Expression profiling of major histocompatibility and natural killer complex genes reveals candidates for controlling risk of graft versus host disease

Abstract

Background: The major histocompatibility complex (MHC) is the most important genomic region that contributes to the risk of graft versus host disease (GVHD) after haematopoietic stem cell transplantation. Matching of MHC class I and II genes is essential for the success of transplantation. However, the MHC contains additional genes that also contribute to the risk of developing acute GVHD. It is difficult to identify these genes by genetic association studies alone due to linkage disequilibrium in this region. Therefore, we aimed to identify MHC genes and other genes involved in the pathophysiology of GVHD by mRNA expression profiling.

Methodology/principal findings: To reduce the complexity of the task, we used genetically well-defined rat inbred strains and a rat skin explant assay, an in-vitro-model of the graft versus host reaction (GVHR), to analyze the expression of MHC, natural killer complex (NKC), and other genes in cutaneous GVHR. We observed a statistically significant and strong up or down regulation of 11 MHC, 6 NKC, and 168 genes encoded in other genomic regions, i.e. 4.9%, 14.0%, and 2.6% of the tested genes respectively. The regulation of 7 selected MHC and 3 NKC genes was confirmed by quantitative real-time PCR and in independent skin explant assays. In addition, similar regulations of most of the selected genes were observed in GVHD-affected skin lesions of transplanted rats and in human skin explant assays.

Conclusions/significance: We identified rat and human MHC and NKC genes that are regulated during GVHR in skin explant assays and could therefore serve as biomarkers for GVHD. Several of the respective human genes, including HLA-DMB, C2, AIF1, SPR1, UBD, and OLR1, are polymorphic. These candidates may therefore contribute to the genetic risk of GVHD in patients.

Figure 1. Induction of a GVHR in BN rat skin explants exposed to PVG lymphocytes.

A summary of the histological GVHR grading of BN skin samples cultured in medium alone, together with syngeneic BN lymphocytes, and together with pre-stimulated allogeneic PVG lymphocytes (n = 12 in each group) is given. The samples represented by closed circles were used for both gene expression profiling and qRT-PCR experiments, whereas the other samples were only used for gene expression profiling. The pair-wise comparison (U test) indicated a significant difference between skin explant cultures with BN and PVG lymphocytes.

Figure 2. Expression profiling of BN skin explant samples exposed to allogeneic (PVG) lymphocytes in comparison to those exposed to syngeneic (BN) lymphocytes.

(A) The log2-fold changes in gene expression of significantly regulated MHC genes (p<0.05) are shown. (B) The log2-fold changes in gene expression of significantly regulated NKC genes (p<0.05) are shown. (C) The log2-fold changes in gene expression of 168 significantly (p<0.05) and strongly (log2-fold change ≥1 or ≤−1) regulated non-MHC and non-NKC genes indicate the range of observed alterations in gene expression levels among the 6342 tested genes. In panels A and B, black bars indicate a strong change (log2-fold change ≥1 or ≤−1), dotted bars alterations below this amplitude, and white bars expression changes that were not detected at a significant level with all, but at least with 50% of the probes present on the array for that gene. When more than one probe indicated a significant change of gene expression the means and standard deviations of the log2-fold changes are shown (see Tables S1, S2, and S3 for further details).

Figure 3. Verification of the regulation in gene expression observed in the microarray experiment by qRT-PCR.

A subgroup of 8 samples used for the microarray experiment (see Fig. 1 ) was analyzed by qRT-PCR for the expression of 10 MHC and 3 NKC genes. The ΔΔct value was calculated, i.e. the Δct (Gapdh – gene of interest) of the allogeneic skin explant samples minus Δct (Gapdh – gene of interest) of the corresponding control sample. The control sample was either a parallel skin explant exposed to syngeneic lymphocytes as in the microarray experiment (syngeneic control, black bars) or a parallel skin explant sample cultured in medium only (medium control, white bars). The means of the ΔΔct values plus standard errors of the mean (SEM) are shown. A positive value indicates an up-regulation of gene expression in the allogeneic samples.

Figure 4. Analysis of T cell infiltration in skin explants.

(A) Analysis of Cd3z gene expression in the same samples as shown in Fig. 3 . (B) Correlation of Cd3z and other gene expression levels (ΔΔct values for allogeneic skin explants minus syngeneic controls) in these samples. Pearson's correlation coefficients (r) and the p-values for the corresponding tests are given above the diagrams. In brackets Spearman's correlation coefficients (r) and the p-values for the corresponding tests are shown.

Figure 5. Induction of a GVHR in a second series of BN (filled circles) and LEW.1N (open circles) rat skin explants.

Skin explants were co-cultured with pre-stimulated allogeneic lymphocytes from rats with a minor (BN lymphocytes and LEW.1N skin), major (LEW.1A (RT1^a) or LEW.1AV1 (RT1^av1) lymphocytes and LEW.1N skin), or a minor and major histoincompatibility (PVG lymphocytes (RT1^c) and BN skin or LOU/C (RT1^u) lymphocytes and LEW.1N skin). A summary of the histological GVHR grading of skin samples cultured in medium alone, together with syngeneic BN or LEW.1N lymphocytes, and together with allogeneic lymphocytes is given.

Figure 6. Verification of gene regulations observed in the microarray experiment by qRT-PCR in an independent set of 17 skin explant assays.

Three samples were derived from skin explant assays with minor (upper panel), 5 with major (middle panel), and 9 with minor and major histoincompatibility (lower panel). The GVHR grading for these samples is shown in Fig. 5 . The expression of 7 MHC and 3 NKC was analyzed by qRT-PCR. The ΔΔct value, i.e. Δct (Gapdh – gene of interest) of the allogeneic skin explant samples minus mean of Δct (Gapdh – gene of interest) of the corresponding control samples (BN or LEW.1N, respectively), was calculated. The control samples were either skin explant samples exposed to syngeneic lymphocytes (syngeneic control) or skin explant samples cultured in medium only without added lymphocytes (medium control) and their GVHR grading is also shown in Fig. 5 . The means of the ΔΔct values plus SEM are shown. A positive value indicates an up-regulation of gene expression in the allogeneic samples.

Figure 7. Analysis of MHC and NKC gene regulation in skin explants exposed to pre-stimulated allogeneic lymphocytes depending on GVHR grading.

The expression of 7 MHC and 3 NKC was analyzed by qRT-PCR. The relative changes of gene expression levels were calculated using a mathematical model for relative quantification of real-time PCR data which also takes into account variations of the amplification efficiencies of different primer pairs . The means plus SEM are shown. A value >1 indicates an up-regulation of gene expression in the allogeneic samples. The control samples were either skin explant samples exposed to syngeneic lymphocytes (syngeneic control, upper panel), skin explant samples cultured in medium only (medium control, mean panel), or freshly frozen healthy skin samples (healthy skin control, lower panel).

Figure 8. Analysis of MHC and NKC gene regulation in GVHD skin lesions from transplanted animals.

BN (RT1ⁿ) rats were transplanted with bone marrow of PVG (RT1^c) rats. Rats that developed acute GVHD were scarified and skin lesions with signs of GVHD were obtained for RNA preparation and histology. The expression of 7 MHC and 3 NKC was analyzed by qRT-PCR using the B2m gene as reference. The relative changes of gene expression levels were calculated . The means plus SEM are shown for skin lesion with grade I and grade II GVHD. A value >1 indicates an up-regulation of gene expression in the allogeneic samples. The control samples were freshly frozen skin samples from healthy BN rats (n = 7).