NFATc1 marks articular cartilage progenitors and negatively determines articular chondrocyte differentiation

The origin and differentiation mechanism of articular chondrocytes remain poorly understood. Broadly, the difference in developmental mechanisms of articular and growth-plate cartilage is still less elucidated. Here, we identified that the nuclear factor of activated T-cells cytoplasmic 1 (NFATc1) is a crucial regulator of articular, but not growth-plate, chondrocyte differentiation during development. At the early stage of mouse knee development (embryonic day 13.5), NFATc1-expressing cells were mainly located in the flanking region of the joint interzone. With development, NFATc1-expressing cells generated almost all articular chondrocytes but not chondrocytes in limb growth-plate primordium. NFATc1-expressing cells displayed prominent capacities for colony formation and multipotent differentiation. Transcriptome analyses revealed a set of characteristic genes in NFATc1-enriched articular cartilage progenitors. Strikingly, the expression of NFATc1 was diminished with articular chondrocyte differentiation, and suppressing NFATc1 expression in articular cartilage progenitors was sufficient to induce spontaneous chondrogenesis while overexpressing NFATc1 suppresses chondrogenesis. Mechanistically, NFATc1 negatively regulated the transcriptional activity of the Col2a1 gene. Thus, our results reveal that NFATc1 characterizes articular, but not growth-plate, cartilage progenitors during development and negatively determines articular chondrocyte differentiation at least partly through regulating COL2A1 gene transcription.


Introduction
The basic mechanism underlying articular cartilage development, particularly the origin and differentiation of articular chondrocytes, remains poorly understood. It is well appreciated that synovial joint tissues, including the articular cartilage, originate from a distinct group of progenitors from those that generate the limb primary cartilaginous anlagen Koyama et al., 2008;Chijimatsu and Saito, 2019. GDF5 is one of the genes widely used for tracking synovial joint and articular cartilage development Decker, 2017. Using reporter mice, multiple groups have displayed that Gdf5-expressing cell lineages form almost all articular chondrocytes Koyama et al., 2008;Rountree et al., 2004;Shwartz et al., 2016;Decker et al., 2017. As GDF5 is expressed in both interzone cells and its flanking cells at the early stage of joint morphogenesis, current results cannot discriminate which site is the origin of articular chondrocytes. Also, GDF5 expression is greatly diminished at the late stage of embryonic development and almost undetectable in articular cartilage in neonatal mice Decker, 2017. Thus, GDF5 cannot be used to track articular cartilage progenitors postnatally.
PRG4 is an articular cartilage progenitor marker in the late stage of synovial joint development Chijimatsu and Saito, 2019;Chagin and Medvedeva, 2017. This gene encodes lubricin, a major component of synovial fluid and responsible for joint lubricity Coles et al., 2010. PRG4 is detected from the stage of joint cavitation and is predominantly expressed in the surficial layer of developed articular cartilage Rhee et al., 2005. Several studies exploited Prg4 CreERT2 reporter mice to track postnatal articular cartilage development and identified this gene as a marker for postnatal and adult articular cartilage progenitors Decker et al., 2017;Kozhemyakina et al., 2015. Since PRG4 starts to predominantly express and function at the late stage of articular cartilage development, it does not label the primary progenitors of articular cartilage. Several other molecules, such as Sox9, Dkk3, and Tgfbr2, were also utilized to track articular cartilage and synovial joint development Shwartz et al., 2016;Decker et al., 2017;Li et al., 2013, but none of these molecules have been shown to specifically and constantly label articular cartilage progenitors and to be able to distinguish the origin of articular chondrocytes.
In addition to the origin of articular chondrocytes, molecular mechanisms regulating articular chondrocyte differentiation remains largely unknown. In particular, the transcriptional regulation of articular eLife digest Within the body are about 300 joints connecting bones together. Many factorsincluding trauma, inflammation, aging, and genetic changes -can affect the cushion tissue covering the end of the bones in these joints known as articular cartilage. This can lead to diseases such as osteoarthritis which cause chronic pain, and in some cases disability.
To treat such conditions, it is essential to know how cells in the articular cartilage are formed during development. In the embryo, most cells come from groups of progenitor cells that are programmed to produce specific types of tissue. But which progenitor cells are responsible for producing the main cells in articular cartilage, chondrocytes, and the mechanisms that govern this transformation are poorly understood.
In 2016, a group of researchers found that the gene for the protein NFATc1, which is important for building bone, is also expressed in a group of progenitor cells at the site where ligaments insert into bone in mice. Inactivation of NFATc1 in these progenitor cells has also been shown to cause abnormal cartilage to form, a condition termed osteochondromas. Building on this work, Zhang, Wang et al.including some of the researchers involved in the 2016 study -set out to find whether NFATc1 is also involved in the normal development of articular chondrocytes.
To investigate, the team used genetically modified mice in which any cells with NFATc1 also had a green fluorescent protein, and tracked these cells and their progeny over the course of joint development. This led them to discover a group of NFATc1-containing progenitor cells that gave rise to almost all articular chondrocytes in the knee joint.
Further experiments revealed that when NFATc1 was removed, this made the progenitors become articular chondrocytes very quickly. In contrast, when the cells had excess amounts of the protein, the formation of articular chondrocytes was significantly reduced. This suggests that the level of NFATc1 governs when progenitors develop into articular chondrocytes.
These findings have provided a way to track the progenitors of articular chondrocytes throughout development and study how articular cartilage is formed. In the future, this work could help researchers develop treatment strategies for osteoarthritis and other cartilage-based diseases. However, before this can happen, further work is needed to confirm that the effects observed in this study also relate to humans. chondrocyte differentiation is far from clear. SOX9 is essential in multiple steps of chondrogenesis, but it was originally and mainly studied in growth-plate chondrocytes Lefebvre and Dvir-Ginzberg, 2017. Although SOX9 is also expressed in articular cartilage and is essential for maintaining adult articular cartilage homeostasis Haseeb et al., 2021, its detailed functions and mechanisms in articular cartilage development remain to be elucidated. Also, SOX9 starts to express in mesenchymal cells from the very early stage of limb development before the cartilage template formation and it alone is not sufficient to induce chondrogenesis Lefebvre and Dvir-Ginzberg, 2017;Akiyama et al., 2005. Therefore, the identification of a core transcriptional regulator of articular chondrocyte differentiation is paramount for understanding the basic mechanism of articular cartilage development and exploring new strategies for treating disorders of articular cartilage.
The nuclear factor of activated T-cells, cytoplasmic 1 (NFATc1) is one of the five members of the NFAT family, which share a similar DNA binding domain of approximately 300 amino acid residues Hogan et al., 2003;Vaeth and Feske, 2018. NFAT signaling plays a broad function in various physiological and pathological processes, including immune cell differentiation and functions, cardiac valve development, and cancer progression and metastasis Hogan et al., 2003;Mancini and Toker, 2009. In the skeletal system, NFATc1 is critical for osteoclast differentiation and functions Aliprantis et al., 2008;Takayanagi, 2007 and is also involved in osteoblast differentiation by cooperating with the Osterix gene Koga et al., 2005. Intriguingly, NFATc1 expression was found in the superficial layers of articular cartilage as well and decreased in human osteoarthritic cartilage Greenblatt et al., 2013. Following these studies, we recently identified a function of NFATc1 in restricting osteochondroma formation from entheseal progenitors Ge et al., 2016, revealing that NFATc1 is a suppressor of chondrogenesis in these cells.
In this study, we unexpectedly found that NFATc1 constantly labels articular cartilage progenitors throughout embryonic development and postnatal growth. The expression of NFATc1 is diminished with articular chondrocyte differentiation, and suppression of NFATc1 in articular cartilage progenitors is sufficient to induce spontaneous chondrocyte differentiation through regulating the transcriptional activity of the Col2a1 gene. These findings provide novel insights into the identity and origin of articular cartilage progenitors and identify a fundamental function of NFATc1 in determining physiological articular chondrocyte differentiation.

Results
Articular cartilage is derived from NFATc1-expressing progenitors Following our previous discovery that NFATc1 identifies entheseal progenitors at the site of ligaments inserted onto the bone Ge et al., 2016, we unexpectedly found that in Nfatc1 Cre ;Rosa26 mTmG/+ dual-fluorescence reporter mice, the majority of articular chondrocytes expressed green fluorescence protein (GFP) at 8 weeks of age [ Figure 1(A and B), 90.55 ± 6.38%, n=5 mice]. As in this genetic reporter mouse line, both Nfatc1-expressing cells and their progenies express GFP, this finding suggests that articular chondrocytes were either expressing Nfatc1 or derived from Nfatc1-expressing progenitors. To clarify the expression pattern of NFATc1 during articular cartilage development, we mapped GFP + cells in Nfatc1 Cre ;Rosa26 mTmG/+ mice at the early stage of knee joint morphogenesis (E13.5), postnatal day 0 (P0), and 2 weeks of age [ Figure 1(A)]. At E13.5, GFP + cells were mainly localized to the flanking region of the joint interzone with only sporadic distribution in the interzone site. We further examined the expression of NFATc1 at this stage by crossing the tamoxifen-induced Nfatc1 CreERT2 mouse line with Rosa26 mTmG/+ mice to generate Nfatc1 CreERT2 ;Rosa26 mTmG/+ reporter mice, in which the real-time expression of NFATc1 could be reflected by GFP shortly after tamoxifen pulse. The localization of Nfatc1-expressing cells surrounding the joint interzone was verified after administering tamoxifen to Nfatc1 CreERT2 ;Rosa26 mTmG/+ mice at E11.5 and sampling at E13.5 [ Figure 1 In neonatal Nfatc1 Cre ;Rosa26 mTmG/+ mice (P0), GFP + cells consisted of a portion of cells in the presumptive articular cartilage site [ Figure 1A]. Strikingly, at 2 weeks of age, most articular chondrocytes turned out to be GFP + , similar to that at 8 weeks of age [ Figure 1(A)]. To clarify the real-time expression of NFATc1 in articular cartilage at 2 weeks and 8 weeks of age, we used Nfatc1 CreERT2 ;Ro-sa26 RFP/+ reporter mice, in which the expression of NFATc1 is reflected by red fluorescence protein (RFP) shortly after tamoxifen administration. Of interest, with tamoxifen pulse, RFP + cells were found Figure 1. Articular cartilage is derived from NFATc1-expressing progenitors. (A) Confocal microscopy images showing the distribution of GFP + cells during articular cartilage development at the knee of Nfatc1 Cre ;Rosa26 mTmG/+ mice (n=5 animals for each age, two knee joints per animal). Arrow indicating the main location of GFP + cells at the knee at embryonic day 13.5 (E13.5). P0, postnatal day 0. (B) Quantification of GFP + cells in the articular cartilage of Nfatc1 Cre ;Rosa26 mTmG/+ mouse knee at 8 weeks of age (n=5 animals, one knee joint per animal). AC, articular cartilage. (C) Representative confocal images demonstrating the distribution of RFP + cells in the articular cartilage at 2 weeks and 8 weeks of age in Nfatc1 CreERT2 ;Rosa26 RFP/+ mice 48 hrs after tamoxifen pulse for 5 consecutive days (n=3 mice for each age, two knee joints per animal). (D) Quantification of RFP + cells in the articular cartilage of Nfatc1 CreERT2 ;Rosa26 RFP/+ mouse knee at 2 weeks and 8 weeks of age (n=3 mice for each age, one knee joint per animal). Notably, there were no GFP + chondrocytes in the primordium of growth-plate cartilage at E13.5 in both Nfatc1 Cre ;Rosa26 mTmG/+ and tamoxifen-induced Nfatc1 CreERT2 ;Rosa26 mTmG/+ mice [ Figure 1(A) and Figure 1-figure supplement 1A], suggesting that NFATc1-expressing cells do not generate the cartilaginous primordium of growth-plate. We did not detect GFP + cells in articular cartilage in Rosa-26 mTmG/+ control mice and Nfatc1 CreERT2 ;Rosa26 mTmG/+ mice without tamoxifen induction [ Figure 1figure supplement 1D], suggesting that there was no Cre leakage in the articular cartilage in these two reporter mouse lines. Together, these results reveal that articular chondrocytes are derived from NFATc1-expressing progenitors and NFATc1 expression is diminished with articular cartilage development.

Colony formation and multipotent differentiation of NFATc1-expressing progenitors
The lineage tracing data in Nfatc1 Cre and Nfatc1 CreERT2 reporter mice suggest that NFATc1 characterizes articular cartilage progenitors. In this context, the fluorescence-labeled cells after tamoxifen-induced recombination in Nfatc1 CreERT2 reporter mice should be able to form in vivo cell clones in or next to the articular cartilage with development. To verify this assumption, we exploited the Nfatc1 CreERT2 ;Ro-sa26 mTmG/+ double-fluorescence reporter mouse line and administered two dosages of tamoxifen to dams at P0 and P1, respectively. One week following the tamoxifen pulse, GFP + cells were detected at the presumptive articular cartilage site at the mouse knee [ Figure 2(A)]. Local GFP + cell clusters with 3-6 cells each could be observed in articular cartilage by 2 weeks and 8 weeks of age [Figure 2(A and B)]. Notably, GFP + cell clusters were also found in the meniscus, synovial lining, and ligament [Figure 2(A and B)], suggesting that NFATc1 also marks progenitor cells for joint tissues other than articular cartilage. Indeed, in Nfatc1 Cre ;Rosa26 mTmG/+ mice, GFP + cells also formed the meniscus, synovial lining, ligament, and primordium of the patella at the knee [ Figure 1(A) and Figure 2

(C)].
To further characterize the colony formation capacity of NFATc1-expressing articular cartilage progenitors, we cultured and sorted GFP + cells and their counterparts (GFPcells) from the knee of neonatal Nfatc1 Cre ;Rosa26 mTmG/+ mice [ Figure 3(A)]. The ex vivo colony formation assay showed that GFP + cells formed remarkably more numerous and larger cell clones in comparison with GFPcells when plated at the same cell densities and cultured for the same time period [ Figure 3 To study the differentiation potentials of NFATc1-expressing articular cartilage progenitors, we put GFP + and GFPcells under chondrogenic, osteogenic, and adipogenic differentiation conditions, respectively. Notably, GFP + cells displayed a much higher potential to differentiate toward chondrocytes, osteoblasts, and adipocytes compared to GFPcells [ Figure 3(C-E)]. A more striking difference was noticed under the context of chondrocyte differentiation: in the 3D cell-pellet culture model, GFP + cells always grew into larger pellets as shown by the diameter of cell pellets and displayed a (E) Immunohistochemistry detecting the expression of NFATc1 during mouse articular cartilage development (n=3 mice for each age, two knee joints per animal). Data are mean ± SD of results from five or three animals; scale bars, 200 μm except for the right three images in (E), 50 μm.
The online version of this article includes the following source data and figure supplement(s) for figure 1: Source data 1. Quantification data for GFP + or RFP + cells in articular cartilage.    Taken together, these results demonstrate the intrinsic capacities of colony formation and multipotent differentiation of NFATc1-expressing articular cartilage progenitors.

Transcriptional profile of NFATc1-enriched articular cartilage progenitors
Next, we sought to dissect the molecular signature of NFATc1-enriched articular cartilage progenitors. In order to minimize the influence of differentiated cells in GFP + and GFPcell populations, singleclone cells were sorted at the first passage (P1), amplified for one more passage, and subjected to transcriptome analysis at P2 [ Figure  Cell surface markers are important in identifying and sorting progenitor or stem cells. Transcriptome analyses showed that both GFP + and GFPprogenitors expressed several surface markers of cells of mesenchymal origin, including Cd9, Sca1, Thy1, Cd73, Cd166, Cd200, and Cd51, but not hematopoietic or endothelial markers Cd11b, Cd45, or Cd31 (Supplementary file 1). When compared with GFPcells, GFP + cells displayed higher expression of Cd105, Cd10, and Cd13 and lower expression of . Combined, these data identify a set of genes preferentially expressed in NFATc1-expressing articular cartilage progenitors and provide a perspective to understand the transcriptional signature of these progenitors.

NFATc1 negatively regulates articular chondrocyte differentiation
The function of NFATc1 in articular cartilage progenitors remains unclear. As aforementioned, the lineage tracing of Nfatc1-expressing cells showed that most articular chondrocytes were GFP + in Nfatc1 Cre ;Rosa26 mTmG/+ mice at 8 weeks of age, but the real-time expression of NFATc1 was confined to the superficial layers of articular cartilage as shown by RFP expression after tamoxifen pulse in being cultured in the chondrogenic differentiation medium for 3 weeks. Isotype as a negative control for COL2A1 antibody. The maximum diameter of cell pellets reflecting the proliferative capacity of GFP + and GFPcells. n=9 with cells from three animals, three replicates for each. (D) Alizarin red staining and gene expression analysis of Ibsp and Sp7 demonstrating the osteogenic potential of GFP + and GFPcells after being cultured in the osteogenic differentiation medium for 4 weeks. n=6 with cells from three animals, two replicates for each, nonparametric Mann-Whitney test for colony counting data, two-way ANOVA followed by Sidak's tests for gene expression data, experiments repeated twice. (E) Oil red O staining and gene expression analysis of Fabp4 and Lpl displaying adipogenesis in GFP + and GFPcells after being cultured in the adipogenic differentiation medium for 10 days. n=6 with cells from three animals, two replicates for each, nonparametric Mann-Whitney test for colony counting data, two-way ANOVA followed by Sidak's tests for gene expression data, experiments repeated twice. (F) Schematic illustration and histology respectively showing transplantation of GFP + cells along with Matrigel matrix underneath the dorsal skin of severe combined immune-deficient mice and the formation of chondrocytes, chondrocyte clusters, and hypertrophic cartilage-like structure (arrows) 4 weeks later. Images are representative of six animals, with GFPcells as the control (results shown in Figure 3-figure supplement 1B). All data are mean ± SD. Scale bars, 400 μm (C), 500 μm (D), 200 μm (E, F).
The online version of this article includes the following source data and figure supplement(s) for figure 3: Source data 1. Data of colony numbers, cell pellet diameters, and qPCR.     These results indicate that the in vivo deletion of Nfatc1 in limb mesenchymal progenitors promotes articular chondrocyte differentiation as well. Note that a previous study showed that Nfatc1 deletion in Col2a1-expressing cells (Nfatc1 Col2a1 mice) does not affect articular cartilage integration or osteoarthritis progression induced by destabilization of the medial meniscus Greenblatt et al., 2013, suggesting that NFATc1 is dispensable in Col2a1-expressing differentiated articular chondrocytes. Therefore, NFATc1 may primarily function in articular cartilage progenitors to regulate articular chondrocyte differentiation.
Based on gene expression changes after deleting or overexpressing NFATc1, Col2a1 turned out to be one of the most significantly changed chondrocyte-related genes we examined [ Figure 5(B, C and F) and reference Ge et al., 2016]. To understand the mechanism of NFATc1 regulating articular chondrocyte differentiation, we performed computational screening and identified a total of 38 potential NFAT binding sites across the upstream 6 k base pairs of exon 1 and the intron 1 of mouse Col2a1 gene [ Figure 5

Discussion
For quite a long period, the interzone cells have been considered as the origin of articular chondrocytes Chijimatsu and Saito, 2019;Rux et al., 2019. Also, progenitor cells in the flanking mesenchyme surrounding the joint interzone were found to migrate into the interzone region and form articular cartilage Shwartz et al., 2016;Decker et al., 2017;Niedermaier et al., 2005;Koyama et al., 2007. The current obscurity in identifying the precise origin of articular chondrocytes could be attributed to the lack of a specific molecular marker to distinguish cells in the flanking region from interzone cells. From this perspective, the limited expression of NFATc1 in the flanking region at the primary stage of mouse knee development [E13.5, Figure 1(A and E) and Figure 1-figure supplement 1A] provides a unique opportunity to track progenitors in the flanking region during articular cartilage development. The progressive pattern of NFATc1-expressing cells contributing to articular cartilage formation (Figures 1-2) suggests that NFATc1-expressing progenitors in the flanking region may represent the origin of articular chondrocytes. Further studies into the spatiotemporal roadmap of NFATc1expressing progenitor cell development are essential to elucidate the landscape of articular cartilage formation. Notably, these results also demonstrate that NFATc1 can constantly track articular cartilage progenitors throughout embryonic development and postnatal growth.
Transcriptome analyses in both single-clone and bulk primary cells reveal that NFATc1-expressing articular cartilage cells enrich several previously identified articular cartilage progenitor cell markers including Osr2, Prg4,Postn,Col3a1,and Gdf6 [Figure 4(C) ;Decker, 2017;Bian et al., 2020], therefore advocating their identity as articular cartilage progenitors. In addition, cell surface molecules including CD105, CD13, CD10, CD9, SCA1, and CD166 can be considered as complementary markers for identifying and screening articular cartilage progenitors. Most impressively, NFATc1-expressing articular cartilage progenitors enrich a set of genes like Fbn2, Piezo2,Dchs1,Enpp1,Gdf6,Fgf9,Trps1,Col27a1,Tgfbr2,Col1a1,and Col1A2 [Figure 4(B)], whose mutations have been linked to a diverse range of human musculoskeletal disorders (https://www.omim.org). These findings suggest that the dysfunction of articular cartilage progenitors may underlie these human musculoskeletal disorders. Of note, we did not detect the expression of Gdf5 in articular cartilage progenitors from neonatal mice, which coincides with previous reports that Gdf5 is greatly diminished with synovial joint development and undetectable in the articular cartilage of neonatal mice Decker, 2017;Francis-West et al., 1999;Merino et al., 1999. However, the enriched expression of Gdf6 may provide a complementary role to Gdf5 [ Figure 4(B); Settle et al., 2003]. Furthermore, based on the distribution of GDF5 + progenitors during synovial joint development and their contribution to the formation of most articular chondrocytes in previous reports Decker, 2017;Decker et al., 2017, NFATc1-expressing articular cartilage progenitors are probably a subset of GDF5 + progenitors. Together, these results provide important insights into the molecular signature of articular cartilage progenitors. induction, n=3). Experiment repeated three times with cells from three animals. (C) Quantification of alcian blue staining and gene expression analysis of Col2a1 and Col10a1 showing decreased chondrogenesis after overexpressing NFATc1 in GFP + cells by infecting a caNFATc1 retrovirus structure. For alcian blue staining, n=6 with cells from three animals, two replicates for each; for gene expression analysis, n=3, experiment repeated twice with cells from two animals. (D) Safranin O staining demonstrating enhanced articular cartilage staining in the hip of Prrx1 Cre ;Nfatc1 fl/fl vs. Prrx1 Cre ;Nfatc1 fl/+ mice at 12 weeks of age. Representative images from five animals in each group were displayed. (E) Representative images of alcian blue staining and polarized light on H&E staining manifesting increased staining (arrows) and thickness of articular cartilage (double arrows) in the knee of Prrx1 Cre ;Nfatc1 fl/fl relative to Prrx1 Cre ;Nfatc1 fl/+ mice at 16 weeks of age. n=5 animals for each group. (F) Quantitative PCR determining the expression of Acan, Col2a1, and Col10a1 genes in articular cartilage of Prrx1 Cre ;Nfatc1 fl/fl relative to Prrx1 Cre ;Nfatc1 fl/+ mice at 8 weeks of age. n=6 animals for each group. (G) Computational screening of NFAT binding sites on mouse Col2a1 promoter and intron 1 sequences by PROMO software recognizing 38 putative NFAT binding sites across -6 k bp of the promoter and intron 1. The transcriptional starting site is counted as +1. Location is given in bp relative to the transcriptional starting site. (H) Cleavage under targets and release using nuclease (CUT&RUN)-qPCR showing the binding of NFATc1 to the promoter region of mouse Col2a1 (chr15: 98004609-98004620, mm10). (I) Luciferase assay of transcriptional activity of Col2a1 after deleting or overexpressing NFATc1 in ATDC5 cells. n=3, experiment repeated three times. All data shown as mean ± SD, two-tailed Student's t-test performed. Scale bar, 200 μm (D), 100 μm (E).
The online version of this article includes the following source data and figure supplement(s) for figure 5: Source data 1. Data of qPCR and luciferase assay.   Notably, a protective role of NFATc1 in osteoarthritic cartilage has been suggested previously, which is based on the decrease of NFATc1 expression in lesional osteoarthritic cartilage in human patients, as well as the severe articular cartilage deterioration in mice with conditional deletion of Nfatc1 driven by the Collagen 2 promoter Cre with a background of Nfatc2 deficiency (Nfatc1 Col2a1 ;Nfatc2 -/mice) Greenblatt et al., 2013;Beier, 2014. While the Nfatc1 Col2a1 ;Nfatc2 -/mice indeed develop osteoarthritic phenotypes, we did not find the typic osteoarthritic phenotype in the follow-up study in mice with deletion of Nfatc1 driven by tamoxifen-induced Aggrecan CreERT2 with the background of Nfatc2 deficiency (Nfatc1 AggrecanCreERT2 ;Nfatc2 -/mice) after administering tamoxifen at different ages. Instead, these animals develop obvious osteochondroma-like lesions at the entheseal site and within ligaments around the joint Ge et al., 2016. Since Col2a1 is also expressed in perichondrial precursors Akiyama et al., 2005;Ono et al., 2014, the accelerated osteoarthritis phenotype in Nfatc1 Col2a1 ;Nfatc2 -/mice could be secondary to the osteochondroma phenotype in these animals, instead of a direct beneficial role of NFATc1 in articular cartilage.
A recent study by Atsuta et al. using limb bud mesenchymal cells showed that the ectopic expression of NFATc1 seems to promote chondrocyte differentiation as shown by the alcian blue staining in micromass-cultured cells Atsuta et al., 2019). While additional experiments are necessary to confirm chondrocyte differentiation at the molecular level, overexpression of NFATc1 in cells from the whole limb bud might not reflect the physiological function of NFATc1 in chondrogenesis because NFATc1 expression is highly confined to articular and perichondrial progenitors during skeletal development according to the lineage tracing data in Nfatc1 Cre and Nfatc1 CreERT2 reporter mice (Figures 1 and 2 and reference Ge et al., 2016). Therefore, this disparity could also be due to the different cell types utilized in these two studies.
The mechanism of regulating NFATc1 expression during articular cartilage development remains unclear. A previous study showed that Notch signaling suppresses NFATc1 expression in ATDC5 cells and primary chondrocytes Zanotti and Canalis, 2013. Furthermore, the Notch signaling can be activated by mechanical loading in mandibular condylar chondrocytes and bone marrow stromal cells Ziouti et al., 2019;Yan et al., 2021. Therefore, it is possible that mechanical loading and Notch signaling act upstream of NFATc1 expression during articular cartilage development. Further studies need to verify this speculation and explore other biochemical, biophysical, and epigenetic pathways regulating NFATc1 expression during articular chondrocyte differentiation. In addition, our transcriptome data show that Nfatc4 is also enriched in articular cartilage progenitors. The function of NFATc4 and its relationship with NFATc1 in regulating articular cartilage formation needs further investigation.
In summary, we have unveiled that NFATc1-expressing progenitors generate articular but not growth-plate chondrocytes during development and identified NFATc1 as a critical negative transcriptional regulator of articular chondrocyte differentiation. Given the importance of NFATc1 in articular chondrogenesis, modulating NFAT signaling in skeletal progenitors may represent a novel, precise strategy for articular cartilage regeneration and treating cartilaginous diseases.

Limitations of the study
There are some important limitations to our present study. Firstly, the GFP + cells we used in this study might contain some cells not expressing NFATc1 but derived from NFATc1-expressing precursors. Given the prominent progenitor cell properties of the GFP + cell population, if these derived cells also display characteristics of progenitor cells, their precursors (NFATc1-expressing cells) will represent a higher hierarchy of progenitors or stem cells, which yet support our conclusions. In fact, NFATc1 is positively expressed in the incipient articular cartilage of neonatal mice [ Fig. 1(E)] and these NFATc1expressing cells can form local cell clusters with articular cartilage development [ Fig. 2(A)], indicating that NFATc1 also marks articular cartilage progenitors in neonatal mice. Secondly, these GFP + cells might include cells from other articular tissues (e.g. meniscus, articular synovium, and ligament). Since all articular tissues are derived from progenitor cells sharing the same molecular markers Koyama et al., 2008;Shwartz et al., 2016 and currently there are no specific markers to distinguish progenitors for these different articular tissues, it was challenging to explicitly isolate articular cartilage progenitor cells at the early stages of development. Future studies will be necessary to dissect the developmental hierarchy of NFATc1-expressing articular cartilage progenitor cells and elucidate the mechanisms determining their differentiation to meniscal, synovial, or ligament cells versus articular chondrocytes. Thirdly, the detailed molecular mechanism of NFATc1 regulating articular chondrocyte differentiation needs further exploration. Future studies combining techniques like ChIP, CUT&RUN, or CUT&Tag with high-throughput DNA sequencing will be able to unveil the transcriptional landscape of NFATc1-regulated articular chondrocyte differentiation. Lastly, the origin of NFATc1 + articular cartilage progenitors should be further explored, which will be critical to better understand the basic mechanism of articular cartilage development and leverage it for articular cartilage regeneration.  Muzumdar et al., 2007, and Gt(ROSA)26Sor tm9(CAG-tdTomato)Hze (Rosa26 RFP ) Madisen et al., 2010 mice were obtained from the Jackson Laboratory. Severe combined immune deficient (SCID) beige mice were acquired from Beijing Vital River Laboratory Animal Technology Co., Ltd. All mice were housed under the standard barrier facility on a 12 hr light/dark cycle with ad libitum access to water and regular chow. All animal studies followed the recommendations in the Guide for the Care and Use of Laboratory Animals of the U.S. National Institutes of Health and were approved by Institutional Animal Care and Use Committee at Capital Medical University (protocol #: AEEI-2022-036). Animals were randomly assigned numbers and evaluated blindly to experimental conditions. Tamoxifen (T5648, Sigma-Aldrich) was administered by intraperitoneal injection at 1 mg/10 g body weight for adult mice and 0.5 mg/10 g body weight for 2-week-old or dam mice. To induce Cre recombination in Nfatc1 CreERT2 ;Rosa26 RFP/+ mice at 2 weeks and 8 weeks of age, tamoxifen was injected for 5 consecutive days, which procedure was confirmed to induce a high recombination efficiency of CreER T2 in mouse articular cartilage Ge et al., 2016. Histology and confocal microscopy imaging Limb samples at different ages (n=3-5 mice for each age, two knee joints per animal) were fixed in 4% paraformaldehyde and processed for serial frozen sections at 8-10 μm thickness. The expression of GFP or RFP was observed using a Leica SP8 confocal microscope.

Continued on next page
For quantifying GFP or RFP positive cells in articular cartilage, the KEYENCE BZ-X710 fluorescence microscope and software were used under the 40 x objective and cell count module. Slides were selected every 100 μm, and five slides were used for each sample (n=5 Nfatc1 Cre ;Rosa26 mTmG/+ mice, three Nfatc1 CreERT2 ;Rosa26 RFP/+ mice). The total number of cells in each tissue was counted by combining the DAPI stain with cell morphology under phase contrast. The number of GFP + or RFP + cells was counted by combining the green or red fluorescence with DAPI. The percentage of GFP + or RFP + cells for each articular tissue was calculated by dividing the number of GFP + or RFP + cells by the total number of cells. The average of five slices from each sample was considered as the percentage of GFP + or RFP + cells in the sample.

Cell culture and sorting
Cells were isolated from the knee of neonatal Nfatc1 Cre ;Rosa26 mTmG/+ mice. Briefly, after removing the skin and most surrounding muscle tissue, the mouse knee was minced into small fragments and digested with 3 mg/ml collagenase type I (Worthington) and 4 mg/ml dispase (Roche) in complete culture media for 15 min at 37°C. After digestion, cells were passed through a 70 μm strainer and cultured for 7 days. GFP + and GFPcells were sorted using a FACSAria fusion or FACSAria II cell sorter (BD Biosciences).

Ex vivo assays of cell colony formation
For cell colony formation assay, single-cell suspensions of GFP + and GFPcells (n=6 with cells from three mice, two replicates for each) were plated in 6-well-plates at indicated densities and cultured for 2 weeks. Cells were fixed in 10% neutral formalin for 15 min and stained in 1% crystal violet for 5 min. Clone numbers (>50 cells) were counted and scored blindly under the microscope.

Ex vivo assays of osteogenic and adipogenic differentiation
The same number of cells were plated in 6-well-plates and cultured for 2 weeks in complete growth media for colony formation. Differentiation media were added to induce osteogenesis (αMEM supplemented with 10% FBS, 10 nM dexamethasone, 50 μg/ml L-ascorbic acid, and 10 mM β-glycerophosphate) for 4 weeks or adipogenesis (αMEM supplemented with 10% FBS, 100 nM dexamethasone, 50 μM indomethacin, and 5 ug/ml insulin) for 10 days. Calcium nodules and fat were visualized by staining with alizarin red and Oil Red O, respectively. Total cell clones with positive staining were blindly counted under the microscope. The expression of osteogenic or adipogenic marker genes was examined by quantitative PCR.
For micromass culture, 2x10 6 cells in 20 ul media were dropped in the center of a well in the 24-well-plate. After incubating for 2 hrs, complete culture media was added and cultured overnight. Chondrogenic media was added the next day to start the induction of chondrogenesis. Chondrogenic differentiation was evaluated by alcian blue staining and the expression of chondrocyte-associated genes.
In vivo cell transplantation GFP + or GFPcells (2.0×10 6 ) mixed with Matrigel (200 μl, BD Biosciences) were injected into the dorsal skin of SCID beige mice (n=6 animals per group). Transplants were harvested after 4 weeks and fixed in 10% formalin for histological analyses. The formation of hypertrophic cartilage-like tissue in GFP + cells was determined by two experienced pathologists based on the hematoxylin stain and the hardness of the tissue when making sections.

Quantitative PCR
Total RNA was isolated using the QIAzol lysis reagent. RNA samples were treated with an RNase-Free DNase Set (QIAGEN), and equal amounts (1 μg) were used for reverse transcriptase reaction using random primers (AffinityScript QPCR cDNA Synthesis Kit). PCR primer sequences are listed in Supplementary file 2. All gene expression was normalized to housekeeping genes Gapdh and presented by 2 -△Ct or 2 -△△Ct (Schmittgen and Livak, 2008).

RNA preparation for transcriptome analysis
Total RNA was isolated from single-cell clones or bulk GFP + and GFPcells (n=3 with cells from three different mice in each group) using QIAzol lysis reagent (Qiagen). DNA was removed using the RNase-Free DNase Set (QIAGEN). RNA was quantified with Nanodrop 2000. The integrity of RNA was evaluated using an Agilent Bioanalyzer 2100 (Agilent Technologies) and agarose gel electrophoresis.

Library preparation and RNA-sequencing (RNA-seq)
One mg of total RNA from each sample was subjected to cDNA library construction using a NEBNext Ultra non-directional RNA Library Prep Kit for Illumina (New England Biolabs). Briefly, mRNA was enriched using oligo(dT) beads followed by two rounds of purification and fragmented randomly by adding the fragmentation buffer. The first-strand cDNA was synthesized using random hexamers primer, after which a custom second-strand synthesis buffer (Illumina), dNTPs, RNase H, and DNA polymerase I were added to generate the second-strand (ds cDNA). After a series of terminal repairs, polyadenylation, and sequencing adaptor ligation, the double-stranded cDNA library was completed following size selection and PCR enrichment.
The resulting 250-350 bp insert libraries were quantified using a Qubit 2.0 fluorometer (Thermo Fisher Scientific) and quantitative PCR. The size distribution was analyzed using the Agilent Bioanalyzer 2100. An equal amount of each RNA-Seq library was sequenced on an Illumina HiSeq 4000 Platform (Illumina) using a paired-end 150 run (2x150 bases).

Bioinformatics analysis
Paired-end clean reads were aligned to mouse genome GRCm38/mm10 using STAR (v2.5) Dobin et al., 2013. HTSeq v0.6.1 was used to count the read numbers mapped to each gene. And then FPKM of each gene was calculated based on the length of the gene and read counts mapped to this gene Trapnell et al., 2010. Differential expression analysis between two groups was performed using the DESeq2 R package (2_1.6.3) Anders and Huber, 2010, which provides statistical routines for determining differential expression in digital gene expression data using a model based on the negative binomial distribution. The resulting P-values were adjusted using Benjamini and Hochberg's approach for controlling the False Discovery Rate (FDR). Genes with an adjusted P-value less than 0.05 found by DESeq2 were assigned as differentially expressed. Gene Ontology (GO) enrichment analysis of differentially expressed genes was implemented by the clusterProfiler R package Yu et al., 2012, in which gene length bias was corrected. GO terms with an adjusted P value less than 0.05 were considered to be significant.

Flow cytometry
The Cells (n=3 with cells from three different mice) were stained with antibodies or IgG isotype controls for 30 min at room temperature. Stained cells were analyzed on a FACSCalibur or BD LSR II flow cytometer (BD Biosciences). Positive cells were gated based on both unstained and isotype-matched IgG-stained cells. Data analysis was performed using FlowJo software (BD Biosciences).

caNFATc1 retrovirus production
The retroviral expression vectors pMSCV-GFP and pMSCV-caNFATc1 have been previously described Horsley et al., 2008. Recombinant retroviruses were produced by co-transfecting either the pMSCV-GFP or pMSCV-caNFATc1 vectors together with retroviral packing plasmids gag/pol and pVSV-G (Addgene) into Phoenix cells using Effectene Transfection Reagent (QIAGEN). Supernatants were harvested and filtered through a 0.45 μm filter 48 hr after transfection. GFP + cells were infected by adding retroviral supernatant with 4 μg/ml polybrene for 72 hr before further analyses.
Searching for NFATc1 binding sites in the Col2a1 gene FIMO Grant et al., 2011 was used to search the mouse genomic sequence of chr15: 97970080-98038617 (coordinates are based on the reference genome mm10), which contains the mouse Col2a1 gene and the upstream and downstream intergenic regions, for the NFATc1 motif occurrence (JASPAR motif ID MA0624.2: https://jaspar.genereg.net/matrix/MA0624.2/) with the following command: fimo -oc . --verbosity 1 --thresh 1.0E-4 NFATc1_ MA0624. 2. meme sequences. fa. As a result, 20 NFATc1 motifs (P-value <0.0001) were detected in this region. The FIMO output (.gff file) was visualized in the UCSC Genome Browser using the mouse reference genome mm10, together with the ENCODE Registry of candidate cis-Regulatory Elements (cCRE) and the regulatory elements from the ORegAnno, a community-driven resource for curated regulatory annotation (ENCODE cCRE and ORegAnno data are built-in annotations available in the UCSC Genome Browser). One of the 20 NFATc1 motif occurrences that overlap with the Col2a1 promoter region is highlighted in Figure 5figure supplement 1B.

CUT&RUN-qPCR
Sorted GFP + Cells (1x10 5 cells/reaction, n=3 with cells from three different mice) were incubated and combined with magnetic beads precoated with concanavalin A (Vazyme). Cells were permeabilized in antibody buffer containing 0.05% digitonin and incubated with Go-ChIP-Grade purified anti-NFATc1 antibody (10 μg/ml, BioLegend) or purified mouse IgG1, κ isotype ctrl antibody (10 μg/ml, BioLegend) overnight at 4℃. After washing in chilled digitonin buffer, cells were incubated with protein G fused Micrococcal Nuclease (pG-MNase, Vazyme) on ice for 1 hr. Subsequently, 100 mM CaCl2 solution was added to activate MNase and cleave chromatin. After stopping buffer was added, DNA fragments in the supernatant were extracted using a Qiagen MinElute PCR purification kit. The following primer sequences were used to amplify the fragment of the Col2a1 gene that contains the potential NFATc1 binding site: forward, 5'-T TTGG AGCG ACCG GGAG CATA T-3 '; reverse, 5'-G GGTC TCTA CCGC TCCC TCA TG-3'. The signals obtained from each immunoprecipitation are expressed as a percent of the total input chromatin and normalized with the Sample Normalization Primer Set for Spike-In DNA (Vazyme).

Luciferase assay
The transcription activity of Col2a1 was determined by co-transfecting ATDC5 cells (Sigma-Aldrich RRID:CVCL_3894), which had been infected with Nfatc1-CRISPR lentivirus or caNFATc1 retrovirus, with 1 g of the reporter plasmid (pGL2B-COL2-6.5E309, a gift from Dr. Mary Goldring) and 10 ng of pRL Renilla luciferase plasmid using the Effectene transfection reagent (Qiagen). Cells were transfected for 8 hrs and then cultured for 24 hrs. The activity of the Firefly luciferase was measured and normalized to Renilla luciferase activity using the Dual-Luciferase Reporter Assay System (Promega).

Statistical analyses
All data are presented as mean ± SD. The normality and the equal variance of data sets were tested using the Shapiro-Wilk test and F test, respectively. Data were determined to be normally distributed and have equal variance unless specified otherwise. Differences between the two groups were evaluated by the two-tailed Student's t-test. For counting data of cell colonies, comparisons were performed using the nonparametric Mann-Whitney test. Analyses of multiple groups were performed using two-way ANOVA followed by Sidak's test for between-group comparisons. All analyses were performed using Prism 9.2.1 (GraphPad). The funders had no role in study design, data collection and interpretation, or the decision to submit the work for publication.