Characterization of Gaucher disease bone marrow mesenchymal stromal cells reveals an altered inflammatory secretome

Abstract

Gaucher disease causes pathologic skeletal changes that are not fully explained. Considering the important role of mesenchymal stromal cells (MSCs) in bone structural development and maintenance, we analyzed the cellular biochemistry of MSCs from an adult patient with Gaucher disease type 1 (N370S/L444P mutations). Gaucher MSCs possessed a low glucocerebrosidase activity and consequently had a 3-fold increase in cellular glucosylceramide. Gaucher MSCs have a typical MSC marker phenotype, normal osteocytic and adipocytic differentiation, growth, exogenous lactosylceramide trafficking, cholesterol content, lysosomal morphology, and total lysosomal content, and a marked increase in COX-2, prostaglandin E2, interleukin-8, and CCL2 production compared with normal controls. Transcriptome analysis on normal MSCs treated with the glucocerebrosidase inhibitor conduritol B epoxide showed an up-regulation of an array of inflammatory mediators, including CCL2, and other differentially regulated pathways. These cells also showed a decrease in sphingosine-1-phosphate. In conclusion, Gaucher disease MSCs display an altered secretome that could contribute to skeletal disease and immune disease manifestations in a manner distinct and additive to Gaucher macrophages themselves.

Figure 1

MSC characterization. (A) Cells were analyzed for surface expression of MSC markers CD44, CD73, CD90, and CD105; endothelial marker CD31; hematopoietic markers CD34 and CD45; and phagocytic lineage markers CD11b, CD115, and CD206 by flow cytometry analysis (leftmost curves correspond to isotypic controls). (B) Photograph of differentiated MSCs grown in osteocytic differentiation medium and adipocytic differentiation medium for 4 weeks, stained, respectively, with Alizarin Red (left) and Oil Red O (right); scale bar = 50 μm. (C) Relative mRNA expression of osteoblastic differentiation markers runt-related transcription factor 2 (RUNX2), bone-specific alkaline phosphatase (ALPL), and osteocalcin (OC). Expression levels in normal MSCs at 4 weeks were used as a reference value. Error bars represent 1 SD. (D) Flow cytometry analysis of passage 3 MSCs stained with CFSE and allowed to proliferate for 7 days before flow cytometry analysis. (E) Scanned autoradiogram of a Western blot performed for APE-1 expression analysis in passage 3 MSCs, with β-tubulin expression used as a loading control.

Figure 2

Lysosomal and GBA studies. (A) Transmission electron microscope photography of passage 5 MSCs, the lysosomes being the dark gray organelles; scale bar = 1 μm. (B) Total lysosomal content assessed by flow cytometry of Lyso Tracker Yellow–stained passage 3 MSCs. (C) Relative mRNA levels for GBA in passage 4 MSCs. (D) Scanned autoradiogram of a Western blot performed for GBA expression analysis on passage 4 MSCs, with β-tubulin expression used as a loading control on the same samples. The large band for GBA is explained by differently glycosylated forms of GBA. (E) GBA enzymatic activity expressed in nanomoles of 4-MU formed per hour per milligrams of total protein content in the sample. **P < .001. (F-G) Sphingolipids were analyzed in normal and Gaucher disease MSCs by high-performance liquid chromatography–tandem mass spectrometry and were normalized for total cellular inorganic phosphate. Shown are glucosylceramide levels (F) and sphingosine-1-phosphate (G). *P < .05. All error bars represent 1 SD.

Figure 3

Cholesterol content and lactosylceramide trafficking. (A) MSCs stained with filipin and observed under a fluorescence microscope; scale bar = 20 μm. (B) Total cholesterol and cholesteryl ester content expressed as nanograms per milligrams of total protein content in the samples. (C) Lysosomes (stained with Lyso Tracker Red) and BODIPY-lactosylceramide (green) after 2 hours of trafficking, with corresponding signal overlap in white, and correlation coefficients in the graph (Manders M1 corresponds to the correlation of the green signal on the red signal, Manders M2 is red on green); scale bar = 20 μm. All error bars represent 1 SD.

Figure 4

Inflammatory signal screening. (A) RT-PCR array results of genes encoding common cytokines. The log10 of the cycle threshold (Ct) were compared in a scatter plot, where circles in the top left represent genes with higher expression in Gaucher disease MSCs and circles in the bottom right represent genes with higher expression in normal MSCs. Cytokines not detected in either sample are not shown. (B) RT-PCR array results of genes encoding inflammatory cytokines. (C) Autoradiogram of relevant portions of “Human Cytokine Antibody Array 8” membranes, which screen 172 cytokines and cytokine receptors. (D) Relative mRNA levels of various genes important for MSC interactions with surrounding cells, in Gaucher disease MSCs compared with normal MSCs. (E) Scanned autoradiogram of a Western blot performed for COX-2 expression analysis on passage 4 MSCs, with β-tubulin expression used as a loading control on the same samples. (F-H) Secretion levels of IL-8, PGE2, and CCL2 by MSCs of different passages. *P < .05. All error bars represent 1 SD.

Figure 5

Reproduction of findings in chemical model. (A) Glucosylceramide levels in untreated and CBE-treated normal hMSCs. (B) S1P levels in untreated and CBE-treated normal hMSCs. (C) Glucosylceramide levels in CBE-treated human and murine MSCs and 293T cells, expressed as molar ratios compared with untreated cells. (D) S1P levels in CBE-treated human and murine MSCs and 293T cells, expressed as molar ratios compared with untreated cells. (E) Scatter plot comparing log2 intensities of detected genes in arrays of untreated and CBE-treated hMSCs, to show the quality of the assay. (F) Relative mRNA levels of selected genes in CBE-treated normal MSCs compared with untreated normal MSCs. *P < .05. All error bars represent 1 SD.