Expansion and characterization of bone marrow derived human mesenchymal stromal cells in serum-free conditions

Bone marrow-derived mesenchymal stromal cells (BM-MSCs) are gaining increasing importance in the field of regenerative medicine. Although therapeutic value of MSCs is now being established through many clinical trials, issues have been raised regarding their expansion as per regulatory guidelines. Fetal bovine serum usage in cell therapy poses difficulties due to its less-defined, highly variable composition and safety issues. Hence, there is a need for transition from serum-based to serum-free media (SFM). Since SFM are cell type-specific, a precise analysis of the properties of MSCs cultured in SFM is required to determine the most suitable one. Six different commercially available low serum/SFM with two different seeding densities were evaluated to explore their ability to support the growth and expansion of BM-MSCs and assess the characteristics of BM-MSCs cultured in these media. Except for one of the SFM, all other media tested supported the growth of BM-MSCs at a low seeding density. No significant differences were observed in the expression of MSC specific markers among the various media tested. In contrary, the population doubling time, cell yield, potency, colony-forming ability, differentiation potential, and immunosuppressive properties of MSCs varied with one another. We show that SFM tested supports the growth and expansion of BM-MSCs even at low seeding density and may serve as possible replacement for animal-derived serum.

Mesenchymal stromal cells (MSCs) have emerged as a new player in the field of regenerative medicine due to its tremendous therapeutic applications. MSCs are plastic adherent, fibroblast-like cells which are mesodermal in origin, positively expressing cell surface markers CD29, CD44, CD49a-f, CD51, CD73, CD90, CD105, CD106, CD166 and Stro-1 while being negative for CD14, CD19, CD34, CD45 and HLA DR expression and exhibit trilineage differentiation in vitro 1 . They are multipotent cells showing self-renewal capacity and are present in multiple tissues including bone marrow, dental pulp, synovial fluid, umbilical cord, cord blood, Wharton's jelly, limbal tissue, adipose tissue, placenta and amniotic fluid 2, 3 . Clinical applications of MSCs are essentially attributed to their key biological properties which include (i) the ability to home to sites of inflammation during tissue injury when intravenously injected, (ii) the capacity to differentiate into several cell types including cartilage, muscle, bone, connective tissue and fat cells (iii) secreting various bioactive molecules thereby stimulating recovery of injured cells and (iv) suppressing inflammation and exhibiting immunomodulatory functions 4 .
Bone marrow-derived MSCs (BM-MSCs) account for ~ 0.001-0.01% of bone marrow mononuclear cells. Due to its low abundance, extensive in vitro culturing and expansion is required to obtain adequate numbers for research or clinical application 5 . The properties of BM-MSCs such as ease of isolation from BM without causing an immunological problem, ability to expand in vitro within a short period of time, bio preservation for point-of-care delivery with minimal loss of potency, and have not shown any reported serious adverse reactions during autologous or allogenic therapy makes them an important tool in regenerative medicine [6][7][8] . Furthermore, BM-MSCs have proven to be safe and are being widely tested in clinical trials, thus representing a powerful therapeutic paradigm for the treatment of a wide range of debilitating diseases 9 . Currently, there are more than 130 registered/ ongoing clinical trials aimed at elucidating the potential of BM-MSCs based cell therapy worldwide (www.clini caltr ials.gov).
However, the CPD of MSCs grown in StemXVivo SFM was significantly lesser (p < 0.01) when compared to control medium at both seeding densities (9.93 and 2.33 respectively) since this media was unable to support the growth of MSCs at P5. The PDT increased with the passage number from P4 to P5 in all media evaluated, irrespective of seeding density, as shown in Fig. 4a (Fig. 4b). On the other hand, PDT of MSCs cultured in StemXVivo SFM was significantly higher than any other media tested (p < 0.05). StemXVivo SFM supported the BM-MSC culture at initial exposure to SFM at P4, but the cells did not proliferate beyond  Immunophenotype of BM-MSCs cultured in serum-free/xeno-free media before and after cryopreservation. The immunophenotypes of BM-MSCs cultured in low serum/ SFM/XFM and control medium at two different seeding densities were analyzed using flow cytometry before and after cryopreservation (n = 2). BM-MSCs showed good expression of CD markers during adaptation to serum-free conditions. The CD marker expression analysis showed low levels (< 5%) of typical MSC negative markers such as CD34, CD45, and HLA-DR and high levels (> 90%) of positive markers such as CD73, CD90, CD105 for MSCs seeded at 1000 cells/cm 2 ( Fig. 5a & Fig. 5b) and < 10% of negative markers and > 90% of positive markers for MSCs seeded at 5000 cells/cm 2 both before and after cryopreservation ( Fig. 5c & Fig. 5d). BM-MSCs showed comparable expression profile of CD markers except for cells cultured in control medium and PLTMax hPL which showed expression of HLA-DR of 7.0-11.76% and 0.67-10% respectively than any of the media tested. RoosterNour-  www.nature.com/scientificreports/ deposits were evident during differentiation of MSCs cultured in RoosterNourish, RoosterNourish-MSC XF, StemMACS-MSC XF, PLTMax hPL, MSC NutriStem XF and was comparable to the control medium; whereas the MSCs grown in MSC NutriStem XF when seeded in 6 well plates were unable to differentiate into chondrocytes. This may be due to cells unable to form micromass when seeded in 6 well plates which require coating with ECM substrate for the attachment of cells ( Fig. 6f3 & l3). To overcome this, chondrogenic micromass pellets were generated by seeding BM-MSCs (2 × 10 5 cells/well) in 96-well V-bottom plates along with the control medium. Differentiation towards the chondrogenic lineage was observed after 14 days of induction as evident by Safranin O staining (Fig. 7a). Overall, BM-MSCs cultured in low serum/SFM/XFM exhibited comparable tri-lineage differentiation potential as evident by their staining characteristics.

Figure 5.
Immunophenotyping of human BM-MSCs cultured in low serum/ SFM/XFM before and after cryopreservation in the two different seeding densities (1000 cells/cm 2 and 5000 cells/cm 2 ). The known positive and negative human MSC specific cell surface antigens were estimated by flow cytometry. Isotope-matched controls were used to determine the non-specific binding by PE and FITC-conjugated antibodies. All the experiments were performed in duplicates and the data is represented as mean ± SD. The expression of positive markers (CD73, CD90 and CD 105) was more than 90% and that of negative markers (CD34, CD45) was less than 5% before and after cryopreservation in the seeding densities of 1000 cells/cm 2 (a & b) and 5000 cells/cm 2 (c & d). However the expression of HLA-DR of BM-MSCs cultured in control medium and PLTMax hPL were 7.0-11.76% ± 0.66 and 0.67-10% ± 0.76 respectively. To determine the impact of the various SFM/XFM under investigation on the differentiation potential, the expression patterns of key differentiation-specific genes were analyzed in MSC's which were subjected to adipogenic, osteogenic and chondrogenic differentiation. There was no significant difference in the cellular properties observed when the cells were cultured at seeding densities of 1000 cells/cm 2 and 5000 cells/cm 2 , hence for further assays cells seeded at 1000 cells/cm 2 was considered. Early commitment factors and late-stage maturation markers were chosen for analysis. As shown in Fig. 7b, cells cultured in SFM/XFM under evaluation successfully showed Figure 6. Differentiation potential of BM-MSCs cultured in vitro in low serum/SFM/XFM and control medium during freeze thaw analysis. BM-MSCs cultured at seeding density of 1000 cells/cm 2 and 5000 cells/cm 2 was cryopreserved. Freeze thaw analysis was carried out wherein the cells were thawed, revived and seeded to test their ability for trilineage differentiation. (a) BM-MSCs grown in SFM/XFM (1000 cells/cm 2 ) consistently differentiated into adipogenic (a1-f1), osteogenic (a2-f2) and chondrogenic (a3-f3) lineages, non-induced controls (a4-f4). b) BM-MSCs grown in SFM/XFM (5000 cells/cm 2 ) SFM differentiated into adipogenic (h1-l1), osteogenic (h2-l2) and chondrogenic (h3-l3) lineages, non-induced controls (h4-l4) except for BM-MSCs cultured in MSC NutriStem XF. Adipogenic differentiation was detected by oil droplet formation, (oil red O staining). Osteogenic differentiation indicated by calcium accumulation (alizarin red staining). Chondrogenic differentiation detected by Safranin O staining of glycosaminoglycan of cartilage. The images were captured at 10X magnification with scale bar of ∼100 μm.  (Fig. 7c). Similar results were obtained for chondrogenic differentiation, wherein cells cultured in PLTMax hPL and RoosterNourish-MSC XF showed increased expression of COL1A1 and COL10A1 (late-stage maturation genes). However, this did not correlate with enhanced upregulation of the chondrogenic commitment factor SOX9 (Fig. 7d). Overall assessment of prototypic early commitment and late-stage genes associated with trilineage differentiation indicated better differentiation induction when MSC's were cultured in RoosterNourish-MSC XF and PLTMax hPL as compared to MSC's cultured in control medium.
Colony forming ability of BM-MSCs expanded in serum-free/xeno-free media. The colonyforming ability of BM-MSCs cultivated in low serum/ SFM/XFM was tested by CFU-F assay using cells seeded www.nature.com/scientificreports/ at 1000 cells/cm 2 post cryopreservation. Typically, MSCs are characterized by their properties of plastic adherence and formation of colonies when plated at low cell densities as determined by CFU-F assay where more than 50 cells are considered as one colony. The CFU-F observed during the culture in various media showed considerable differences in colony morphology. Morphology and number of CFU-F in StemMACS-MSC XF, PLTMax hPL were smaller, few, and dispersed. MSC colonies in RoosterNourish, RoosterNourish-MSC XF were large, varied in shape with more number of colonies. The colonies in the control medium were more in number, larger, and merged (Fig. 8a). The mean CFU-F of cryopreserved MSCs post revival (P6) was 24 ± 2.64, 9 ± 2.08, 7 ± 1.15, 4 ± 0.6 for RoosterNourish, RoosterNourish-MSC XF, StemMACS-MSC XF and PLTMax hPL respectively when compared to control medium (25 ± 2.52) as shown in Fig. 8b. BM-MSCs in MSC NutriStem XF did not show any colonies even after 21 days of culture. The colony forming ability of MSCs cultured in SFM was significantly reduced when compared to the control and RoosterNourish medium (p < 0.008).
Level of VEGF secretion of BM-MSCs cultured in serum-free/xeno-free media. Growth factor/ cytokine secretion is one of the characteristics of MSCs that plays a crucial role in cell engraftment, neovascularization, and wound healing and which may be directly linked to the potency of MSCs. One such growth factor is VEGF, which when present at a high level improves the therapeutic efficacy of MSCs.

BM-MSCs cultured in serum-free/xeno-free media retain immune suppression ability.
Apart from their trilineage differentiation potential, the essential feature that makes MSC's useful for clinical applications is their immunomodulatory ability. The immunomodulatory property of MSC's was studied using an in vitro Mitogen induced lymphoproliferation assay wherein MSC's ability to suppress phytohaemagglutinin (PHA) stimulated T cell blast formation was investigated. As shown in Fig. 8d, cells cultured in SFM/XFM retained the ability to suppress T cell blast formation. More than 60% T cell blast suppression was noted in cocultures of MSCs cultured in all the SFM/XFM tested.

Discussion
MSCs isolated from different tissues are generally expanded in vitro using animal-derived growth serum (FBS). However, the use of MSCs cultured in FBS containing media for clinical applications poses potential risk and safety concerns due to the possibility of exposure to infectious agents and animal antigen 33 . Lot-to-lot variability and cost-intensive pre-testing of FBS samples raise concern for its usage in stem cell therapy. A few studies have employed a defined low-serum containing medium for MSC growth demonstrating efficient proliferation and maintenance of MSC characteristics leading to long-term consistency of results 37,38 . To eliminate the use of animal-derived growth supplements, several research laboratories have developed "in-house" SFM for MSC culture with a wide range of media supplements and culture ingredients and investigated MSC properties with variable results 21,39-43 . However, the biosafety of the media components in these disclosed formulations is not ensured and warrants further analysis. Moreover, some of the studies have shown differences in the growth of MSCs derived from different tissues of the same species 16,28,31 or same tissue of different species when cultured in defined SFM indicating serum-free culture conditions are tissue and species-specific and underlines the need for optimization for MSCs from different origin and/or relevant animal species 30  The proliferation index in co-cultures is relative to proliferation in PHA blasts (cultured in the absence of MSCs) which was considered as 100% (e) To determine IDO enzyme activity, the culture supernatant of MSC: PHA activated PBMC co-cultures of different media groups was assayed by spectrophotometric detection of the kynurenine concentration (tryptophan metabolite which is a product of IDO catabolism) and is represented as Kynuerurine units (μM). IDO secretion was higher in co-cultures cultured in SFM/XFM when compared to serum-containing medium. www.nature.com/scientificreports/ in supporting the growth of BM-MSCs initially cultured under serum conditions. Hence, the present study was undertaken to investigate the use of six different commercially available media for BM-MSC expansion and compared the MSC culture characteristics and functions with serum-containing medium. As new generations of SFM are emerging from various companies, these data may be substantial to address the specific challenges associated with the expansion of MSCs for therapeutic use during the transition from serum-containing media to serum-free media. The novelty of this study compared to earlier studies is that: (1) this is a comprehensive study evaluating six different commercially available low serum/SFM/XFM for its suitability in supporting the growth and expansion of BM-MSCs initially cultured under serum conditions until passage 3. More interestingly, the study provides experimental evidences and baseline data to support the growth and expansion of BM-MSCs using SFM/XFM from the cell banks that are already created using serum containing media without additional evaluation to test the suitability of the medium, (2) two different seeding densities of MSCs have been evaluated to determine the suitability of the medium to support growth and expansion of BM-MSCs. (3) BM-MSCs used in the present study were produced by pooling cells from 3 different donors at passage 2, which reduces individual variability in properties such as rate of proliferation, secretory activity, and differentiation, resulting in a product with robust paracrine activity.
Earlier studies have reported that the lower seeding density (50-1000 cells/cm 2 ) supports the efficient proliferation of MSC cultures in serum media 54 . Nevertheless, the seeding densities for MSC cultures must be optimized to minimize growth in patches observed when the seeding density is below the suboptimal range 6 . Abrahamsen et al., (2002) reported that lower seeding densities are perhaps better for expansion of MSCs since less contact inhibition is observed which is regulated by the Wnt signaling pathway 55 . In the present study, we carried out the comprehensive characterization of six different commercially available media through the passive adaptation method of frozen cultures, wherein the BM-MSCs were isolated and cultured in the serum-containing medium until passage 3 was subsequently cultured in low serum/SFM/XFM for the next 2 passages with seeding densities of 1000 cells/cm 2 and 5000 cells/cm 2 . BM-MSCs showed similar proliferation potential, expansion efficiency, and cell viability (> 85%) in RoosterNourish, RoosterNourish-MSC XF, PLTMax hPL, StemMACS-MSC XF and MSC NutriStem XF when compared to control medium at both seeding densities and were chosen for further testing to determine the functional characteristics. Although StemXVivo SFM showed comparable cell numbers at P4, it was unable to support growth at P5 at both seeding densities. MSCs cultured with the seeding densities of 1000 cells/cm 2 and 5000 cells/cm 2 demonstrated almost equivalent cell yield and viability, surface marker expression, and tri-lineage differentiation. These results suggested that even at a lower seeding density of 1000 cells/cm 2 , MSCs were able to proliferate efficiently in SFM while maintaining its functional characteristics.
The maintenance of a steady PDT is crucial for the expansion of MSCs for therapeutic use. Most of the serumfree formulations tested claim equivalent cell population doubling time and preservation of MSC characteristics when compared with serum-containing medium 22,51,56,57 . Growth kinetics of BM-MSCs in SFM evaluated in the present study showed varied PDT ranging from 21.89 to 45.7 h except for StemXVivo SFM and this is in concurrence with previous studies reporting similar PDT of MSCs cultured in SFM 26,28,31,48 . However, the performance with each of the medium varied with one another and with control medium which may be attributed to the differences in the composition of the medium 45,58,59 . Notably, BM-MSCs grown in RoosterNourish, RoosterNourish-MSC XF, StemXVivo SFM, StemMACS-MSC XF and MSC NutriStem XF seemed to be similar in size to one another with an average cell diameter higher than control medium, while PLTMax hPL cultured BM-MSCs displayed average cell diameter similar to control medium. The BM-MSCs grown in SFM appeared larger and thicker when compared to the serum containing medium. Whether these morphological characteristics are related to cell expansion capacity is elusive.
Immunophenotype analysis before and after cryopreservation revealed that BM-MSCs cultured in SFM expressed higher MSC specific surface markers CD73, CD90, and CD105, whereas their expression levels of hematopoietic cell markers (CD34, CD45) and HLA-DR molecules were very low satisfying the phenotypic criteria for describing hMSCs 60 . One interesting observation is the expression of HLA-DR in RoosterNourish, RoosterNourish-MSC XF, StemXVivo SFM, StemMACS-MSC XF and MSC NutriStem XF cultured BM-MSCs was low compared to that of PLTMax hPL and serum-supplemented cultures. A few studies have shown that BM-MSCs cultured in serum-supplemented medium with bFGF tend to express HLA-DR 54,61 . In this direction, serum-free conditions would be safer for large-scale manufacturing leading to lower HLA-DR expression.
To assess the multipotentiality of BM-MSCs grown in low serum/SFM/XFM, cryopreserved cells were induced to differentiate into trimesenchymal lineages, and relative expression of the representative osteogenic, chondrogenic and adipogenic genes were quantified. MSCs cultured in different SFM/XFM exhibited trilineage differentiation potential, albeit to varying levels. BM-MSCs cultured in RoosterNourish and RoosterNourish-MSC XF certainly differentiated towards trimesenchymal lineages and was comparable to the serum-containing medium. Overall, higher level of differentiation induction was observed when MSC's were cultured in RoosterNourish-MSC XF and PLTMax hPL compared to MSC's cultured in other media. Furthermore, to evaluate the potential of cells expanded in SFM, human MSCs cultured in low serum/SFM/XFM were placed into a CFU-F assay. The CFU-F efficiency of BM-MSCs cultured in SFM/XFM media was significantly less when compared to the control medium, whereas colony-forming ability of BM-MSCs grown in RoosterNourish (containing 1% FBS) was comparable to the control medium.
Recent studies have detailed the importance of determining the potency of human MSCs to predict their therapeutic efficacy and assessing the paracrine factors and immunomodulatory status of MSCs to develop better standards [62][63][64] . Moreover, it is reported that MSC potency may be directly linked to their secretory profile 65,66 . VEGF is one such factor that is secreted by MSCs and is essential for osteogenic differentiation and angiogenesis 67 . Our previous research findings established VEGF as a surrogate potency marker for BM-MSCs and through the VEGF quantitation data obtained from a large number of production batches; we estimated the level of www.nature.com/scientificreports/ VEGF ≥ 2 ng/ml/million BMMSCs to qualify for producing angiogenic activity. Also, we established a high correlation of VEGF levels with in vitro endothelial function 68 . In the present study, the level of VEGF secreted by cryopreserved MSCs cultured in different media was estimated. MSCs cultured in PLTMax hPL secreted a higher amount of VEGF (> 2 ng/ml/million cells); while MSCs cultured in other media secreted VEGF < 2 ng/ ml/million cells. The reason for the lower level of VEGF secretion by MSCs when cultured in SFM needs to be further investigated. Furthermore, BM-MSCs cultured in SFM/XFM retained their immunosuppressive potential. Several secretory molecules have been proposed to play a crucial role in regulating the immunosuppressive effects of MSCs, including indoleamine 2,3-dioxygenase, nitric oxide, prostaglandin E2, transforming growth factor beta and so on 34 . Besides, we quantified the levels of IDO secretion in BM-MSCs and it was observed that cells cultured in SFM/XFM secreted significantly higher level of IDO compared to the control medium. Immunosuppression activity is a multifactorial process; hence, the exact mechanism through which the SFM/ XFM grown cells render immunosuppression remains to be determined.
In conclusion, our research findings successfully demonstrated the growth of pre-isolated MSCs (P3) in low serum/serum-free conditions for two subsequent passages at the lower seeding density of 1000 cells/cm 2 while retaining their functional properties. This is the first comprehensive study reporting the extensive characterization of BM-MSCs in six different commercially available low serum/SFM/XFM along with serum-containing medium with low and high seeding densities. There was no significant difference in the expression of MSC specific markers among each of the media tested. On the other hand, the cell yield, population doubling time, potency in terms of VEGF secretion, colony forming ability, differentiation potential to mesodermal lineage and immunosuppressive potential of MSCs varied with one another. The present study showed that among the media evaluated, RoosterNourish-MSC XF and PLTMax hPL would be preferable in terms of cell yield, preserving MSC characteristics and reduced overall costs. MSC NutriStem XF and StemXVivo SFM requires extra step of coating the culture surface which may not be feasible during large scale production of MSCs for clinical applications. However, by eliminating the use of FBS it is certain that these SFM tested accomplish the key requirements in terms of multi-passage expansion of MSCs. This study provides useful information for the researchers from different laboratories to conduct studies using suitable SFM without need for additional step of evaluation. The most suitable media for MSC expansion should be chosen accordingly as different media produce MSCs with different properties. Future optimization studies and assessment of cell performance are required in large scale culture system before employing these media for therapeutic applications with the requirement to achieve cGMP and regulatory compliance.

Materials and methods
Isolation and culturing of BM-MSCs. Bone marrow was aspirated from healthy donors of age group between 18 and 40 years, after receiving written informed consent and approval in accordance to the Manipal University Ethics Committee Guidelines (Protocol Number: SRPL/CLI/07-08/001). All the experimental protocols were approved by the Institutional Committee for Stem Cell Research. Human mesenchymal stromal cells was isolated by density gradient centrifugation as described previously 69 . Cells were plated in T-75 flasks with DMEM-KO (Thermo Fisher Scientific, USA) containing 10% FBS (Hyclone, Waltham, USA), 2 mM L-Glutamax (Invitrogen, USA) and maintained at 37 °C incubator with 5% humidified CO 2. Media change was performed to remove non-adherent cells at defined intervals. The MSCs were grown in serum-containing medium up to passage 3 (P3) and are derived from multiple donor pool (3 different donors).
Sub culturing and expansion of BM-MSCs in different serum-free media. Human MSCs that are cryopreserved at P3 were thawed at 37 °C in a water bath, revived with pre-warmed culture medium, centrifuged at 1200 rpm for 10 min. The pellet obtained was then dislodged, resuspended in DMEM-KO, and seeded (P4) at low (1000 cells/cm 2 ) and high seeding densities (5000 cells/cm 2 ) in T-75 flasks (Thermo Fisher Scientific, USA) with low serum/SFM/XFM as listed in Table-1 by employing passive adaptation method 45 . The T-75 flasks were coated with MSC attachment solution (Biological Industries, USA) for MSCs cultured in NutriStem XF Medium and human fibronectin (R&D Systems, Minneapolis, USA) for MSCs cultured in StemXVivo human mesenchymal stem cell expansion media according to manufacturer's instructions. Other SFM tested are coating-free media. The composition of this media is proprietary and is not available publicly. MSCs cultured in DMEM-KO medium supplemented with 10% FBS, 2 mM L-Glutamax, 1X Pen-Strep (Invitrogen, USA) and 2 ng/ml bFGF (Invitrogen, USA) served as the control (hereafter referred as control medium). The cultures were screened at regular intervals to check confluency, morphological, and phenotypic characteristics. Once the cultures attained 80-90% confluency, the cells were harvested by enzymatic digestion with 0.25% trypsin-EDTA (Gibco, USA) at 37 °C for 2-3 min followed by neutralization of trypsin activity. The resulting cell suspension was expanded in the next passage (P5) in CellSTACK-1 (CS-1) chamber with vent caps (Corning, New York, USA) at low and high seeding densities. Media change was performed when the cultures attained 55-60% confluency and were harvested at 80-90% confluence. The cells were washed two times with Dulbecco's phosphate-buffered saline (DPBS, Gibco, USA) for 2-3 min. The BM-MSCs were trypsinized by addition of pre-warmed 0.25% trypsin-EDTA and incubation for 2-3 min at 37 °C followed by neutralization of trypsin activity by using the respective media. The cells were then centrifuged at 1200 rpm for 10 min, followed by DPBS wash and centrifugation. The cells thus obtained were resuspended in Cryostor 5 freezing media (BioLife Solutions, USA) and filled in 1.8 ml Nunc vials at a density of 5 × 10 6 cells/ml. The vials were then freezed in a freezing container (Nalgene Mr. Frosty, Sigma, USA) at a cooling rate of 1 °C/min in a − 80 °C freezer. After 12-16 h, the frozen vials were stored in liquid nitrogen below − 150 °C. www.nature.com/scientificreports/ Population doublings. To study the growth kinetics, pre-isolated BM-MSCs (cultured in serum-containing medium up to passage 3) grown in low serum/SFM/XFM from P3 onwards were used. The population doublings (PDs) and doubling time (PDT) of BM-MSCs cultured in these media and control medium were estimated from P3 to P5 and harvesting was carried out after reaching 80-90% confluency in each passage.

Immunophenotype analysis by flow cytometry.
To examine the effect of freezing and thawing on MSC functions, the cryopreserved BM-MSCs (P5) were thawed quickly at 37 °C, revived with Plasmalyte-A (Baxter, USA). The cell count and viability were assessed using Vi-CELL XR Cell Viability Analyzer (Beckman Coulter, USA) which works on trypan blue dye exclusion principle. The resulting cell suspension was centrifuged, resuspensded in respective media and used for analyzing CD marker expression, differentiation potential, potency and ability to form CFU-F. The StemXVivo SFM was unable to support the growth of MSCs at P5 in both the seeding densities tested and hence freeze thaw analysis of MSCs cultured in StemXVivo SFM was not performed. Differentiation potential. The tri-lineage differentiation capacity of BM-MSCs cultured in low serum/ SFM/XFM or control medium was evaluated during freeze-thaw analysis. Recovered BM-MSCs (7 × 10 5 cells) were cultured in the respective media in 37 °C incubator with 5% CO 2 and seeded into StemPro Osteogenesis, Adipogenesis or Chondrogenesis Differentiation medium (Gibco, USA) in 6-well plates as per the manufacturers protocol. Complete media change was performed during the culture period once in 3 to 4 days till they attain 90% confluency. The cells were then induced for adipogenesis and osteogenesis; screened at regular intervals for differentiation. Once the differentiation is observed, staining of cells was performed with 2% Alizarin red S (Sigma-Aldrich, USA) to monitor osteogenic differentiation and 6.3% Oil Red O (Sigma-Aldrich, USA) for adipogenic differentiation. For chondrogenic differentiation, chondrogenic micromass pellets were generated by seeding 1 × 10 6 recovered cell suspension in chondrogenic differentiation medium in 6-well plates and/or 2 × 10 5 cells/well in 96-well V-bottom plates (CELLSTAR, Greiner Bio-One GmbH) after centrifugation at 1200 rpm for 5 min, and placed in an incubator maintained at 37 °C with 5% CO 2 . Medium change was performed once in every 3 days. The cells were screened at regular intervals and once the differentiation was observed, the micromass pellet cultures were stained with 1% Safranin-O (Sigma, USA). The differentiated cells were visualized after fixing with 10% formalin and washing with DPBS. The images were captured using IX-71 inverted microscope (Olympus, Japan) under 10X objective. The uninduced cultures expanded in the respective media served as control.
Quantitative gene expression analysis. Total RNA was extracted using RNA isoplus (Takara Bio) and cDNA was synthesized using PrimeScript RT-PCR kit (Takara Bio). Gene expression analysis was carried out using SYBR green (TB green Premix Ex Taq II ,Takara Bio) on Step One plus qPCR instrument (Applied Biosystems) using gene specific primers as listed in Table 2. Relative mRNA expression levels were normalized to GAPDH and ΔCt was calculated. ΔΔCt was calculated by subtracting ΔCt of test with ΔCt of the calibrator (uninduced MSC's). The fold change was calculated using the formula 2 −ΔΔCt and represented as log 10 (fold change) after calculating the mean values of triplicates.
Colony forming unit-fibroblast (CFU-F) assay. Recovered BM-MSCs (total of 500 cells) were plated in respective media and control medium (n = 2 per group) in a 60 mm cell culture dish (Thermo Fisher Scientific, USA). The plates were stained with 0.5% crystal violet solution (Sigma, USA) 14 days post incubation and visible colonies were counted. The data was reported as average number of colonies for each group and represented as mean ± SD.
Vascular endothelial growth factor secretion (Freeze-thaw analysis). Conditioned medium from BM-MSCs was prepared and the amount of VEGF secreted was estimated as described previously 68 . In brief, the recovered cells (total 1 × 10 6 cells) were seeded in T-75 flasks in respective low serum/SFM/XFM media and control medium. The cultures were incubated at 37 °C incubator and 5% CO 2 . After 72 h, the conditioned medium was collected; filtered using a 0.22 μm syringe filter (Merck-Millipore, NJ, USA), aliquoted and stored at -80 °C until use.
Enzyme-linked immunosorbent assay. Human angiogenic factor (VEGF) secreted by MSCs was estimated by enzyme-linked immunosorbent assay (ELISA) using Human Quantikine ELISA kit (R&D Systems, Minneapolis, USA) according to the manufacturer's protocol. The respective media were used as controls. In brief, 200 μL of conditioned media was added to 96-well microplates that were pre-coated with monoclonal antibody specific for human VEGF and incubated for 2 h. After washing with wash buffer, enzyme-linked polyclonal antibody specific for human VEGF was added to each well, followed by incubation for 2 h and washing to remove unbound antibody-enzyme reagent. The substrate solution was added and incubated for 30 min, and www.nature.com/scientificreports/ the reaction was terminated by the addition of the stop solution. The level of VEGF was estimated by measuring the optical density at 450 nm using a VersaMax microplate reader (Molecular Devices, USA). The cells were harvested after 72 h by enzymatic digestion using 0.25% Trypsin-EDTA followed by neutralization with respective media. The cell count was estimated using the Vi-CELL XR Cell Viability Analyzer. The samples were assayed in duplicates and the amount of VEGF secreted was represented as ng/ml/ million cells.

Mitogen induced lymphoproliferation assay and BrdU incorporation assay. BM-MSC's were
seeded at a density of 20 × 10 3 cells per well of a 96 well plate in their respective media and allowed to adhere overnight. On the subsequent day they were treated with Mitomycin C (10 µg/ml) for 2 h, washed with PBS to remove traces of mitomycin and recovered in their respective media for 24 h. PBMC's were isolated using lymphoprep (Axis shield) as per manufacturer's instructions. The buffy coat layer was collected, washed thrice with PBS and then resuspended in respective SFM/XFM except for the control group where the PBMC's were resuspended in RPMI supplemented with 10% FBS. In order to stimulate T cell blast formation, PHA (20 µg/ml) was added to the PBMC suspension after which 1 × 10 5 PBMC's were added to MSC's to obtain MSC: PBMC ratio of 1:5 as previously optimized. The co-culture was continued for 72 h and BrdU priming was performed 12 h before lymphoproliferation assessment. After 72 h, BrdU incorporation assay (BrdU Cell proliferation kit, Calbiochem) was performed as per the manufacturer's instructions. PBMC alone, Mitomycin C treated MSC alone and PBMC co-cultured with MSC served as background controls whereas PHA treated PBMC's served as positive controls.
The extent of T cell blast suppression was calculated by determining the level of BrdU incorporation compared to the PHA treated PBMCs which was considered as 100%.
Indoleamine 2,3 dioxygenase assay. Freshly collected supernatants from co-cultures were treated with half the volume of 30% trichloroacetic acid, vortexed and then centrifuged at 10,000 rpm for 5 min. The resulting supernatant was collected to which equal volume of Ehrlich reagent (100 mg in 5 ml glacial acetic acid) was added, incubated for 10 min at room temperature after which absorbance was measured at 490 nm. IDO activity was assessed by measuring the kynurenine concentration as determined by plotting the absorbance values against the kynurenine standard curve.
Statistical analysis. Quantitative data were reported as mean ± standard deviation (SD). Analysis of variance (ANOVA) and unpaired t test were performed using GraphPad Prism Software (www.graph pad.com).
Results were considered statistically significant if p value was < 0.05.

Data availability
The datasets generated and analyzed during the current study are available from the corresponding author on reasonable request. www.nature.com/scientificreports/ Publisher's note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creat iveco mmons .org/licen ses/by/4.0/.