Portal de Periódicos da CAPES

Anomalies in conventional T and invariant natural killer T‐cell populations in Fabry mice but not in Fabry patients

2008; Wiley; Volume: 143; Issue: 4 Linguagem: Inglês

10.1111/j.1365-2141.2008.07380.x

1365-2141

Andrea Balreira, M. Fátima Macedo, Cristina Girão, Lorena G. Rodrigues, João Paulo Oliveira, Clara Sá-Miranda, Fernando A. Arosa,

Lysosomal Storage Disorders Research

Fabry disease (FD), an X-linked sphingolipidosis, results from the defective activity of α-galactosidase A, leading to lysosomal accumulation of globotriaosylceramide and digalactosylceramide primarily in vascular endothelium, smooth-muscle cells and epithelial cells. A mouse model of FD showed accumulation of globotriaosylceramide in several organs, namely in kidneys and liver (Ohshima et al, 1997) and has been used recently to reveal a reduction in CD1d-restricted invariant natural killer T (iNKT) cells in liver, spleen and thymus (Gadola et al, 2006). However, whether lipid accumulation is associated with anomalies of conventional T cells remains uncertain. This knowledge may help understanding how the immunological system deals with lipid overload and may give clues to develop novel immunotherapies (Cox, 2003). Thus, endogenous glycosphingolipids are thought to bind to CD1d and be presented to human and mice iNKT cells endowed with highly regulatory properties (Bendelac et al, 2007). In this context, we have previously shown that accumulation of glucosylceramide in Gaucher disease patients results in overt anomalies of the immunological system, namely in T and iNKT lymphocyte populations akin to upregulated expression of CD1d and major histocompatibility complex (MHC) class II molecules on monocytes (Lacerda et al, 1999; Balreira et al, 2005). We performed a comprehensive analysis of T and iNKT cell populations in Fabry mice and FD patients. All but one FD patients were undergoing enzyme replacement therapy (ERT) with recombinant α-galactosidase A (Fabrazyme®, Genzyme, Cambridge, MA, USA), ranging between 5 and 58 weeks of therapy. T and iNKT cells were analysed in peripheral blood samples from FD patients and healthy controls upon consent. Hepatic and splenic T and iNKT cells were analysed in Fabry mouse and C57BL/6J mouse males, killed at 12 and 24 weeks of age. Liver and spleen mononuclear cells were isolated following a previously described protocol with minor modifications (Goossens et al, 1990). Human and mice cells were stained as described (Balreira et al, 2005) and immediately acquired without fixation in a FACScalibur (Becton Dickinson, Mountain View, CA, USA). For each sample, 50 000 events were acquired and analyzed using the flowjo software (TreeStar, Inc, Ashland, OR, USA). P-values <0·05 were considered to be significant. In Fabry mice, liver CD3+CD4+ T cells were significantly decreased while CD3+CD8+ T cells were increased, and these differences were observed both at 12 and 24 weeks of age (Fig 1A), leading to a twofold decrease in the CD4+/CD8+ T-cell ratios between Fabry and control mice in the liver. Less obvious imbalances in CD3+CD4+ and CD3+CD8+ T cells were observed in the spleen of Fabry mice (Fig 1B). Analysis of Vα14+ T cells in spleen and liver by using PBS57-loaded CD1d tetramers (kindly provided by the National Institutes of Health tetramer core facility) revealed low numbers of CD3+Vα14+ T cells in the liver of Fabry mice when compared to control mice due mostly to low numbers of CD4+Vα14+ T cells (Fig 1C). The anomalies in liver Vα14+ T cells were maintained after correction to the percentage of CD4+ T cells and were not observed in the spleen (data not shown). Surprisingly, a similar analysis of T and iNKT cell populations in peripheral blood samples of FD patients revealed a complete absence of anomalies when compared to healthy controls (Fig 1D). The only anomaly found in FD patients was a statistically significant up-regulation of cell surface MHC-class II molecules on peripheral blood monocytes (Fig 1E). Finally, analysis of the percentages of CD3int and CD3high T cells in the liver revealed a marked decrease in CD4+CD3int T cells in Fabry mice when compared to control mice (Fig 2A), that resulted in a two-fold decrease in the CD3int/CD3high T-cell ratio (Fig 2B). Imbalances in conventional T and invariant natural killer T cells in Fabry mice. (A and B) Results show the percentage of total CD4+ and CD8+ T cells in liver (A) and spleen (B) in each age group of mice studied. Each point represents a different individual. (C) Results show the percentage of liver CD4+ T cells within Vα14+CD3+ T cells for each group. Each point represents a different individual. (D) Results show the percentage of peripheral blood T-cell subsets in healthy controls and FD patients. Values represent mean ± SD for each group studied. (E) Expression of MHC class II expression in Fabry patient monocytes. Histograms show mean fluorescence intensity (MFI) values for MHC class II expression from a representative control (thick line) and a FD patient (shaded area). Background staining with mouse immunoglobulins is included (thin line). Statistical significant differences (t-test) are indicated. Anomalies in hepatic CD3int T cells from Fabry mice. (A) Histograms show MFI values of CD3 molecules in liver T cells from a representative 12-weeks old C57BL/6J and Fabry mice. CD3 cell surface expression was analyzed in total liver lymphocytes, and then in CD3+CD4+ T cells and CD3+CD8+ T cells. Dotted lines define CD3int from CD3high populations. A decrease in the percentage of CD3int can be clearly observed within CD4+ T cells in Fabry mice. (B) Results show the ratio (mean ± SD) of liver CD3int/CD3high T cells for both age groups. Statistical significant differences (t-test) are indicated. n.s., not significant; w, weeks. Overall, we showed that Fabry mice displayed overt anomalies in liver T and iNKT cell populations, which is in agreement with previous studies (Gadola et al, 2006). However, in our study we did not find significant differences in splenic iNKT cells and the reduction in liver iNKT cells was not as dramatic as that described by Gadola et al (2006). These differences may probably arise from the use of different CD1d tetramers (PBS57 vs. α-galactosylceramide) and/or the age of the mice under study. Thus, as shown here, some differences tend to disappear with age. Importantly, we showed that the low numbers of hepatic T lymphocytes in Fabry mice is mostly due to overt anomalies in the percentage of CD3int T cells (Fig 2B). These T cells are abundant in the liver of mice and encompass both CD1d-restricted and CD1d-unrestricted T cells with regulatory properties (Abo, 2001). These results suggest that globotriaosylceramide accumulation in mice impact on both conventional T cells and CD1d-restricted iNKT cells, negatively affecting CD4+ T cells in liver and spleen. The observed increase in CD8+ T cells may result from homeostatic changes or from globotriaosylceramide-induced CD8+ T cell increase or both. In any case, these imbalances may compromise CD4+ T-cell-mediated responses under pathological situations. Notably, no anomalies in T and iNKT cells were observed in Fabry patients; MHC class II molecules upregulation in monocytes was the only anomaly detected. MHC-class II upregulation has been associated with disease-associated inflammatory status (Jeyakumar et al, 2003; Balreira et al, 2005), which can be associated with the accumulation of sphingolipids in class II-enriched endolysosomal compartments. The lack of overt T-cell anomalies in Fabry patients was not related to the fact that they were undergoing ERT, as no significant changes in the percentage of T and iNKT cells were observed between patients at different times after ERT (data not shown). The absence in humans of equivalent anomalies seen in mice may be the result of the marked differences between the innate and adaptive arms of the immunological system of these two species (Mestas & Hughes, 2004) and should be taken into consideration. The authors would like thank the collaboration of all subjects included in this study and to acknowledge the NIH Tetramer Facility for CD1d tetramers complexed with PBS57. We also like to thank Professor R.O. Brady for kindly providing the Fabry disease mice. AB and LGR were funded by fellowships from Fundação para a Ciência e Tecnologia (SFRH/BD/19496/2004 and SFRH/BPD/25333/2005 respectively). This work is part of the PhD thesis of AB and was supported by a grant from Genzyme.