Hybrid complexes of high and low molecular weight hyaluronan delay in vitro replicative senescence of mesenchymal stromal cells: a pilot study for future therapeutic application

Effects of HA formulations on cell cycle, apoptosis, and senescence

We incubated MSCs with media containing the different HA formulations for one, two, three and seven days. On these cells, we performed several biological assays. Here we present data following three days of treatment. These results overlapped those obtained at the other time points (data not shown).

Ki67 immunostaining is currently used to detect the percentage of cycling (G₁; S; G₂/M) and resting cells (G₀) that are Ki67 positive and negative, respectively. Nevertheless, cycling cells must have high translational levels in order to produce enough proteins for cell growth and proliferation. A key protein involved in the regulation of protein translation is the ribosomal protein S6 (RPS6), a component of the 40S ribosomal subunit. In cells with active protein production, RPS6 is phosphorylated via the mTOR pathway. The level of phosphorylated RPS6 (pRPS6) is elevated in cycling cells and is low or negligible in resting (G₀) cells. The combination of these two proteins allows the identification of cycling cells (Ki67(+) pRPS6(+) and quiescent cells (Ki67(-) pRPS6(-)) [11,12].

We evaluated the effect of the several HA formulations on MSC cultures by determining the Ki67 pRPS6 immunostaining in every tested condition. Cells were incubated for three days in media supplemented with the different HA formulations. In this way, we identified the actual proportion of cycling cells (Fig. 1A).

MSCs cultivated in L-HA either at 0.16% or 0.5% concentration showed a significant reduction in cycling cells compared with the controls, and this was accompanied by a concomitant increase in quiescent cells (Fig. 1A). No substantial modifications in the number of cycling and resting cells occurred in cultures treated with H-HA (Fig. 1A). Cells incubated with 0.16% or 0.5% of HCC showed a decline in the percentage of cycling cells and an increment of resting cells (Fig. 1A).

Flow cytometry analysis of MSC cultures evidenced that all the experimental conditions, but 0.16% HCC, modified the pattern of cell cycle profiles with respect to control conditions (Fig. 1B). For example, we detected a general increase in the percentage of G₂/M cells and reduction of S phase cells in 0.16% H-HA. The observed changes were not statistically significant; nevertheless, this may suggest that HA treatments may partially alter the in vitro MSC cycling process.

The apoptosis process was not affected by incubation with the different HA mixtures (Fig. 2A) while senescence was noticeably reduced (Fig. 2B), as detected by annexin V flow cytometry analysis and in situ beta-galactosidase assay, respectively. In particular, HCC treatment reduced more than three times the percentage of senescent cells. Ki67 and pRPS6 immunostaining confirmed the decline in senescence. Senescent cells have a high metabolic activity, are permanently arrested, and hence are Ki67(-) pRPS6(+). Indeed, in HCC treated cells, we observed a decline of Ki67(-) pRPS6(+) cells compared with the controls. In detail, in cells incubated with 0.16% HCC, we detected 18.4% Ki67(-) pRPS6(+) cells while in the control culture this value was 31.1%. Also, the treatment with 0.5% HCC reduced the number of Ki67(-) pRPS6(+) cells (20.4% versus 31.1%). It must be underlined that leaving the cell cycle and entering quiescence is a progressive process. For this reason, within a Ki67(-) pRPS6(+) cell population, some cells are true senescent cells (i.e., cell cycle arrested cells with a very active metabolism), others are becoming quiescent and still have a residual pRPS6 staining. Nevertheless, reduction of this cell population lends further credit to the beta-galactosidase data showing a decrease in senescence following treatment of MSCs with the HCC mixture.

We then evaluated the long-term effects of 0.16% HCC on MSCs since this was the experimental condition with the best biological performance (no alteration in cell cycle profile, no change in cell death percentage, and a huge reduction in senescence). Indeed, after 30 days of incubation in 0.16% HCC, the percentage of senescent cells was still low (9.0% +/- 1.6%) compared with controls (39.1% +/- 2.8%).

Stemness and differentiation potential

We evaluated how HA mixtures could affect stemness and multipotentiality of MSCs. To this end, we carried out a CFU assay on the MSC cultures to test their clonogenicity (i.e., their ability to expand at a single-cell level), which is a fundamental feature of self-renewing stem cells. We also evaluated if the tripotential differentiation capacity of MSCs was influenced by HA formulations. Adipo-chondro-osteo differentiation in the presence of each of the tested compounds allowed us to analyze this aspect [13]. The number of MSC clones was reduced after incubation with all the analyzed HA mixtures (Fig. 3). The reduction in CFU clones that we observed prompted us to evaluate if this decrement affected, in the same manner, all the different progenitors or if there was a bias toward a specific lineage-committed clone. MSCs, which we cultivated with the different HA formulations, were then incubated in osteogenic, adipogenic, and chondrogenic differentiating media. These three differentiation pathways were preserved in the presence of the HA mixtures as detected by Alizarin Red S, Oil Red O, and Alcian Blue stainings, respectively (Fig. 4A-B-C). To get further insights on the differentiation process, we analyzed by qRT-PCR the expression levels of several differentiation markers. The mRNA levels of PPAR-gamma (PPARG) and CEBP-alpha (CEBPA), two markers of adipogenesis, were upregulated in the presence of 0.16% HCC compared with the control differentiation cultures. In the other experimental conditions, no significant differences were observed, or in some cases, we evidenced a decline in CEBP-alpha levels (Fig. 4A).

Four markers of osteogenesis (SPP1, BGLAP, COL1A1, RUNX2) were significantly downregulated in all tested experimental conditions compared with controls (Fig. 4B).

A strong increase of COL11A1, a marker of chondrogenesis, was detected in 0.16% HCC and in cells incubated with 0.16% H-HA. Aggrecan (ACAN), another marker of chondrocyte differentiation, showed an increase in 0.16%, 0.5% HCC, and 0.5% H-HA (Fig. 4C). We also analyzed Collagen 2 (COL1A2), SOX9 and MMP-13 to further confirm effects HA complexes on chondrogenis. COL1A2 was upregulated in HCC (0.16% - 0.5%) and L-HA (0.5%). MMP13 showed a decline in all the tested experimental conditions. We barely detected SOX9 and this did not allow to perform quantification (Fig. 4C).

Production of hybrid cooperative complexes

HCCs = Hybrid cooperative complexes, obtained through the patented NAHYCO technology [15], are commercialized by IBSA. This technology is based on thermal procedures for the formation of hybrid cooperative complexes of hyaluronic acid, starting from an initial mixture of an equal amount (ratio 1:1) of H-HA (Mw = 1200±100 kDa Mw/Mn = 1.4) and L-HA (Mw = 100±10kDa, Mw/Mn = 1.4). The experimental concentration used was 32g/L: 32 mg H-HA + 32 mg L-HA in 2 mL volume, provided in prefilled syringes.

In our experiments, we used H-HA = pharmaceutical grade, highly purified linear hyaluronan of 1200 ±100 kDa MW (Altergon Italia, Italy) and L-HA = pharmaceutical grade highly purified linear hyaluronan of 100 ± 10 kDa MW (Altergon Italia, Italy).

Human MSC cultures

After obtaining bone marrow from healthy donors who had provided their informed consent, we separated cells on a Ficoll density gradient (GE Healthcare, Italy) and collected and washed the mononuclear cell fraction in phosphate-buffered saline (PBS). We seeded 1–2.5 × 10⁵ cells/cm² in modified Eagle’s medium (alpha-MEM) containing 10% fetal bovine serum (FBS) and basic fibroblast growth factor (bFGF). After 72 hr, we discarded non-adherent cells and cultivated adherent ones to confluency. We then further propagated cells for the assays reported below. Cells were used at passage 2 or 3.

Treatment with hyaluronan complexes

We prepared L-HA, H-HA and HCC mixtures and mixed them with culture media obtaining low viscous solutions. The final concentration of each hyaluronan mixture in the viscous solution was either 0.16% or 0.5%. We seeded 5 × 10³ cells/cm² in alpha-MEM containing 10%FBS and bFGF. After 24 hr, medium was replaced with the above indicated viscous solutions and cells were incubated for different time points.

Cell cycle analysis and immunodetection

For cell cycle analysis and immunostaining, cells were collected and fixed in 70% ethanol followed by PBS washes, and finally, they were dissolved in a hypotonic buffer containing propidium iodide. Samples were acquired on a Guava EasyCyte flow cytometer (Merck Millipore, Italy) and analyzed with a standard procedure using EasyCyte software. We used primary antibodies to detect Ki67 (sc23900, SantaCruz Biotech, CA, USA) and pRPS6 (MC27, Millipore, Italy). Cells were then incubated with corresponding secondary antibodies that were FITC or TRITC conjugated. For each sample, 5,000 cells were evaluated on the Guava instrument. For cell cycle analysis, cells were dissolved in a hypotonic buffer containing propidium iodide. Samples were acquired on a Guava EasyCyte flow cytometer (Merck Millipore, Milano, Italy) and analyzed with a standard procedure using EasyCyte software.

In vitro osteogenic differentiation and Alizarin Red S staining

Osteogenic differentiation was performed by culturing the MSCs with DMEM medium (EuroClone, Pero, Italy) supplemented with 10% FBS (EuroClone, Italy), 0.05mM ascorbic acid (Sigma-Aldrich, MO, USA), 10mM β-glycerophosphate (Sigma-Aldrich, Saint Louis, MO, USA), and 100nM dexamethasone (Sigma-Aldrich, MO, USA) for 21 days, with changes of medium every 3 days. Differentiation was performed either in presence or in absence of the above indicated hyaluronan complexes. Cultures were stained with Alizarin Red S (Sigma-Aldrich, MO, USA) to visualize calcium sediments that were acquired by microscope.

In vitro adipocyte differentiation and Oil Red Oil staining

MSCs were seeded at a density of 1.5x10⁴ in six-well plates and grown in a standard DMEM medium. At 70-80% confluence, the medium was replaced with an adipogenic induction medium composed of high-glucose DMEM (EuroClone, Italy) supplemented with 10% FBS, 1 mM dexamethasone (Sigma-Aldrich, MO, USA), 10 µg/mL insulin (Sigma-Aldrich, MO, USA), 0.5mM 3-isobutyl-1-methylxanthine (Sigma-Aldrich, Saint Louis, MO, USA), and 200 µM indomethacin (Sigma-Aldrich, Saint Louis, MO, USA). Cells were cultured in this medium for 21 days.Differentiation was performed either in presence or in absence of the above indicated hyaluronan complexes.

Adipogenic differentiation was confirmed on day 21 using an Oil Red O stain (Sigma-Aldrich, Saint Louis, MO, USA) as an indicator of intracellular lipid accumulation. Briefly, cells were washed twice in PBS, fixed with 4% formaldehyde for 10 min at room temperature, rinsed once with 3% isopropanol, and stained with Oil Red O staining solution. Then, cells were rinsed with water and photographed under the microscope.

In vitro chondrogenic differentiation and Safranin O staining

Chondrogenic differentiation was performed in 3D culture. Briefly, 1x10⁵ cells were seeded as a pellet in 96 round bottom multi-wells and cultured in achondrogenic medium composed of DMEM, 1% FBS, 50 nM ascorbate-2-phosphate (Sigma-Aldrich, MO, USA), 0.1 mM dexamethasone (Sigma-Aldrich, Saint Louis, MO, USA), and 10 ng/mL human transforming growth factor (hTGF)-β1 (Preprotech, UK). The medium was replaced every 3 days. Differentiation was performed either in presence or in absence of the above indicated hyaluronan complexes.

After 21 days, Safranin O staining was performed to detect glycosaminoglycan formation on the cell surfaces, (Sigma-Aldrich, MO, USA) In short, cell pellets were fixed in cold (4°C) acetone:methanol solution for 60 min and then included in cryostat embedding medium (Bioptica, Italy) for cryosectioning. Pellet sections were incubated at room temperature in 1% Safranin O solution for 30 min followed by three rinses in 3% acetic acid for 2 min each. After rinsing in deionized water for 2 min, the surfaces were allowed to dry for imaging.

Colony Forming Units assay (CFU)

MSC cultures were obtained as described above. Cultures were expanded to 70–80% confluence. On these cells, we carried out a CFU assay. Briefly, we plated 1,000 cells in 10 cm culture dish and incubated for 14 days in a growth medium. Subsequently, the medium was discarded, and colonies were fixed with 100% methanol for 10 minutes at -20°C. Colonies were then stained 0.01% (w/v) crystal violet (Sigma-Aldrich, MO, USA) in 25% methanol in PBS for 30–60 min. For every experimental condition, we counted the number of colonies in culture dishes at light microscope.

Annexin-V assay

Apoptotic cells were detected using Nexin V kit (Millipore, Italy) on a Guava EasyCyte flow cytometer, following the manufacturer’s instructions. The kit utilizes two separate dyes (Annexin V and 7AAD) to identify a broad spectrum of apoptotic and non-apoptotic cells. Annexin V (red) binds to phosphatidylserine on the external membrane of apoptotic cells, while 7AAD (blue) permeates and stains DNA of late-stage apoptotic and dead cells. Staining allows the identification of three cell populations: non-apoptotic cells (Annexin V- and 7AAD-); early apoptotic cells (annexin V+ and 7AAD-); and late-apoptotic or dead cells (Annexin V+ and 7AAD+). In our experimental conditions, early and late apoptotic cells were grouped.

In situ senescence-associated beta-galactosidase assay

Cells were fixed using a solution of 2% formaldehyde and 0.2% glutaraldehyde. After this, cells were washed with PBS and then incubated at 37°C for at least 2 hr with a staining solution (citric acid/phosphate buffer (pH 6), K₄Fe(CN)₆, K₃Fe(CN)₆, NaCl, MgCl₂, and X-Gal). The percentage of senescent cells was calculated by the number of blue, b-galactosidase-positive cells out of at least 500 cells in different microscope fields as already reported.

RNA extraction, RT-PCR, and real-time PCR

We extracted total RNA from cell cultures using Omnizol (EuroClone, Italy), according to the manufacturer’s protocol, and measured mRNA levels by RT-PCR amplification.

We used sequences of mRNAs from the Nucleotide Data Bank (National Center for Biotechnology Information, MD, USA) to design primer pairs for real-time RT-PCR reactions (Primer Express, Applied Biosystems, Italy); primer sequences are available upon request. We used appropriate regions of HPRT and/or GAPDH cDNA as controls and ran real-time PCR assays on an Opticon 4 machine (MJ Research, MA, USA). We carried out reactions according to the manufacturer’s instructions using a SYBR green PCR master mix and used the 2^-ΔΔCT method as a relative quantification strategy for quantitative real-time PCR data analysis.

Statistical analysis

We evaluated statistical significance using analysis of variance, followed by Student’s t and Bonferroni’s tests. For data with continuous outcomes, we used mixed-model variance analysis and, in any case, analyzed all data with GraphPad Prism version 5.01 (GraphPad, La Jolla, CA, USA).