The oncometabolite R-2-hydroxyglutarate produced by mutant IDH dysregulates the differentiation of human mesenchymal stromal cells and induces DNA hypermethylation


 Background: Isocitrate dehydrogenase (IDH1/2) gene mutations are the most frequently observed mutations in cartilaginous tumors. The mutant IDH causes elevation in the levels of R-enantiomer of 2-hydroxylglutarate (R-2HG). Mesenchymal stromal cells (MSCs) are reasonable precursor cell candidates of cartilaginous tumors. This study aimed to investigate the effect of oncometabolite R-2HG on MSCs.Methods: Human bone marrow MSCs treated with or without R-2HG at concentrations 0.1 to 1.5mM were used for experiments. Cell Counting Kit-8 was used to detect the proliferation of MSCs. To determine the effects of R-2HG on MSC differentiation, cells were cultured in osteogenic, chondrogenic and adipogenic medium. Specific staining approaches were performed and differentiation-related genes were quantified. Furthermore, DNA methylation status was explored by Infinium 450 K arrays. Real-time PCR was applied to examine the signaling component mRNAs involved in.Results: R-2HG showed no influence on the proliferation of human MSCs. R-2HG blocked osteogenic differentiation, whereas promoted adipogenic differentiation of MSCs in a dose-dependent manner. In addition, R-2HG inhibited chondrogenic differentiation of MSCs, but increased the expression of genes related to chondrocyte hypertrophy in a lower concentration (1.0 mM). Moreover, R-2HG induced a pronounced DNA hypermethylation state of MSCs. Ingenuity pathway analysis showed Sonic Hedgehog (Shh) signaling as the most enriched signaling pathway. Further data indicated that R-2HG decreased the mRNA levels of Shh and Gli1, indicating Shh signaling inhibition.Conclusions: The oncometabolite R-2HG dysregulated the chondrogenic and osteogenic differentiation of MSCs possibly via induction of DNA hypermethylation, improving the role of R-2HG in cartilaginous tumor development.

prepared from MSCs treated either in the absence or presence of 1.0 mM R-2HG for 6 days using EZ DNA methylation Kit (Zymo Research, D5002, USA). A total amount of 500 ng of DNA was bisulfite converted and subsequently processed for hybridization onto an Infinium Human Methylation 450 Bead Array (Illumina, San Diego, CA, USA) under the manufacturer's instructions. This array can interrogate 27,578 CpG dinucleotides encompassing 14,495 genes. In brief, the DNA was mixed with bisulfite, and the nonmethylated C nucleotides were converted to U (T), whereas the methylated C nucleotides remained to be unaffected. Subsequently the bisulfite-treated DNA was amplified, fragmented, and hybridized to locus-specific oligonucleotides on the BeadArray. C or T nucleotides were detected by fluorescence signaling in order to obtain the single-nucleotide extension of the DNA fragments. The results were interpreted as a ratio (β value) of methylated signal (C) when compared with the sum of methylated and unmethylated signal (C-T) for each locus, where 0 was regarded as fully unmethylated DNA and 1 as fully methylated DNA.

Heat maps
The heat maps were designed by Mev software. The Euclidean distance within the two groups of samples was calculated using the average linkage measure [the mean of all pair-wise distances (linkages) between the members of the two concerned groups]. Gene annotation and enrichment analyses were performed by KEGG databases using the DAVID Bioinformatics Resources (http://david.abcc.ncifcrf.gov/) interfaces and WebGestalt (http://bioinfo.vanderbilt.edu/webgestalt/), respectively.

Gene pathway analysis
To determine the biological processes enriched within genes of differential methylation in the comparisons, we uploaded the gene lists into the Ingenuity Pathway Analysis (IPA; Ingenuity Systems, Redwood City, CA, USA). Each gene symbol was linked to its corresponding gene object in the Ingenuity Pathways Knowledge Base. Then the IPA integrates the genes and molecules that share part of the same biological functions or regulatory networks interacting together. The over-represented cellular and molecular functions were ranked according to the calculated P-value.

Statistical analysis
The results are expressed as mean ± standard error (SE), each performed in duplicates. Statistical analysis was performed by analysis of variance (ANOVA). All analyses used SPSS software (Paris, France). A p-value of < 0.05 was considered significant.

1.
R-2HG did not influence the proliferation and phenotype of human MSCs The effect of R-2HG on the proliferation of MSCs was examined by CCK-8 assay. As shown in Figure   1A, R-2HG showed no affect on the proliferation of MSCs at concentrations 0.1mM, 0.5 mM, 1mM or 1.5mM.
The expression of surface antigens of MSCs was analyzed using flow cytometry. As shown in Figure   1B, R-2HG had no influence on the immunophenotype of MSCs, shown as positive for CD105, CD90 and CD73 and negative for CD34, CD45 and HLA-DR.

R-2HG inhibits osteogenic differentiation of MSCs
Osteogenic differentiation in MSCs in the presence of R-2HG (1, 1.5mM) showed a dose dependent impaired calcification when compared to MSCs in the absence of R-2HG. Alizarin red staining revealed a low extent of mineralization with less detectable bone nodules in R-2HG treated MSCs when compared to those in controls ( Figure 2A). To further investigate the effects of R-2HG on MSC differentiation, we analyzed the relative mRNA expression levels of osteoblast-specific transcription factors (LPL and Osterix) and osteoblastic markers (IBSP and BGLAP). The results showed that R-2HG reduced the expression level of both early (LPL and IBSP) and late (Osterix and BGLAP) osteoblast differentiation-related genes significantly, which is consistent with the results in the functional assays ( Figure 2B).

R-2HG inhibits chongenetic differentiation of MSCs, but promotes the expression of genes related to chondrocyte hypertrophy.
To evaluate the effect of R-2HG on chongenetic differentiation property of MSCs, the cells were made into cell pellets of high-density and then were induced for chondrogenesis for 21 days. As is shown in Figure 3A, the physical dimension of the pellets in the presence of 1.5mM R-2HG showed marked decrease compared to those in the absence of R-2HG. Morphologically, matrix deposition as well as collagen 2a (COL2a) staining in cell pellets showed decreased growth in the presence of 1.0mM R-2HG ( Fig 3B). The pellets of MSCs in the presence of 1.5mM R-2HG failed to undergo immunohistochemistry. The expression of chondrogenic markers including Sox9 and Col2a demonstrated down-regulation in MSCs treated with R-2HG at 1.0mM and 1.5mM. However, hypertrophic markers including Runx2 and Col10a1 were up-regulated in 1.0mM R-2HG treated group, while down-regulated in 1.5mM treated group ( Fig 3C). These data confirmed that R-2HG suppresses chongenetic differentiation of human MSCs, but might promote the onset of chondrocyte hypertrophy in lower concentrations.

R-2HG promote the adipogenic differentiaion of MSCs
Next, the effect of R-2HG on adipocytic differentiation was evaluated. MSCs with 100% confluence were induced in adipogenic medium. As is shown in Figure 4A, R-2HG (at 1 and 1.5mM) promoted adipogenic differentiation of MSCs which is measured by increased lipid vacuoles (oil red O staining).
Furthermore, R-2HG enhanced the relative mRNA expression of adipocyte-specific transcription factors (C/EBPα and Pparg2) and the marker genes (adiponectin and aP2), supporting the above functional results ( Figure 4B).

R-2HG induced a pronounced DNA hypermethylation state of MSCs.
R-2HG affects histone modification and DNA methylation. DNA methylation is considered as a critical epigenetic modification that regulates the differentiation of stem cells, and so the changes in DNA methylation of MSCs exposed to R-2HG were explored.
As is shown in Figure 5A, R-2HG treated MSCs showed a profound DNA hypermethylation at CpG islands when compared with control. 154 differentially methylated CpGs between the two groups were identified after analyzing the data . A more detailed analysis of the differential distribution pattern of DNA methylation revealed wide-spread global changes, equally affecting all chromosomes ( Figure 5B). In R-2HG group , hypermethylation was found in 117 genes and hypomethylation in 37 genes (Supplemental Table S1). In addition, the most significantly hypermethylated genes in R-2HG treated samples included stem cell differentiation regulators such as GFI1, GEFT and RUNX1.In order to gain deep insights into the mechanism of aberrant DNA methylation, IPAwas performed. The data implicated several signalling pathways involved in MSC differentiation including the Sonic Hedgehog (Shh), insulin/insulin-like growth factor and Wnt signal pathways ( Figure 5C). Taken together, the oncometabolite R-2HG induced a pronounced DNA hypermethylation state of MSCs. 6.

R-2HG decreased the expression of Sonic Hedgehog signal components
As is shown in Figure 5C, the Shh signaling is considered as the most enriched pathway identified by IPA assay, and regulates cell differentiation. Shh signaling is initiated through the binding of Shh To confirm the results of IPA and elucidate the mechanism further, the expression of components of Shh signaling was investigated by RT-PCR. Our data showed that Shh ligand, a secreted glycoprotein that activates Shh pathway, decreased in a dose-dependent manner in MSCs treated by R-2HG ( Figure 6). In addition, GLI1, a key marker of Shh signaling, was down-regulated significantly in R-2HG These results revealed that the oncometabolite R-2HG induced by mutant IDH mutation blocked the osteogenic and chongenetic differentiation, while promoted the adipogenic differentiaion of MSCs.
The underlying mechanisms might be associated with hypermethylation of stem cell differentiation related genes, such as Shh signaling.

Discussion
Some metabolites play a critical role as regulators of some important enzymes in various biological pathways. According to recent studies, metabolic alterations promote the initiation and development of malignant cells. R-2HG that is produced by mutant IDH proteins is regarded as a prototype of these oncometabolites, and a serious of studies have proved the role of R-2HG in malignant transformation [13,25]. Elevated levels of R-2HG that are caused due to mutations in IDH1 and IDH2 are frequently shown (up to 87%) in enchondromas [4]. Impaired differentiation by R-2HG has been reported in central nervous system and during hematopoietic differentiation processes [13,25]. We therefore examined the effects of R-2HG on the characteristics, especially on the differentiation properties of human MSCs, which presumably act as precursors of cartilaginous tumors.
The results of the present study showed that R-2HG impaired the calcification of MSCs and reduced the expression of both early and late osteoblast differentiation-related genes in a dose-dependent inhibitor, enasidenib (AG-221), in patients with relapsed or refractory IDH2-mutated AML has been approved by FDA [41]. The development of IDH inhibitors is an emerging treatment option for patients with chondrosarcoma.

Conclusion
In conclusion, the present study results showed that R-2HG impaired osteogenic and chondrogenic  The expression of Sonic Hedgehog signaling in MSCs exposed to R-2HG. After 6 days of treatment, quantitative RT-PCR was performed to evaluate Gli1, Gli2, Gli3, Shh, Smo and Ptch1. Data are presented as mean ± S.D. and performed in triplicate from an experiment representative of three independent experiments. * P<0.05 vs. the group of MSCs in absence of R-2HG. ** P<0.01 vs. the group of MSCs in absence of R-2HG.

Supplementary Files
This is a list of supplementary files associated with this preprint. Click to download.