Altered Adipose-Derived Stem Cell Characteristics in Macrodactyly

Macrodactyly is a congenital disease characterized by aggressive overgrowth of adipose tissue in digits or limbs frequently accompanied with hyperostosis and nerve enlargement; its pathological mechanism is poorly understood. Adipose-derived stem cells (ASCs) have been extensively studied in tissue engineering and regenerative medicine as an ideal alternative substitute for bone marrow-derived mesenchymal stem cells (BM-MSCs), but their pathological role is largely unknown. In this study, ASCs from macrodactyly adipose tissues (Mac-ASCs) were isolated and compared to ASCs derived from the normal abdominal subcutaneous adipose tissue (Sat-ASCs) for cell morphology, surface marker expression, proliferation rate, and tri-lineage differentiation potential. Despite similar cell morphology and cell surface marker expression, Mac-ASCs showed higher cell proportion in the S phase and increased proliferation compared with Sat-ASCs. Moreover, osteogenic and chondrogenic differentiation capacities were enhanced in Mac-ASCs, with reduced adipogenic potential. In addition, the expression levels of adipogenic genes were lower in undifferentiated Mac-ASCs than in Sat-ASCs. These findings unraveled enhanced proliferation activity, a regression in the differentiation stage, and greater potentiality of ASCs in macrodactyly, which could contribute to hyperostosis and nerve enlargement in addition to adipose tissue overgrowth in patients.


Results
The adipose tissue from macrodactyly has normal histology. To assess whether there was histological abnormalities in the adipose tissue from macrodactyly, we compared H&E stained adipose tissues from macrodactyly and normal abdominal subcutaneous tissues. The size of the adipocytes showed no significant differences between the two groups ( Fig. 1C-E), indicating a normal histology for the adipose tissue in macrodactyly.

Mac-ASCs express common ASC cell surface markers.
To evaluate whether there was any anomaly of ASCs in macrodactyly, we isolated ASCs from adipose tissues of macrodactyly. ASCs from normal abdominal subcutaneous adipose tissues were also obtained for comparison. Interestingly, Mac-ASCs and Sat-ASCs both displayed a spindle shape in the preliminary culture and subcultures, and cell morphology showed no significant change at up to five passages ( Fig. 2A and B). Next, the expression of cell surface markers was assessed in Mac-ASCs and Sat-ASCs. While mock staining showed no signal, up to 90% cells in both Mac-ASC and Sat-ASC groups showed positive staining for the ASC-specific cell surface markers CD105, CD29, and CD90, but the cells were negative for the BM-MSC-specific cell surface marker CD106 (Fig. 2C), consistent with previous findings 9 . These results suggested that there is no differences between the two groups concerning these markers.

Mac-ASCs show enhanced cell proliferation.
To assess the proliferative activity of Mac-ASCs, cells were seeded in 96-well plates and evaluated by the MTT assay at different time points. The growth curves showed that Mac-ASCs had a significantly higher proliferation rate compared with Sat-ASCs (Fig. 3A). To explore the mechanism underlying the differential proliferation rates, cell cycle distribution and apoptosis were assessed by flow cytometry. Mac-ASCs showed increased cell proportion in the S phase and decreased G0/G1 population compared with Sat-ASCs. Cell proportion in the G2/M phase were similar between Mac-ASCs and Sat-ASCs ( Fig. 3B-D). Moreover, the cell proportions in both early and late apoptosis were similar in Mac-ASCs and Sat-ASCs ( Fig. 3E and F). Thus, Mac-ASCs had increased cell proportion in the S phase and enhanced proliferation.
Altered differentiation potential of Mac-ASCs. We then assessed the differences in multi-linage differentiation potential between Mac-and Sat-ASCs. Adipogenic induction was successful because a significant accumulation of lipid droplets stained by Oil Red O was observed in induced Sat- (Fig. 4C) and Mac-ASCs (Fig. 4D), but not in non-induced counterparts ( Fig. 4A and B, respectively). However, the lipid droplets were much less intense and smaller in Mac-ASCs (Fig. 4D) than in Sat-ASCs (Fig. 4B) after induction. To confirm the reduced adipogenic differentiation of Mac-ASCs, the mRNA expression levels of Adiponectin, C/EBPα, and PPARγ were assessed by qRT-PCR; these genes are late, medium, and early white adipose tissue markers, respectively 10-12 . The adipocytes size in six random fields, each from three macrodactyly samples and three controls, were measured and quantified by using the ImagePro software (Version 4.0 Analytik, Germany). Unpaired t-test was applied to assess the significance of difference between the two groups.
Consistent with the phenotypic observation, the mRNA levels of all three genes were significantly lower in Mac-ASCs than in Sat-ASCs after adipogenic induction (Fig. 4E).
Next, we performed osteogenic differentiation experiments. After 14 days of osteogenic differentiation, the cells were stained with Alizarin Red, a specific marker of bone nodules. Intense Alizarin Red staining was observed in induced Sat-ASCs ( Fig. 4H) but not in non-induced counterparts (Fig. 4F), indicating an efficient osteogenic differentiation. Unexpectedly, induced Mac-ASCs exhibited even much stronger Alizarin Red staining (Fig. 4I), suggesting enhanced osteogenic differentiation compared with Sat-ASCs. Consistently, qRT-PCR analysis showed a dramatically increased mRNA expression of the osteogenic marker osteocalcin 13 and moderately increased AKP and RUNX2 mRNA levels 13 , in differentiated Mac-ASCs compared with the levels obtained for Sat-ASCs (Fig. 4J).
Then, we assessed whether chondrogenic differentiation potential was also affected in Mac-ASCs. The protein levels of Collagen II, a chondrogenic differentiation marker, were comparable between induced Mac-ASCs and Sat-ASCs as indicated by IHC (Fig. 4K-N). However, qRT-PCR analysis revealed higher mRNA expression levels of Col2A1 and Aggrecan, another chondrogenic marker 13 in induced Mac-ASCs than in the Sat-ASC group (Fig. 4O). The expression of another chondrogenic marker, Sox9 13 was unchanged (Fig. 4O).
Undifferentiated Mac-ASCs show reduced adipogenic gene expression. The reduced adipogenic, and elevated osteogenic and chondrogenic differentiation potentials of Mac-ASCs may reflect a regression in the adipogenic differentiation commitment. To test this hypothesis, the expression levels of adipogenic genes in undifferentiated Mac-ASCs were evaluated. qRT-PCR analysis revealed that the mRNA expression levels of Adiponectin, C/EBPα, and PPARγ were indeed significantly lower in undifferentiated Mac-ASCs than in Sat-ASCs (Fig. 5), suggesting that Mac-ASCs were less adipogenic differentiation-committed. These findings indicated a regression in the differentiation stage where the cells came to have greater potentiality in Mac-ASCs.

Discussion
Since their first identification in 2001, ASCs have been extensively studied as an ideal substitute for BM-MSCs in tissue engineering and regenerative medicine. However, their pathological role remains elusive. Macrodactyly is a progressive congenital anomaly characterized by aggressive overgrowth of the adipose tissue, frequently accompanied with hyperostosis and nerve enlargement in digits or limbs 1-3 . Its pathological mechanism is poorly understood. In this study, we compared ASCs derived from adipose tissues of macrodactyly and normal abdominal subcutaneous adipose tissues for their properties. Despite similar cellular morphology and cell surface marker expression, Mac-ASCs displayed increased cell proportion in the S phase and enhanced proliferation compared with Sat-ASCs. Meanwhile, elevated osteogenic and chondrogenic, but attenuated adipogenic differentiation potential were found for Mac-ASCs.
Because it is impossible to obtain adipose tissues from normal hands, abdominal subcutaneous adipose tissues obtained in necessary defatting of abdominal skin for grafting operation were selected as normal controls. Human ASCs derived from different anatomic regions have been compared for their proliferative activity and differentiation potential. Schipper et al. 14 reported that ASCs derived from various subcutaneous regions, including the arm, thigh, inguinal region, and abdomen, display comparable proliferation rates and adipogenic differentiation potential. Jurgens et al. 15 demonstrated that ASCs derived from the subcutaneous regions of the abdomen and hip/thigh show comparable proliferation rates, with no significant differences in osteogenic or chondrogenic differentiation potential. Therefore, the altered proliferation activity and differentiation potential of Mac-ASCs are likely not attributable to anatomical differences but reflect the pathological condition of macrodactyly.
The altered differentiation potential of Mac-ASCs was intriguing. Preadipocytes isolated from subcutaneous depots have a higher adipogenic potential than those obtained from omental or mesenteric depots, suggesting that the subcutaneous tissue produces progenitor cells more committed to mature adipocytes 14 . Mac-ASCs displayed enhanced osteogenic and chondrogenic, but reduced adipogenic differentiation potential, suggesting a regression in the adipogenic differentiation commitment of Mac-ASCs. This notion was supported by the decreased expression of adipocyte-specific genes in undifferentiated Mac-ASCs. While tissue overgrowth in macrodactyly is probably a result of elevated ASC proliferation activity, bone hyperostosis and nerve enlargement could be a consequence of the elevated potential of Mac-ASCs. If this is the case, removing ASCs by surgical intervention, such as defatting, could be beneficial in controlling not only the overgrowth of adipose tissue but also the bone hypertrophy and nerve enlargement in the affected digits.
Recent studies have associated macrodactyly with somatic activating mutations of PIK3CA 16,17 . The latter gene encodes the p110a catalytic subunit of PI3K, an essential component of the RTK-PI3K-AKT signaling pathway that is critical for cellular growth and metabolism 18 . PIK3CA mutations frequently occur in many human cancer types, and have been established as causative and key driver of cancer 19 . Interestingly, the expression of H1047R (one of the most frequent mutations of PIK3CA) in lineage-committed basal or luminal cells of the adult mouse mammary gland evokes cell dedifferentiation into a multipotent state that contributes to tumor heterogeneity 20 . The potential functional impact of PIK3CA mutations on the alteration of Mac-ASC characteristics is of particular interest, and should be addressed in future studies.

Methods
Samples. This study was approved by the ethics committee of Shanghai 9th People's Hospital. Adipose tissue samples were obtained from defatting surgery of macrodactyly. Normal controls were abdominal subcutaneous adipose tissues from necessary defatting of abdominal skin for grafting operation. Informed consent was obtained from all patients. Eight samples from macrodactyly patients (six males and two females, each from the hand or foot) and six samples from normal controls (two males and four females, all from the abdomen) were collected; the donors ranged from 2 to 12 years old. H&E staining. The surgically removed adipose tissue samples were fixed in 4% paraformaldehyde, followed by H&E staining, as previously described 21, 22 . ASC isolation and culture. ASCs were isolated from adipose tissues of macrodactyly and normal abdominal subcutaneous adipose tissues as described previously 23 . Briefly, the adipose tissue was minced under sterile conditions, and digested with 0.05% Collagenase II (Sigma-Aldrich, St. Louis, USA) in serum-free DMEM (Hyclone, Logan, USA) at 37 °C for 2-3 h. After digestion, the fat tissue was filtered and centrifuged (1500 rpm at 37 °C for 5 min). The cells were re-suspended in low glucose DMEM (Hyclone, Logan, USA) containing 10% FBS (Gibco, Grand Island, USA) and 100 μg/mL penicillin-streptomycin (Gibco, Grand Island, USA), and cultured at 37 °C in a 5% CO 2 incubator (HEPA class 100, Thermo, Waltham, USA). When the cells became confluent, they were detached with 0.25% trypsin-EDTA (Gibco, Grand Island, USA) and subcultured at the same density for passage.
Immunostaining of cell surface markers and flow cytometry. Immunostaining was performed as previously described 9,24 . Briefly, Mac-ASCs and Sat-ASCs (P1) were harvested by centrifugation, washed once with PBS, and re-suspended in PBS. Approximately 10 5 cells in 1 mL of PBS were incubated with 5 μL of anti-CD106, anti-CD105, anti-CD29, or anti-CD90 antibodies (eBioscience, CA, USA) for 20 min. Then, the cells were washed and re-suspended in 1 mL PBS, and the expression of cell surface markers was analysed by flow cytometry.
Histochemical and immunohistochemical staining. Cells in 6-well plates were fixed with 4% paraformaldehyde for 30 min and washed in PBS 3 times. The adipogenic differentiation group was stained with Oil Red O to evaluate lipid droplets. The osteogenic differentiation group was stained with Alizarin Red for calcium deposit assessment. Immunohistochemical (IHC) staining of Collagen II was performed for the chondrogenic differentiation group. For IHC analysis, the samples were blocked with 3% BSA, and anti-Collagen II primary antibody (rabbit Collagen II, ab34712, 1:200, Abcam, Cambridge, UK) was applied and incubated at 4 °C overnight. After three washes in PBS, horseradish peroxidase (HRP) conjugated detection antibody (GAR, K55007, 1:1, Dako, Denmark) was added with diaminobenzidine tetra-hydrochloride (DAB) as the chromogen. Sections were counterstained with hematoxylin. The slides were analyzed under a light microscope (Eclipse 90i, Nikon, Tokyo, Japan).
RNA extraction and quantitative real-time polymerase chain reaction (qRT-PCR). As previously described 10 , total RNA was extracted using TRIzol Reagent (Life Technologies, CA, USA) without detaching the cells. Reverse transcription was performed from 2 μg of total RNA, with oligo (dT) and Revert Aid Reverse Transcriptase (Thermo Scientific Inc.) following the manufacturer's instructions. The mixture was then incubated at 30 °C for 10 min, 42 °C for 60 min, 95 °C for 5 min, and 5 °C for 5 min. cDNA was amplified using Power SYBR Green PCR (TIANGEN) on a Real-time thermal cycler (Biosystems 7500 Fast Real-time PCR System, Life Technologies, CA, USA); each measurement was performed in triplicate. Relative gene expression levels were obtained by the comparative Ct (cycle threshold) method. Adiponectin, C/EBPα, and PPARγ were used to evaluate the adipogenic potential of cells; AKP, Osteocalcin and Runx2 were employed to assess the osteogenic potential; Aggrecan, SOX9 and Col2A1 were used to determine the chondrogenic potential. GAPDH was used as an internal control. The primers used for qPCR are listed in Table 1. The experiments were performed at least in triplicate.

Statistical analyses.
Data are mean ± standard deviation (SD). Statistical analyses were performed with the GraphPad Prism 5 software (GraphPad Software Inc., CA, USA). Unpaired t-test was applied to assess the difference between adipocytes size of macrodactyly and control tissues. Student's t-test was applied to assess cell proliferation and cell cycle data. The differences among multiple groups in gene expression patterns of adipogenic, osteogenic and chondrogenic induction were evaluated by two-way ANOVA. P < 0.05 was considered statistically significant. Data Availability. No datasets were generated or analysed during the current study.

Ethics, consent and permissions. This study was approved by the ethics committee of Shanghai 9th
People's Hospital (reference: 201580). Informed consent was obtained from all patients, or legal parents or guardians of children. All methods were performed in accordance with the relevant guidelines and regulations. Consent to publish. We have obtained consent from all patients, or legal parents or guardians of children, to report and publish individual patient data.