Clinical significance of hyaluronan levels and its pro-osteogenic effect on mesenchymal stromal cells in myelodysplastic syndromes

Background Hyaluronan (HA), a major component of the extracellular matrix, has been proven to play a crucial role in tumor progression. However, it remains unknown whether HA exerts any effects in myelodysplastic syndromes (MDS). Methods A total of 82 patients with MDS and 28 healthy donors were investigated in this study. We firstly examined the bone marrow (BM) serum levels of HA in MDS by radioimmunoassay. Then we determined HA production and hyaluronan synthase (HAS) gene expression in BM mesenchymal stromal cells (MSC) and mononuclear cells derived from MDS patients. Finally, we investigated the effects of HA on osteogenic differentiation of MSC. Results The BM serum levels of HA was increased in higher-risk MDS patients compared to normal controls. Meanwhile, patients with high BM serum HA levels had significantly shorter median survival than those with low HA levels. Moreover, the HA levels secreted by MSC was elevated in MDS, especially in higher-risk MDS. In addition, HAS-2 mRNA expression was also up-regulated in higher-risk MDS-MSC. Furthermore, we found that MSC derived from MDS patients with high BM serum HA levels had better osteogenic differentiation potential. Moreover, MSC cultured in HA-coated surface presented enhanced osteogenic differentiation ability. Conclusions Our results show that elevated levels of BM serum HA are related to adverse clinical outcome in MDS. Better osteogenic differentiation of MSC induced by HA may be implicated in the pathogenesis of MDS.


Background
Myelodysplastic syndromes (MDS) represent a diverse group of myeloid clonal disorders characterized by ineffective hematopoiesis, one or more cytopenias and potential progression to acute myeloid leukemia (AML). The precise mechanisms leading to the development of MDS are incompletely understood, however, the bone marrow (BM) niche may play an important role in the development, progression and response to treatment of MDS [1,2]. The BM niche is mainly composed of mesenchymal stromal cells (MSC), endothelial cells, immune cells and other non-cell component such as various cytokines and extracellular matrix (ECM) [3].
Hyaluronan (HA), a major component of the ECM, is a member of the glycosaminoglycan polysaccharide family composed of repeating disaccharides of N-acetylglucosamine and glucuronic acid. HA is not only a structurally important molecule, but also has the potential to modify many cellular behaviors such as adhesion, proliferation, differentiation and migration [4,5]. The effects of HA on cell behaviors are mediated by the HA receptors such as CD44, RHAMM, LYVE-1, layilin and HARE [6,7]. In particular, HA production is increased in many malignant tissues and elevated HA levels have been shown to enhance tumor cell invasion, migration and proliferation [8,9]. In addition, HA can be released by many cell types,

Open Access
Journal of Translational Medicine both stromal cells and hematopoietic cells [10]. HA is expressed on human BM sinusoidal endothelium and endosteum, the regions where MSC are also abundant [11]. Recent research suggested that the HA could provide a protective niche for MSC, supporting the maintenance of their 'stemness' [12]. Several other reports supported a critical role of HA in the physiopathology of hematological malignancies, such as multiple myeloma (MM) [13,14], AML [15]. However, it is still unclear whether HA exerts any effects in MDS.
In the present study we analyzed BM serum levels of HA from MDS patients and normal controls by radioimmunoassay, and concluded that higher-risk MDS patients had high BM serum levels of HA. The patients with high BM serum HA levels were correlated with poor prognosis. The HA production and hyaluronan synthase 2 (HAS-2) gene expression were elevated in higher-risk MDS-MSC. Moreover, we demonstrated that HA could facilitated osteogenic differentiation of MSC. MDS with high BM serum HA levels had better osteogenic differentiation potential of MSC. Our findings supported an important regulatory role for HA in the pathophysiology of MDS.

Patients
A total of 82 patients with MDS and 28 healthy donors from our own center between Jun 2011 and March 2014 were investigated in this study. All patients were untreated when they were recruited into this study. MDS was diagnosed in accordance with the minimum diagnostic criteria established by the conference on MDS (Vienna, 2006) [16].

Measurement of the HA levels
The bone marrow serum and cell culture medium samples were centrifuged at 8000 g for 10 min. The supernatant was used to determine the total HA concentration. The HA levels were determined via radioimmunoassay [17].

Isolation and culture of MSC
Following the isolation by density centrifugation, the BM mononuclear cells (MNC) were seeded at a concentration of 1 × 10 6 cells/mL and cultured in Human Mesenchymal Stem Cell Growth Medium (Cyagen Biosciences Inc, Guangzhou, China) at 37 °C with 5% CO 2 in a humidified atmosphere as previously described [18]. The supernatant containing non-adherent cells was removed and medium was changed every 3 days. MSC used in all experiments were derived from passages 2-4. To fulfill the criteria of the International Society for Cellular Therapy, MSC were evaluated by flow cytometry for the absence of CD34, CD45 antigens and the presence of CD73, CD90, CD105 and CD166 [19].

Real-time quantitative polymerase chain reaction
RNA from MSC was purified using the RNeasy Mini Kit (QIAGEN, Germany) according to the manufacturer's instructions. cDNA was prepared using the First Strand cDNA Synthesis Kit (Fermentas, Burlington, Canada) following the manufacturer's protocol. PCR was performed on an ABI 7500 real-time PCR machine (Applied Biosystems). The primer sequences of runt related transcription factor 2 (RUNX-2), bone sialoprotein (BSP), alkaline phosphatase (ALP), Type1 collagen (COL-1), osteopontin (OPN), osteocalcin (OCN) and hyaluronan synthase 1/2/3 (HAS-1/2/3) are listed in Table 1. GAPDH served as reference control, and differences in mRNA expression levels were calculated as fold changes by the 2 −△△Ct method.

Osteogenic differentiation assay
MSC were seeded at 3 × 10 4 cells/well in 6-well plate in Human Mesenchymal Stem Cell Osteogenic Differentiation Medium (Cyagen Biosciences Inc, Guangzhou, China), and the medium was replaced every 3 days. After 3 weeks differentiation, cells can be fixed and stained with Alizarin red, visualized using light microscopy. The bound staining was eluted with 10% (wt/vol) cetylpyridinium chloride (sigma), and alizarin red-S in samples was quantified by measuring absorbance at 572 nm. before use. After the HA solution was applied to a 6-well plate surface, the coated substratum was kept to dried at 45 °C for 30 min. We used HA coated surfaces at a concentration of 30 μg/cm 2 .

Statistical analysis
All statistical analyses were performed using the Graph-Pad Prism 5.01. The statistical differences between groups were determined by the two-tailed unpaired Student's t test. Kaplan-Meier curves were used for analysis of overall survival (OS) and time to AML progression. The data were presented as mean ± SEM, p < 0.05 was considered statistically significant.

Patient general features
In total, 82 patients were enrolled in this study. The follow-up cutoff date was defined as the end of June 2015.
The median age was 58 years old (range 21-83 years), and the male-to-female ratio was 1.  Table 2).

Levels of HA were elevated in BM serum of higher-risk MDS patients
Firstly we analyzed BM serum levels of HA from 82 MDS patients and 28 normal controls by radioimmunoassay.  Fig. 1c). We also found that there was no obvious correlation between HA levels and percentage of BM blast cells (p = 0.45, r = 0.088).

Patients with high BM serum HA levels had significantly shorter median survival
According to the BM serum levels of HA, all the patients were divided to two groups: high HA (> 100 μg/L) and low HA (0 to 100 μg/L). As shown in Fig. 1d, the median survival of patients with high HA levels was 18 months and the median survival of patients with low HA levels did not reach. The median survival between the two groups was statistically different (HR: 2.149, 95% CI 1.125 to 4.105, p = 0.021). However, no significant impact was seen on time to AML progression (HR: 2.372, 95% CI 0.724 to 7.772, p = 0.154; Fig. 1e).

HA secreted by MSC was elevated in MDS, especially in higher-risk MDS
We next measured the HA levels in culture medium supernatants from MNC by radioimmunoassay. However, there was no difference between MDS patients and normal controls (Fig. 2a). We also examined HA levels in culture medium supernatants from MSC, we found that the HA levels were elevated in culture medium supernatants from MDS-MSC compared to normal controls (1121 ± 29.99 vs. 969.1 ± 54.12, p = 0.013), especially in higher-risk MDS-MSC (1159 ± 42.64 vs. 969.1 ± 54.12, p = 0.022; Fig. 2b). Furthermore, we determined the mRNA expression of HAS-1/2/3 in MDS-MNC. The results showed that the expression of HAS-1 in MDS-MNC was higher than normal controls (p = 0.027), and lower-risk MDS-MNC also exhibited significant higher HAS-1 expression than normal controls (p = 0.011; Fig. 2c). The expression of HAS-2 was up-regulated in higher-risk MDS-MNC compared to normal controls (p = 0.004) and lower-risk MDS-MNC (p = 0.002; Fig. 2d). No significant difference in HAS-3 expression was observed between the MNC of normal controls and MDS patients (Fig. 2e). We also investigated the mRNA expression of HAS-1/2/3 in MDS-MSC. HAS-2 mRNA expression was 2.3-fold greater in higher-risk MDS-MSC compared with normal MSC (p = 0.003), meanwhile, HAS-2 mRNA expression was significantly increased in higher-risk MDS-MSC compared to lower-risk MDS-MSC (p = 0.019; Fig. 2g). However, there was no difference between any groups in HAS-1 or HAS-3 (Fig. 2f, h).

MDS with high BM serum HA levels had better osteogenic differentiation function in MSC
In a next step we selected ten patients with high BM serum levels of HA and ten patients with low BM serum levels of HA,then investigate the osteogenic differentiation potential of MSC between two groups. After 21 days of osteogenic induction differentiation, relative calcium production by MSC from MDS patients with low BM serum levels of HA was reduced compared to MDS patients with high BM serum levels of HA (p = 0.022; Fig. 3a, b). In addition, mRNA expression of genes associated with osteogenic differentiation in MSC was investigated, the results showed that OPN mRNA expression was significantly increased in MSC from patients with high BM serum levels of HA (p = 0.010). However, the mRNA expression of the other genes (RUNX-2, BSP, ALP, COL-1, OCN) investigated were no obvious difference between two groups (Fig. 3c).

HA induced MSC osteogenic differentiation
Finally, to further study the involvement of HA in the regulation of osteogenic differentiation of MSC, we performed an osteogenic induction culture of MSC with low BM serum HA levels on surface with and without HA-coating, and evaluated the mineralization at day 7, 14, and 21, respectively. As shown in Fig. 4a, mineralization was evident from day 14, which was confirmed by alizarin red S staining. Compared with normal controls, the relative calcium production increased in MSC with HA-coated surface at day 14 and day 21 (p = 0.013, p = 0.002, respectively; Fig. 4b). Next we investigated the expression of genes associated with osteogenic differentiation in MSC following osteogenic induction at day 0, 7, 14, and 21. MSC with HA-coated demonstrated higher ALP gene expression compared with normal controls at each of the time points (day 7, p = 0.002; day 14,  16:234 p = 0.003; day 21, p = 0.008, respectively). The results also showed that the expression levels of COL-1 was increased in MSC with HA-coated at day 7 compared to normal controls (p = 0.046). Moreover, MSC with HAcoated showed the higher OPN expression than normal controls at day 21 (p = 0.002; Fig. 4c).

Discussion
There is increasing evidence that HA production is elevated in tumors and may play an important role in tumor progression [9,22,23]. However, the clinical implication of HA levels in MDS remains unclear. HA has been always detected in various body fluids, such as serum, lymph, urine, and pleural fluid [24]. To better understand the effect of HA in the BM niche of MDS, we examined the HA levels in bone marrow serum of MDS patients in this study. The main finding was that higher-risk MDS patients showed significantly elevated BM serum HA levels, as compared to normal controls. Earlier study evaluated the HA distribution in AML/MDS by histochemical stain,and showed that AML patients exhibited stronger HA staining, four of 8 MDS patients showed higher HA staining compared to normal controls [15]. An interesting observation in the present study is that an inverse correlation between OS and BM serum HA levels. However, BM serum HA levels were not associated with time to AML progression. In addition, the cell surface receptor, CD44 has been shown to be important in malignant cell adhesion, survival, migration, and invasion. Serum CD44 was reported slightly increased in MDS patients [25]. Another study also demonstrated that elevated serum CD44 levels were associated with shorter  16:234 survival in MDS patients [26]. These results indicate that HA-CD44 signals are activated in MDS, suggesting their critical role in the pathogenesis of MDS.
Elevated BM serum HA levels which may be caused by increased HA production by the malignant cells themselves or surrounding stromal cells. Moreover, HA biosynthesis is regulated by three transmembrane glycosyltransferase isoenzymes: HAS-1, HAS-2 and HAS-3 [27]. Therefore, we characterized HA production and HAS-1/2/3 gene expression in MNC and MSC. We examined HA levels in culture medium supernatants from MSC,we have shown that the levels of HA secreted by MSC was increased in MDS, especially in higherrisk MDS, up-regulated mRNA expression of HAS-2 in higher-risk MDS-MSC might explain the overproduction of HA. In addition, we also found the HA could secret by MNC, although the MDS-MNC exhibited significant higher mRNA expression of HAS-1/2,this did not affect the HA production in MNC from MDS patients. All these data indicate that MSC play a prominent role in the augmentation of HA levels in the BM serum of higherrisk MDS.
It is well known that lower-risk MDS and higher-risk MDS possess different biological characteristics, lowerrisk MDS present a trend to BM failure, while higherrisk MDS are more likely to convert into leukemia. In this study we found the obvious differences in the BM serum levels of HA between lower-risk and higher-risk MDS. Our previous study also showed the differences in osteogenic state between lower-risk and higher-risk MDS-MSC, the osteogenic differentiation potential of lower-risk MDS-MSC were impaired,but those of higherrisk MDS-MSC were relatively normal [18]. Thus we speculate that the distinct HA levels in lower-risk and