HA is a major component of the extracellular matrix, taking part in regulating variable biological process including cell proliferation, differentiation and migration (Lee JY and Spicer AP 2000). As previous study reported, HA has the ability to promote the differentiation of hAMSCs into chondrocytes and repair cartilage injuries combined with hAMSCs . However, as the main receptor of HA in cells, CD44’s influence on the differentiation of hAMSCs into chondrocytes induced by HA is not yet known. In this study, we also found that HA can induce hAMSCs to differentiate into chondrocytes, including the production of type II collagen and the accumulation of glycosaminoglycans (Fig. 1A and 1B). In addition, the chondrocyte-related genes SOX9 and ACAN also increased significantly after HA treatment (Fig. 2A and 2B). But the transcription level of COL2A1 was down-regulated after HA treatment (Fig. 2C). The reason may be that COL2A1 has multiple transcripts. Only one of the transcripts was detected in our research, which resulted in a difference from the protein expression level. Furthermore, we also explored the role of CD44 in the differentiation of hAMSCs into chondrocytes induced by HA. Inhibition of CD44 can significantly reduce the differentiation-promoting effect of HA, and this effect may be related to Erk and Smad signals.
There is a close connection between CD44 and HA. As previous reported, HA signals through CD44 to regulate NSC quiescence and differentiation (Su W et al. 2017). In addition, Hyaluronan initiates chondrogenesis mainly via CD44 in human adipose-derived stem cells (Wu SC et al. 2013). Therefore, we detected the effect of CD44 on HA in promoting the differentiation of hAMSCs into chondrocytes. Type 2 collagen and aggrecan are two important markers of chondrocytes, the production of them is significantly reduced after blocking CD44 (Fig. 1A and 1B). At the same time, we also found that the transcription level of SOX9 has also been down-regulated (Fig. 2A). Sox9 is an essential transcription factor for maintaining the cartilage phenotype and chondrogenesis and has a positive regulatory effect on the expression of COL2A1 and ACAN (Tew SR et al. 2005). The transcription levels of ACAN and COL2A1 also appeared to be reduced accompanied by a decrease of SOX9 in our results (Fig. 2B). Therefore, CD44 blockade is very likely to inhibit the differentiation-promoting effect of HA and reduce the expression of COL2A1 and ACAN by down-regulating the transcription level of SOX9. It should be noted that CD44 should play a positive regulatory role in the differentiation process of HA. Interestingly, the transcription level of COL2A1 was decreased in HA group compared normal group. One hand, the transcription level of a single transcript may not reflect the expression of protein due to the inconsistent sequence of multiple transcripts. On the other hand, after that stimulation, most protein levelchanges were determined by mRNA level changes, with the exception of a small number of proteins that were primarily upregulated via inducing translation of pre-existing transcripts (Liu Y et al. 2016). HA may directly up-regulate the protein level of COL2A1 by inducing translation of pre-existing transcripts. Therefore, the level of mRNA shows a downward trend. But these were only our speculations, the specific regulation mechanism needs further research to explain.
TGF-β/Smad signal pathway and Erk signal have been implicated in MSC chondrogenesis, and there is cross-talk between TGF-β/Smad and integrin/FAK/ERK signaling pathways in regulating hypertrophy of MSC chondrogenesis (Zhang T et al. 2015). In addition, HA activates the RASL11B gene to potentiate the chondrogenic differentiation of hAMSCs via the activation of Sox9 and Erk/Smad signaling (Luo Y et al. 2020). Therefore, Smad signal and Erk signal were detected in our study. The results showed that the transcription levels of Smad2 and Erk2 changed after CD44 was inhibited, but Smad5 was only down-regulated after HA treatment, and CD44 had no effect on it (Fig. 3), We speculate that HA may regulate the Smad1/5/9 signal through other pathways but not CD44. In addition, the transcription level of Erk1 is not regulated by HA. Subsequently, Western-blot results showed that the phosphorylation levels of Erk1/2 and Smad2 were significantly increased after HA treatment, while CD44 blockade inhibited the phosphorylation of Erk1/2 and Smad2 (Fig. 3), but the levels of total protein of Erk1/2 and Smad2 at 24h were not change, which was inconsistent with the change of 48h transcription level. We considered that this may be due to the temporality of protein expression. In the early stage of differentiation, the increase of protein phosphorylation promotes the increase of total protein during the differentiation process to maintain the level of protein phosphorylation, and finally promotes the differentiation of chondrocytes. Using inhibitors and agonists of Erk1/2 and Smad2 signals to explore the relationship between Erk1/2 and Smad signals in the process of CD44 regulation, we found that there is a mutual regulation between Erk1/2 and Smad signals. After inhibiting or activating Erk1/2, as the phosphorylation of Erk1/2 changes, Smad2 also shows a corresponding change trend (Fig. 5A-5C). After inhibiting the activity of Smad2, the activity of Erk1/2 was also inhibited (Fig. 5A and 5B). However, after treatment with Smad2's agonist TGF-β there was no positive effect on the activity of Smad2, but an inhibition effect was produced (Fig. 5C). This may be the excessive activation of Smad2 signal by TGF-β, leading to negative feedback regulation of the TGF-β signal pathway. There are two important Smad inhibitors Smad6 and Smad7 in the negative feedback regulation pathway, which inhibited Smad signaling on multiple levels by inhibiting C-terminal Smad receptor phosphorylation by inducing receptor dephosphorylation and degradation, by inhibiting Smad receptor DNA binding, and by inhibiting complex formation between Smad4 and Smad receptors (Thielen NGM et al. 2019). Among the two Smad inhibitors, Smad7 is mainly responsible for inhibiting the activation of Smad2/3. Therefore, we speculate that the addition of TGF-β may promote the expression of Smad7, resulting in a decrease in the phosphorylation level of Smad2(Fig. 6). In addition, TGF-β induces mesenchymal stem cells to differentiate into chondrocytes by regulating Smad2, Smad3 and Smad4, and Smad3 and Smad4 play a leading role in the differentiation process. Overexpression of Smad2 can promote cartilage formation, while overexpression of Smad3 can inhibit the differentiation of mesenchymal stem cells into chondrocytes (de Kroon LM et al. 2017)
. The different roles that Smad2 and Smad3 played in the differentiation process of mesenchymal stem cells are the reason why we chose to detect Smad2 separately instead of Smad2/3. In summary, CD44 mediates HA and activates Smad2 signal and Erk1/2 signal in the process of HA promoting the differentiation of hAMSCs into chondrocytes. There is an interaction between Smad2 and Erk1/2 and excessive activation of Smad2 cause negative feedback regulation of Smad signal.