Identification of Surface Antigens That Define Human Pluripotent Stem Cell-Derived PRRX1+Limb-Bud-like Mesenchymal Cells

Stem cell-based therapies and experimental methods rely on efficient induction of human pluripotent stem cells (hPSCs). During limb development, the lateral plate mesoderm (LPM) produces limb-bud mesenchymal (LBM) cells that differentiate into osteochondroprogenitor cells and form cartilage tissues in the appendicular skeleton. Previously, we generated PRRX1-tdTomato reporter hPSCs to establish the protocol for inducing the hPSC-derived PRRX1+ LBM-like cells. However, surface antigens that assess the induction efficiency of hPSC-derived PRRX1+ LBM-like cells from LPM have not been identified. Here, we used PRRX1-tdTomato reporter hPSCs and found that high pluripotent cell density suppressed the expression of PRRX1 mRNA and tdTomato after LBM-like induction. RNA sequencing and flow cytometry suggested that PRRX1-tdTomato+ LBM-like cells are defined as CD44high CD140Bhigh CD49f−. Importantly, other hPSC lines, including four human induced pluripotent stem cell lines (414C2, 1383D2, HPS1042, HPS1043) and two human embryonic stem cell lines (SEES4, SEES7), showed the same results. Thus, an appropriate cell density of hPSCs before differentiation is a prerequisite for inducing the CD44high CD140Bhigh CD49f− PRRX1+ LBM-like cells.


Introduction
Damaged cartilage is a hallmark of osteoarthritis and rheumatoid arthritis damage cartilage. Although cartilage transplantation therapy aims to regenerate damaged cartilage [1,2], current methods, including microfracture and osteochondral autograft transplantation, are invasive and burdensome for patients.
Surface antigens are cell membrane-localized proteins whose expression patterns vary with cell type [19][20][21][22][23]. Identifying these surface antigens can help assess the frequency of specific cell populations and to purify them. Before chondrogenesis, the CD271 + CD73 − population should be purified from hPSC-derived NC-like cells, and then the CD271 + CD140A + CD73 + CD13 − CDH2 − population can be used to test the chondrogenic capacity of NC-like-derived progenitor cells [10]. Previously, we established a protocol that not only induces almost all hPSC-derived LPM cells into PRRX1 + LBM-like cells but also stably expands them with a high chondrogenic capacity [14]. However, the downregulation of PRRX1 causes the loss of their chondrogenic capacity, indicating that surface antigens defining PRRX1 + LBM-like cells immediately prior to expansion should be identified to exclude the possible contamination of low chondrogenic cells.
In this study, we found that a high density of pluripotent cells dramatically suppresses PRRX1 expression after LBM-like induction. Using a PRRX1-tdTomato reporter hPSC line, we showed that CD44 high CD140B high CD49f − can be used to identify PRRX1-tdTomato + LBM-like cells. Finally, we confirmed that PRRX1 + LBM-like cells derived from other hPSC lines are also CD44 high CD140B high CD49f − .

Effect of Pluripotent Cell Density before LBM-like Induction on PRRX1 Expression
Previously, we ontogenically induced PRRX1 + LBM-like cells from hPSCs and established their expansion method (Figure 1a) [14]. Although almost all LPM cells became PRRX1 + LBM-like cells following our protocol, conditions that decrease their induction efficiency have not been identified. We cultured PRRX1-tdTomato reporter hPSCs for two different time periods-three days (the condition previously used [14]) and seven daysduring the pluripotent state to determine their effect on inducing PRRX1 + LBM-like cells. Seven-day cultures yielded higher cell densities than three-day cultures (Figure 1b). Pluripotent: 7 days showed higher cell density than Pluripotent: 3 days at Day 0 (pluripotent state). Following LBM-like cell (Day 4) induction under both conditions, PRRX1 expression dramatically decreased after seven days of culture ( Figure 1c). Pluripotent: 7 days-derived cells were called PRRX1 low Day 4 cells hereafter, as induction of PRRX1-tdTomato + LBM-like cells was also suppressed in Pluripotent: 7 days (Figure 1d-f).

Identification of Surface Antigens Defining PRRX1 + LBM-like Cells
RNA-seq analysis of global gene expression was performed to identify surface antigens that distinguish PRRX1 + LBM-like cells from PRRX1 low Day 4 cells. Volcano plot analysis was performed to find differences in surface antigen expression between PRRX1 + LBM-like cells and PRRX1 low Day 4 cells (Figure 2a). We identified 21 genes upregulated in PRRX1+ LBM-like cells and 20 genes upregulated in PRRX1 low Day 4 cells that encode cell surface markers. Flow cytometry analysis showed that CD44 (HCAM) and CD140B (PDGFRB) were highly expressed in conjunction with PRRX1-tdTomato (Figure 2b,c). In contrast, CD49f (ITGA6) expression was detected only in PRRX1 low Day 4 cells.   ExpLBM with high chondrogenic capacity is assessed by CD90 high CD140B high CD82 low , but surface antigens that define PRRX1 + LBM-like cells have not been identified. (b) Phase images at pluripotent state just before LBM-like induction. PRRX1-tdTomato reporter hPSCs were seeded and maintained on culture dishes for 3 or 7 days. Note that cells at Pluripotent: 7 days had high cell density.

Identification of Surface Antigens Defining PRRX1 + LBM-like Cells
RNA-seq analysis of global gene expression was performed to identify surface antigens that distinguish PRRX1 + LBM-like cells from PRRX1 low Day 4 cells. Volcano plot analysis was performed to find differences in surface antigen expression between PRRX1 + LBM-like cells and PRRX1 low Day 4 cells (Figure 2a). We identified 21 genes upregulated in PRRX1+ LBM-like cells and 20 genes upregulated in PRRX1 low Day 4 cells that encode cell surface markers. Flow cytometry analysis showed that CD44 (HCAM) and CD140B (PDGFRB) were highly expressed in conjunction with PRRX1-tdTomato (Figure 2b,c). In contrast, CD49f (ITGA6) expression was detected only in PRRX1 low Day 4 cells.

Detection of CD44 high CD140B high CD49f − LBM-like Cells Derived from hiPSC Lines
We used four hiPSC lines, including 414C2 (the parental line of PRRX-tdTomato reporter hiPSC), 1383D2, HPS1042, and HPS1043, to analyze the expression of CD44, CD140B, and CD49f after LBM-like induction. Each cell line showed that CD44 and CD140B were highly expressed in LBM-like cells (Pluripotent: 3 days) but CD49f was only expressed in PRRX1 low Day 4 cells (Pluripotent: 7 days) (Figures 3a and 4).

Detection of CD44 high CD140B high CD49f − LBM-like Cells Derived from hiPSC Lines
We used four hiPSC lines, including 414C2 (the parental line of PRRX-tdTomato reporter hiPSC), 1383D2, HPS1042, and HPS1043, to analyze the expression of CD44, CD140B, and CD49f after LBM-like induction. Each cell line showed that CD44 and CD140B were highly expressed in LBM-like cells (Pluripotent: 3 days) but CD49f was only expressed in PRRX1 low Day 4 cells (Pluripotent: 7 days) (Figures 3a and 4).

Detection of CD44 high CD140B high CD49f − LBM-like Cells Derived from hESC Lines
Next, we tested two hESC lines, including SEES4 and SEES7, to analyze the expression of CD44, CD140B, and CD49f after LBM-like induction. Each cell line showed that CD44 and CD140B were highly expressed in LBM-like cells (Pluripotent: 3 days), but CD49f was only expressed in PRRX1 low Day 4 cells (Pluripotent: 7 days) (Figures 3b and 4).

Detection of CD44 high CD140B high CD49f − LBM-like Cells Derived from hESC Lines
Next, we tested two hESC lines, including SEES4 and SEES7, to analyze the expression of CD44, CD140B, and CD49f after LBM-like induction. Each cell line showed that CD44 and CD140B were highly expressed in LBM-like cells (Pluripotent: 3 days), but CD49f was only expressed in PRRX1 low Day 4 cells (Pluripotent: 7 days) (Figures 3b and  4).

Discussion
hPSC-based therapies show great potential for clinical applications, particularly cartilage regeneration. However, methods must be established to assess and increase the purity of differentiated cells to avoid teratoma formation in vivo [24]. hPSCs have high glycolytic activity, and their metabolism gradually shifts to oxidative phosphorylation

Discussion
hPSC-based therapies show great potential for clinical applications, particularly cartilage regeneration. However, methods must be established to assess and increase the purity of differentiated cells to avoid teratoma formation in vivo [24]. hPSCs have high glycolytic activity, and their metabolism gradually shifts to oxidative phosphorylation during differentiation [25,26]. Several groups use metabolic shifts to kill undifferentiated hPSCs and increase the purity of differentiated cells [27,28], but surface antigens that enable us to understand their induction efficiency have not been identified. Here, we developed a surface antigen-based quality assurance to test the purity of hPSC-derived LBM-like cells. Although our previous study showed that almost all hPSC-derived LPM cells become PRRX1 + LBM-like cells [14], the expression of PRRX1 significantly decreased in response to the longer expansion time of hPSCs before LBM-like induction (Figure 1). Colonies of hPSCs at Pluripotent: 7 days were large and highly condensed, unlike Pluripotent: 3 days (Figure 1b). Cell density determines stem cell fate, as preculture at high cell density also changes cell sensitivity to cytokine stimulation [29][30][31][32]. Our results indicated that high cell density in the pluripotent state negatively affects the induction of PRRX1 + LBM-like cells from LPM cells.
RNA sequencing and flow cytometry identified hPSC-derived PRRX1 + LBM-like cells as CD44 high CD140B high CD49f − (Figures 2-4). Freshly isolated mouse limb bud mesenchymal cells express surface antigens such as Sca1, CD105, CD90, CD73 [17]. Although the human ortholog of mouse Sca1 does not exist, we found that LBM-like cells express CD90 but do not express CD73 and CD105 (data not shown). CD44 is a hyaluronan receptor expressed by several stem and differentiated cells [33,34]. Although CD44 is highly expressed in the apical ectodermal ridge to support the proliferation of LBM cells [35], a small population of LBM cells is also CD44 + [17]. CD44 not only is a target gene in the WNT/beta-catenin signaling pathway but also increases its activity [36]. Our previous study showed that WNT/beta-catenin signaling upregulates PRRX1 expression, so CD44 may help LPM cells differentiate into LBM-like cells.
CD140B is a receptor for platelet-derived factors, and its surface level on Expandable LBM-like cells (ExpLBM) decreases with the downregulation of PRRX1 and chondrogenic capacity [14]. During development, LPM cells differentiate into cardiac mesoderm or LBM cells [12]. Consistent with ExpLBM, CD140B levels in LBM-like cells correlate positively with PRRX1 levels, suggesting that PRRX1 transcriptionally activates CD140B. CD49f is a marker of stem/progenitor cell populations [37], including cardiac stem cells, and was expressed here only in PRRX1 low Day 4 cells. High cell density inhibits WNT signaling to promote cardiomyocyte differentiation from hiPSCs [38] and the expression level of PRRX1 in cardiac mesoderm is lower than that of forelimb mesoderm [18,39], indicating that high cell density may have promoted the cardiac mesoderm-directed differentiation of LPM cells.

Materials and Methods
Cell culture. The hPSCs were cultured and maintained using StemFit (AK02N, Ajinomoto, Tokyo, Japan). Before reaching subconfluency, the cells were dissociated with TrypLE Select (Thermo Fisher, Waltham, MA, USA)/0.25 mM EDTA and suspended in StemFit containing 10 µM Y27632 (Wako, Tokyo, Japan). The cells (1 × 10 4 ) were then suspended in StemFit containing 10 µM Y27632 and 8 µL iMatrix511 (human laminin-511 E8 fragment, Nippi, Tokyo, Japan) and added to a 6-cm dish. The culture medium was replaced on the next day with fresh StemFit without Y27632. The medium was changed every two days until the next passage. The following hPSC lines were used: 414C2 hiPSCs were provided by the Center for iPS Cell Research and Application, Kyoto University, Japan; 1383D2 hiPSCs, HPS1042 hiPSCs, HPS1043 hiPSCs, SEES4 hESCs, and SEES7 hESCs were donated by RIKEN BRC (Tokyo, Japan).
Differentiation of LBM-like cells. The differentiation of hPSCs was performed as described previously [14]. In brief, hPSCs (3 × 10 4 ) were suspended in 1 mL of StemFit containing 10 µM Y27632, and 4 µL of iMatrix511 were added to a 3.5-cm culture dish. The culture medium was replaced the next day with fresh StemFit without Y27632. After culturing for two or six days, the cells were differentiated to mid-primitive streak, LPM and LBM-like cells as shown in Figure 1b.
Real-time quantitative RT-PCR (qPCR). Total RNA was extracted using an RNeasy kit (QIAGEN, Germantown, MD, USA), and cDNA was synthesized using M-MLV Reverse Transcriptase (Thermo Fisher, Waltham, MA, USA) and oligo-dT primers (Sigma-Aldrich, St. Louis, MO, USA). cDNAs were then used as templates for qPCR analysis with gene-specific primers. qPCR was performed using an AriaMX Real-Time PCR System (Agilent, Santa Clara, CA, USA). The cycle parameters were as follows: denaturation at 95 • C for 30 s, annealing at 62 • C for 30 s, and elongation at 72 • C for 30 s. The expression level of each gene was calculated using the 2 −∆∆Ct method. The primer sequences are as follows; ACTB Forward-5 AGAAAATCTGGCACCACACC3 , ACTB Reverse-5 AGAGGCGTACAGGGATAGCA3 , PRRX1 Forward-5 TGATGCTTTTGTGCGA-GAAGA3 , PRRX1 Reverse-5 AGGGAAGCGTTTTTATTGGCT3 .
Flow cytometry. Following dissociation with Accutase (Thermo Fisher, USA), 1 × 10 5 cells were suspended in 100 µL of 2% FBS/PBS and incubated with fluorophore-conjugated antibody (200×) for 1 h on ice. The cells were then washed, and fluorescence was detected using a CytoFLEX S flow cytometer (Beckman Coulter, Brea, CA, USA). Data were analyzed using FlowJo v10. RNA-sequence analysis. Total RNA was extracted using an RNeasy kit (Qiagen, USA), and sequencing libraries were prepared using a KAPA RNA HyperPrep Kit with RiboErase (HMR) (Kapa Biosystems, Wilmington, MA, USA) and a SeqCap Adapter Kit (Set A or Set B, Roche) following the manufacturer's instructions. The sequencing libraries were transferred to a GENEWIZ and loaded onto a HiSeq 2500 system (Illumina, San Diego, CA, USA) for sequencing. All sequence reads were extracted in FASTQ format using the CASAVA 1.8.4 pipeline. Trimmomatic (version 0.36) was used to remove adapters and filter raw reads <36 bases as well as leading and trailing bases with < quality 20. Filtered reads were mapped to hg19 using HISAT2 software (version 2.1.0). Raw counts for each gene were based on sense-strand data obtained using featureCounts software from the Subread package. RUVSeq (release 3.10) was used for further normalization to account for sample variations. Differentially expressed genes were identified using DESeq2 analysis with a threshold of padj < 0.01 and abs (log 2 FC) > 1. Principal component analysis and a heatmap of gene expression levels of each differentially expressed marker were analyzed using the R prcomp function and hclust function (R version 3.6.1). The raw and processed RNA-seq data were deposited in the NCBI GEO database under the accession number (GSE197373).
Study approval. The Ethics Committee of Okayama University Graduate School of Medicine, Dentistry and Pharmaceutical Sciences, approved the experimental protocols for studies of human subjects. Written informed consent was provided by each donor.
Statistical analysis. Data were analyzed using Prism 9. All data were acquired by performing biological replicates of three independent experiments and are presented as the mean ± standard error of the mean. Statistical significance was determined using a two-tailed t-test using the Bonferroni method.

Conclusions
Our study showed that the cell density of hPSCs should be optimized before differentiation. Although human mesenchymal stromal cells (hMSCs) are widely used for cartilage regeneration, their invasive isolation from different tissues and donor-dependency inhibit stable results [40][41][42][43]. Our differentiation protocol mimics the human developmental process and provides stable results, including quality assurance of chondrogenesis, which are seldom achieved using hMSC-based current therapies. This study supports hPSC-based skeletal regeneration or disease therapy and offers novel insights on developmental biology and stem cell research.  Takarada). These funders had no role in the study design, data collection and analysis, decision to publish, or preparation of the manuscript. Data Availability Statement: All data generated during and/or analyzed during this study are included in this published article.