In this study, we functionally identified three sensors of acidic pHe in DAOY MB cells: ASICs, TMEM206, and OGR1. In contrast to the previously reported presence of ASIC1a and OGR1 [6, 16], the expression of functional TMEM206 in DAOY cells had not been documented until now. In addition, we found a prominent expression of BKCa channels, which were secondarily activated by acid via OGR1. The outwardly rectifying currents of TMEM206 and BKCa channels were most easily discerned at unphysiologically positive membrane potentials where strong acidification elicited currents with a complex pattern (Fig. 1 and Fig. 3). Nevertheless, also at the resting membrane potential, pH 5.0 elicited cell volume changes, which were sensitive to inhibitors of the three acid-sensors and of BKCa (Fig. 6), suggesting their contribution to the response of DAOY cells to strong acidification. Moreover, pH 4.5 even induced cell death, which was sensitive to the same inhibitors (Fig. 7). Whether cell death was related to the cell volume changes and whether both phenomena occur in MB in situ remains unknown, requiring further investigation.
In the past, acid-sensitive Cl− currents have been identified in many mammalian cells [37] and in 2019, TMEM206 was recognized as the channel responsible for these currents [10, 26]; however its precise functions under physiological and pathological conditions remain elusive. Previous studies have linked its activation during strong acidic “shock” treatments to rapid changes in cell volume and a necrosis-like cell death in HEK cells, HeLa cells, and neurons [10, 11, 38]. Our own results in DAOY cells support these observations (Fig. 6, Fig. 7). While the activation threshold of TMEM206 is low at room temperature (~ pH 5.5), it increases at body temperature (~ pH < 6) [39], suggesting that cell death induction by activation of TMEM206 might be relevant for a number of pathologic conditions such as ischemic stroke or tumors. Nonetheless, it is plausible that TMEM206 is involved in other functions within MB cells. For example, it has recently been demonstrated that TMEM206 plays a key role in macropinosome shrinkage, an important mechanism for tumor growth [13]. And in human osteosarcoma cells, overexpression of TMEM206 is linked to an increase in migration and proliferation in vitro [40]. Uncovering the exact role of TMEM206 in MB cells warrants further study.
To date, BKCa channels have not been studied in MB, but they have been extensively studied in other brain tumors, particularly in gliomas, where they may promote proliferation and migration [41]. Interestingly, the facilitation of migration through Ca2+-activated K+ channels is achieved largely by cell shrinking [42]. This effect becomes apparent only when these channels are activated by an increase in [Ca2+]i [43, 44]. Notably, glioma cells express specific BKCa splicing variants that are particularly sensitive to [Ca2+]i [41, 45, 46]. Recent findings, however, suggest that in glioblastoma stem cells, BKCa channels are not involved in the aggressive migration at acidic pH in vitro [47], indicating that they do not universally play a role in the migration of cancer cells. Whether BKCa channels might promote migration in MB cells as well, for instance under mildly acidic conditions through an OGR1-mediated activation, remains an open question that needs further investigation. Furthermore, Ca2+-activated K+ channels such as BKCa and IKCa appear to be important for the volume regulation of glioma cells [48].
Concerning acid-induced changes in cell volume, it has been hypothesized previously that in HEK cells an acidic stimulation quickly activates ASICs, depolarizing the cells through a rapid Na+-influx and thus providing the initial depolarization needed for activation of TMEM206 [10, 11]. Our observations support this hypothesis (Fig. 6). The combined effect of the ASIC-mediated Na+-influx and the TMEM206-mediated Cl−-influx likely caused DAOY cells to swell at pH 5.0. However, OGR1-mediated activation of BKCa also contributed to the initial swelling, albeit to a small degree (Fig. 6c). This was surprising, considering that K+-efflux is expected to lead to cell shrinking and regulatory volume decrease (RVD) [49, 50]. The changes in DAOY cell volume after acidic stimulation appeared to result from a complex interplay between several different ion channels. The slope of the subsequent shrinking was also influenced by all three channels (ASIC1a, TMEM206 and BKCa). A potential causal link between the initial swelling and subsequent shrinking, maybe as a form of RVD, would explain why inhibiting the swelling also reduced the subsequent shrinking. Activation of BKCa channels was likely the main cause of the subsequent shrinking since they allow K+-efflux. The shrinking was likely delayed by ~ 40 s because BKCa channels are activated with a similar delay (Fig. 1d, Fig. 3b), due to the delayed elevation of [Ca2+]i by OGR1 activation (Fig. 5). Thus, we propose that Cl−-flux via TMEM206 contributes to the initial swelling as well as to the delayed shrinking of DAOY cells under strong acidification by “following” a leading cation, either Na+ (ASIC1a) or K+ (BKCa). It is reasonable to assume that other ion channels and transporters are involved in this complex mechanism as well, for instance TRPC4 channels that have been shown to elevate [Ca2+]i under acidic stimuli in cooperation with OGR1 [51].
Little is known about how ASIC1a can affect brain tumors in vivo and the existing data is conflicting [52]. While some studies suggest that ASICs can influence the proliferation and migration of glioma cells [53, 54], newer studies performed in glioblastoma tumorspheres propose an ASIC1a-mediated activation of a non-canonical necroptosis pathway [9]. DAOY cells evade this cell death by a low expression of RIP3, an important protein involved in the necroptosis pathway [6]. Our study suggests a possible novel function of ASIC1a in tumors, where it could synergize with other acid sensors on the plasma membrane, to affect the cell volume.
Two hours exposure to a strongly acidic environment resulted in the death of DAOY cells (Fig. 7), which was reduced by the inhibition of ASIC1a, TMEM206, BKCa, or OGR1. Given the short time frame and the involvement of ion channels, it is likely that a necrosis-like cell death mechanism, influenced by changes in cell volume, occurred. A link between cell swelling and necrosis has been repeatedly documented in other studies [55, 56] and TMEM206 has been previously associated with cell volume alterations and cell death in HeLa cells, HEK cells, and neurons [10, 11, 38], although, so far, not in combination with other ion channels. But it is unclear whether the changes in cell volume caused the cell death of DAOY cells in our experiments.
While DAOY cells serve as a model for SHH MB, UW228 are a model for WNT MB [57]. SHH MB is characterized by an increased expression of ASIC1a and a reduced expression of ASIC2 and this expression pattern is recapitulated in DAOY cells, suggesting their suitability for studying the role of ASICs in SHH MB [6]. In contrast, ASICs are absent in UW228 cells, rendering them an unsuitable model for studying the role of ASICs in MB [6] (Table 1). In the current study, we found that DAOY and UW228 cells both expressed TMEM206, but the mRNA expression and current amplitude of TMEM206 was ~ 2-fold lower in UW228 than in DAOY cells. Overall, TMEM206 exhibits similar expression level in MB and normal tissue (Fig. 8a), but its expression is more pronounced in WNT than SHH MB (Fig. 8d). Thus, the lower expression of TMEM206 in UW228 than in DAOY cells does not recapitulate the distinct expression patterns observed in the two MB subtypes (Table 1).
Table 1
qualitatively summarizes the expression of the GOIs in SHH MB and WNT MB, and in UW228 and DAOY cells. Expression data for MB tissue is based on the Cavalli dataset, expression data for the cells on functional data and qPCR. Data for ASIC1a are from [6]. “+” denotes normal expression, “-“ downregulation” and “++” upregulation.
| ASIC1a | TMEM206 | BKCa | OGR1 |
WNT MB | ++ | ++ | + | + |
UW228 | - | + | + | - |
SHH MB | ++ | + | ++ | ++ |
DAOY | ++ | ++ | ++ | ++ |
A possible involvement of BKCa in MB tumor biology is supported by the fact that BKCa is considerably less expressed in MB compared to normal tissue (Fig. 8a). Like TMEM206, both DAOY and UW228 cells expressed BKCa, but the mRNA expression and current amplitude were ~ 10-fold lower in UW228 cells than DAOY cells. Interestingly, WNT tumors exhibited the lowest amount of KCNMA1 mRNA compared to more aggressive MB subgroups (Fig. 8d). Thus, the lower expression of BKCa in UW228 than in DAOY cells accurately reflects the distinct expression patterns in the two MB subtypes (Table 1).
The analysis of the Cavalli microarray dataset revealed that OGR1 is highly expressed in SHH MBs compared to other MB subgroups (Fig. 8d). Our finding that DAOY cells strongly express OGR1, while UW228 cells do not (Fig. 5), suggests that DAOY cells are representative for SHH MB concerning OGR1 expression (Table 1). In a stem cell-rich DAOY MS culture, the expression of OGR1 increased even further (3-fold). The role of OGR1 in cancer stem cells remains unknown. Surprisingly, the survival rate of MB patients expressing high levels of OGR1 was increased in the Cavalli microarray dataset (Fig. 8c). Previously, it was shown that OGR1 activates the MEK/ERK pathway in DAOY cells through acid-induced elevation of [Ca2 + i] [16], which could promote proliferation and tumor growth. Additionally, OGR1 increases expression of TRPC4 channels in DAOY cells [51]. In the same study, direct activation of TRPC4 channels promoted migration, however indirect activation through acid-induced OGR1-activation did not. Moreover, when investigating the effects of OGR1 on other cancer types, conflicting data emerge depending on the specific tumor type. While some studies suggest that OGR1 exhibits pro-tumor effects, for instance in melanoma, pancreatic cancer and colorectal cancer [58–60], others propose the opposite [61–63]. Given that MB patients expressing high levels of OGR1 exhibit a higher survival rate (Fig. 8c), the role of OGR1 in MB might be anti-tumoral. However, the high expression could also be an epiphenomenon that is not causally linked to the survival. Nevertheless, the high expression of OGR1 in adult MB patients (Fig. 8b) was striking, because MB in adult patients is very rare and about 75% of adult MBs belong to the SHH group and metastasis almost never occurs [64]. Therefore, the potential anti-tumoral role of OGR1 in SHH MB merits further investigation. The high expression of both BKCa and OGR1 in DAOY cells (Table 1) indicates that DAOY cells are a suitable model for studying their functional relationship in SHH MB.
In summary, we characterized the acid sensors ASIC1a, TMEM206, OGR1 and BKCa channels in MB cells and our results indicate that DAOY cells are a valuable model to study their functional interrelationship, including their role in volume regulation and cell death.