Supplemental materials for: Hauri-Hohl et al, Vol 109, Issue 9, 4080-4088 -- Blood

Blood, Vol. 109, Issue 9, 4080-4088, May 1, 2007

Donor T-cell alloreactivity against host thymic epithelium limits T-cell development after bone marrow transplantation
Blood Hauri-Hohl et al. 109: 4080

Supplemental materials for: Hauri-Hohl et al, Vol 109, Issue 9, 4080-4088

Files in this Data Supplement:

  • Table S1. TEC injury in GVHD is associated with enhanced expression of IFN-γ-inducible genes (PDF, 25 KB) -
    Oligonucleotide microarray analysis was used to detect global gene expression in purified CD45−I-Ab+MTS24− TECs (pooled from 15 mice/experiment) at day 11 after GVHD induction, as described in Figure S4. Gene expression profiles in mice given allogeneic transplants (B6→B6D2F1; b→bd) were compared to those given syngeneic transplants (BDF1→BDF1; bd→bd; set as 1.0). The table shows the 15 genes (accession nos. from the National Center for Biotechnology Information (NCBI; http://www.ncbi.nlm.niah.gov) whose relative expression levels were 3.0. Another 37 genes were up-regulated between 2.5- and 3.0-fold and 123 genes were found to be down-regulated by 2.5-fold (data not shown). GeneChip Mouse Expression Set MOE430A (Affymetrix UK Ltd, High Wycombe, UK) were used and experiments were carried out in triplicate. aGVHD was found to activate specific genetic programs in TECs. The observed pattern of changes in TEC gene expression resembled that of another major target organ of GVHD, the liver, where alterations in the expression of IFN-–inducible genes were among the most profound changes caused by GVHD.1

  • Figure S1. TEC lines and primary TECs used for this study (PDF, 121 KB) -
    (A-C) TEC lines. (A) Unmanipulated cortical TEC1-2 and medullary TEC2-3 cell lines express very low cell surface levels of MHC class I and II but up-regulate this expression following 72 hours of exposure to rmIFN-. (B). Medullary TEC3-10 and cortical TEC1C9 cells were cultured for 72 hours with 103 U/mL IFN- and STAT-1 Tyr701 phosphorylation was detected in both lines by immunoblot analysis. (C). TEC2-3, TEC3-10 and TEC1C9 were cultured without (□) or with 103 U/mL rmIFN- (gray box). Apoptotic TECs were assessed as in Figure 2. The data expand information shown in Figure 2 and substantiate that there are no essential differences between cortical and medullary TEC lines. (D) Primary adult TECs. MHC class I (H-2Kb) and II (I-Ab) expression on thymic total stromal cell preparation from adult B6 mice (panels i-ii) differs from expression in TEC lines. Panel iii depicts MHC class I expression on gated CD45−I-Ab+ cells. Panels iv and v show the purified populations of primary CD45−I-Abint+high TEC (iv) and CD45+I-Ab+ hematopoietic pAPCs (v) as they were used for data shown in Figures 2A and 4D. The epithelial cells were isolated by flow cytometry to a purity of 99.97%. (E) Fetal TECs. Thymic total stromal cell preparations were generated from BDF1 embryos at day 13 (E13) of gestation (top) and were enriched by negative selection for CD45 and 6 days of 2-deoxyguanosine-treatment (purity of 99.9%; bottom). Identical results were obtained for purification of fetal TECs from STAT1+/+ and STAT-1−/− mice (E14) used for Figure 3. FI indicates fluorescence intensity. (F) T cell-deficient B6.nu/nu mice (H-2b) carrying a heterotopic thymus derived from E13 fetal BDF1 TECs (H-2bd) received mature T cells from BDF1 (i-v) and B6.CD45.1 (H-2b, vi-xi) donors, respectively. Macroscopic (i and vi), immunohistochemical (ii-v and vii-x) and flow cytometric analyses (xi) of ectopic thymi were performed 3 weeks later. The vasculature was analyzed by labeling endothelial cells with anti-CD31 mAb. ERTR7 was used to stain fibroblasts and reactivity to the UEA1 lectin was used as an additional marker to distinguish TECs. Two individual experiments were performed, with 4 animals per group and experiment, yielding comparable results and data from one experiment are shown. The photomicrographs were taken from one representative animal; sc indicates subcapsular area of heterotopic thymus; m, medullary area; ki, kidney. These grafting experiments established that placement of highly purified fetal TECs into allogeneic nude mice results in a regularly structured ectopic thymus, which supports regular T-cell development. Signs of surgery-related inflammation were not detected at 5 weeks after grafting (that is, the time point of alloreactive T-cell infusion). Both TEC composition and thymocyte development were unaffected in mice that received BDF1 T cells, corresponding to the fact neither the host lymphocytes nor the engrafted TECs were allogeneic to the infused T-cells (i-v). In contrast, the transfer of B6.CD45.1+ T cells resulted in an extensive infiltration of the grafted thymus tissue (xi), a decrease in its size as well as gross pathologic changes indicative of an intragraft immune response (vi-x). The ingrowing syngeneic vasculature and fibroblasts within the heterotopic thymus appeared vital and were thus unlikely targets of infused alloreactive T cells, limiting the alloreaction to the engrafted TECs. However, allogeneic TECs were both diminished in number and profoundly disorganized (viii-x). These extensive changes to the stromal compartment correlated with a lack in the support of regular thymocyte maturation (data not shown). Because the B6.CD45.1+ T cells were alloreactive to the transplanted TECs but nonresponsive to all other cells of the recipient, including nonepithelial thymic stromal cells, these results indicated that TECs were sufficient in vivo to induce thymic GVHD under transplant conditions where a MHC disparity was restricted exclusively to the thymic epithelial compartment.

  • Figure S2. Characterization of BM chimeras used for this study (PDF, 193 KB) -
    (B6→BDF1) BM chimeras were generated as described in the text and remained untreated, denoted (b⇒ bd), or received B6.CD45.1 T cells, denoted 45.1+→(b⇒ bd). (A) Donor/host chimerism was tested 4 months after generation of (b⇒ bd) mice. Populations from separate primary and secondary lymphoid cell compartments were analyzed by flow cytometry for the presence of residual recipient hematopoietic cells (H-2bd). Host (Kbd) and donor (Kb) cells were analyzed for their expression of the surface markers CD11b and CD11c. Results were compared with naïve, untransplanted B6 mice. Residual host-type DCs and macrophages (mφ) were characterized by a H-2Kb+H-2Kd+CD11c+CD11b+ and H-2Kb+H-2Kd+CD11b+CD11c− phenotype, respectively. Mice presenting a donor-type DC chimerism of 98% were selected for further transplantation experiments. An aggregate of 11 experiments was performed, with at least 10 mice each. FI indicates log fluorescence intensity. The data substantiate that it was very unlikely that residual host pAPCs were responsible for donor T-cell activation in 45.1+→(b⇒ bd) mice (Figure 5) since numbers and frequencies of host-type DCs and macrophages were very low or virtually undetectable in lymphoid organs of these chimeric mice. This high extent of donor chimerism contrasted with result of a previous study by Shlomchik and coworkers.2 These authors reported incomplete replacement of host APCs to be a possible cause for development of (mild and delayed) GVHD in the minor histocompatibility antigen-mismatched transplantation model used. (B) 45.1+→(b⇒ bd) mice developed thymic GVHD (Figures 5 and 6) but did not show signs of GVHD in the classical target organs, skin and liver. Sections of livers and skin from the back were stained with hematoxylin and eosin to detect typical GVHD-associated changes at 2 weeks after donor T-cell infusion. Control BM chimeric mice did not receive B6.CD45.1 T cells. As one positive control, a representative liver section of a mouse with aGVHD (b→bd) is given at the bottom of the figure. These data demonstrated that in this model aGVHD specifically targets the thymus. (C) Detection of residual radioresistant host-type CD11b+ and CD11c+ cells in the skin of (b⇒ bd) BM chimeras. These cells are shown in the flow cytometry plot on the left and the percentages are given in the bar graph (gray box). Despite the presence of host-type DC, no skin GVHD developed in these mice (data not shown). A cohort of chimeras was also exposed to UV light (Waldmann UV 3003 k source, t = 3.16 minutes, 0.4 J/cm2 back and front) to further deplete skin pAPC, as previously described by Merad et al.3 The data show that 4 weeks after UV treatment, host-type DCs were eliminated to a substantial degree (■) and were gradually replaced by donor BM-derived DCs. The right panels show donor T-cell infiltration and development of host CD4+CD8+ thymocytes in UV-treated mice (b⇒ bd) BM, which had received B6.CD45.1 T cells. There was no statistical difference to chimeric mice with aGVHD but not treated with UV light.

  • Figure S3. Comparison of thymic architecture in chimeric and nonchimeric mice with aGVHD (PDF, 277 KB) -
    Confocal microscopic analysis of thymic sections at 2 weeks after infusion of donor T cells was performed in 5 groups: syngeneically transplanted BDF1 mice (bd→bd, panels i-iii), allogeneically transplanted BDF1 mice (b→bd, panels iv-vi), (b⇒ bd; panels vii-ix) and 45.1+→(b⇒ bd) BM chimeras (panels x-xii) as described in Figures 5 and 6 and in Figure S2. Panels xiii-xv show data from “reverse” BM chimeric mice, that is, 45.1+→(bd⇒b) mice. Top two rows display thymic architecture (as assessed by staining with hematoxylin) at 2 microscope magnification levels. Bottom row: Analysis of TEC architecture. The use of antibodies to K5, K18, and MTS10 allowed the distinction of 4 TEC subpopulations: major cortical TEC (K5−K18+MTS10−, blue), minor cortical TEC (K5+K18+MTS10−, white), major medullary TEC (K5+K18−MTS10+, pale yellow), and K5+K18−MTS10− TEC (green). Four individual experiments were performed, with 2 to 4 animals per group and experiment. These experiments yielded comparable results, and therefore a representative photomicrograph of one mouse for each group is shown. Cells positive for K18 are typical for the cortical region, whereas K5+ cells are representative for the thymic medulla. These gross morphologic and immunohistochemical analyses further substantiated that total body irradiation to generate chimeric mice had not significantly altered the thymic architecture at the time point of donor T-cell transfer (hat is, at least 120 days after BMT to generate chimeras; panels vii-ix).

  • Figure S4. Caspase-12 is a putative mediator of TEC death during aGVHD (PDF, 333 KB) -
    (A) To determine gene expression profiles in TECs, pure populations of CD45−I-Ab+MTS24− cells were isolated (R4; from a total of 15 mice/group) by flow cytometry from mice undergoing allogeneic (b→bd) and syngeneic (bd→bd) transplantation (day 11 after GVHD induction), respectively. Following reverse transcription of 20 ng optimal quality RNA, the cDNA was subjected to 2 rounds of amplification by in vitro transcription with synthetic oligo(dT)24/T7 RNA polymerase promoter primer and T7 polymerase (Affymetrix Gene Chip Eukaryotic Small Sample Labeling Assay Version II, 2003). Biotin-labeled and fragmented cRNA was hybridized to DNA microarrays (GeneChip Mouse Expression Set MOE430A, 22,960 genes, Affymetrix). Microarrays were scanned using a GeneArray Scanner and Fluidics Station 400 (Affymetrix). Microarray data were normalized and modeled using both the perfect match-mismatch and perfect match only algorithms in dCHIP (http://www.dchip.org). Experiments were carried out in triplicate. The lower panel depicts data from this oligonucleotide microarray analysis as a comparison between allogeneically and syngeneically transplanted mice whereby bd→bd was set as 1.0 (dashed line). We found expression of the cysteinyl aspartate proteinase caspase-12, a gene associated with epithelial cell apoptosis in response to IFN-,4 to be up-regulated within 11 days after induction of acute GVHD. (B) Confirmation of caspase-12 mRNA up-regulation by quantitative real-time PCR analysis of primary CD45−I-Abint+high TECs from mice with GVHD (b→bd, day 11) and without GVHD (bd→bd). The latter were used as calibrator (1.0, dotted line). The following primers were used: caspase-12 forward GTTTATGTCCCATGGCATCC, caspase-12 reverse AGGCCTGCATGATGAGAATC; GAPDH forward ACCATGTAGTTGAGGTCAATGAAGG, GAPDH reverse GGTGAAGGTCGGTGTGAACG (Sigma-Genosys, Cambridge, United Kingdom). Data from 2 experiments are shown. (C) Expression of caspase-12 in thymic sections from naïve BDF1 mice (confocal microscopy panels i and iii) was compared to its expression in mice with acute GVHD (b→bd; panels ii and iv). Colocalization of TUNEL, K18, and caspase-12 denotes apoptotic TECs expressing caspase-12 protein (white color, iv, arrows). (D) Enhanced expression of caspase-12 protein in TEC1-2 in response to 500 U/mL rmIFN-. FI indicates fluorescence intensity. Exposure to exogenous IFN- for 72 hours also induced in TEC1-2 the cleavage of procaspase-12 to its catalytically active form, a step that is necessary for propagating the apoptotic cascade. The inset shows this cleavage which results in an approximate 42-kDa fragment (with 2 splice variants as described by Kalai et al4). These experiments were repeated 3 times. Since the expression levels of other caspases (ie, caspases-1, -2, -3, -4, -5, -6, -7, -8, -9, and -14) were unchanged in primary TECs challenged by acute GVHD (A), caspase-12 may therefore serve as a putative mediator for induction of TEC apoptosis to be inhibited by Q-VD-Oph (Figure 2D). A causative role for caspase-12 would be in keeping with previous data demonstrating its definitive role in the apoptosis of lung epithelium.5 It remains to be investigated, however, whether this pathway is definitely involved in TEC death.

    REFERENCES

    1. Ichiba T, Teshima T, Kuick R, et al. Early changes in gene expression profiles of hepatic GVHD uncovered by oligonucleotide microarrays. Blood. 2003;102:763-771.
    2. Shlomchik WD, Couzens MS, Tang CB, et al. Prevention of graft-versus-host disease by inactivation of host antigen-presenting cells. Science. 1999;285:412-415.
    3. Merad M, Hoffmann P, Ranheim E, et al. Depletion of host Langerhans cells before transplantation of donor alloreactive T cells prevents skin graft-versus-host disease. Nat Med. 2004;5:510-517.
    4. Kalai M, Lamkanfi M, Denecker G, et al. Regulation of the expression and processing of caspase-12. J Cell Biol. 2003;162:457-467.
    5. Bitko V, Barik S. An endoplasmic reticulum-specific stress-activated caspase (caspase-12) is implicated in the apoptosis of A549 epithelial cells by respiratory syncytial virus. J Cell Biochem. 2001;80:441-454.