levels of intra-hepatic iron (measured by the modified method of Scheuer). The highest bars correspond to the most influential genes {HAMP, TFRC, CDHR2). The reference line is the threshold of statistical significance as determined by Goeman's Globaltest.

Figure 3. This figure shows that the serum levels of hepcidin (the peptide encoded by the gene HAMP) are significantly increased in tolerant recipients as compared with non-tolerant liver recipients.

Figure 4.

(A) This figure shows that the serum levels of ferritin (a marker of systemic body iron stores) are significantly higher in tolerant than in non-tolerant liver recipients.

(B) This figure shows the distribution of ferritin serum levels among tolerant (TOL) and non-tolerant (Non-TOL) recipients: ferritin <12 ng/mL (depleted iron stores), ferritin 12-30 ng/mL (reduced iron stores), ferritin >30 ng/mL (replete iron stores). As shown, none of the tolerant patients exhibit depleted iron stores, and only a minority exhibit reduced iron stores. In contrast, approximately 30% of non-tolerant patients exhibit abnormally low systemic iron stores. Figure 5. This figure shows that the area of liver tissue sections that stains positive for phosphorylated STAT3 is significantly greater in tolerant (TOL) than in non-tolerant (Non-TOL) liver recipients.

EXAMPLES

Example 1. Patient population and study design.

Blood and liver biopsy specimens were collected from a group of liver transplant recipients enrolled in a prospective European Commission supported multi-center clinical trial of immunosuppressive drug withdrawal in liver transplantation (Title: Search for the immunological Signature of Operational Tolerance in Liver Transplantation; clinicaltrials.gov identification NCT00647283). Inclusion criteria were the following: 1) >3 years after transplantation; 2) stable liver function and no episodes of rejection during the 12 months prior to inclusion; 3) no history of autoimmune liver disease; 4) pre-inclusion liver biopsy without significant abnormalities (no signs of acute or chronic rejection according to Banff criteria; absence of portal inflammation in >50% of portal tracts; absence of central perivenulitits in >50% of central veins; absence of bridging fibrosis or cirrhosis). In enrolled recipients immunosuppressive drugs were gradually weaned until complete discontinuation over a 6-9 month period and then followed-up for 12 additional months. Patients were considered as tolerant if no rejection episodes occurred during the entire duration of the study and no significant histological changes were noted in a liver biopsy obtained at the end of the 12-month follow-up period. Patients undergoing acute rejection during the study were considered as non-tolerant. Out of the 102 recipients enrolled in the trial 79 (33 tolerant and 46 non-tolerant) were included in the current study. Blood and liver biopsy specimens available for the study were obtained before immunosuppressive drugs were discontinued from both tolerant (TOL, n=33) and non-tolerant (Non-TOL, n=46) recipients, at the time of rejection from non-tolerant recipients (n=14), and at the end of the study in tolerant recipients (n=4). In addition, liver tissue samples were also obtained from the following patient groups: a) liver transplant recipients with chronic hepatitis due to recurrent hepatitis C virus infection (HEPC, n=12); b) liver transplant recipients with typical acute cellular rejection taking place during the immediate post-transplant period (REJ, n=9); c) liver transplant recipients under maintenance immunosuppression with normal liver function and normal liver histology 1 year after transplantation (CONT-Tx, n=8); and d) non-transplanted patients undergoing surgery for colorectal liver metastases (CONT, n=10). Participating recipients were enrolled from Hospital Clinic Barcelona (Spain), University Tor Vergata Rome (Italy) and University Hospitals Leuven (Belgium). The study was approved by the institutional review boards of the three participating institutions and written informed consent was obtained from all study patients. Clinical and demographic characteristics of patients included in the study are summarized in Table 1 and an outline of the study design is depicted in Figure 1. A detailed description of the patient population and clinical outcomes immunosuppression withdrawal clinical trial will be reported elsewhere.

Example 2. Liver biopsy specimens and histological assessment.

Liver biopsies were performed percutaneously under local anaesthesia. A 2-3 mm portion of the needle biopsy liver cylinder was immediately preserved in R Alater reagent (Ambion, Austin, USA), kept at 4°C for 24h and then cryopreserved in liquid nitrogen after removal of the RNAlater reagent. The remaining cylinder was fromalin- fixed and paraffin-embedded. In CONT patients surgical liver biopsies of non- tumoral livers were obtained and processed as previously described. For histological assessment 3 μιη thick slides were stained using hematoxylin-eosin and Masson's trichrome for connective tissue analysis. The histological examinations were performed by the same pathologist who was blinded to all clinical and biological data. The following histopathological items were evaluated and scored semiquantitatively: 1) number of complete portal tracts; 2) number of central veins; 3) overall parenchymal architecture; 4) lobular inflammation; 5) central vein perivenulitis; 6) portal tract inflammation; 7) bile duct lesions; 8) bile duct loss; 9) presence of portal vein branches; 9) portal fibrosis; 10) perisinusoidal fibrosis. Example 3. RNA extraction and processing.

For total RNA extraction cryopreserved liver tissue samples were homogenized in TRIzol reagent (Invitrogen, San Diego, CA, USA) using pestle and nuclease-free 1 ,5ml reaction tubes (Ambion). Total RNA was then extracted following the manufacturers guidelines and quality was assessed with the Agilent 2100 Bioanalyzer (Agilent Technologies, Santa Clara, USA).

Example 4. Illumina microarray experiments.

One hundred and five liver RNA samples (20 TOL, 32 Non-TOL, 14 Non-TOL-Rej, 12 HEPC, 9 REJ, 8 CONT-Tx and 10 CONT; all of them from Hospital Clinic Barcelona) were processed into cRNA and hybridized onto Illumina HumanHT-12 Expression BeadChips containing 48,771 probes corresponding to 25,000 annotated genes (Illumina, Inc. San Diego, CA, USA). Expression data was computed using BeadStudio data analysis software (Illumina, Inc.) and subsequently processed employing quantiles normalisation using the Lumi bioconductor package [6]. Next, we conducted a conservative probe-filltering step excluding those probes with a coefficient of variation of 5%, which resulted in the selection of a total of 33,062 probes out of the original set of 48,771.

Example 5. Affymetrix microarray experiments.

In a selected group of 10 TOL and 10 Non-TOL recipients microarray experiments were replicated onto Affymetrix Human Genome U133 Plus 2.0 arrays covering 47,000 annotated genes by 54,675 probes (Affymetrix, Inc, Santa Clara, CA, USA) and which comprises commercially available nucleic acid probes. Additional Affymetrix experiments were conducted in employing RNA extracted from 4 TOL- Post liver samples (4 of the 10 TOL recipients from whom liver biopsy tissue obtained 12 months after complete drug withdrawal was also available). Gene expression data were normalized using the guanidine-cytosine content-adjusted robust multiarray algorithm, which computes expression values from probe-intensities incorporating probe-sequence information. Thereupon, we employed a conservative probe- filtering step excluding probes not reaching a log ₂ expression value of 5 in at least one sample, which resulted in the selection of a total of 18,768 probes out of the original set of 54,675.

Example 6. Microarray gene expression data analysis.

To identify genes differentially expressed between the different microarray study groups we employed Significant Analysis of Microarray (SAM). SAM uses modified t test statistics for each gene of a dataset and a fudge factor to compute the t value, thereby controlling for unrealistically low standard deviations for each gene. Furthermore SAM allows control of the false discovery rate (FDR) by selecting a threshold for the difference between the actual test result and the result obtained from repeated permutations of the tested groups. For the current study we employed SAM selection using FDR<10% and 1000 permutations. To graphically represent global gene expression differences between the different study groups, the entire filtered probe list was used to perform a correspondence analysis as implemented in the between-group-analysis (BGA) function included in the made4 package (Culhane AC, et al. Bioinformatics 2005). This method is capable of visualizing high-dimensional data such as multiple gene expression measurements in a 2D graph in which the areas delimited by the ellipses represent 95% of the estimated binormal distribution of the sample scores in the first and second axes. To determine whether the group of genes of interest was significantly associated with clinical outcome (tolerance versus rej ection) , clinical variables (donor and recipient age, gender, type of immunosuppressive therapy, time since transplantation) and histologic features (iron content), we employed the Globaltest software (Goeman, et al. Bioinformatics 2004). A syntax within this software was also employed to correct the association found between gene expression and clinical outcome for the possible confounding effects of nuisance clinical covariates found to be statistically different between tolerant and non-tolerant recipe

nts.

Example 7. Quantitative real-time PCR (qPCR) experiments.

To validate the microarray expression results the expression patterns of a group of 104 target genes and 3 housekeeping genes (Supplementary Table 1) were measured employing the ABI 7900 Sequence Detection System and TaqMan LDA microfluidic plates (Applied Biosystems, Carlsbad, USA), which comprises commercially available oligonucleotide primers, on a subgroup of 48 recipients (18 TOL and 31 Non-TOL; all of them from Hospital Clinic Barcelona). In addition, qPCR experiments were performed in an independent group of 10 TOL and 1 1 Non-TOL recipients provided by University Tor Vergata Rome and University Hospitals Leuven and from whom microarray data were not available. Target genes were selected based on: 1) Illumina and Affymetrix microarray experiment results; 2) blood transcriptional biomarkers previously described by our group as being associated with liver operational tolerance (M.Martinez-Llordella et al. J Clin Invest 2008); and 3) prominent immunoregulatory genes described in the literature. DNA was removed from total RNA preparations using Turbo DNA- free DNAse treatment (Ambion), and RNA was then reverse transcribed into cDNA using the HighCapacity cDNA Reverse Transcription Kit (Applied Biosystems). To quantify transcript levels target gene Ct values were normalized to the housekeeping genes to generate ACt values. The results were then computed as relative expression between cDNA of the target samples and a calibrated sample according to the AACt method. The following three samples were employed as calibrators: 1) pooled R A from the 8 CONT-Tx samples; 2) pooled R A from the 10 CONT samples; and 3) commercially available liver RNA {Human Liver Total RNA, Ambion).

Example 8. Identification and validation of gene classifiers.

To develop biopsy-based qPCR gene expression classifiers to predict the success of immunosuppression withdrawal we conducted an exhaustive search for predictive models employing the linear discriminant analysis and logistic regression algorithms implemented in the misclassification penalized posterior (MiPP) software. MiPP is based on a stepwise incremental classification modelling for discovery of the most parsimonious diagnosis models and employs a double cross-validation strategy. First, to obtain the optimal models while avoiding the pitfalls of a large screening search, we conducted a 10-fold cross validation step on the training set of 18 TOL and 31 Non-TOL liver recipients from Hospital Clinic Barcelona. Next, random splitting cross-validation of the diagnosis models was conducted on the whole data set (which included the 56 samples from Barcelona and the 21 samples from Rome and Leuven) by repeatedly partitioning it into training set (2/3) and independent test set (1/3) for external model validation. For each model identified in the training set the optimal probability cut-off of tolerance was computed through a ROC analysis. The use of a large number of random splits of test and training sets allowed us to obtain confidence bounds on the accuracy of the diagnosis. On the basis of these confidence bounds, the diagnosis performance and mean misclassification error rates were obtained for each of the candidate classifiers. To demonstrate that the performance of the models was not center-dependent, we then computed SN, SP, NPV, PPV and overall error rates for the samples collected from Barcelona and those obtained from Rome and Leuven. Example 9. Serum hepcidin measurements.

Serum samples were obtained from the 64 enrolled liver recipients at baseline using BD vacutainer SST II (BD Bioscience, Franklin Lakes, USA) and stored at -80°C. Quantitative serum hepcidin measurements were conducted by a combination of weak cation exchange chromatography and time-of- flight mass spectrometry (TOF MS). For quantification a hepcidin analogue (synthetic hepcidin-24; Peptide International Inc.) was imployed as internal standard. Peptide spectra were generated on a Microflex LT matrix-enhanced laser desorption/ionisation TOF MS platform (Bruker Daltonics). Serum hepcidin-25 concentrations were expressed as nmol/L. The lower limit of detection of this method was 0.5 nM; average coefficients of variation were 2.7% (intra-run) and 6.5% (inter-run). The median reference level of serum hepcidin- 25 is 4.2 nM, range 0.5- 13.9 nM.

As shown in Figure 3 the serum levels of hepcidin (the peptide encoded by the gene HAMP) are significantly increased in tolerant recipients as compared with non- tolerant liver recipients and, therefore, the level of hepcidin is a valuable marker for diagnosis and/or prognosis of the tolerant state of a patient to be submitted to liver transplantation. Example 10. Comparative assay between the method for diagnosis of tolerance in liver transplantation comprised in the state of the art and the method of the invention.

The method of the state of the art comprises the measure in peripheral blood samples of the expression of a group of genes (KLRF1, PTGDR, NCALD, CD 160, IL2RB, PTCH1, ERBB2, KLRB1, NKG7, KLRD1, FEZ1, GNPTAB, SLAMF7, CLIC3, CX3CR1, WDR67, MAN1A1, CD9, FLJ14213, FEM1C, CD244, PSMD14, CTBP2, ZNF295, ZNF267, RGS3, PDE4B, ALG8, GEMIN7) different from that presented in the present invention. Moreover the present example shows that the method of the invention has a higher discriminative power.

The method of the state of the art gives rise to the following results:

A) Combination oiFEMIC and IL8: SN=63%; SP=80.77%; ER= 7.08%; PPV=77.68%; NPV=72.41%

B) Combination of KLRF1 and SLAMF7

SN= 27.7%, SP 92.31%, ER= 37.5%, PPV=73.68%, NPV=72.41%

C) Combination of KLRF1 and IL2RB

SN= 50%, SP= 84.62%, ER=31.25%, PPV= 73.33%, NPV=66.67%

SN: sensitivity

SP: specificity

ER: error rate

PPV: positive predictive value

NPV: negative predictive value

However, the method of the present invention based on the measure of the genes comprised in Tables 3, 4 and 6 in liver tissue of the same 48 patients (the three best performing models were selected for this comparative assay) led to the following results:

A) Combination of TFRC, CDHR2, HMOX1, MIF, HAMP, IFNG, PEBP1, SLC5A12, ADORA3:

SN=77.27%, SP=96.15%, ER=12.5%, PPV=94.44%, NPV=83.33%

B) Combination of TFRC, PEBP1, MIF, ADORA3

SN=90.91%, SP=84,62%, ER=12.5%, PPV=83.33%, NPV=91.67%

C) Combination of TFRC, IFNG, HAMP, CDHR2

SN=77.27%, SP=88.46%, ER=16.67%, PPV=85%, NPV=82.14%

SN: sensitivity

SP: specificity

ER: error rate

PPV: positive predictive value

NPV: negative predictive value Therefore, it was concluded that the measurement of the expression of the genes comprised in the present invention in liver tissue samples, appears at least as accurate, than the measurement of the genes comprised in the state of the art in peripheral blood, in order to identify the liver recipients who can successfully leave the immunosuppressive medication because they are tolerant to the transplantation.

Example 11. Serum ferritin measurements.

Serum samples were obtained from the 64 enrolled liver recipients at baseline using BD vacutainer SST II (BD Bioscience, Franklin Lakes, USA) and stored at -80°C. Serum ferritin measurements were conducted by an automated ELISA method. Ferritin serum levels correlated with serum hepcidin. Thus, ferritin serum levels were significantly higher in tolerant (TOL) than in non-tolerant (Non-TOL) recipients (Figure 4A). Absent iron stores (serum ferritin <12 ng/mL, a highly specific indicator of iron deficiency) were exclusively observed among Non-TOL recipients (Figure 4B). The association between either hepcidin or ferritin and tolerance was not confounded by recipient age, time from transplantation or baseline immunosuppressive therapy, as demonstrated by their independent predictive value in a logistic regression multivariable analysis.

Since, as shown in Figure 4, the serum levels of ferritin are significantly increased in tolerant recipients as compared with non-tolerant liver recipients, the level of ferritin is a valuable marker for diagnosis and/or prognosis of the tolerant state of a patient to be submitted to liver transplantation.

Example 12. Phospho-Stat3 immunostaining. Given the capacity of hepcidin via phosphorylation of Janus kinase 2 (Jak2) to activate the transcription factor signal transducer and activator of transcription 3 (Stat3) (De Domenico et al. J Clin Invest 2010), we quantified phophorylated Stat3 in a subset of liver biopsies from 12 tolerant (TOL) and 13 non-tolerant (Non-TOL) patients. Paraffin-embedded liver biopsy sections were deparaffinized and antigen retrieval was performed by boiling for 15 min in lmM EDTA (pH 8.0). Subsequently, sections were treated with background reducing reagents (DAKO, Glostrup, Denmark) and stained with phospho-STAT3 (Tyr 705) rabbit monoclonal antibody (Cell Signalling Technology, Danvers, MA, USA) following the providers instructions. Immunostainings were developed employing DAB substrate (DAKO) and counterstained using Hematoxilin. Immages were assessed using a Nikon Eclipse E600 microscope and analySIS software and were analyzed using Image J Software (NIH, Bethesda, USA) by threshold controlled reduction of blue channel, followed by binarization and measurement of residual stained pixels. Strong phospho-Stat3 staining was almost exclusively noted in hepatocyte nuclei, which clustered in randomly distributed foci along the biopsy sections. In contrast, staining was faint and much less frequent in the nuclei of infiltrating mononuclear leukocytes, macrophages and endothelial cells.

In comparison to non-tolerant patients (Non-TOL) samples, tolerant patients (TOL) liver biopsies exhibited a significantly increased hepatocyte phospho-Stat3 staining (Figure 5). So, the level of phospho-Stat3 is a valuable marker for diagnosis and/or prognosis of the tolerant state of a patient to be submitted to liver transplantation.

Example 13. Assessment of liver iron content.

To estimate the magnitude of intrahepatic iron stores we stained 3 μιη thick liver slides with Perls staining. Iron content was semiquantitatively assessed employing both Scheuer's modified and Total Iron Score (TIS) scoring systems as described (Deugnier et al. Hepatology 1993; Scheuer et al. J Pathol Bacteriol 1962). In addition, hepatocyte and mesenchymal (endotelial and Kupffer cell) iron staining was separately scored. Mild peri-portal hepatocyte iron deposition was noted in the majority of liver samples collected from tolerant (TOL) patients. In contrast, no stainable iron was observed in the liver biopsies from non-tolerant (Non-TOL) recipients (Figure 1A). These differences were exclusively due to hepatocyte, rather than mesenchymal iron accumulation. The differences in iron stores between TOL and Non-TOL could be used to discriminate between the two groups of patients regardless of the time elapsed since transplantation or the type of immunosuppressive drugs administered (Figure IB).

Since, as shown in Figure 1, the liver iron content is significantly higher in the livers of tolerant patients who may successfully leave the immunosuppressive therapy (TOL) than in those non-tolerant patients where this is not possible (Non-TOL), the level of iron may be used as a valuable marker for the diagnosis and/or prognosis of the tolerant state of a patient to be submitted to liver transplantation.

REFERENCES

I . Lerut, J., and Sanchez-Fueyo, A. 2006. An appraisal of tolerance in liver transplantation. Am J Transplant 6: 1774-1780.

2. N. Najafian et al, "How can we measure immunologic tolerance in humans?" J.

Am. Soc. Nephrol. 2006, vol. 17, pp. 2652-63.

3. Newell et al, "Tolerante assays: measuring the unknown", Transplantation 2006, vol. 81, pp. 1503-9.

4. J. Cai et al, "Minor H antigen HA-l-specific regulator and effector CD8+ T cells, and HA-1 micro chimer ism, in allograft tolerance", J. Exp. Med. 2004, vol. 199, pp. 1017-23.

5. E. Jankowska-Gan E et al, "Human liver allograft acceptance and the "tolerance assay", Hum. Immunol. 2002, vol. 63, pp. 862-70.

6. P. Sagoo et al, "Development of a cross-platform biomarker signature to detect renal transplant tolerance in humans", J.Clin. Invest. 2010, vol. 120, pp. 1848-61.

7. S. Brouard et al, "Operationally tolerant and minimally immunosuppressed kidney recipients display strongly altered blood T-cell clonal regulation", Am. J. Transplant. 2005, vol. 5, pp. 330-40.

8. G.V. Mazariegos et al, "Dendritic cell subset ratio in peripheral blood correlates with successful withdrawal of immunosuppression in liver transplant patients",

Am. J. Transplant. 2003, vol. 3, pp. 689-96.

9. Y. Li et al, "Analyses of peripheral blood mononuclear cells in operational tolerance after pediatric living donor liver transplantation", Am. J. Transplant. 2004, vol. 4, pp. 2118-25.

10. M. Martinez-Llordella et al. Multiparameter of immune profiling of operational tolerance in liver transplantation. Am. J. Transplant. 2007, vol. 7, pp.309- 19.

I I . I.Puig-Pey et al. Characterization of gammadelta T cell subsets in organ transplantation. Transplant. Int. 2010. Epub.

12. S. Brouard et al. Identification of a peripheral blood transcriptional biomarker panel associated with operational renal allograft tolerance. Proc. Natl. Acad. Sci. U

S A. 2007, vol. 104, pp. 15448-53. K.A. Newell et al. Identification of a B cell signature associated with renal transplant tolerance in humans. J. Clin. Invest. 2010, vol. 120, pp. 1836-47.

Y.M. Deugnier et al. Differentiation between heterozygotes and homozygotes in genetic hemochromatosis by means of a histological hepatic iron index: a study of 192 cases. Hepatology 1993, vol. 17, pp. 30-4.

P.J. Scheuer et al. Hepatic pathology in relatives of patients with haemochromatosis. J. Pathol. Bacteriol. 1962, vol. 84, pp. 53-64.

A.C. Culhane et al. MADE4: an R package for multivariate analysis of gene expression data. Bio informatics. 2005, vol. 21, pp. 2789-90.

J.J.Goeman et al. A global test for groups of genes: testing association with a clinical outcome. Bioinformatics. 2004, vol. 20, pp. 93-9.

De Domenico, I., et al. Hepcidin mediates transcriptional changes that modulate acute cytokine-induced inflammatory responses in mice. The Journal of clinical investigation 120, 2395-2405 (2010).