Genetic variants in Major Histocompatibility Complex-linked genes Associate with Pediatric Liver Transplant Rejection

doi:10.1053/j.gastro.2008.05.080

Gastroenterology. Author manuscript; available in PMC 2010 October 18.

Published in final edited form as:

Gastroenterology. 2008 September; 135(3): 830–84010.

Published online 2008 June 3. doi: 10.1053/j.gastro.2008.05.080

PMCID: PMC2956436

NIHMSID: NIHMS69061

Genetic variants in Major Histocompatibility Complex-linked genes Associate with Pediatric Liver Transplant Rejection

Rakesh Sindhi, MD, Brandon W. Higgs, PhD, Daniel E. Weeks, PhD, Chethan AshokKumar, PhD, Ronald Jaffe, MB.BCh, Cecilia Kim, MS, Patrick Wilson, BS, Nydia Chien, MSN, Joseph Glessner, BS, Anjan Talukdar, MD, George Mazariegos, MD, M. Michael Barmada, PhD, Edward Frackleton, BA, Nancy Petro, BS, Andrew Eckert, MS, Hakon Hakonarson, MD, PhD, and Robert Ferrell, PhD

Hillman Center for Pediatric Transplantation, Children’s Hospital of Pittsburgh, Pittsburgh, Center for Applied Genomics, Children’s Hospital of Philadelphia, Philadelphia, and the Department of Human Genetics and Biostatistics, Graduate School of Public Health, University of Pittsburgh, Pittsburgh.

Address correspondence to: Rakesh Sindhi, MD, FACS. Hillman Center for Pediatric Transplantation Children’s Hospital of Pittsburgh 3705 Fifth Avenue Pittsburgh, PA 15213. Phone/Fax: (412-692-6110/6116 ; Email: Rakesh.Sindhi/at/chp.edu

The publisher's final edited version of this article is available at Gastroenterology

See other articles in PMC that cite the published article.

Publisher's Disclaimer

Abstract

Background/Aims

Limited access to large samples and independent replication cohorts precludes genome-wide association (GWA) studies of rare but complex traits. To localize candidate genes with family-based GWA, a novel exploratory analysis was first tested on 1,774 major histocompatibility complex single nucleotide polymorphisms (SNPs) in 240 DNA samples from 80 children with primary liver transplantation (LTx), and their biological parents.

Methods/Results

Initially, 57 SNPs with large differences (p<0.05) in minor allele frequencies were selected, when parents of children with early rejection (Rejectors) were compared with parents of Non-Rejectors. In hypothesis-testing of selected SNPs, the gamete competition statistic identified the minor allele G (ancestral allele T) of the SNP rs9296068, near HLA-DOA, as being significantly different (p=0.018) in parent-to-child transmission between outcome groups. Subsequent simple association testing confirmed over- and under-transmission of rs9296068 based on 1) the most significant differences between outcome groups, of 1,774 SNPs tested (p=0.002), and 2) allele (G) frequencies that were greater among Rejectors (51.4 vs. 36.8%, p=0.015), and lower among Non-Rejectors (26.8 vs. 36.8%, p=0.074), compared with 400 normal control Caucasian children. In early functional validation, a) Rejectors demonstrated significant repression of the first HLA-DOA exon closest to rs9296068, and b) Rejectors with the risk allele showed 3-fold greater intragraft content of B-lymphocytes, whose antigen-presenting function is inhibited selectively by HLA-DOA, compared with Rejectors without the allele.

Conclusions

The minor allele of the SNP rs9296068 is significantly associated with LTx rejection, and with enhanced B-lymphocyte participation in rejection, likely due to a dysfunctional HLA-DOA gene product.

Introduction

In children, liver transplantation (LTx) is usually performed for multiple congenital liver diseases, and results in highly variable outcomes (1). Acute cellular rejection (ACR) occurs in 50% and post-LTx lymphoma-like malignancy affects 2-10% (2). Genetic variants, exemplified most commonly by single nucleotide polymorphisms (SNPs) are a significant basis for individual variation (3). However, the majority of the >10 million catalogued SNPs do not alter gene function (4). Also, promising associations involving discrete SNPs are rarely reproduced (5-7). Population stratification, or over-representation of one ethnicity within an outcome group may allow a SNP representing this ethnicity to be viewed as the outcome-specific SNP (8). Family-based association studies can minimize stratification (9). In this design, parental genotype serves as the control, and transmission frequency of a SNP from heterozygous parents to affected offspring is compared with the expected 50% transmission frequency. Significant deviations reflect transmission disequilibrium, and are more likely to represent disease-specific variation than variations due to ethnicity. The known location of such a SNP is then used to identify a potential candidate gene, and a causal variant in proximity. Increasingly, genome-wide association (GWA) studies are being conducted, recognizing that most disease traits are complex, and likely originate from an interaction between multiple causal variants within multiple genes, and the environment.

Several limitations preclude application of GWA in transplantation, especially among pediatric LTx recipients. First, simultaneous testing of large numbers of SNPs necessitates stringent multiple-testing correction. Under these conditions, statistical power requirements mandate a large subject sample, or a two-step study in which candidate SNPs identified in a screening sample are replicated in a second step. Accruing a homogeneous validation cohort can take several years, because approximately 550 pediatric LTx are performed annually in the United States, and are distributed among >50 transplant centers (10). Secondly, the numbers of candidate SNPs in an association study may not exceed false-discovery thresholds. From such a limited list of candidates, any true-positive association may at best account for large effects.

We report for the first time, a novel, multi-step approach to candidate gene localization, which incorporates preliminary functional validation in the same test cohort. First, a two-tier, family-based association method identifies the candidate gene or locus. Next, transmission distortion in that locus is confirmed by allele frequency comparisons with a large control group. Because SNPs within non-coding regions can influence gene expression and splicing at distances of up to 100 kb, differential regulation and alternative splice variants of candidate genes is evaluated with probes specific to exons of whole gene transcripts in the first of two preliminary functional validation tests (11, 12). In the second, the candidate is characterized in diseased tissue with immunohistochemistry. Because our sample size is small, and consists of 80 case-parent-parent trios, our association method is based on the following biologic assumptions: A) If a SNP is strongly associated with an inherited trait, it should demonstrate differences in parental allele frequencies (PAF) between the two groups, B) In family-based association testing, a genetic variant that is associated with rejection/non-rejection should show over-transmission in one outcome group and under-transmission in the other in the gamete competition test (GC). The GC statistic evaluates transmission disequilibrium using full pedigree data, similar to the well-known Transmission Disequilibrium Test (TDT) (8), but is based on a likelihood ratio test (13). The GC statistic also handles missing data and allows haplotype-based analysis. C) In a confirmatory step, potential candidates should also demonstrate i) significant differences in allele frequency when the two outcome groups are compared with each other, and ii) evidence of overtransmission in one group and undertransmission in the other when compared individually, with a large cohort of 400 normal Caucasian children.

We asked whether rejection/non-rejection outcomes are associated with genetic variants in the major histocompatibility complex (MHC) region. The 1,774 MHC-SNPs evaluated here are a subset of 550,000 genome-wide SNPs recently characterized in our test population. MHC-SNPs have been analyzed separately at first, to explore the utility of our statistical method, which is based in part on screening/testing approaches for quantitative traits proposed by others, by applying it to a candidate region with known impact in organ transplantation (14-17). The results show that the minor allele G at the SNP locus rs9296068, which is significantly associated with rejection outcomes, localizes to the 5′ UTR (untranslated) region of the HLA-DOA gene. Differential regulation of HLA-DOA, which is selectively expressed in B-lymphocytes, is suggested by associated repression of the first HLA-DOA exon, and significantly higher B-lymphocyte content in allograft biopsies from Rejectors (18, 19).

Methods

Patient Population

All studies were performed with approval from the University of Pittsburgh’s Institutional Review Board. DNA was extracted from 3 ml whole blood samples from 80 primary pediatric LTx recipients ages 0-22 years, and their biological parents (240 DNA samples). All children received previously described Tacrolimus monotherapy after steroid-free induction with rabbit, anti-human thymocyte globulin (rATG, Genzyme, Cambridge, MA) (20). Children who experienced biopsy-proven ACR within the first 60 days after LTx were termed Rejectors. Normal controls consisted of 400 disease-free Caucasian children with no prior or ongoing use of imunosuppressants, or transplantation. In addition, all 400 samples were selected for the absence of large copy number variants that could suggest a disease association. This normal control population was recruited at the Center for Applied Genomics at the Children’s Hospital of Philadelphia (CAG-CHOP).

Genotyping

was performed with 550,000 genome-wide SNP loci with the HumHap550k SNP bead array (Illumina, San Diego, CA.). Analysis from this experiment was restricted to 1,813 SNPs covering 168 MHC genes on chromosome 6p, of which 1,774 passed our quality control criteria.

DNA extracted by the Gentra Purigene system (Minneapolis, MN) was activated, and SNPs were characterized using manufacturer’s protocol. The Illumina Beadstation GX software was used to extract genotype calls (Illumina, San Diego, CA.) (21).

Statistical Methods

A total of 1,813 SNPs from the MHC region of the HumHap550k SNP array (Illumina, San Diego, CA.) were identified for analysis. Genotype data from the array were merged with pedigree and phenotype data and quality controlled for the following: Mendelian errors (recoded to missing), Hardy-Weinberg equilibrium (SNPs with deviations at p<.001 removed), missing rate across patients (>10% removed), minimal minor allele frequency (SNPs with MAF <1% removed), and low genotyping rate for an individual (<10% removed in the simple association test) using the PLINK v0.99r software package (22). This filtering reduced the number of SNPs to 1,774 total. A two-sample proportions statistic (two-sided) was used to determine PAF differences (p-value cutoff <0.05) between parents of Rejectors and Non-Rejectors. This screening step yielded a subset of MHC-SNPs for subsequent hypothesis testing using the Gamete Competition (GC) statistic, as implemented in the software package Mendel v7.0.0 (23). In this approach, a segregation parameter τ is estimated by maximum likelihood from the family data, and a likelihood ratio test is then applied. For a two-allele locus with alleles 1 and 2, the probability that a heterozygous parent with genotype 1/2 transmits allele 1 is defined as Pr (1/2 -> 1)= τ₁/( τ₁+ τ₂); the null hypothesis of Mendelian transmission corresponds to τ₁ = τ₂ = 1 (14). The transmission parameter τ_k is set to 1 for the most frequent allele k, while the remaining τ_k values (for all k not equal to the most frequent allele) are estimated by maximum likelihood. In our application, we assumed symmetric transmission: that is, an allele that is over-transmitted to Rejectors is assumed to be similarly under-transmitted to Non-Rejectors. Results are summarized as the p-value and chi-squared statistic (with 95% confidence bounds) for the PAF comparison, and the estimated segregation parameter (τ), the allele frequency, and the p-value of the GC statistic (see Appendix 1). Note that this screening/testing approach markedly reduces multiple testing problems because the PAF test (which only uses parental genotypes) is independent of the gamete competition test (which depends on how often specific alleles are transmitted to the children). Thus, since the PAF test alone is used for SNP selection, we only have to adjust for the actual number of GC tests conducted on the much smaller number of selected SNPs.

Confirmatory association testing with 400 normal Caucasian controls

Comparisons of MAF for significant MHC-SNPs were conducted, between Rejectors and normal controls, and between Non-Rejectors and normal controls (chi-square test).

Validating candidacy of the HLA-DOA locus with differential gene splicing patterns

The Affymetrix Human Exon 1.0 ST array was used to measure differential splicing patterns in archived RNA isolated from 29 of 80 children. Probe summaries for both the genes and exons were computed using the Affymetrix Power Tools (APT) software and ‘rma-sketch’ normalization method. The gene-level normalized intensities (NI) were computed with the MiDAS algorithm and both the splicing index (SI) (24) and Student’s t-test p-values (two-sided) were computed on the NI values in R (25). Principal components analysis (PCA) was calculated separately on exon-level and gene-level intensities and was used to remove 3 outlier samples, for a remaining total of 26 samples-11 Rejectors and 15 Non-Rejectors. To remove low expressed gene-level probes, those with values less than 3.5 (log₂ scale) in >50% of the samples in either group were filtered out leaving 17,242 of the original 22,011 probes. Those gene-level probes that were highly differentially expressed between groups were also removed using a fold change threshold of log₂ (10). This step accounts for the tendency for the gene-level probe set intensities in each group to be “disproportionately affected by background noise or saturation” (Affymetrix Technical notes). Exon-level probes were filtered based on APT’s detection above background (DABG) p-value greater than 0.05 in n-1 samples leaving 218,402 of the original 287,329 exon-level probes.

After gene and exon probe filtering, 7 gene-level probes (and 67 corresponding exon-level probes for the genes) within a window size of 200 kb that included the HLA-DOA gene on chromosome 6 were extracted and examined for differential splicing (positions 33,000,000-33,200,000). This was conducted to restrict the analysis to only the immediate region surrounding the HLA-DOA gene.

Relating intragraft content of B-lymphocytes, to rs9296068

Because the HLA-DOA gene inhibits class II antigen presentation selectively in B-cells, we asked whether B-cells were present in rejecting allograft biopsies, and whether the B-cell content of allograft biopsies varied with the presence or absence of the risk allele of rs9296068. Slides were re-cut from stored tissue blocks for 36 of 80 children, with 4μm thick tissue sections, stained in batches on a Ventana immunostainer using mouse monoclonal anti CD79a (DAKO M7050 at 1:100, Ventana CCI mild, 32 minutes at 32C,iView DAB detection), as well as their positive and negative biological staining controls for each batch. Hematoxylin counterstain was applied after DAB detection. The surface marker CD79a identifies cells of B-cell/plamacytoid lineage.

Cell counting was done blinded to outcome, and prior to histologic re-review of the tissues. Duplicate counts were done on each 10^th case to assess variance. Portal and lobular regions of allografts were optically scored separately, at a magnification of ×400 and ×200, respectively. B-cell counts were expressed as total number per portal area, and total number per lobule (Figure 3).

Figure 3

a). Significant correlation exists between B-lymphocyte content in liver grafts with ACR, and numbers of risk allele G at rs9296068 (r²=0.194, p=0.021). Accompanying micrographs show ACR, with portal areas containing brown, horseradish-peroxidase-stained (more ...)

Results

Demographics

Early ACR occurred at mean±SD 32.5±4.2 days after LTx in 37 children who were termed Rejectors. Rejectors and Non-Rejectors (n=43) were similar with respect to age (median±SEM 6±1.1 vs 4.6±0.9 years, p=NS), male: female gender distribution (16:21 vs 21:22, p=NS), and etiology of liver failure requiring LTx (Table 1). The racial distribution in the Rejector and Non-Rejector groups was (Caucasian: African-American: Other=31:2:4 vs 42:0:1, respectively). Also, there were no differences between Rejectors and Non-Rejectors, in donor-recipient matching at the HLA-A, HLA-B, and HLA-DR loci (1.4±1.2 vs 1.8±1.2 antigens, p=NS), or disease severity as reflected in the Pediatric End-Stage Liver Disease score (PELD, 25.5±13.2 vs 24±14.2, P=NS). The sample size was too small to evaluate the effect of disease, age at transplant, and immunosuppression on rejection outcomes, or whether the highly associated SNP(s) were independent predictors of rejection outcome.

Table 1

Etiology of liver disease leading to LTx in 80 children.

Fifty-seven SNPs passed our screening test due to large differences in MAF, when parents of Rejectors were compared with parents of Non-Rejectors (Appendix 1). When the GC statistic was applied to these 57 SNPs, only one SNP, rs9296068, in the 5′ flanking UTR of HLA-DOA demonstrated differences between groups with p<0.05 in parent-to-child transmission (p=0.0183, Table 2). Specifically, the minor allele G was transmitted more frequently to Rejectors, and less so to Non-Rejectors from biological parents. The physical position of the implicated SNP is 33096673 (Build 35). Figure 1 summarizes the results of the PAF comparison for the entire set of 1,774 MHC SNPs (upper panel), and the results of the GC test applied to the 57 selected SNPs for Rejectors and Non-Rejectors (lower panel).

Table 2

Only one SNP, rs9296068 near the HLA-DOA gene, met criteria for selection and hypothesis testing in our two-tier approach. This SNP demonstrated a large difference (p=0.041) in MAF when parents of Rejectors were compared with parents of Non-Rejectors, (more ...)

Figure 1

Upper panel shows −log₁₀(p-values) for comparison of minor allele frequencies of 1,776 SNPs in parents of Rejectors, and parents of Non-Rejectors. Fifty seven SNPs show large differences (p<0.05, −log₁₀(p-value)>1.3) in (more ...)

In the confirmatory association testing step, 77 SNPs showed significant differences in allele frequencies when Rejectors were compared directly with Non-Rejectors (p≤0.05). Among these 77 SNPs, additional between-group comparisons showed 39 SNPs to be significantly different among Rejectors, and 19 SNPs to be significantly different among Non-Rejectors, when each group was compared separately with 400 normal controls (Appendix 2). In direct comparisons, the differences between Rejectors and Non-Rejectors were most significant (p=0.002) for the SNP rs9296068 (Table 3). When compared with normal controls, the minor allele (G) of rs9296068 was more commonly seen among Rejectors (36.7% vs 51.4%, p=0.015), but less commonly among Non-Rejectors (36.7 % vs 26.8%, p=0.074). Only one other SNP rs9276994 shows similar differences in distribution, when Rejectors are compared with Non-Rejectors (48.6% vs 28%, p=0.009), and when normal controls are compared either with Rejectors (37.7% vs 48.6 %, p=0.074) or with Non-Rejectors (37.7% vs 28%, p=0.083) (Table 4). In all, five of 14 top-ranked discriminatory SNPs localized to the 5′ flanking UTR of HLA-DOA (Table 4).

Table 3

Confirmatory association testing shows that MAF for rs9296068 are greater among Rejectors, and less among Non-Rejectors, compared with a large cohort of 400 normal Caucasian children. Together, these results confirm parent-to-child over-transmission among (more ...)

Table 4

Results of confirmatory association testing for SNPs, ordered by physicial position, which showed the most significant differences in MAF (p≤0.010) between Rejectors and Non-Rejectors (last column). Of 14 such SNPs, five are near HLA-DOA. Among (more ...)

The first HLA-DOA exon is repressed in Rejectors

Following probe filtering (explained above) and retention of only those genes with at least one exon with p<0.05, two genes remained, HLA-DOA and HLA-DPA1 (Appendix 3). The first HLA-DOA exon between physical positions 33,085,305-33,085,362 was repressed among Rejectors, as suggested by −1.53-fold decreased expression, compared with Non-Rejectors (Figure 2). This repressed exon lies immediately adjacent to the 5′ flanking UTR of HLA-DOA, which contains the risk allele rs9296068 at position 33,096,673, and is ≈11.3 kb removed from the risk allele. In contrast, ≈47.8 kb, and three HLA-DPA exon transcripts separate rs9296068 from the discriminatory HLA-DPA exon at position 33,144,437-33,144,544, whose expression is −1.30-fold lower among Rejectors, compared with Non-Rejectors.

Figure 2

Splicing index values, defined as the fold change between mean gene-level normalized exon intensities in the Rejector and Non-Rejector groups and 95% confidence intervals for the 10 exons that map to the HLA-DOA gene. The physical position ranges of each (more ...)

Intragraft B-cell content is higher in Rejectors with the risk allele of rs9296068 (Figure 3)

The risk allele (G) of rs9296068 was present in seven of 10 liver allograft biopsies from Non-Rejectors (allele frequency=0.4), and 19 of 26 Rejectors (allele frequency=0.5). B-cell content (median±SEM) was low at 1.1 cell per portal area in Non-Rejectors. Among Rejectors, the risk allele (G) was absent in seven, heterozygous in 12, and homozygous in seven. B-cell content in respective allograft biopsies from these Rejectors was 4±2.6 cells, 11±3.3 cells, and 16±4.7 cell per portal area, and showed significant correlation with the risk allele (r²=0.194, p=0.021, Figure 3). Also, differences between B-cell content were significant when Rejectors without the risk allele (n=7, 4±2.6 B-cells/portal area) were compared with Rejectors who were either homozygous or heterozygous for the risk allele of rs9296068 (n=17, 11.54±2.4 cells/portal area, p=0.022). The lobular areas contained an average of 0-3 B-cells per lobule in both Rejectors and Non-Rejectors, and showed no differences.

Subanalysis in Caucasian children to evaluate population stratification

The HapMap database shows rs9296068 (G) allele frequencies of 31% in Caucasians (CEU), and 67.8% in Yoruba (YRI) populations. To avoid false positive association in case-control comparisons, trios for two African-Americans and all six Non-Caucasians were excluded, five new trios from Caucasians children with LTx added, and both, the PAF comparison and simple case/control association were recalculated for 35 Rejectors, and 42 Non-Rejectors. Four SNPs adjacent to HLA-DOA remain discriminatory (p<0.05) in comparison of Rejectors with Non-Rejectors. Among them, rs6457699, rs9276994, and rs9296068 are ranked 3, 8 and 9 among the top ten SNPs, with all showing greater MAF among Rejectors and less among Non-Rejectors, compared with 400 normal Caucasian children (Appendix 4). In the PAF test, higher MAF among parents of Rejectors approached significance for rs6457699 (p=0.055) and rs9276994 (p=0.077) but not for rs9296068 (40 vs 34.7%, p=0.43). This observation suggests that the Rejector group is relatively underpowered to detect MAF differences for the rs9296068 SNP, even though the GC test confirms significant differences in parent-to child transmission between groups for its risk allele (p=0.004). Finally, both rs6457699, and rs9276994 are situated ≈4kb and ≈6kb upstream of the first HLA-DOA exon, and the GC test was significant for rs6457699 (p=0.04), implicating HLA-DOA once again (Appendix 4).

The GC test, unlike a case/control association, is based on distortions in allele transmission from parent to child, and so should be less sensitive to population substructure. It is of interest to assay whether or not transmission distortion occurs at the SNP of interest in the HapMap non-disease reference population (26). Therefore, the GC statistic was calculated for both alleles of rs9296068 for 30 YRI families (90 individuals), and 20 CEU families (90 individuals) using genotype and pedigree files from the Hapmap database. In both populations, the risk allele failed to show preferential transmission compared with its complementary allele, as suggested by GC p-values >0.05. Therefore, the null hypothesis of Mendelian transmission cannot be rejected. This sub-analysis supports our conclusions from GC testing of 72 Caucasian and 8 Non-Caucasian trios.

Discussion

Our multi-step approach identifies the HLA-DOA gene, and the B-lymphocyte in which it is exclusively expressed, as plausible candidates contributing to pediatric LTx rejection. Adding early functional validation to family-based association in the same dataset is especially useful, because it obviates the need for immediate replication in an independent cohort, whose accrual may take several years in the case of rare disease traits. This novel approach was motivated by failure when we first performed genome-wide association with the unmodified TDT applied to all 550,000 SNPs (on-going study and data not shown). All SNPs showing significant p-values failed to remain significant after Bonferroni correction for multiple hypothesis testing.

Because SNP reduction with PAF comparisons and the GC test are independent, the combined p-value for r9296068 is a multiple of values from both procedures (0.041*0.0183=0.00075). However, this would not be significant after multiple-testing correction for 1,774 SNPs. Therefore, transmission characteristics of rs9296068 were confirmed by showing that allele frequencies in Rejectors and Non-Rejectors were respectively, significantly greater (51.4 vs 36.8%, p=0.015), and less (26.8 vs 36.8%, p=0.074) than those in the normal control population of 400 Caucasian children (Table 3). Finally, in direct comparisons of allele frequencies between Rejectors and Non-Rejectors, the 5′ flanking UTR of HLA-DOA was represented by five SNPs among 14 top-ranked SNPs (p≤0.01) (Table 4), of which rs9296068 achieved the highest p-value (0.002), and another, rs9276994, also showed evidence of over-transmission in Rejectors (p=0.074), and under-transmission in Non-Rejectors (p=0.082), even though parental allele frequencies were similar. Significantly, all five highly-ranked SNPs, beginning with rs6457699 and ending with rs9277015, were present upstream, in an ≈11.6 kb segment, at a distance of ≈4-15 kb from the first HLA-DOA exon (Figures (Figures22 and and3).3). Promoter function has been predicted for a highly conserved ≈10kb region immediately upstream of the first HLA-DOA exon, with Caucasians demonstrating a linkage disequilibrium (LD) block encompassing the rs9296068 locus (Figure 4) (27). Because 3 of 5 discriminatory SNPs localize to this putative promoter, altered transcription factor binding, and differential regulation of HLA-DOA gene expression can be postulated. Significant repression of the first HLA DOA exon among Rejectors in our study supports this view. Decreased HLA-DOA gene expression is also seen with increasing numbers of the risk allele of rs9296068 for both Caucasian and Yoruba populations in public databases (28).

Figure 4

The HLA-DOA gene is transcribed from the minus strand. The upstream 5′ flanking UTR region defined by five discriminatory SNPs in simple association testing of Rejectors vs Non-Rejectors lies between rs9296068 to the right and rs6457699 to the (more ...)

We interpret our observed associations as suggestive of a causal role for the HLA-DOA gene, and for the B-lymphocyte, rather than a causal role for the SNP itself; however causality remains unproven until definitive studies identify a true causal variant, and the mechanism by which it might alter HLA-DOA gene expression and B-cell function. For the association itself, all statistical tests used sequentially in the current study are necessary because neither simple association testing nor family-based association testing generates, by itself, a list of SNPs larger than the expected false positive prediction (p=0.05 times 1,774 SNPs=89 false-positives). The HLA-DOA gene, and its adjacent region, is of interest for several reasons. Compared with one of its 3′ neighbors, HLA-DPB1, which is highly polymorphic, and can influence immunological outcomes in bone-marrow, corneal and renal transplantation, the HLA-DOA gene is relatively non-polymorphic, and inhibits class II antigen presentation in mature B-cells; but its role in organ transplantation is unknown (29-31, 19, 20). Our findings lead us to speculate that a missing exon transcript may produce a dysfunctional HLA-DOA gene product, which facilitates rejection by failing to inhibit antigen presentation by B-lymphocytes. The resulting increased B-lymphocyte participation in the rejection process is seen as nearly three-fold higher intragraft B-lymphocyte content in rejecting liver grafts in our study, when the minor allele G of rs9296068 was present, compared with rejecting allografts from children without this allele. Among pediatric LTx, prior supportive evidence also includes greater resistance of activated B-lymphocytes to immunosuppression, as well as a relative excess of B-lymphocytes during the risk period for early rejection (18, 32). B-lymphocyte-rich infiltrates, as well as B-lymphocyte-specific gene expression products have already been demonstrated in renal allografts with steroid-resistant acute cellular rejection (33). While our observations are preliminary, they suggest that the HLA-DOA gene and its vicinity be mapped further to identify novel causal variants.

We acknowledge several limitations. The heterogeneous diagnoses precipitating LTx are potential sources of stratification, although no single disease dominated either outcome group. For example, all 5 children with PSC carried the risk allele, with four experiencing rejection, despite statistically similar proportions of PSC in either group (4/37 Rejectors vs 1/43 Non-Rejectors, p=NS). Second, our SNP reduction step which relies on PAF comparisons, is biased by inclusion of eight Non-Caucasians trios, seven in the Rejector group. In accepting rs9296068 as a candidate, an illustrative subanalysis shows that our conclusions would be unchanged if an adequately powered Caucasian sample was available (Appendix 5). For these reasons, we have also relied on functional studies, coupled with allele-specific decrease in HLA-DOA expression in public databases for both CEU and YRI, to suggest biological relevance for the risk allele. We hope to relate differences in HLA-DOA exon repression to allelic variations at rs9296068 in a larger future sample. The small numbers of Rejectors (n=11) in whom HLA-DOA exon repression was shown, could not be divided further for adequately powered correlations with allelic variants in this study.

We conclude that our combination of methods can identify biologically relevant associations in small populations. We propose to validate our conclusions and address the primary power limitations of this pilot study by extending it to multicenter LTx populations.

Acknowledgments

Support: 5RO1AI073895-01, Children’s Hospital of Pittsburgh Research Foundation, and Hillman Foundation of Pittsburgh.

Special thanks: Timothy Billiar and Roger Oxendale.

Appendix 1

Appendix 1

p-values for parental allele frequency comparison, and for the GC statistic comparing prent-to-child transmission of SNPs between Rejectors and Non-Rejectors

The minor allele for each snp is identified in the allele column, and the minor allele frequencies (MAF)s in both the parents and offspring for each group (NR and R) are provided for this allele.

For each SNP in the GC model, the transmission parameter tau _k is set to 1 for the most frequent allele k, while the remaining tau_k values (for all k not equal to the most frequent allele) are estimated by maximum likelihood.

SNP	Closest Gene	Distance to gene	Position	SNP type	Parental allele frequency comparison, p- value	Lower conf. Interval	Higher conf. Interval	Rejector Parents MAF	Non-Rejector Parents MAF	Rejector MAF	Non-Rejector MAF	Rejector vs. Non-Rejector GC p value	tau_k	Allele frequency in group k	Allele	distance to exon boundary
rs4713213	OR5V1	0	29446941	INTRONIC	0.046	0.0027	0.2463	50.8%	38.3%	44.4%	40.2%	0.285	0.7749	44.0%	1	−9
rs4598109	OR5V1	0	29452683	INTRONIC	0.028	0.0146	0.2546	46.2%	32.7%	38.6%	35.7%	0.144	0.7032	38.8%	1	−9
rs238883	OR5V1	0	29454205	INTRONIC	0.033	−0.2547	−0.0115	38.6%	52.0%	43.1%	46.4%	0.097	1.5377	46.0%	2	−9
rs238882	OR5V1	0	29454342	INTRONIC	0.012	0.0324	0.2711	48.5%	33.3%	41.7%	35.7%	0.228	0.7473	40.7%	1	−9
rs238872	OR5V1	0	29459852	INTRONIC	0.040	0.0066	0.251	53.8%	40.9%	50.0%	46.4%	0.180	0.7316	47.2%	1	−9
rs3094556	OR5V1	0	29461387	INTRONIC	0.004	0.0571	0.2963	58.7%	41.0%	55.6%	46.4%	0.290	0.7794	49.7%	2	−9
rs3094551	OR5V1	0	29462778	INTRONIC	0.010	−0.2759	−0.0379	36.2%	51.9%	40.0%	46.4%	0.303	1.2708	44.4%	1	−9
rs3094550	OR5V1	0	29462788	INTRONIC	0.009	−0.28	−0.0399	36.6%	52.6%	40.0%	47.6%	0.382	1.2278	45.2%	2	−9
rs2517817	HLAH_HUMAN	671	29967496	DOWNSTREAM	0.032	−0.1501	−0.009	4.6%	12.5%	2.9%	13.8%	0.273	0.6088	10.0%	1	−9
rs7758512	Q6ZU40_HUMAN	0	30078568	INTRONIC	0.034	0.0049	0.1686	15.9%	7.2%	14.7%	12.2%	0.374	0.7223	12.1%	2	−9
rs9261129	Q6ZU40_HUMAN	0	30087558	INTRONIC	0.042	0.0024	0.1605	14.4%	6.3%	14.3%	11.9%	0.405	0.735	11.8%	2	−9
rs9261277	ZNRD1	0	30139070	INTRONIC	0.044	0.0014	0.16	15.2%	7.1%	14.3%	10.5%	0.631	0.8399	11.5%	2	−9
rs1264701	TRIM31	−4318	30174337	DOWNSTREAM	0.027	−0.2051	−0.0144	13.1%	24.1%	9.7%	22.1%	0.514	0.816	19.2%	1	−9
rs3734838	TRIM31	0	30188210	NON_SYNONYMOUS_CODING	0.042	0.0005	0.1417	11.5%	4.4%	10.0%	2.3%	0.719	1.1603	8.3%	1	−9
rs9261441	TRIM31	10891	30199737	INTERGENIC	0.012	0.019	0.1779	16.2%	6.3%	15.3%	3.6%	0.372	1.3678	10.8%	1	−9
rs2022065	TRIM10	0	30229439	3PRIME_UTR	0.045	0.0021	0.224	34.1%	22.8%	35.7%	23.2%	0.638	1.1264	28.2%	1	88
rs1345229	TRIM26	1191	30290374	UPSTREAM	0.008	0.0226	0.1713	14.2%	4.5%	13.9%	2.4%	0.341	1.4499	9.2%	1	−9
rs9357097	TRIM39	−9500	30393100	INTERGENIC	0.015	−0.2417	−0.0275	19.2%	32.7%	16.7%	32.1%	0.480	0.8097	27.2%	1	−9
rs2516649	GNL1	−16814	30604861	INTERGENIC	0.040	−0.2333	−0.0069	25.0%	37.0%	23.6%	35.7%	0.994	0.9958	31.2%	2	−9
rs2844713	GNL1	0	30627237	INTRONIC	0.043	−0.2307	−0.0052	26.8%	38.6%	27.8%	36.0%	0.476	1.1975	33.2%	1	−9
rs4248154	NM_001010909	44941	31110595	INTERGENIC	0.017	−0.1853	−0.02	8.1%	18.4%	11.1%	20.2%	0.522	0.7948	14.1%	1	−9
rs2523849	C6orf15	−53952	31133030	INTERGENIC	0.031	0.0079	0.1963	22.4%	12.2%	24.3%	13.1%	0.734	1.1123	17.6%	2	−9
rs2428514	C6orf15	−51487	31135495	INTERGENIC	0.044	0.0015	0.1804	19.9%	10.8%	20.0%	12.2%	0.910	0.9627	15.6%	1	−9
rs2517403	C6orf15	−11994	31174988	INTERGENIC	0.048	0.0013	0.2352	41.0%	29.2%	42.9%	34.5%	0.669	0.9029	35.3%	2	−9
rs2844635	C6orf15	−3522	31183460	DOWNSTREAM	0.040	0.0055	0.2384	41.0%	28.9%	42.9%	35.0%	0.603	0.8816	35.1%	2	−9
rs9295957	Q6H1K9_HUMAN	11917	31265572	INTERGENIC	0.046	−0.1814	−0.0039	10.6%	19.9%	10.6%	17.1%	0.498	1.2478	15.6%	1	−9
rs7745906	1C07_HUMAN	−32518	31311987	INTERGENIC	0.037	−0.1749	−0.0076	8.8%	18.0%	7.4%	18.6%	0.960	1.0194	14.2%	1	−9
rs2074488	1C07_HUMAN	524	31348410	UPSTREAM	0.038	−0.156	−0.0066	6.0%	14.1%	8.6%	14.0%	0.386	1.4316	10.6%	1	−9
rs9366778	1C07_HUMAN	29266	31377152	INTERGENIC	0.039	−0.2425	−0.0071	33.1%	45.6%	34.7%	41.7%	0.446	1.2061	40.3%	1	−9
rs406936	NP_008860.4	0	32041140	INTRONIC	0.048	−0.1643	−0.0031	8.1%	16.5%	5.6%	15.0%	0.780	0.9085	12.5%	1	−9
rs454212	NP_008860.4	0	32042351	INTRONIC	0.028	−0.1715	−0.0123	6.4%	15.5%	6.1%	12.2%	0.948	1.0266	11.3%	1	−114
rs387608	STK19	0	32049536	INTRONIC	0.048	−0.1643	−0.0031	8.1%	16.5%	5.6%	15.0%	0.780	0.9085	12.5%	1	−9
rs2269429	TNXB	0	32137161	NON_SYNONYMOUS_CODING	0.044	−0.1475	−0.0044	5.2%	12.8%	5.6%	9.8%	0.400	1.3939	9.3%	1	−9
rs204899	TNXB	0	32165605	INTRONIC	0.012	−0.1571	−0.0214	3.7%	12.7%	2.8%	8.8%	0.657	1.1994	8.6%	1	−9
rs3115553	C6orf10	−14648	32353805	INTERGENIC	0.049	0.0004	0.2025	26.8%	16.7%	25.7%	23.2%	0.298	0.7368	21.9%	1	−9
rs9268384	C6orf10	0	32444564	NON_SYNONYMOUS_CODING	0.045	−0.2416	−0.004	33.8%	46.1%	35.7%	43.9%	0.962	0.9873	40.3%	2	5
rs4424066	C6orf10	2096	32462406	UPSTREAM	0.039	−0.2411	−0.0078	30.6%	43.0%	26.4%	37.8%	0.798	1.0642	37.2%	2	−9
rs3817973	BTNL2	−635	32469089	DOWNSTREAM	0.022	−0.2518	−0.0207	29.4%	43.0%	25.7%	36.9%	0.707	1.0954	37.0%	1	−9
rs3793126	BTNL2	0	32479597	INTRONIC	0.044	−0.2128	−0.0048	17.7%	28.6%	18.1%	25.6%	0.727	1.1156	23.8%	2	−9
rs86567	HLA-DOA	0	33084737	INTRONIC	0.037	−0.2434	−0.0089	31.6%	44.2%	31.9%	48.8%	0.292	0.7825	37.8%	2	−9
rs6457699	HLA-DOA	4258	33089625	UPSTREAM	0.031	0.0123	0.2547	53.7%	40.4%	52.9%	30.5%	0.072	1.5742	45.9%	1	−9
rs6933994	HLA-DOA	9731	33095098	INTERGENIC	0.042	0.0053	0.2485	58.1%	45.4%	55.7%	36.3%	0.173	0.7236	49.2%	1	−9
rs6457702	HLA-DOA	10660	33096027	INTERGENIC	0.009	−0.2802	−0.0399	33.3%	49.3%	37.5%	52.5%	0.757	0.9284	42.4%	1	−9
rs9296068	HLA-DOA	11306	33096673	INTERGENIC	0.041	0.0053	0.2447	46.3%	33.8%	50.0%	26.8%	0.018	1.8513	38.9%	2	−9
rs986521	COL11A2	0	33244123	INTRONIC	0.038	−0.2075	−0.0074	16.2%	26.9%	15.7%	30.0%	0.243	0.7138	21.8%	2	−151
rs9277932	COL11A2	0	33249231	INTRONIC	0.028	−0.25	−0.0154	30.9%	44.2%	35.3%	45.1%	0.813	1.0611	38.0%	1	−26
rs2855425	COL11A2	0	33252351	INTRONIC	0.040	−0.221	−0.0067	20.9%	32.3%	18.6%	32.9%	0.336	0.7725	26.6%	2	−125
rs6531	RXRB	0	33271429	SYNONYMOUS_CODING	0.036	−0.2241	−0.0088	20.6%	32.2%	18.6%	32.9%	0.342	0.776	26.4%	2	28
rs211474	DAXX	29822	33428591	INTERGENIC	0.045	0.0029	0.2401	46.3%	34.2%	43.1%	32.1%	0.926	1.0226	39.5%	1	−9
rs211455	KIFC1	−31102	33436496	INTERGENIC	0.019	0.023	0.2572	45.7%	31.7%	38.9%	29.1%	0.824	0.9483	37.8%	1	−9
rs211452	KIFC1	−28639	33438959	INTERGENIC	0.020	0.0223	0.263	52.2%	38.0%	43.1%	34.9%	0.624	0.8897	43.7%	2	−9
rs211457	KIFC1	0	33473618	INTRONIC	0.025	−0.1556	−0.0122	5.1%	13.5%	5.6%	14.6%	0.580	0.7843	9.5%	1	−166
rs3116713	PHF1	0	33490266	NON_SYNONYMOUS_CODING	0.025	−0.1445	−0.0116	3.7%	11.5%	4.2%	12.2%	0.537	0.7515	7.9%	2	33
rs211456	ZBTB9	0	33497359	INTRONIC	0.041	0.006	0.2481	54.5%	41.8%	47.2%	39.5%	0.718	0.9181	46.7%	1	−9
rs396746	NP_997380.1	0	33665023	INTRONIC	0.032	−0.1845	−0.0101	10.1%	19.9%	13.9%	20.7%	0.709	1.1293	15.1%	1	−9
rs169737	ITPR3	−12767	33684555	INTERGENIC	0.024	−0.2185	−0.0165	16.2%	27.9%	21.4%	25.6%	0.269	1.3651	22.0%	1	−9
rs12529825	ITPR3	0	33725381	INTRONIC	0.012	−0.1571	−0.0214	3.7%	12.7%	5.6%	14.6%	0.683	0.8395	8.5%	2	−9

Appendix 2

Appendix 2

List of 77 SNPs with significant differences in allele frequencies between Rejectors (R, n=37) and Non-Rejectors (NR, n=43), between Rejectors and 400 normal control Caucasian children, and Non-Rejectors vs 400 normal control Caucasian children. When compared with normal controls, only rs9296068 shows allele frequencies which are greater in Rejectors, and less among Non-Rejectors.

rs926068 also shows the greatest differences in allele frequencies between Rejectors and Non-Rejectors, and is one of five SNPs representing the HLA-DOA gene for which allele frequency differences between groups (R vs NR) are≤0.01.

					Allele frequency			Rejector vs Non-Rejectors Chi-sq			Rejectors vs normal controls Chi-sq		Rejectors vs Normal Controls Chi-sq
SNP	Closest Gene	Distance to gene	SNP type	Position	Rejectors	Non-Rejectors	Normal Control	SNP rank	p-value	odds ratio	p-value	odds ratio	p-value	odds ratio
rs9296068	HLA-DOA	11306	INTERGENIC	33096673	51.40%	26.80%	36.70%	1	0.002	2.888	0.0152	1.822	0.074	0.6311
rs6457699	HLA-DOA	4258	UPSTREAM	33089625	54.30%	31.70%	46.70%	2	0.005	2.558	0.226	1.353	0.009	0.5288
rs602875	HLA-DRB1	16048	INTERGENIC	32681607	42.90%	21.90%	23.80%	3	0.006	2.667	0.0004	2.401	0.707	0.9003
rs1264583	HLAH_HUMAN	10138	INTERGENIC	30401462	14.30%	2.40%	5%	4	0.007	6.667	0.001	3.167	0.299	0.475
rs3094097	Q6ZU40_HUMAN	0	INTRONIC	30741854	14.30%	2.40%	9.70%	5	0.007	6.667	0.227	1.543	0.028	0.2314
rs1345229	TRIM26	1191	UPSTREAM	30290374	14.30%	2.40%	11.90%	6	0.007	6.667	0.553	1.237	0.009	0.1855
rs3134879	TRIM39	−1138	UPSTREAM	30123884	26.60%	48.70%	41.10%	7	0.007	0.3807	0.023	0.519	0.192	1.363
rs2517904	DHX16	0	INTRONIC	29976963	2.90%	15.80%	7.50%	8	0.007	0.1561	0.147	0.3618	0.009	2.317
rs9366752	C6orf12	0	3PRIME_UTR	30132656	34.30%	15.80%	20%	9	0.008	2.769	0.005	2.087	0.368	0.7536
rs1264581	TRIM39	0	SYNONYMOUS_CODING	30405484	16.20%	3.70%	5.30%	10	0.009	5.082	0.0003	3.455	0.524	0.6799
rs9276994	HLA-DOA	6866	INTERGENIC	33092233	48.60%	28.10%	37.70%	11	0.009	2.423	0.075	1.557	0.083	0.6428
rs9277015	HLA-DOA	15877	INTERGENIC	33101244	44.30%	24.40%	37%	12	0.009	2.464	0.227	1.353	0.023	0.5493
rs6933994	TRIM31	10891	INTERGENIC	33095098	57.10%	36.30%	46%	13	0.01	2.345	0.614	0.8808	0.002	2.065
rs9261441	HLA-DOA	9731	INTERGENIC	30199737	15.70%	3.70%	13.60%	14	0.01	4.91	0.623	1.184	0.01	0.2412
rs2394255	HLA-A	9079	INTERGENIC	30057800	58.60%	37.80%	48.60%	15	0.012	2.326	0.11	1.494	0.062	0.6423
rs532098	HCG9	3625	DOWNSTREAM	32686030	30.90%	51.20%	45.60%	16	0.012	0.4255	0.019	0.5327	0.332	1.252
rs9468829	NP_003888.2	36906	INTERGENIC	30857212	21.40%	7.30%	12.10%	17	0.012	3.455	0.024	1.994	0.204	0.5773
rs16896742	HLA-DRB1	20471	INTERGENIC	30030719	46.90%	26.80%	34.20%	18	0.012	2.406	0.042	1.695	0.177	0.7043
rs2516684	RPP21	48325	INTERGENIC	30470974	17.10%	4.90%	13.10%	19	0.014	4.034	0.345	1.369	0.031	0.3394
rs12206499	HCG9	−5765	INTERGENIC	30045106	37.10%	19.50%	25.70%	20	0.015	2.438	0.0377	1.709	0.219	0.7011
rs6904029	HCG9	−809	UPSTREAM	30051046	37.10%	19.50%	25.70%	21	0.015	2.438	0.039	1.704	0.215	0.699
rs3823355	HCG9	0	NON_SYNONYMOUS_CODING	30050062	37.10%	19.50%	25.90%	22	0.015	2.438	0.041	1.693	0.207	0.6945
rs6933546	HLAH_HUMAN	671	DOWNSTREAM	33103992	42.90%	24.40%	36.60%	23	0.016	2.325	0.301	1.298	0.027	0.5582
rs2517817	HLA-DOA	18625	INTERGENIC	29967496	2.90%	14.50%	7.90%	24	0.016	0.1791	0.138	0.3548	0.048	1.982
rs3869070	Q6ZU40_HUMAN	0	INTRONIC	30131847	55.90%	36.60%	44.70%	25	0.018	2.196	0.077	1.564	0.156	0.7123
rs2647044	HLA-G	23348	INTERGENIC	32775888	22.70%	8.70%	9.90%	26	0.019	3.067	0.001	2.655	0.726	0.8655
rs2975033	BAT2	0	INTRONIC	29930240	35.70%	18.70%	25.60%	27	0.019	2.407	0.066	1.612	0.176	0.6698
rs9267522	BAT2	0	NON_SYNONYMOUS_CODING	31711749	27.10%	12.20%	13%	28	0.019	2.682	0.001	2.493	0.836	0.9295
rs3117583	BAT3	0	INTRONIC	31727555	27.10%	12.20%	13%	29	0.019	2.682	0.001	2.493	0.836	0.9295
rs3130618	BAT4	0	NON_SYNONYMOUS_CODING	31740113	27.10%	12.20%	13%	30	0.019	2.682	0.001	2.493	0.836	0.9295
rs3115663	HB25_human	21592	INTERGENIC	31709822	27.10%	12.20%	13.10%	31	0.019	2.682	0.001	2.466	0.812	0.9193
rs4711207	Q6ZU40_HUMAN	0	INTRONIC	30113733	34.40%	17.50%	22.80%	32	0.02	2.469	0.036	1.776	0.28	0.7193
rs2040450	RPP21	19669	INTERGENIC	30442318	16.20%	4.90%	12.70%	33	0.022	3.763	0.42	1.321	0.037	0.3509
rs4259245	ITPR3	0	INTRONIC	33732199	27.10%	45.10%	35.70%	34	0.022	0.4531	0.148	0.6695	0.093	1.478
rs2107202	TRIM40	0	INTRONIC	30213722	20	36.60%	25.90%	35	0.025	0.4333	0.279	0.7162	0.037	1.653
rs915664	DDR1	−54188	INTRONIC	30902596	35.70%	19.50%	30.60%	36	0.025	2.292	0.377	1.259	0.036	0.5492
rs3129763	HA25_HUMAN	−14209	INTERGENIC	32698903	38.60%	21.90%	22.90%	37	0.025	2.233	0.003	2.11	0.84	0.9452
rs3893464	HCG9	−7642	INTERGENIC	30043229	37.10%	54.90%	50.60%	38	0.029	0.4859	0.03	0.5763	0.463	0.843
rs9261301	RNF39	0	INTRONIC	30149538	52.90%	35.40%	44.40%	39	0.03	2.049	0.171	1.405	0.117	0.6859
rs9267546	LY6G6D	−1245	UPSTREAM	31781415	17.60%	6.20%	9.30%	40	0.03	3.214	0.026	2.102	0.371	0.6541
rs387608	HCG9	0	INTRONIC	32049536	5.70%	17.10%	15%	41	0.031	0.2944	0.033	0.3434	0.618	1.167
rs406936	TRIM39	−9500	INTERGENIC	32041140	5.70%	17.10%	14.70%	42	0.031	0.2944	0.037	0.3503	0.574	1.19
rs9357097	NP_008860.4	0	INTRONIC	30393100	17.10%	32.50%	30.90%	43	0.031	0.4297	0.0155	0.4615	0.776	1.074
rs2394250	STK19	0	INTRONIC	30051635	51.40%	34.10%	41.60%	44	0.031	2.042	0.111	1.485	0.189	0.7272
rs2257914	TRIM40	0	5PRIME_UTR	30228542	10%	23.20%	17.50%	45	0.032	0.3684	0.108	0.5238	0.203	1.422
rs2021723	TRIM10	0	3PRIME_UTR	30211902	10%	23.20%	17.30%	46	0.032	0.3684	0.119	0.533	0.182	1.447
rs2844776	TRIM26	0	INTRONIC	30279806	15.70%	30.50%	20.10%	47	0.033	0.4251	0.374	0.74	0.028	1.741
rs2395175	HLA-DRA	−2621	UPSTREAM	32513004	4.30%	14.60%	14.20%	48	0.033	0.2612	0.02	0.2714	0.907	1.039
rs2187668	HA25_HUMAN	0	INTRONIC	32713862	20.60%	8.50%	8.60%	49	0.034	2.778	0.0013	2.747	0.978	0.9888
rs9267911	TRIM31	0	INTRONIC	32313088	58.60%	41.50%	47.30%	50	0.035	1.996	0.069	1.578	0.317	0.7908
rs2523990	NOTCH4	13266	INTERGENIC	30185208	58.60%	41.50%	47.40%	51	0.035	1.996	0.0722	1.57	0.307	0.7868
rs12212092	HCG9	−9631	INTERGENIC	30236421	12.90%	3.70%	6.30%	52	0.036	3.885	0.037	2.19	0.338	0.5635
rs6457109	TRIM10	0	NON_SYNONYMOUS_CODING	30041240	18.60%	7.30%	9.90%	53	0.036	2.889	0.0233	2.082	0.455	0.7205
rs9277027	HLAH_HUMAN	−5770	INTERGENIC	33106216	30%	15.80%	30.40%	54	0.037	2.275	0.948	0.9824	0.006	0.4319
rs2523809	HLA-DOA	20849	INTERGENIC	29957598	8.60%	20.70%	11.80%	55	0.037	0.3585	0.424	0.7041	0.019	1.964
rs375912	NR_002139.1	−14455	INTERGENIC	33124706	32.90%	18.30%	36.50%	56	0.039	2.186	0.543	0.8514	0.001	0.3895
rs2517930	HLA-DPA1	−15618	INTERGENIC	29853054	37.10%	21.90%	30.80%	57	0.039	2.101	0.271	1.329	0.097	0.6325
rs4713411	NM_001010909	29501	INTERGENIC	31095155	28.10%	44.90%	43.40%	58	0.04	0.4807	0.018	0.5114	0.796	1.064
rs1264567	TRIM40	−11178	INTERGENIC	30474079	20%	8.50%	18.90%	59	0.041	2.679	0.818	1.075	0.02	0.4011
rs213213	RPP21	51430	INTERGENIC	33291708	21.40%	36.60%	26.70%	60	0.041	0.4727	0.332	0.7468	0.058	1.58
rs1419675	RING1	3232	DOWNSTREAM	30200686	21.40%	36.60%	25.90%	61	0.041	0.4727	0.413	0.7813	0.037	1.653
rs6910071	C6orf10	0	INTRONIC	32390832	7.14%	18.30%	19.10%	62	0.043	0.3436	0.0126	0.3253	0.855	0.9467
rs86567	HLA-DOA	0	INTRONIC	33084737	31.40%	47.60%	38.10%	63	0.043	0.5053	0.267	0.7439	0.095	1.472
rs3734838	RNF39	12015	INTERGENIC	30188210	10.30%	2.40%	8.50%	64	0.044	4.59	0.622	1.229	0.052	0.2676
rs6909253	TRIM31	−6114	INTERGENIC	30163622	52.90%	36.60%	37.30%	65	0.044	1.943	0.01	1.889	0.906	0.9719
rs9261394	TRIM31	0	NON_SYNONYMOUS_CODING	30172541	52.90%	36.60%	38%	66	0.044	1.943	0.015	1.829	0.801	0.9413
rs1003581	GABBR1	0	INTRONIC	29648183	9.10%	21.20%	19.90%	67	0.045	0.3706	0.032	0.4019	0.778	1.085
rs594223	C6orf125	0	INTRONIC	33775943	21.40%	9.80%	17.90%	68	0.045	2.523	0.46	1.253	0.063	0.4967
rs9468692	RNF39	1571	UPSTREAM	30227869	14.30%	4.90%	7.10%	69	0.046	3.25	0.032	2.173	0.445	0.6685
rs1150735	TRIM10	0	3PRIME_UTR	30153178	27.10%	42.70%	40%	70	0.046	0.5003	0.034	0.5588	0.637	1.117
rs1052486	BAT3	0	NON_SYNONYMOUS_CODING	31718665	40%	56.40%	43.70%	71	0.046	0.5152	0.008	1.934	0.987	0.9961
rs1264701	TRIM31	−4318	DOWNSTREAM	30174337	10%	21.90%	22.40%	72	0.048	0.3951	0.015	0.3842	0.921	0.9726
rs1077393	BAT3	0	INTRONIC	31718508	40%	56.10%	44.10%	73	0.048	0.5217	0.01	1.899	0.969	0.991
rs986521	COL11A2	0	INTRONIC	33244123	15.70%	29.30%	20.10%	74	0.048	0.4506	0.374	0.74	0.053	1.642
rs9262138	DHX16	0	NON_SYNONYMOUS_CODING	30735846	10%	2.40%	5.90%	75	0.049	4.444	0.17	1.78	0.196	0.4005
rs9267665	ZBTB12	1087	UPSTREAM	31978835	10%	2.40%	4.30%	76	0.049	4.444	0.029	2.503	0.43	0.5632
rs7745906	1C07_HUMAN	−32518	INTERGENIC	31311987	7.35%	18.30%	13%	77	0.05	0.3545	0.177	0.5311	0.182	1.498

Appendix 3

Appendix 3

Summary statistics for differentially expressed exons between Rejectors and Non-rejectors

Exon ID	Transcript ID	Exon fold change^†	Exon p-value^†	Transcript fold change^‡	Transcript p-value^‡	Transcript/SNP name	Exon start/SNP location	Exon stop
2950312	2950307	1.063	0.555	−1.083	0.326	HLA-DOA	33,081,985	33,082,504
2950313	2950307	1.122	0.397	−1.083	0.326	HLA-DOA	33,082,838	33,082,870
2950314	2950307	1.069	0.372	−1.083	0.326	HLA-DOA	33,082,939	33,082,970
2950317	2950307	−1.065	0.638	−1.083	0.326	HLA-DOA	33,083,111	33,083,291
2950322	2950307	−1.060	0.395	−1.083	0.326	HLA-DOA	33,083,777	33,083,802
2950323	2950307	−1.011	0.905	−1.083	0.326	HLA-DOA	33,083,818	33,083,847
2950324	2950307	1.017	0.84	−1.083	0.326	HLA-DOA	33,083,857	33,083,987
2950325	2950307	1.127	0.383	−1.083	0.326	HLA-DOA	33,084,068	33,084,099
2950326	2950307	−1.048	0.748	−1.083	0.326	HLA-DOA	33,085,222	33,085,260
2950327	2950307	−1.532	0.0153	−1.083	0.326	HLA-DOA	33,085,305	33,085,362
	untranslated					rs6457699	33,089,625
	untranslated					rs9276994	33,092,233
	untranslated					rs6933994	33,095,098
	untranslated					rs9296068	33,096,673
	untranslated					rs9277015	33,101,244
2950331	2950329	1.129	0.545	−1.166	0.419	HLA-DPA1	33,140,788	33,140,818
2950332	2950329	−1.161	0.517	−1.166	0.419	HLA-DPA1	33,140,836	33,140,861
2950333	2950329	−1.246	0.204	−1.166	0.419	HLA-DPA1	33,140,925	33,141,014
2950338	2950329	−1.302	0.0128	−1.166	0.419	HLA-DPA1	33,144,437	33,144,544
2950340	2950329	−1.114	0.31	−1.166	0.419	HLA-DPA1	33,144,774	33,144,804
2950341	2950329	−1.014	0.875	−1.166	0.419	HLA-DPA1	33,144,844	33,144,876
2950342	2950329	1.030	0.792	−1.166	0.419	HLA-DPA1	33,144,929	33,144,963
2950343	2950329	−1.114	0.479	−1.166	0.419	HLA-DPA1	33,144,969	33,144,999
2950345	2950329	1.057	0.565	−1.166	0.419	HLA-DPA1	33,145,413	33,145,477
2950346	2950329	1.346	0.113	−1.166	0.419	HLA-DPA1	33,145,573	33,145,641
2950348	2950329	1.035	0.723	−1.166	0.419	HLA-DPA1	33,149,238	33,149,262

^†Exon fold change, or splicing index, is calculated as the ratio of mean gene-level normalized intensities between rejectors and non-rejectors. Negative values indicate exon skipping or repression, whereas positive values indicate exon enrichment. The p-values are calculated using a Student’s two-sample t -test on gene-level normalized intensities. Orange cells indicate p<.05.

^‡Transcript fold change is calculated as the ratio of the mean gene-level intensities between rejectors and non-rejectors. All exons within a transcript have identical gene-level intensities, but different exon-level intensities. The p-values are calculated using a Student’s two-sample t -test on gene-level intensities.

Appendix 4

Appendix 4

Top-ranked SNPs with p<0.01 from simple association testing with Rejectors (n=35) vs Non-Rejectors (n=42), all of whom are Caucasians

						MAF	MAF	MAF	REJ vs. NONREJ				REJ vs. CONTROLS				NONREJ vs. CONTROLS
SNP	rank	physical position	type	Closest gene	Distance to gene	Rejectors	Controls	Non-Rejectors	P-VALUE	OR	L95	U95	Chi-Sq P-value	OR	L95	U95	Chi-Sq P-value	OR	L95	U95

rs2975033	4	29930240	INTERGENIC	HLA-G	23348	36.36	25.62	14.86	0.0034	3.273	1.451	7.382	0.05727	1.659	0.9801	2.807	0.04007	0.5068	0.262	0.9804
rs12206499	10	30045106	INTERGENIC	HCG9	−5765	36.36	25.69	16.22	0.0065	2.952	1.332	6.544	0.05914	1.653	0.9765	2.797	0.07133	0.5598	0.2957	1.06
rs3823355	11	30050062	UPSTREAM	HCG9	−809	36.36	25.87	16.22	0.0065	2.952	1.332	6.544	0.06405	1.637	0.9675	2.77	0.06659	0.5545	0.2929	1.049
rs6904029	12	30051046	NON_SYNONYMOUS_CODING	HCG9	0	36.36	25.75	16.22	0.0065	2.952	1.332	6.544	0.06059	1.648	0.9738	2.788	0.06976	0.5581	0.2948	1.056
rs2394255	2	30057800	DOWNSTREAM	HCG9	3625	59.09	48.62	33.78	0.0027	2.831	1.423	5.631	0.1021	1.526	0.9166	2.542	0.01447	0.5391	0.3266	0.8901
rs3869070	6	30131847	INTRONIC	Q6ZU40_HUMAN	0	56.25	44.75	32.43	0.0049	2.679	1.339	5.358	0.0755	1.587	0.9502	2.652	0.04098	0.5926	0.3572	0.9832
rs9366752	1	30132656	3PRIME_UTR	C6orf12	0	33.33	20	12.16	0.0026	3.611	1.521	8.575	0.01061	2	1.165	3.433	0.1024	0.5538	0.27	1.136
rs6909253	15	30163622	INTERGENIC	RNF39	12015	54.55	37.25	32.43	0.0083	2.5	1.258	4.968	0.005529	2.021	1.22	3.35	0.4111	0.8086	0.4868	1.343
rs9261394	16	30172541	INTERGENIC	TRIM31	−6114	54.55	38	32.43	0.0083	2.5	1.258	4.968	0.008155	1.958	1.181	3.245	0.344	0.7832	0.4716	1.301
rs2523990	13	30185208	INTRONIC	TRIM31	0	60.61	47.38	37.84	0.0071	2.527	1.278	4.997	0.0387	1.709	1.023	2.854	0.1156	0.6762	0.4143	1.104
rs1345229	18	30290374	UPSTREAM	TRIM26	1191	15.15	11.87	2.703	0.0086	6.429	1.354	30.53	0.4332	1.325	0.654	2.685	0.01625	0.2061	0.04976	0.8539
rs9357097	7	30393100	INTERGENIC	TRIM39	−9500	15.15	30.95	36.11	0.0051	0.3159	0.1382	0.7224	0.006962	0.3984	0.1999	0.7937	0.3663	1.261	0.7619	2.087
rs1264583	19	30401462	UPSTREAM	TRIM39	−1138	12.12	5	1.351	0.0095	10.07	1.224	82.82	0.01509	2.621	1.172	5.86	0.1556	0.2603	0.03527	1.921
rs1264581	5	30405484	SYNONYMOUS_CODING	TRIM39	0	14.06	5.29	1.351	0.0041	11.95	1.47	97.1	0.004299	2.93	1.356	6.329	0.1354	0.2453	0.03327	1.808
rs3095150	17	31040511	INTERGENIC	DPCR1	10534	45.45	42.8	24.32	0.0086	2.593	1.263	5.32	0.6759	1.114	0.6723	1.844	0.002013	0.4295	0.248	0.744
rs6457699	3	33089625	UPSTREAM	HLA-DOA	4258	53.12	46.75	28.38	0.0031	2.86	1.414	5.786	0.3256	1.291	0.775	2.15	0.002381	0.4513	0.2672	0.7623
rs9276994	8	33092233	INTERGENIC	HLA-DOA	6866	46.88	37.75	24.32	0.0055	2.745	1.332	5.658	0.1487	1.455	0.8726	2.426	0.02181	0.53	0.3058	0.9186
rs6933994	14	33095098	INTERGENIC	HLA-DOA	9731	56.25	45.99	33.33	0.0072	2.571	1.282	5.156	0.7293	0.9134	0.5468	1.526	0.0007677	2.349	1.411	3.909
rs9296068	9	33096673	INTERGENIC	HLA-DOA	11306	46.88	36.75	24.32	0.0055	2.745	1.332	5.658	0.1074	1.519	0.9105	2.533	0.0328	0.5532	0.3191	0.959

Parental Allele test and Gamete Competition (GC) test for the same snps as above

The 5′UTR flanking region of HLA-DOA is represented by 4 SNPs, of which rs6457699 lies roughly 4kb from the first HLA-DOA exon, and also shows differences in parental allele comparisons, which approach significance at p=0.055 (yellow cell).

rs9296068 is more frequent in parents of Rejectors, compared with parents of Non-Rejectors. However, differences are not statistically significant, due to reduced power in the Rejector group, and among Rejector parents, as a result of removing seven non-Caucasians, and adding 5 caucasian trios.

Appendix 5

						Parental Allele test			MAF-Parents	MAF-Parents	GC test
SNP	rank	physical position	type	Closest gene	Distance to gene	Chi-Sq p-value	LCI	UCI	Rejectors	Non-Rejectors	P-VALUE

rs2975033	4	29930240	INTERGENIC	HLA-G	23348	0.08	−0.01	0.21	30.95	21.05	0.075
rs12206499	10	30045106	INTERGENIC	HCG9	−5765	0.22	−0.04	0.19	31.75	24.34	0.0126
rs3823355	11	30050062	UPSTREAM	HCG9	−809	0.19	−0.04	0.19	32.03	24.34	0.026
rs6904029	12	30051046	NON_SYNONYMOUS_CODING	HCG9	0	0.22	−0.04	0.19	31.75	24.34	0.026
rs2394255	2	30057800	DOWNSTREAM	HCG9	3625	0.42	−0.07	0.18	47.62	42.11	0.004
rs3869070	6	30131847	INTRONIC	Q6ZU40_HUMAN	0	0.46	−0.07	0.18	45.16	40.00	0.014
rs9366752	1	30132656	3PRIME_UTR	C6orf12	0	0.42	−0.06	0.16	26.98	22.08	0.006
rs6909253	15	30163622	INTERGENIC	RNF39	12015	0.16	−0.03	0.21	45.24	36.18	0.041
rs9261394	16	30172541	INTERGENIC	TRIM31	−6114	0.19	−0.04	0.21	45.97	37.33	0.039
rs2523990	13	30185208	INTRONIC	TRIM31	0	0.10	−0.02	0.23	54.10	43.33	0.076
rs1345229	18	30290374	UPSTREAM	TRIM26	1191	0.01	0.02	0.18	14.52	4.61	0.35
rs9357097	7	30393100	INTERGENIC	TRIM39	−9500	0.00	−0.29	−0.08	14.17	32.89	0.734
rs1264583	19	30401462	UPSTREAM	TRIM39	−1138	0.26	−0.03	0.11	8.73	4.67	0.138
rs1264581	5	30405484	SYNONYMOUS_CODING	TRIM39	0	0.25	−0.03	0.11	9.68	5.33	0.0959
rs3095150	17	31040511	INTERGENIC	DPCR1	10534	0.35	−0.06	0.18	39.52	33.33	0.008
rs6457699	3	33089625	UPSTREAM	HLA-DOA	4258	0.06	0.00	0.25	52.42	40.13	0.045
rs9276994	8	33092233	INTERGENIC	HLA-DOA	6866	0.08	−0.01	0.23	44.44	33.33	0.063
rs6933994	14	33095098	INTERGENIC	HLA-DOA	9731	0.12	−0.02	0.23	55.56	45.27	0.074
rs9296068	9	33096673	INTERGENIC	HLA-DOA	11306	0.40	−0.07	0.18	40.32	34.67	0.004

Footnotes

Presented in part at the American Transplant Congress, May 7, 2007, San Francisco, CA.

NCBI Data: http://www.ncbi.nlm.nih.gov/projects/geo/query/acc.cgi?acc=GSE11361 Accession No.: GSE11361

Conflict of interest: None. Study subjects informed.

This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

References

1. Martin SR, Atkison P, Anand R, Lindblad AS., SPLIT Research Group Studies of Pediatric Liver Transplantation 2002: patient and graft survival and rejection in pediatric recipients of a first liver transplant in the United States and Canada. Pediatr Transplant. 2004;8(3):273–83. [PubMed]

2. Jain A, Mazariegos G, Kashyap R, Kosmach-Park B, Starzl TE, Fung J, Reyes J. Pediatric liver transplantation. A single center experience spanning 20 years. Transplantation. 2002;73(6):941–7. 27. [PMC free article] [PubMed]

3. www.ncbi.nlm.nih/projects/SNP/snp_summary.cgi.

4. Ireland J, Carlton VEH, Falkowski M, Moorhead M, Tran K, Useche F, Hardenbol P, Erbilgin A, Fitzgerald R, Willis TD, Faham M. Large-scale characterization of public database SNPs causing non-synonymous changes in three ethnic groups. Human Genetics. 2006;119:75–83. [PubMed]

5. Hutchinson IV, Pravica V, Sinnot PJ. Genetic regulation of cytokine synthesis: consequences on acute and chronic organ allograft rejection. Graft. 1998;1:15.

6. Daly AK, Day CP, Donaldson PT. Polymorphisms in Immunoregulatory genes: Toward Individualized Immunosuppressive therapy? Am J Pharmacogenomics. 2002;2(1):13–23. [PubMed]

7. Hirschhorn JN, Lohmueller K, Byrne E, Hirschhorn K. A comprehensive review of genetic association studies. Genet Med. 2002;4:45–61. 2002. [PubMed]

8. Cardon LR, Palmer LJ. Population stratification and spurious allelic association. Lancet. 2003;361:598–604. 2003. [PubMed]

9. Spielman RS, McGinnis RE, Ewens WJ. Transmission test for linkage disequilibrium: the insulin gene region and insulin-dependent diabetes mellitus (IDDM) Am J Hum Genet. 1993;52:506–516. [PubMed]

10. OPTN Annual Report 2006.

11. Hull J, Campino S, Rowlands K, Chan MS, Copley RRR, Taylor MSS, Rockett K, Elvidge G, Keating B, Knight J, Kwiatkowski D. Identification of Common Genetic Variation That Modulates Alternative Splicing. PLoS Genetics. 2007;3(6):e99. 2007. [PubMed]

12. Chen Y, Zhu J, Lum PY, Yang X, Pinto S, MacNeil DJ, Zhang C, Lamb J, Edwards S, Sieberts SK, Leonardson A, Castellini LW, Wang S, Champy MF, Zhang B, Emilsson V, Doss S, Ghazalpour A, Horvath S, Drake TA, Lusis AJ, Schadt EE. Variations in DNA elucidate molecular networks that cause disease. Nature. (Online) doi:10.1038/nature06757. [PMC free article] [PubMed]

13. Sinsheimer JS, Blangero J, Lange K. Gamete-competition models. Am J Hum Genet. 2000;66:1168–1172. [PubMed]

14. Ionita-Laza I, McQueen MB, Laird NM, Lange C. Genome-Wide Weighted Hypothesis Testing in Family-Based Association Studies, with an application to a 100K scan. The American Journal of Human Genetics. 2007;volume 81:000. (2007) preprint, May 10, 2007. [PubMed]

15. Shih MC, Whittemore AS. Tests for genetic association using family data. Genet Epidemiol. 2002;22(2):128–45. [PubMed]

16. Van Steen K, McQueen MB, Herbert A, Raby B, Lyon H, Demeo DL, Murphy A, Su J, Datta S, Rosenow C, Christman M, Silverman EK, Laird NM, Weiss ST, Lange C. Genomic screening and replication using the same data set in family-based association testing. Nat Genet. 2005;37(7):683–91. [PubMed]

17. Sheldon S, Poulton K. HLA typing and its influence on organ transplantation. Methods Mol Biol. 2006;333:157–74. [PubMed]

18. Sindhi R, Magill A, Abdullah A, Seward J, Tresgaskes M, Bentlejewski C, Zeevi A. Enhanced Donor-specific Alloreactivity occurs independent of immunosuppression in children with early liver allograft rejection. Am J Transplant. 2005;5:96–102. [PubMed]

19. Denzin LK, Sant’Angelo DB, Hammond C, Surman MJ, Cresswell P. Negative regulation by HLA-DO of MHC class II-restricted antigen processing. Science. 1997;278(5335):106–9. [PubMed]

20. Roucard C, Thomas C, Pasquier MA, Trowsdale J, Sotto JJ, Neefjes J, van Ham M. In vivo and in vitro modulation of HLA-DM and HLA-DO is induced by B lymphocyte activation. Journal of Immunology. 2001;167(12):6849–58. [PubMed]

21. Shen R, Fan JB, Campbell D, Chang W, Chen J, Doucet D, Yeakley J, Bibikova M, Garcia Wickham E, McBride C, Steemers F, Garcia F, Kermani BG, Gunderson K, Oliphant A. High-throughput SNP genotyping on universal bead arrays. Mutation Research. 2002;573:70–82. [PubMed]

22. Purcell S, Neale B, Todd-Brown K, Thomas L, Ferreira MAR, Bender D, Maller J, de Bakker PIW, Daly MJ, Sham PC. PLINK: a toolset for whole-genome association and population-based linkage analysis. American Journal of Human Genetics. (in press) http://pngu.mgh.harvard.edu/purcell/plink/

23. Lange K, Cantor R, Horvath S, Perola M, Sabatti C, Sinsheimer J, Sobel E. Mendel version 4.0: A complete package for the exact genetic analysis of discrete traits in pedigree and population data sets. Amer J Hum Genetics. 2001;69(supplement):A1886.

24. Srinivasan K, Shiue L, Hayes JD, Centers R, Fitzwater S, Loewen R, Edmondson LR, Bryant J, Smith M, Rommelfanger C, Welch V, Clark TA, Sugnet CW, Howe KJ, Mandel-Gutfreund Y, Ares M., Jr Detection and measurement of alternative splicing using splicing-sensitive microarrays. Methods. 2005;37(4):345–59. [PubMed]

25. R Development Core Team R: A language and environment for statistical computing. R Foundation for Statistical Computing; Vienna, Austria: 2007. ISBN 3-900051-07-0, URL http://www.R-project.org.

26. The International HapMap Consortium A haplotype map of the human genome. Nature. 2005;437:1299–320. [PMC free article] [PubMed]

27. Zhao T, Chang LW, McLeod HL, Stormo GD. PromoLign: A Database for Upstream Region Analysis and SNPs. Human Mutation. 2004;23:534–539. [PubMed]

28. Ge D, Zhang D, Need AC, Martin O, Fellay J, Telenti A, Goldstein DB. WGAViewer: A Software for Genomic Annotation of Whole Genome Association Studies. Genome Res. 2008 In press. [PubMed]

29. Loiseau P, Esperou H, Busson M, Sghiri R, Tamouza R, Hilarius M, Raffoux C, Devergie A, Ribaud P, Socie G, Gluckman E, Charron D. DPB₁ disparities contribute to severe GVHD and reduced patient survival after unrelated donor bone marrow transplantation. Bone Marrow Transplantation. 2002;30(8):497–502. [PubMed]

30. Munkhabat B, Hagihara M, Sato T, Tsuchida F, Sato K, Shimazaki J, Tsubota K, Tsuji K. Association between HLA-DPB1 matching and 1-year rejection-free survival in high-risk corneal transplantation. Transplantation. 1997;63(7):1011–6. [PubMed]

31. Mytilineos J, Deufel A, Opelz G. Clinical relevance of HLA-DPB locus matching for cadaver kidney retransplants: a report of the Collaborative Transplant Study. Transplantation. 1997;63(9):1351–4. [PubMed]

32. Talukdar A, AshokKumar C, Farrar J, Wilson P, Janakiramanan A, Tregaskes M, Sindhi R. Lymphocyte subset reconstitution in pediatric liver recipients induced with steroid-free rabbit anti-human thymocyte globulin. Pediatric Transplantation. (In press) [PMC free article] [PubMed]

33. Sarwal M, Chua MS, Kambham N, Hsieh SC, Satterwhite T, Masek M, Salvatierra O., Jr. Molecular heterogeneity in acute renal allograft rejection identified by DNA microarray profiling. New England Journal of Medicine. 2003;349(2):125–38. [PubMed]