Effect of age and gender on cell proliferation and cell surface characterization of synovial fat pad derived mesenchymal stem cells


  • Emma Fossett,

    1. University College London Institute for Orthopaedics and Musculoskeletal Sciences, Royal National Orthopaedic Hospital, Stanmore HA7 4LP, United Kingdom
    Search for more papers by this author
  • Wasim S. Khan,

    Corresponding author
    1. University College London Institute for Orthopaedics and Musculoskeletal Sciences, Royal National Orthopaedic Hospital, Stanmore HA7 4LP, United Kingdom
    • University College London Institute for Orthopaedics and Musculoskeletal Sciences, Royal National Orthopaedic Hospital, Stanmore HA7 4LP, United Kingdom. T: +44 (0) 7791 025554; F: +44 (0) 20 8954 2301
    Search for more papers by this author
  • Umile Giuseppe Longo,

    1. Department of Orthopaedics, Royal Free Hampstead, NHS Trust, London, United Kingdom
    Search for more papers by this author
  • Peter J. Smitham

    1. University College London Institute for Orthopaedics and Musculoskeletal Sciences, Royal National Orthopaedic Hospital, Stanmore HA7 4LP, United Kingdom
    Search for more papers by this author


Cell based therapies are being investigated for biological repair of a variety of disorders. Previous work has shown that mesenchymal stem cells (MSCs) from older patients have reduced proliferation rates. As age is associated with greater musculoskeletal morbidity, e.g., osteoarthritis, an optimal MSC expansion strategy is required for older patients. In this in vitro study we investigate how age and gender affect MSC proliferation rate and cell surface characterization, as well as identify a relationship between seeding density and proliferation that could be applied to therapeutic MSC uses. Synovial fat pad derived MSCs were isolated and expanded from 14 patients undergoing total knee replacements. The cells were seeded at densities between 50 and 10,000 cells/cm2 and cell proliferation studies, flow cytometry, and cell surface staining were performed. Females were found to have consistently higher cell proliferation and cell surface marker expression. The cell surface marker CD105 had a constant expression irrespective of age. A statistically significant inverse relationship was found between seeding densities and cell proliferation rates. This study has shown that patient characteristics do effect cell proliferation rate and cell surface characterization, but as seeding density has a significant relationship with proliferation rate, it can be altered, possibly along with other cell culturing strategies, to compensate for the effects of patient factors on MSCs. We have also shown that gender affects cell proliferation and cell surface characterization, something most previous studies may have failed to identify as they group male and female patients together. © 2012 Orthopaedic Research Society. Published by Wiley Periodicals, Inc. J Orthop Res 30:1013–1018, 2012

Stem cells are an undifferentiated cell population capable of self-renewal and differentiation down one or more lineages.1 As ethical issues surround the use of embryonic stem cells, there is an increasing interest in the potential of adult stem cells. Adult mesenchymal stem cells (MSCs), present in many tissues in small numbers, are required to restore normal function via repair and regeneration of tissues.2 These cells have the capacity for proliferation and multilineage differentiation in vitro,3 and are seen as a valuable source for tissue regeneration and cellular therapy. Since Friedenstein et al.4 isolated MSCs from bone marrow, other sources such as adipose,5 synovium,6 and synovial fat pad7 have also been reported to contain MSCs. Studies have shown that the synovial fat pad derived MSCs have many advantages over MSCs derived from bone marrow and other tissues; synovial fat pad derived cells yield a higher number of MSCs that have a higher proliferation capacity,8–10 and are associated with low donor site morbidity and are easily accessible via arthroscopy.7

As MSCs are found in low numbers in adult tissue, expansion in vitro is required to reach the desired numbers for laboratory investigations or clinical applications. Although some patient factors such as chronic disease have been shown to adversely affect the yield and proliferation rate of MSCs,11 we have yet to develop a clear understanding of how age and gender affect MSCs. This issue will become more important as we develop cell-based musculoskeletal strategies for managing an increasingly ageing population. Many have reported an inversely proportional relationship between proliferation rate and age,12–14 however, others have reported no relationship.15–17 We have previously investigated proliferation rate of synovial fat pad MSCs in two groups of patients with a mean age of 57 and 86, finding no significant difference in cell proliferation between both groups at five time points at days 2,4, 6, 8, and 10.18

There are factors that we can control that may increase the yield and proliferation rate of MSCs such as the seeding density, type of media, and growth factors. These expansion strategies could potentially be used to offset any adverse effects of ageing and gender. Although limited, the literature suggests that seeding density does have an effect on cell proliferation rate. Both et al.19 found that bone marrow derived MSCs (BMSCs) seeded at 100 cells/cm2 reached their target of 200 million cells, 4.1 days faster than cells that were seeded at 5,000 cells/cm2. A similar relationship was found by Lode et al.20 who investigated the effect of seeding density on three dimensional scaffolds. Although limited, studies looking at synovial fat pad MSCs observed results that were consistent with those from BMSCs.9

Characterization of cells helps to show the origin of the cell and its potential for differentiation.18, 21 We have also previously looked at the effect of age on cell surface expression of synovial fat pad derived MSCs and found no differences between two groups with a mean age of 57 and 86.18 Studies on BMSCs have also showed no significant difference in the expression of cell surface markers with age.13 Others, however, have reported significant age-related reductions in expression of CD44, CD90, CD105, and STRO-1.22 If consistently expressed markers can be found, they can reliably be used to isolate MSC populations regardless of age and gender of donor.

There is little literature on the effects of gender on MSC proliferation potential and cell surface characterization as many papers group results from male and female patients together. MSC culturing techniques such as duration and concentration of collagenase, and best harvesting site for maximal yield vary between males and females.23 These differences suggest that there could also be a difference in proliferation rate and cell surface characterization of MSC from male and female donors.

This study will look at how seeding densities affect cell proliferation of synovial fat pad derived MSCs, and how age and gender affect proliferation rate and cell surface characterization.


Synovial fat pad tissue was obtained from 14 patients undergoing total knee replacements. Ethical approval was obtained from the Centre of Research: Ethical Campaign, Royal National Orthopaedic Hospital, Stanmore, UK, Institutional Review Board prior to performing the study. Patients were fully informed of the use of their tissue and all gave consent for its use.

Cell Isolation and Culture

Approximately 5 ml of synovial fat pad tissue was dissected and MSCs were isolated using 0.2% collagenase type I (Sigma-Aldrich, Gillingham, UK) for 3 h, at 37°C, with constant agitation on a tube roller. The collagenase was neutralized with 10 ml basic medium consisting of low glucose Dulbecco's Modified Eagle Medium (DMEM- Sigma-Aldrich) supplemented with 20% fetal calf serum (FCS- First Link, Birmingham, UK) and 1% penicillin/streptomycin (P/S, 1,000 units/ml penicillin and 1,000 µg streptomycin- Gibco, Invitrogen Incorporated, Paisley, UK). The released cells were sieved (70 µm mesh), washed in basic medium and centrifuged at 2,000 rpm for 5 min separating MSCs and floating adipocytes which were then removed. The cell suspension was plated in a T75 polystyrene tissue culture flask (Corning Incorporated, Corning, NY) with 10 ml basic medium in a humidified incubator with 5% CO2, 20% oxygen at 37°C (Passage 1). Medium was changed three times a week.

Cell Proliferation Rates

Following trysinisation, passage 2 cells were seeded at 50, 250, 500, 1,000, 2,500, 5,000, 7,500, and 10,000 cells/cm2 in 12-well plates (Orange Scientific, Belgium) with four repetitions for each density. They were maintained in 1 ml medium, which was changed twice a week. The dye, Alamar Blue (AbD Serotec, MorphoSys, Kidlington,UK), was used in a 1:10 dilution with phenol red and L-glutamine free DMEM (Sigma-Aldrich) and 1% P/S to measure proliferation rate. The basic media was removed from each well and 1 ml Alamar Blue solution was added. Following incubation at 37°C for 3 h 15 min, 100 µl was removed from each well and underwent spectrophotometric analysis in a Fluorskan Ascent Microplate Fluorometer (Thermo labsystems, Basingstoke, UK). The programme “Ascent Research Edition” version 1.1.1 was used, with a fluorescence excitation wavelength set at 510 nm and a fluorescence emission wavelength of 590 nm. The remaining Alamar Blue solution was removed from the wells and 1 ml of basic media was placed in return. This assay was performed on days 0, 3, 6, 10, 14, 17, and 21. This equation was used to calculate population doublings per day: Population doubling per day = (ln (N/N0))/t, where N = number of cells on desired day, N0 = number of cells on day 0, and t = number of days.

Cell Surface Epitope Characterization

Confluent passage 2 cells were stained for a panel of antibodies for cell surface epitopes using flow cytometry and cell surface staining. This included antibodies against the following: CD34 (marker for hematopoetic cells), CD44 (hyaluronan receptor), CD80 (marker of activated B-cells and monocytes), and CD90 (Thy-1) from abcam, Cambridge, UK, CD45 (receptor for leucocyte common antigen), CD54 (ICAM-1), CD73 (receptor involved in B cell activation), CD106 (VCAM-1), and CD166 (ALCAM) from Novocastra, STRO-1 (marker for BMSCs) from BD Biosciences, UK, IgG from Santa Cruz Biotechnology, Santa Cruz, CA and CD105 (SH2 or endoglin) courtesy of J. Dudhia, Royal Veterinary College, London, UK. For flow cytometry, ½ million cells were used for each antibody. Three controls were used: One with no primary or secondary antibody, one with only the primary (IgG) and one with only the secondary. Cells in monolayer were detached with trypsin (Gibco, Invitrogen Corporation, Paisley, UK, 0.05% trypsin and 0.53mM EDTA), washed and incubated with primary mouse antibodies (1:100 dilution) for 30 min, followed by fluorescein isothiocyanate conjugated (FITC) anti-mouse IgM secondary antibody (Sigma-Aldrich) in a 1:100 dilution for 30 min. Cells were then rewashed and assayed in a flow cytometer (FACS Calibur, BD Biosciences) using the programme, Cell Quest Pro.

Passage 2 cells from samples 5 and 10 were also looked at using cell surface staining. Confluent cells initially seeded at 30,000 cells/cm2 on 22 × 26 mm cover slips, were used. A control using only the primary antibody (IgG) was used. Cells were incubated for 40 min with the primary mouse antibodies (1:100 dilution), washed and then incubated for a further 40 min with 500 µl (1:100 dilution) of FITC-conjugated anti-mouse IgM secondary antibody. The cells were then rewashed before being incubated with 50 µl Hoechst stain (1:100 dilution) for 5 min to stain the nuclei. Images were captured using a fluorescent microscope (Olympus XM10, Olympus, UK).

Statistical Analysis

Experimental data was analyzed using SPSS Statistical Software (Version 17) and Microsoft Excel (Microsoft, UK). Mean values were calculated to compare groups using a t-test when comparing two groups (gender) and an ANOVA and Bonferonni post-hoc multi-comparisons analyses for more than two groups (seeding density). Correlation coefficients were generated to investigate the effects of age. Regression analysis was also used to investigate how a change in seeding density affects proliferation rate. Statistical Significance was set at p < 0.05.


The mean age of the 14 patients was 67 years (range 50–83 years, SD 11 years), and six patients were female and eight were male. Five milliliters of tissue yielded around 7.5 million cells that were plated at a concentration of 100,000 cells per cm2 in a T75 flask. Almost 90% of the cells that were obtained from the tissue digest and initially plated were subsequently removed when the medium was changed at 48 h. Cells isolated from all donors attached to the polystyrene surface of the tissue culture flasks and gradually spread reaching confluence by day 7–10 with no significant differences in the time to reach confluence. The number of cells obtained from 5 ml of tissue after 2 passages was about 10 million, i.e., a confluent T225 flask. All experiments were performed on passage 2 cells to ensure a sufficient number of homogenous cells.

Seeding Densities and Proliferation

The highest population doublings per day (0.329) was observed at the lowest seeding density of 50 cells/cm2. The lowest population doublings per day (0.155) was observed at the highest seeding density of 10,000 cells/cm2. Figure 1 shows that as seeding density increases, population doublings per day, and hence proliferation rate decreases. There was no statistical difference in population doublings between seeding densities of 2,500, 5,000, 7,500, and 10,000 cells/cm2 that had results of 0.180, 0.180, 0.170, and 0.155, respectively (p = 1.000 ANOVA and Bonferonni test). Statistically significant (p < 0.05) and extremely statistically significant (p < 0.001) differences were observed between many of the densities when compared with lower densities of 50 and 250 cells/cm2 as shown in Figure 1. Regression analysis performed on each patient's data, showed to what extent variation in the seeding density affects cell population doublings. All patients had a negative coefficient showing that population doublings decrease with increasing cell seeding densities (p < 0.05). Correlating the regression coefficient with age (Fig. 2) showed that as age increases, regression coefficients get closer to zero, and therefore proliferation rates become less responsive to change in seeding density (Correlation Coefficient = 0.161, p = 0.618). Figure 3 shows a trend that MSCs from female patients (n = 5) have a higher population doublings per day than males (n = 7) at all seeding densities. However, no statistically significant differences (p < 0.05, student's paired t-test) were observed.

Figure 1.

The effect of seeding density on mean population doublings per day. Data is mean ± standard error of the mean (n = 5–12), *p < 0.05, **p < 0.001 (ANOVA with Bonferroni's correction).

Figure 2.

This scatter plot correlates the regression coefficient, or the relationship between increasing seeding density on population doublings, with age (Correlation coefficient = 0.161, p = 0.618).

Figure 3.

A bar chart showing how gender affects population doublings up to day 21. Data is combined mean of all male (n = 7) and all female (n = 5) patients ± standard error of the mean.

Cell Surface Characterization

Flow cytometry (Fig. 4) showed that synovial fat pad derived MSCs stained strongly for CD44 (86.76%), CD73 (64.49%), and CD105 (28.78%). Some expression of CD54 (14.58%) and CD106 (5.38%) was also seen. All other markers showed negative expression below 1.8% fluorescence. Passage 2 cells were also stained with the same panel of antibodies and observed with a fluorescent microscope. Figure 5 shows that cells were again found to stain strongly for CD44, CD73, and CD105 in concordance with flow cytometry analysis. Results from the flow cytometry analysis for each patient were correlated against the age and showed no significant correlation. Figure 6 shows a trend that females (n = 4) had a higher mean expression of all markers than males (n = 7), and there was a statistically significant difference between the expression of STRO-1 (p = 0.039, student's t-test), but not for any of the other cell surface markers (p > 0.05).

Figure 4.

This bar chart shows the percentage fluorescence for a range of cell surface markers. Data is mean ± standard error of the mean (n = 14).

Figure 5.

Cell surface staining on synovial fat pad derived MSCs was performed using a panel of antibodies and FITC conjugated secondary antibody (green), and DAPI (blue). Results showed strong staining for CD44, CD73 and CD105, and negative staining for CD45, CD54, CD80, CD90, CD106, CD166, and STRO-1. No staining was observed for the IgG control. [Color figure can be seen in the online version of this article, available at http://wileyonlinelibrary.com/journal/jor]

Figure 6.

This bar chart shows the mean cell surface marker fluorescence for male (n = 8) and female (n = 6) patients. Data is mean ± standard error of the mean, *p < 0.05.


Cell Proliferation Rate

In our study, cells seeded at the lowest seeding density of 50 cells/cm2 underwent the greatest cell proliferation, and increased in number by 890-fold over 21 days. This was significantly greater than cells seeded at 10,000 cells/cm2 that only increased by 22-fold. These results imply that cells seeded at a lower density undergo rapid expansion compared to higher densities. There were many other statistically significant differences between seeding densities showing that seeding density does have a significant effect on proliferation rate. The lower proliferation rate of cells seeded at a higher density could be due to contact inhibition. These results are consistent with previous studies on seeding density for synovial fat pad, bone marrow, and adipose derived MSCs.10, 18, 19 Finding the relationship between seeding density and optimal cell proliferation is useful in both laboratory investigations as well as potential clinical applications allowing the cell culturing procedure to be less time consuming, with a lower risk of cell culture contamination, and more cost effective.

It is not known whether ageing of MSCs is caused by extrinsic or intrinsic factors. Zhou et al.14 suggested that intrinsic factors such as an increased number of MSCs positive for senescence-associated β-galactosidase (SA- β-gal) with age, together with increased expression of p53 and its pathway genes (p21 and BAX) may be responsible for mediating reduced proliferation in MSCs from older patients, by inducing senescence. Extrinsic factors such as a reduced synthesis of proteoglycans and glycosaminoglycans in the surrounding tissue of MSCs reduces cell proliferation and viability in vivo24 and accumulation of advanced glycosylated end products (AGEs) inhibit proliferation of MSCs by activating apoptosis and reactive oxygen species (ROS) production.25

Although there was a general trend with an age-related decline in population doublings at lower seeding densities, and an age-related increase in population doublings at higher seeding densities, none of these results were significant. Older patients had a smaller regression coefficient than the younger patients investigated, suggesting that MSCs from older patients show less of a change in population doublings than younger patients with increasing seeding density. This suggests that the correct seeding density is more important in younger patients. An age related decline in BMSC proliferation has been reported in the literature,12–14 but this was not correlated with seeding density, and a more in-depth study to investigate this complex issue is needed.

For cells plated at 50 cells/cm2, females had an 890-fold increase in cell number up to 21 days and males only had a 367-fold increase. Although females were found to have higher population doublings at all seeding densities, none of the differences were statistically significant. Although this study has not been able to determine conclusively whether age and gender affects proliferation rate, some interesting patterns have been identified that will guide future research.

Cell Surface Characterization

Flow cytometry analyses on passage 2 MSCs from all patients showed the expression of CD44, CD73, and CD105. Previous studies have also found that synovial fat pad derived MSCs express CD44 and CD105.7, 9 CD73 is the receptor for a molecule involved in B cell activation, and although it is known to be expressed on adipose and BMSCs,26, 27 its expression in synovial fat pad cells has not previously been reported. The low expression of CD34 and CD45 (hematopoietic markers) suggests that there was no contamination with hematopoietic cells during isolation. In this study we had a lower expression of CD54, CD90, CD106, CD166, and STRO-1 than what has previously been observed.7, 9, 18 Discrepancies could be due to the cells being of different passages,9 and different incubation durations, antibody concentrations, and antibody sources.18 Our study also used a high concentration of FCS in culture medium (20%) which has been shown to inhibit expression of some surface antigens.28 We found no significant age-related changes in cell surface expression, similar to our previous results.18 Stolzing et al.22, however, found an age-related decline in the expression of CD90, CD105, and STRO-1, and an increase in the expression of CD44 in BMSCs.

Although this study found a higher expression of all investigated cell surface markers in females, only the expression of STRO-1 was found to be significantly greater than males. This was an interesting finding as other studies have not looked at the effects of gender on cell surface characterization. As androgens have been suggested to have an inhibitory effect and estrogens an excitatory effect on MSCs, and as estrogens are responsible for up regulating ERa and ERb receptor expression on embryonic stem cells,23 it is possible that patient characteristics such as gender account for the variable results in MSC surface expression, sometimes even for cells from the same source, noted in the literature.

Future work is planned looking at the effect of smaller seeding densities on proliferation rate, and using larger number of cell surface markers in age and sex matched patients.


This study has shown that lower seeding densities are associated with higher proliferation rates and this has implications for the rapid expansion of synovial fat pad MSCs for therapeutic use. We have shown that patient characteristics do affect cell proliferation rate and cell surface characterization, but as seeding density has such a significant relationship with proliferation rate, it can be altered, possibly along with other cell culturing strategies, to compensate for the effects of patient factors on MSCs.


The authors would like to thank and are grateful for the funding from the Goldberg Schachmann & Freda Becker Memorial Fund Award, from University College London. They would also like to thank Ayad Eddaoudi and Ambika Angheluta from Camilla Botnar Labs, Great Ormond Street Hospital, London, for their help with the flow cytometry analyses.