The meniscus, a crescent-shaped fibrocartilaginous tissue in the knee, plays roles in load distribution, stability, and lubrication of the knee joint [1]. Degenerative tears of the meniscus occur frequently in middle-aged or older persons, and symptomatic degenerative tears are a common clinical problem [2]. When surgery is required, the usual choice is arthroscopic partial meniscectomy rather than meniscal repair, because of the poor healing of the meniscus [3]. However, meniscectomy increases the risk of osteoarthritis [4].

A degenerative meniscus tear can be successfully repaired with additional cell therapy. The use of mesenchymal stem cells (MSCs) isolated from the synovium is attractive for this purpose because of their high proliferative and chondrogenic potential [5]. For example, a porcine study demonstrated that transplantation of synovial MSCs could promote the healing of meniscal tear repairs that generally do not heal [6]. A clinical study confirmed that transplantation of synovial MSCs following surgical repair improved the clinical symptoms in patients with complex degenerative tears of the meniscus [7]. In both these studies, transplantation was achieved by placing a suspension of synovial MSCs onto the repaired meniscus for 10 min [6, 7]. The number of synovial MSCs adhering to the meniscus correlated with the therapeutic effect of the cells on the meniscus injury [8], but the exact number of cells that adhered to the repaired meniscus in 10 min was not determined. The aim of the present study was therefore to quantify the proportion of synovial MSCs that adhere to the meniscus following their placement as a cell suspension.

The adhesion of synovial MSCs from a cell suspension onto the meniscus implies that the cells are likely to undergo dynamic morphological changes, such as those reported by Rajaraman et al., who examined adhesion of WI-38 lung tissue fibroblasts by scanning electron microscopy (SEM). They found that the initially spherical cells developed “bleb”-like vesicles, microspikes, and a flattened central mass as they adhered to a glass surface [9]. However, unlike WI-38 fibroblasts, MSCs are a heterogeneous population with different surface epitopes, colony formation ability, and differentiation potential [10]. Therefore, MSCs may have a heterogeneous morphology when observed by SEM. A second aim of the present study was therefore to clarify the morphological heterogeneity of synovial MSCs and how this may change during adhesion to the meniscus. A third aim was to speculate the mechanism that allows adhesion of MSCs to the meniscus, as some MSCs adhere almost immediately upon placement on the meniscus. The findings obtained in this study contribute to understand the morphological features of synovial MSC adhesion to the meniscus and provide hints to increase the number of adherent MSCs so that clinical outcomes can be improved.

Materials and methods

Human synovial MSCs

This study was approved by the Medical Research Ethics Committee of Tokyo Medical and Dental University, and informed consent was obtained from all study subjects. Human synovium was harvested from the knees of 14 female donors with osteoarthritis during total knee arthroplasty operations. The average age and standard deviation (SD) of the patients was 73 ± 11 years. The synovium was minced and digested in a solution of 3 mg/mL collagenase (Sigma-Aldrich Japan, Tokyo, Japan) at 37 °C for 3 h, and the digested cells were filtered through a 70 μm cell strainer (Greiner Bio-One GmbH, Frickenhausen, Germany). The obtained nucleated cells were suspended in α-MEM medium (Thermo Fisher Scientific, Rockford, IL, USA) supplemented with 1% antibiotic/antimycotic (Thermo Fisher Scientific) and 10% fetal bovine serum (Thermo Fisher Scientific), and cultured in a cell culture incubator (Astec Co. Ltd, Fukuoka, Japan) in 5% CO2 at 37 °C for 14 days. When the MSCs reached 70–80% confluence, synovial MSCs of passage 0 were harvested for this study. For cell morphology, 1000 cells at passage 1 were cultured in a 60 cm2 dish for 14 days. Colony morphology was examined by staining the cells with crystal violet.

Flow cytometry

Surface markers of synovial MSCs from four donors were examined by flow cytometry on a FACS Verse instrument (Becton, Dickinson and Company, NJ, USA). The cells in Hank’s balanced salt solution were suspended at a density of 5 × 105 cells/mL in FACS buffer and stained for 30 min with the following antibodies: CD44 (PE-Cy7), CD45 (APC-H7), CD73 (V450), CD90 (PE), and CD105 (APC) (all from Becton Dickinson). Data were analyzed using FlowJo software (Tree Star Software, CA, USA) [11].

Differentiation assays

Chondrogenesis was examined by suspending 1.25 × 105 synovial MSCs in 0.5 mL chondrogenic induction medium consisting of DMEM (Thermo Fisher Scientific) containing 10 ng/mL transforming growth factor-β3 (TGF-β3, Miltenyi Biotec, Bergisch Gladbach, Germany), 500 ng/mL bone morphogenetic protein 2 (BMP-2, Medtronic, Minneapolis, MN, USA), 40 μg/mL proline, 100 nM dexamethasone, 100 μg/mL pyruvate, 50 μg/mL ascorbate-2-phosphate, and 50 mg/mL 1%ITS Premix (Becton Dickinson, San Jose, CA, USA). The cells were pelleted by centrifugation at 500 × g for 10 min and then cultured for 21 days. After 21 days, the pellets were sectioned and stained with toluidine blue (Fujifilm Wako Pure Chemical Corporation, Osaka, Japan).

Adipogenesis was determined by suspending 100 synovial MSCs in a 60 cm2 dish and culturing for 14 days in culture medium to produce cell colonies. The adherent cells were further cultured in adipogenic induction medium consisting of α-MEM supplemented with 100 nM dexamethasone, 0.5 mM isobutylmethylxanthine (Sigma-Aldrich), and 50 mM indomethacin (Fujifilm, Wako Pure Chemical Corporation) for an additional 21 days. Adipocytes were stained with oil red O (Muto Pure Chemicals, Tokyo, Japan).

Calcification was studied by plating 100 synovial MSCs in a 60 cm2 dish and culturing for 14 days in culture medium to allow the formation of cell colonies. The adherent cells were further cultured in calcification induction medium consisting of α-MEM supplemented with 50 μg/mL ascorbic acid 2-phosphate (Fujifilm Wako Pure Chemical Corporation), 10 nM dexamethasone (Fujifilm Wako Pure Chemical Corporation), and 10 mM β-glycerophosphate (Sigma-Aldrich). After 21 days, calcification was assessed by alizarin red staining (Merck Millipore, Billerica, MA, USA).

Time course of synovial MSCs adhesion to the meniscus

Fresh porcine knees from six-month-old animals were purchased from Shibaura Zoki Co., Ltd (Tokyo, Japan), and the medial and lateral menisci were excised (Fig. 1). The surface of each porcine meniscus was abraded to reproduce degenerative meniscus. The meniscus was cut into a cylindrical shape 12 mm in diameter (Ichinen MTM Corporation, Osaka, Japan).

Fig. 1: Analysis of adhesion of human synovial MSCs onto the porcine meniscus.
figure 1

(i) A porcine meniscus was removed. (ii) The surface was abraded. (iii) The meniscus was cut into a cylindrical shape. (iv) A cell suspension containing 106 synovial MSCs in 100 μL PBS was placed on the meniscus. (v) Immediately after placement and at 10, 60 min and at 6, 24 h after the placement, the meniscus was washed with 1 mL PBS for the morphological analyses. A meniscus with no cell suspension placement served as the control. Non-adherent cells were also counted to evaluate the adherent cell number.

Full size image

A cell suspension containing 106 synovial MSCs in 100 μL PBS was placed on the meniscus. The meniscus was then washed with 1 mL phosphate-buffered saline (PBS) immediately after placement and then at 10, 60 min and 6, 24 h after MSC placement. The non-adherent cells in these washes were counted to calculate the numbers of adherent cells.

Histological analysis

A cell suspension containing 106 synovial MSCs in 100 μL PBS was placed on the meniscus for the desired period and the meniscus was then analyzed histologically. The menisci (n = 6) were fixed in 4% paraformaldehyde overnight, and then paraffin embedded, sectioned, and stained with hematoxylin/eosin for examination by light microscopy (BZ‐X700; Keyence Co., Ltd, Osaka, Japan).

For immunostaining, paraffin-embedded sections were deparaffinized and treated for 10 min at 37 °C with pepsin (Thermo Scientific) for epitope retrieval. The sections were then blocked with Protein Block Serum-Free (Dako by Agilent Technologies, Santa Clara, CA, United States), washed three times, and reacted with human vimentin (1:200; clone No. 3B4, Dako) overnight at 4 °C. The sections were again washed three times and incubated with Alexa Fluor 488-conjugated secondary antibodies (1:500; Thermo Fisher Scientific, Waltham, MA, USA) for 2 h at 4 °C, followed by counterstaining with 4,6-diamidino-2-phenylindole (DAPI, Wako) [12].

Scanning electron microscopy (SEM)

The menisci were fixed in 2.5% glutaraldehyde for 2 h and washed overnight in 0.1 M phosphate-buffered saline at 4 °C. The specimens were then postfixed with 1% osmium tetroxide for 2 h at 4 °C and dehydrated in graded ethanol solutions. After exchanging with 3-methyl butyl acetate and critical point drying, the specimens were coated with platinum. The surface was observed by SEM (S-4500; Hitachi Ltd, Tokyo, Japan). Synovial MSCs in the floating condition were also observed.

Cell classification by SEM

Synovial MSCs in the flattened state were classified as having round, semi-flat, and flat morphology. Round cells were defined as cells with a spherical cell morphology. Semi-flat cells were defined as cells that were thick, with partially smooth contours, and that lacked a ring-shaped spot pattern. Flat cells were defined as cells that were thin, with not entirely smooth contours, and that had a ring-shaped spot pattern (Supplementary Fig. 1).

The MSC cell processes were categorized by the presence or absence of microspikes and by the presence or absence of pseudopodia following the method of Adams [13]. Microspike positive cells were defined as cells that contained at least three microspikes. Pseudopodia positive cells were defined as cells that contained at least one pseudopodium.

The surface features of synovial MSCs were classified as blebbed, ruffled, and smooth, based on the reports by Rajaraman et al. [9] and Adams [13]. The cell that blebs occupied the largest area was defined as the bleb-dominant cell. Ruffling-dominant cell and smooth-dominant cell were defined in the same way.

For quantification, 50 cells were randomly selected from each of the six donors at each time period by one author (YS) and were classified by another author (SS).

Statistical analysis

All data were statistically evaluated with the Kruskal–Wallis test and Dunn’s multiple comparison test using GraphPad Prism 6 (GraphPad Software, CA, USA). Data were expressed as average ± SD. P values < 0.05 were considered statistically significant.

Results

MSC characteristics

We first examined whether the human synovial cells we prepared had the characteristics of MSCs. Synovial cells were spindle shaped (Fig. 2a) and formed cell colonies 14 days after the initial plating (Fig. 2b). They stained positive for CD 44, 73, 90, and 105 and negative for CD45 (Fig. 2c). They showed chondrogenesis, adipogenesis, and calcification (Fig. 2d). These features confirmed that the synovial cells we prepared had characteristics of MSCs [14].

Fig. 2: Confirmation of stem cell characteristics of a human synovial MSCs preparation.
figure 2

a Cell morphology. b Colony morphology. c Representative histograms for stem cell surface markers. d Multidifferentiation.

Full size image

Time course of changes in the proportions of adhered MSCs

In our clinical transplantation procedure, we place a synovial MSC suspension onto the repaired meniscus for 10 min. We confirmed that 10 min was an adequate period to allow cell adhesion by conducting a time course to examine the proportions of MSCs that adhered to the meniscus. By 10 min, the proportion of adhered cells was 33% (Fig. 3). An unexpected finding was that 28% of the synovial MSCs had already adhered to the meniscus immediately after the placement of the cell suspension onto the meniscus. The proportion of adhered cells gradually increased with time and was significantly increased after 6 h. It reached 96% by 24 h.

Fig. 3: Time course for adhesion of human synovial MSCs onto an abraded porcine meniscus.
figure 3

A synovial MSC suspension was placed onto the meniscus, and the numbers of adhering cells were calculated by counting the non-adherent cells washed from the meniscus. Data are expressed as the average ± standard deviation (n = 6). The proportion of adhered cells was significantly increased after 6 and 24 h.

Full size image

Histological images

Abrasion of the porcine meniscus caused a roughening of the normal smooth surface (Fig. 4a). Histological staining revealed the presence of some cells adhering to this roughened surface by 60 min after cell placement. The cell numbers on the surface of the meniscus increased after 6 h and were further increased at 24 h. Immunostaining revealed that the adhered cells were of human, not porcine, origin (Fig. 4b).

Fig. 4: Histological images of human synovial MSCs during adhesion to an abraded porcine meniscus.
figure 4

a HE staining showing human synovial MSCs adhering to the porcine meniscus after the placement of the MSC suspension. b Immunostaining after 24 h showing stem cell markers. (i) Human vimentin staining confirming the human origin of the adhered cells. (ii) DAPI staining of the nuclei. (iii) Merged images.

Full size image

SEM overview images

The histological analysis did not confirm the cell adhesion within 10 min observed in the in vitro time course study, so we undertook further examinations using SEM. Again, the abraded meniscus did not show the smooth surface of the normal meniscus, but instead had a fibrillated appearance. Low-magnification SEM images revealed several round cells on the meniscus immediately after the placement of the cell suspension, and the numbers of these cells increased after 10 min, with further increase after 60 min. A mixture of round cells and semi-flat cells covered the meniscus by 6 h. Flat cells had spread over the entire surface by 24 h (Fig. 5).

Fig. 5: SEM images of human synovial MSCs during adhesion onto an abraded porcine meniscus.
figure 5

A human synovial MSC suspension was placed onto the meniscus, and the surface of the meniscus was observed by SEM. The most representative area is magnified and shown in the smaller panel. The cells attached to the meniscus in increasing numbers and their morphology changed with time.

Full size image

SEM images of synovial MSCs in the flattened state

Low-magnification SEM images also showed changes in the morphology of the adhered cells with time. We used the SEM images of synovial MSCs to quantify the process of cell flattening. The morphology of the synovial MSCs was classified into three types: round, semi-flat, and flat cells (Fig. 6a, Supplementary Fig. 1). The proportion of round cells was 100% immediately after placement onto the meniscus, and it decreased to 21% after 6 h and finally to 10% after 24 h (Fig. 6b). The proportion of semi-flattened cells increased to 36% after 6 h, and then decreased to 9% after 24 h. The proportion of flattened cells increased to 43% after 6 h and to 80% after 24 h.

Fig. 6: SEM images of morphological changes in human synovial MSCs during adhesion to an abraded porcine meniscus.
figure 6

The synovial MSC suspension was placed onto the meniscus, and the morphological changes in the MSCs were followed over time. a Representative SEM images showing synovial MSCs undergoing a flattening process. b Time course showing changes in the ratios of round, semi-flattened, and flattened cells. The numbers of round cells significantly decreased at 24 h, the numbers of semi-flat cells significantly increased at 6 h, and the numbers of flat cells significantly increased at 24 h. Overall, 50 cells were analyzed from each donor, and the results were plotted individually. The average ± SD for six donors is also shown. *p < 0.05 by Kruskal–Wallis test, followed by Dunn’s multiple comparisons.

Full size image

SEM images of cell processes of synovial MSCs

The dynamic changes in cell morphology over time suggested that cell processes, known to be involved in adhesion, had also changed. Synovial MSCs could be divided into four types, depending on the presence of pseudopodia or microspikes, in the floating state and immediately after, 10 min after, and 60 min after the suspension was placed onto the meniscus (Fig. 7a). The proportion of the cells with pseudopodia was zero in floating state and immediately after placement, but it slightly increased to 3% after 10 min and then rapidly increased to 33% after 60 min. The proportion of cells with microspikes was 36% in the floating state and increased to 76% immediately after placement, followed by a decrease to 60% after 10 min and a further decrease to 27% after 60 min (Fig. 7b).

Fig. 7: SEM images of the formation of cell processes in human synovial MSCs during adhesion to an abraded porcine meniscus.
figure 7

The synovial MSC suspension was placed onto the meniscus, and the synovial MSCs were examined for cell processes. a Representative SEM images of synovial MSCs with/without microspikes (white triangles) and with/without pseudopodia (white arrows). b Time course for the establishment of microspikes and pseudopodia. The cells with microspikes but without pseudopodia significantly increased immediately after the cell suspension was placed and significantly decreased after 60 min. The cells with pseudopodia significantly increased after 60 min irrespective of the presence of microspikes. Overall, 50 cells were analyzed from each donor, and the results were plotted individually. The average ± SD for six donors is also shown. *p < 0.05 by Kruskal–Wallis test, followed by Dunn’s multiple comparisons.

Full size image

SEM images of synovial MSC cell surfaces

The MSCs were a heterogeneous population with different surface epitopes [10]; therefore, they were hypothesized to have a heterogeneous morphology. In addition, some types of MSCs might be more prone to adhesion. We therefore quantified the heterogeneity of the MSCs based on SEM analysis before and after adhesion to the meniscus. The surface of synovial MSCs could be classified into three types: blebbed, membrane ruffled, and smooth (Fig. 8a, Supplementary Fig. 2). Some cells had only one type of surface and others were a mixture of two or three types. The proportion of the cells in which the bleb was dominant was 60–70% at 10 min after the placement of the cell suspension on the meniscus and decreased to 27% at 60 min (Fig. 8b). Membrane ruffling was dominant in 10–20% of the cells until 60 min. A smooth surface was dominant in 14% of the cells in the floating state and this proportion gradually increased to 49% at 60 min after placement.

Fig. 8: SEM images of the cell surface of human synovial MSCs during adhesion to an abraded porcine meniscus.
figure 8

We quantified the heterogeneity of the human MSCs by SEM analysis before and after adhesion to the meniscus. a Representative SEM images of human synovial MSCs. The surfaces were classified as blebbed, membrane ruffling, and/or smooth areas. A portion of each area is shown. Blebs are shown as black arrows, membrane ruffling as open arrows. b Changes in the ratio of cells with bleb-dominant, membrane-ruffling-dominant, and smooth-dominant areas over time. Bleb-dominant cells significantly decreased and smooth-dominant cells significantly increased after 60 min. Overall, 50 cells were analyzed from each donor, and the results were plotted individually. The average ± SD for six donors is also shown. *p < 0.05 by Kruskal–Wallis test, followed by Dunn’s multiple comparisons.

Full size image

SEM images of the synovial MSC suspension immediately after placement on the meniscus

The mechanisms underlying the adhesion of synovial MSCs onto the meniscus immediately after the placement of the cell suspension were investigated by examining high magnification SEM images. Two main findings were revealed: the presence of microspikes caused the synovial MSCs to frequently become trapped by the meniscal fibers (Fig. 9a). The bodies of the synovial MSCs were also occasionally trapped in the meniscal fibers (Fig. 9b, Supplementary Fig. 3).

Fig. 9: SEM images obtained immediately after the placement of a human synovial MSCs suspension onto an abraded porcine meniscus.
figure 9

a Microspikes of synovial MSC are trapped by the meniscal fibers. b Bodies of synovial MSCs are trapped in the meniscal fibers.

Full size image

Discussion

In this study, we analyzed the adhesion of synovial MSCs onto the meniscus by SEM observations. Immediately after the cell suspension was placed onto the meniscus, ~28% of the synovial MSCs had already adhered to the meniscus. The proportion of adhered cells then gradually increased with time and finally reached 96% after 24 h. Initially, the cells were round, but they became semi-flattened after 6 h and fully flattened after 24 h. Cells with pseudopodia were not initially observed, but a few appeared 10 min after MSC placement and their proportion increased to 32% at 60 min after placement. Bleb-dominant cells were found in proportions of 60–70% after the first 10 min and their proportion decreased to 27% after 60 min. Membrane-ruffling-dominant cells had a proportion of 10–20% after the first 60 min. The proportion of smooth-dominant cells was initially 10–20% immediately after placement, but this increased to 49% after 60 min.

The proportion of adhered cells increased with time, possibly due to pseudopodia formation in the synovial MSCs. Indeed, the proportion of cells with pseudopodia was zero in the floating state and just after placement; it slightly increased to 3% after 10 min, but then rapidly increased to 33% after 60 min. Some reports have indicated that the pseudopodia perform the function of cell adhesion [13].

Ten minutes after the cell suspension was placed onto the meniscus, the proportion of adhered cells was only 28%. This was lower than the proportion of 60% that adhered to a cartilage defect [15] and may reflect differences in cell concentration. The concentration was 107 cells in 100 μL PBS for the previously studied cartilage defect but only 106 cells in 100 μL PBS for the meniscus in the present case. Our previous results showed that cells at high concentration tended to adhere to each other [16], and this promoted adhesion to cartilage defects and menisci [17].

The adhered MSCs showed other morphological differences, including the presence of microspikes. These are slender cytoplasmic projections that extend from migrating cells and have roles in migration, sensing, and cell–cell interaction [9, 18]. Miyashita et al. showed that long microspikes in N1E-115 neuroblastoma cells were positively immunostained with CD47 antibody and that forced expression of CD47 induced the formation of these long microspikes [19]. Several other reports have proposed an involvement of microspikes in early cell adhesion. For example, Adams reported the protrusion of microspikes with filopodia from the cell surface during initial matrix attachment [13]. Similarly, Witkowski et al. reported the appearance of microspikes in human diploid cells within 10 min after the attachment of the cells to a glass surface [20]. Rajaraman et al. reported the presence of microspikes in 22% of WI-38 fibroblasts 10 min after adhesion onto a glass dish [9]. In the present study, microspikes were observed in 36% of the floating synovial MSCs and in 76% of the MSCs immediately after placement on the meniscus. We also showed that microspikes could be physically trapped in the meniscal fibers. The previous reports and our current results provide indirect evidence that microspikes could play an important role in the initial attachment of the MSCs to the roughened surface of the abraded meniscus.

Previous reports indicated that pseudopodia perform an exploratory role for the formation of transient adhesions during migration [13, 21]. In the present study, the proportion of adhered cells increased from 28% at the initial placement to 40% at 60 min after placement, while the proportion of MSCs with pseudopodia increased from 0 to 54% in the same time interval. This indicated that pseudopodium formation was involved in adhesion once the synovial MSCs were physically trapped on the meniscus surface.

MSCs have been reported to exist as heterogeneous populations [10], but no detailed reports have appeared regarding diversity in their cell surface morphology. The presence of blebs has been previously reported by transmission electron microscopy observations of bone marrow MSCs [22], while light microscopy has revealed ruffling [23]. In the present study, the cell surfaces of the synovial MSCs in their round morphological state could be classified into three types: blebbed, membrane ruffled, and smooth. Some MSCs had only one type of surface, while others had a mixture of two or three types. These differences in morphology imply differences in cell function, so clarification of the heterogeneous functions of MSCs could be interesting from the viewpoint of differences in their cell surface morphology.

We propose three clinical relevance from the current study. First, if synovial MSCs with microspikes can be isolated, the proportion of MSCs that adhere to the meniscus will increase, resulting in improved clinical outcome because the more MSCs adhered, the better the meniscus was regenerated in our experimental study [8]. Second, 10 min is practically sufficient for cell adhesion to the meniscus because the proportion of adherent cells increased only 8% from 10 to 60 min. Third, meniscus abrasion will improve cell adhesion because it results in entrapment of the cells.

Our study has three limitations. First, meniscal injury consisted of a model with surface abrasion but without tearing. This differs from the clinical situation because synovial MSC transplantation is used clinically for complex degenerative tears [7]. Second, we used human synovial MSCs as donors but porcine menisci as recipients, so this is a xenotransplantation model. The porcine meniscus closely approximates the human meniscus in size, histology, and biomechanics [24, 25], but some immunological responses might occur. Third, this study was done ex vivo, and in vivo studies may give different results.

In conclusion, ~30% of the MSCs placed as a suspension adhered to the meniscus and the proportion of adhered cells increased with time. The cell morphology was changed dynamically for 24 h during the adhesion of synovial MSCs onto the meniscus. In their round morphological state, synovial MSCs were heterogeneous from a morphological point of view. Microspikes in the synovial MSCs appear to be important in the initial attachment, followed by pseudopodia in adhesion to the meniscus.