CXCL4 Contributes to the Pathogenesis of Chronic Liver Allograft Dysfunction

Abstract

Chronic liver allograft dysfunction (CLAD) remains the most common cause of patient morbidity and allograft loss in liver transplant patients. However, the pathogenesis of CLAD has not been completely elucidated. By establishing rat CLAD models, in this study, we identified the informative CLAD-associated genes using isobaric tags for relative and absolute quantification (iTRAQ) proteomics analysis and validated these results in recipient rat liver allografts. CXCL4, CXCR3, EGFR, JAK2, STAT3, and Collagen IV were associated with CLAD pathogenesis. We validated that CXCL4 is upstream of these informative genes in the isolated hepatic stellate cells (HSC). Blocking CXCL4 protects against CLAD by reducing liver fibrosis. Therefore, our results indicated that therapeutic approaches that neutralize CXCL4, a newly identified target of fibrosis, may represent a novel strategy for preventing and treating CLAD after liver transplantation.

Figure 1

CXCL4 is associated with CLAD in liver transplant patients. (a) The serum concentration of CXCL4 was significantly increased in patients with CLAD (67.3 ± 3.1 ng/mL) compared to healthy controls (25.1 ± 2.6 ng/mL, P = 0.026). (b) CXCL4 mRNA was significantly increased in subjects with CLAD versus individuals without CLAD (P = 0.019).

Figure 2

The characteristics of rat CLAD models. CLAD, allogeneic BN-to-Lewis rat liver transplantation with low-dose Tacrolimus Induction. Control, BN-to-BN rat liver transplantation with low-dose Tacrolimus Induction. (a) Recipient survival in different rat orthotopic liver transplantation models. The survival time was shorter in the CLAD group (62.8 ± 12.6 days) than in the control group (105.1 ± 19.3 days, P = 0.003). (b, c) Liver allograft function was evaluated by serum levels of AST and TBIL detected on POD 60. All CLAD liver grafts had moderate to severe dysfunction: AST 441.6 ± 101.6 mg/dL, TBIL 236.5 ± 91.6 mg/dL, but no apparent alterations were observed in Control group (AST 62.76 ± 15.6 mg/dL, P < 0.001; TBIL 43.3 ± 10.2 mg/dL, P < 0.001). (d–g) The pathologic characteristics of low-dose Tacrolimus-induced rat CLAD liver grafts at POD 60. HE stain ((d, e) ×200), Masson stain ((f, g) ×200). The pathology of the CLAD grafts was characterized by the loss of portal vein and loss of bile ducts in portal areas, intimal fibrous thickening without elastosis in hepatic arteries, and intraluminal fibrosis of the central veins with perivenular fibrosis, indicating the development of CLAD. No apparent histologic alterations were observed in the control group.

Figure 3

Quantitative proteomic analysis CLAD liver allografts. iTRAQ quantitative proteomics analysis was performed using protein samples from BN-to-BN liver allografts (control) and BN-to-Lewis liver allografts (CLAD). The proteins were analyzed with 95% or greater confidence as determined by ProteinPilot Unused scores (≤1.3). The following parameters were applied: an unpaired t-test with P value < 0.05, a false discovery rate (FDR) of all peptide and protein identifications <1%, and max missed cleavages and an enrichment score = 2. There was a ≥1.50-fold change for upregulation or ≤0.67-fold change for downregulation. (a) GO term analyses were performed using the online tool from DAVID Bioinformatics Resources (version 6.7). Both gene numbers and P values of specific gene groups were listed. (b) The top ranked 26 CLAD-associated proteins identified by functional annotation clustering/iTRAQ quantitative proteomic analysis. (c) The identified CLAD-associated proteins of liver allografts were categorized by subcellular distributions and functions based on PANTHER and Uniprot KB sources.

Figure 4

Validating CLAD-associated proteins in recipient rat liver allografts. (a) Immunohistochemistry analysis showed that CXCL4 detected on POD 60 was significantly expressed in CLAD liver allografts, compared with the control. (b) Validation by Western blot analysis on POD 60 showed that CXCL4, CXCR3, EGFR, JAK2, STAT3, and Collagen IV were significantly overrepresented in all CLAD liver grafts as compared with the control.

Figure 5

CXCL4 shRNA reduces the expression of fibrogenesis-associated proteins in isolated HSCs. (a) The isolated HSCs were transfected with CXCL4 shRNA for 72 hours. Western blot analysis of HSCs with antibodies against CXCR3, EGFR, JAK2, STAT3, and Collagen IV. (b) Quantitative presentation of these fibrogenesis-associated proteins. The CXCR3, EGFR, JAK2, STAT3, and Collagen IV levels in the CXCL4 shRNA-treated group decreased as a percent of their respective controls (untreated group). The data are presented as the mean ± SD, n = 3 liver grafts per group. ^∗ P < 0.05 versus control group.

Figure 6

Blockade of CXCL4 protects recipient rat livers from CLAD. For the blockade experiment, either CXCL4mab (1 mg/kg) or physiological saline (control) was delivered to recipient rats (BN-to-Lewis) through the tail vein once every week from POD 5 for 2 months. (a) Survival time was more prolonged among recipient rats in the CXCL4mab-treated group than in the physiological saline group (P < 0.001). (b) Liver allograft function was evaluated by serum levels of AST and TBIL detected on POD 60. The AST levels in the CXCL4mab-treated group were significantly lower (105.1 ± 20.3 mg/dL) than those in the physiological saline group (468.6 ± 133.5 mg/dL, P < 0.001). A similar trend was observed for TBIL levels between the CXCL4mab-treated group (71.5 ± 19.7 mg/dL) and the physiological saline group (260.7 ± 84.6 mg/dL, P < 0.001). (c, d) Liver fibrosis of CLAD graft recipient rats was detected on POD 60 by Masson stain. CXCL4mab significantly alleviates recipient rat liver fibrosis after transplantation when compared with those in the physiological saline group.