Potentiation of ryanodine receptor–mediated calcium release by MAPK is responsible for epidermal transformation and carcinogenesis Authors: Pengcheng Wang, Kristan H. Cleveland, Ayaz Shahid, Chathurika Rathnayaka, Farideh Amirrad, Zhenpeng Song, John Paul Estillore, Ruiwu Wang, Thomas G. Back, S. R. Wayne Chen, Ying Huang Published: 2025 ABSTRACT Epidermal growth factor (EGF) induces anchorage-independent growth in promotion-sensitive (P+) mouse epidermal cell model JB6 primarily through activation of the MAPK/ERK signaling pathway. The β-blocker carvedilol inhibits EGF-promoted JB6 transformation, but the underlying mechanism is unknown. Since carvedilol suppresses overactivated ryanodine receptors (RyRs) independently of its adrenergic blocking effects, we hypothesized that EGF-promoted transformation requires RyR-mediated calcium (Ca2+) release and that carvedilol inhibits transformation via targeting RyRs. All RyR subtypes were present in epidermis and strongly upregulated by ultraviolet (UV) radiation, as demonstrated in an RyR2-tdTomato reporter mouse model. In vitro, EGF induced ERK phosphorylation and RyR2 upregulation and increased RyR agonist 4-chloro-m-cresol (4-CMC)-evoked Ca2+ release, which is inhibitable by structurally divergent RyR stabilizers and inhibitors of MAPK and PLC, but not by most β-blockers. Expression of constitutively active K-RAS and MEK-1 or UV also potentiated 4-CMC-evoked Ca2+ release. RyR agonists and the Ca2+ ionophore ionomycin promoted JB6 transformation while RyR stabilizers, the intracellular Ca2+ chelator BAPTA/AM, and inhibitors of MAPK and PLC blocked transformation. The RyR shRNAs abolished the transformation-inhibitory effect of carvedilol. The IC50 values of five carvedilol derivatives for suppressing RyR-mediated Ca2+ release positively correlated with the IC50 values for transformation inhibition. In vivo, UV-induced DNA damage and skin inflammation were enhanced by topical 4-CMC treatment but attenuated in the RyR2-E4872Q knock-in mice in which RyR2 activity is reduced. Human skin tissue microarray analysis confirmed spatial colocalization of phospho-ERK and RyR2 in the same tumor areas. Thus, potentiation of RyR-mediated Ca2+ release by MAPK is an important pathway leading to carcinogenesis. Carvedilol is a β-adrenergic receptor nonselective blocker with antioxidant and anti-inflammatory properties (1–3). In previous studies, carvedilol prevented chemical carcinogen and ultraviolet (UV) radiation–induced skin inflammation and cancer in mice (4, 5). It also blocked epidermal growth factor (EGF)-mediated transformation of JB6 Cl 41–5a (also named JB6 P+ or JB6) cells, a promotion-sensitive murine epidermal cell line commonly used to study the mechanisms of skin tumor promotion or neoplastic transformation (4–6). Carvedilol also prevented tobacco carcinogen benzo(a)pyrene-mediated bronchial epithelium transformation and lung carcinogenesis (7). These preclinical findings are corroborated by a large population-based cohort study, in which long-term carvedilol use was correlated with a reduction in the cumulative incidence of cancer (8). Interestingly, most β-blockers examined in the JB6 cell model did not show antitransformation activity (9), suggesting that carvedilol’s cancer preventive effect is unrelated to its β-blocking activity. Mechanistic studies further demonstrated that α- or β-blockade is not required for the antitransformation activity of carvedilol on JB6 cells (9). Independent of its α/β-adrenergic blocking and antioxidant activity, carvedilol has been shown with inhibitory activity against a family of intracellular calcium (Ca2+) release channels, the ryanodine receptors (RyRs), which are expressed in the sarcoplasmic/endoplasmic reticulum membranes (10). RyRs have three isoforms—RyR1 (dominant in skeletal muscle), RyR2 (dominant in cardiac muscle), and RyR3 (broader but lower-level expression)—with dysfunction linked to disorders like muscle diseases, cardiac arrhythmias, and neurodegeneration (11). The homo-tetramer structure of RyRs is highly unstable and can be destabilized by mutations, posttranslational modifications, or epigenetic regulation in response to extracellular stimuli, causing channel overactivation and leakage (12–14). Importantly, RyRs are considered druggable, and several small molecules, including FDA-approved drugs for noncancerous diseases such as carvedilol and dantrolene, can stabilize the over-active RyR channel and prevent the “leaky” channel activity without completely blocking their normal functions (10, 15, 16). We hypothesized that the anticancer activity of carvedilol involves targeting RyRs. The present study examined the functional role of RyRs and associated Ca2+ release in epidermal neoplastic transformation, as well as their involvement in the antitransformation effects of carvedilol. The hypothesis is further strengthened by the expression and functional analysis of RyRs in response to UV radiation, the most well-established carcinogen and promoter for skin carcinogenesis. Our findings demonstrated that RyR-mediated Ca2+ release is required for neoplastic transformation of JB6 cells and that EGF (and UV) potentiates RyR-mediated Ca2+ release by activating the MAPK/ERK pathway. This potentiation may be an important mechanism leading to carcinogenesis not only of skin but also of other organs. Results Expression and Localization of RyR2 in Mouse Skin. Expression of RyRs was examined in the skin of SKH-1 hairless mice, a commonly used mouse strain in photobiology and skin cancer research (17). First, a RyR2-reporter mouse model of the SKH-1 background, in which the tdTomato fluorescent protein was inserted into the endogenous RyR2 gene locus, was generated (SI Appendix, Fig. S1). Strong red fluorescence signals were detected in the heart (positive control) from heterozygous and homozygous mice (Fig. 1A). We performed fluorescence imaging analysis of skin sections from RyR2-tdTomato mice to directly visualize the location of RyR2 in the skin, where the red signal level was significantly higher in the homozygous skin compared to the wild-type (WT) controls (Fig. 1B). The red signal was mainly localized in the epidermal layer of the skin. In a separate study, the WT and RyR2-tdTomato heterozygous mice were exposed to a single-dose UV radiation [224 mJ/cm2, the minimum erythema dose for the SKH-1 strain (17)], and the skin was collected 24- or 48 h after UV exposure. Interestingly, the RyR2-tdTomato signal in the epidermal keratinocytes strongly increased and a large number of RyR2-expressing cells (likely immune cells) infiltrated into the dermis after UV exposure (Fig. 1C). Similar observation was obtained with immunofluorescence (IF) analysis of RyR2 in the skin of the WT SKH-1 mice where UV radiation increased RyR2 expression in the epidermis (based on its colocalization with Pan-keratin positive cells) (SI Appendix, Fig. S2) and dermis (SI Appendix, Figs. S2 and S3). To confirm the involvement of other subtypes, RyR1 and RyR3, IF analysis was conducted using respective antibodies. As can be seen from SI Appendix, Figs. S4 and S5, expression of RyR1 and 3 in the skin was also strongly induced by UV. Expression of RyRs in Keratinocyte Cell Lines JB6 and HaCaT. To determine whether epidermal keratinocytes express any of the subtypes, the expression levels of RyR1, 2, and 3 in the murine epidermal cells JB6 and human keratinocytes HaCaT were evaluated using conventional RT-PCR. JB6 cells expressed the mouse RyR2 and RyR3. HaCaT cells expressed the human RyR1 and RyR2 abundantly, but RyR3 was undetectable (Fig. 2A). The positive controls (reference cDNAs) displayed the expected bands confirming that the negative PCR results for RyR subtypes are due to a lack of mRNA expression. Although different epidermal keratinocyte cell lines express different subtypes, RyR2 is the common subtype expressed. To determine whether the tumor promoter EGF treatment modulates the mRNA expression of RyRs, the JB6 cells were treated with EGF (10 ng/mL) for 2 or 6 h. RT-PCR analysis indicated that for all three subtypes of RyRs, EGF failed to alter mRNA expression (SI Appendix, Fig. S6). The same investigation was conducted in HaCaT cells, in which the treatment of EGF for up to 6 h failed to change the mRNA levels of RyRs. To verify the negative results, qRT-PCR for RyR2 was conducted for JB6 cells and the data confirmed there was no significant difference before and after EGF treatment even at the longer time point (24 h). ICC analysis using antibodies specific for ERK1/2 and phospho-ERK1/2 (pERK1/2) indicates that treatment with 10 ng/mL EGF for 6 h markedly increased ERK1/2 phosphorylation and promoted its nuclear translocation, as evidenced by a significant elevation of nuclear phospho-ERK1/2 levels, indicating robust activation of the MAPK pathway (Fig. 2B). EGF treatment induced a strong and significant increase of the protein expression of RyR2 in the JB6 cells, suggesting that RyR2 is upregulated at the posttranscriptional levels (Fig. 2C). Tissue Microarray Analysis of Phospho-ERK1/2 and RyR2. We used immunohistochemistry (IHC) to evaluate the expression of pERK1/2 and RyR2 in human skin cancer on a tissue microarray which contains 18 SCC tissue spots and 21 melanoma tissue spots. Interestingly, the IHC images demonstrate spatial colocalization of pERK1/2 and RyR2 in the same tumor areas in both SCC and melanoma (Fig. 3). Scoring from the SCC spots identified a positive correlation between pERK1/2 and RyR2 (Pearson’s r = 0.7171; 95% CI, 0.3761 to 0.8870) (Fig. 3A). A positive correlation was also observed between pERK1/2 and RyR2 in melanoma (Pearson’s r = 0.7412; 95% CI, 0.4552 to 0.8886) (Fig. 3B). The staining images and expression scores for all tissue spots can be found in SI Appendix, Tables S1 and S2. These pilot human data, although with small sample sizes, suggest that there is a cross-talk between MAPK/ERK and RyR2 in human skin tumors. RyR Function in JB6 and HaCaT Cells Based on Ca2+ Flux Assays. The function of RyRs in epidermal cell lines was investigated by evaluating the intracellular free calcium (Ca2+) concentration in response to a commonly used RyR agonist 4-chloro-m-cresol (4-CMC), which was added into one well at a time by automatic injection. In JB6 cells, 4-CMC injection evoked an intracellular Ca2+ release in a dose-dependent manner (Fig. 4A): for doses at or lower than 0.2 mM, Ca2+ release was undetectable, while at doses higher than 0.5 mM, 4-CMC strongly increased intracellular Ca2+ release. Similar effects were observed in HaCaT cells (SI Appendix, Fig. S7): at doses of 2 mM and less, there was barely any detectable Ca2+ release, while at doses of and higher than 5 mM, 4-CMC evoked a strong and significant intracellular Ca2+ release. Unlike the peaks obtained in JB6 cells, the Ca2+ peaks in HaCaT cells did not drop much after 200 s, suggesting that at higher concentrations, Ca2+ release may come from sources other than RyRs, because 4-CMC has been reported as an inhibitor of the sarcoplasmic endoplasmic reticulum Ca2+-ATPase (SERCA) Ca2+ pumps at higher concentrations (18). Effects of EGF Treatment on 4-CMC-Evoked Ca2+ Release. We next determined whether adding EGF by automatic injection to the well was able to induce Ca2+ release in JB6 cells. As shown in SI Appendix, Fig. S8, adding EGF (10 ng/mL) into the cells failed to evoke immediate increase of intracellular Ca2+. Higher doses of EGF (100 ng/mL) were also examined, which failed to induce any Ca2+ release but instead slightly reduced the level of intracellular Ca2+. In JB6 cells, although 0.2 mM 4-CMC alone barely evoked any Ca2+ release (Fig. 4A), pretreatment of the cells with 10 ng/mL EGF for 6 h strongly increased 0.2 mM 4-CMC-induced Ca2+ release (Fig. 4B). Time course study indicated that the effects of EGF treatment were time dependent and that treatment with EGF for 30 min was sufficient to sensitize the cells for 4-CMC-induced Ca2+ release, but incubation with EGF shorter than 30 min had no effect (Fig. 4C). In HaCaT cells, similar effects of EGF were observed (SI Appendix, Fig. S9): 2 mM 4-CMC alone was unable to evoke intracellular Ca2+ release. However, the Ca2+ transient after incubation with EGF at 10 ng/mL for 6 h was greatly increased compared with the control group with 4-CMC injection only. EGF treatment for 30 min was also found to be sufficient to sensitize the cells for 4-CMC-induced Ca2+ release, but coinjection of EGF and 4-CMC at the same time failed to evoke Ca2+ release. Given that RyR potentiation is observed within 30 min, it is plausible that EGF-induced kinase signaling contributes to this rapid response. Effects of RyR Stabilizers, β-Blockers, and Inhibitors of MAPK and PLC on EGF/4-CMC-Induced Ca2+ Release. In JB6 cells, adding known RyR stabilizers, dantrolene (DAN, 100 μM), JTV519 (JTV, 10 μM), and carvedilol (CAR, 10 μM) attenuated the effects of EGF on 4-CMC-evoked Ca2+ release, although the effects of dantrolene were insignificant (Fig. 4 D–F). Similarly, in HaCaT cells, coincubation of the cells with EGF and dantrolene, JTV519, and carvedilol strongly and significantly reduced the effects of EGF (SI Appendix, Fig. S10). In JB6 cells, we evaluated additional RyR stabilizers on 4-CMC-induced Ca2+ release in the presence of EGF (10 ng/mL), and the results are summarized in Table 1. The R- and S- enantiomers of carvedilol showed similar inhibitory effects on EGF/4-CMC-induced Ca2+ release as the racemic carvedilol, consistent with previous studies that both enantiomers are RyR stabilizers (15). The same studies were performed with several β-blockers. Interestingly, most β-blockers that do not exhibit RyR-modulatory activity did not block the effect of EGF, except that atenolol reduced EGF-induced Ca2+ release slightly but significantly (Table 1). Notably, nebivolol, another β-blocker reported to have RyR stabilizing activity (19), was the most potent one in reducing EGF-induced Ca2+ release among all β-blockers tested (Table 1). Most RyR stabilizers tested significantly inhibited EGF-induced Ca2+ release, except for dantrolene. Dantrolene at 100 μM modestly reduced the effects of EGF in JB6 cells (Table 1), but such effects on HaCaT were stronger and significant (SI Appendix, Fig. S10). To explore the involvement of signal transduction pathways in EGF-induced effects on 4-CMC-evoked Ca2+ release, the classical MAPK/ERK inhibitor U0126 was incubated with EGF in JB6 cells. Interestingly, U0126 showed the highest potency among all compounds tested in reducing the Ca2+ release–inducing effects of EGF (Table 1); at 1 μM it abolished the effect of EGF by 95%. The inhibitors of phospholipase C (PLC) D609 and U73122 were also incubated with EGF, and the results showed that both PLC inhibitors strongly and significantly attenuated the potentiating effects of EGF (Table 1). These data indicate that EGF potentiates RyR-mediated Ca2+ release through an MAPK/ERK- and PLC-dependent pathway. Effects of Activating the MAPK Pathway by an Alternative Approach on 4-CMC-Evoked Ca2+ Release. Since EGF activates the MAPK pathway through activating RAS, we next evaluated the effects of a constitutively active RAS mutation. JB6 cells were transiently transfected with a GCaMP plasmid and a K-RAS-G12V plasmid, and 4-CMC-evoked Ca2+ release was measured in GCaMP-positive cells using a calcium imaging assay based on fluorescent microscopy. As expected, expression of the constitutively active K-RAS-G12V strongly potentiated 4-CMC-evoked Ca2+ release, without the need of adding EGF (Fig. 5A). Similarly, cotransfection of GCaMP and a constitutively active form of MEK-1 (caMEK-1) resulted in a significant increase of 4-CMC-evoked Ca2+ release (Fig. 5B). We also cotransfected a dominant-negative mutant form of MEK1 (dnMEK1), but there was no significant difference in Ca2+ release between the dnMEK1-transfected cells and the control cells transfected with GCaMP alone (Fig. 5A). These results indicate that activation of the MAPK pathway is sufficient to potentiate 4-CMC-stimulated Ca2+ release. Since it is well known that UV activates MAPK, we next examined the effects of UV radiation on RyR-mediated Ca2+ release. The GCaMP-transfected JB6 cells were exposed to UV radiation in a range of doses (5 to 50 mJ/cm2) and 6-h later the cells were injected with 0.1 mM 4-CMC to evoke RyR-Ca2+ release. Calcium imaging analysis based on fluorescent microscopy demonstrated a significantly increased level of Ca2+ release in GCaMP-positive cells that were exposed to 15 or 50 mJ/cm2 while the effects of other doses of UV were insignificant (Fig. 5C). Interestingly, in previous studies, 15 mJ/cm2 was used as the dose to transform JB6 cells (5, 20). These data indicate that UV, by activating the MAPK, potentiates RyR-mediated Ca2+ release. Effects of RyR Stabilizers, Agonists, and MAPK/PLC Inhibitors on Neoplastic Transformation of JB6 Cells. Previous studies demonstrated that the racemic carvedilol and its enantiomers R- and S-carvedilol blocked EGF-induced neoplastic transformation of JB6 cells (4, 21). It is interesting to examine whether such activity is also possessed by other RyR stabilizers. JTV519, a pan-RyR stabilizer structurally unrelated to carvedilol (22), displayed the same antitransformation effect on JB6 (IC50 was estimated as 0.48 μM) (Fig. 6A). JTV519 was able to block transformation at nontoxic concentrations since the IC50 value for its cytotoxicity (SRB assay) was >10 μM (Fig. 6A). Dantrolene, another pan-RyR stabilizer with relatively higher potency for RyR1/3, also showed antitransformation effects at nontoxic concentrations (IC50 was estimated as 1 μM) (Fig. 6B). S107, a JTV519 analog more specific for RyR2 (23), showed the highest potency among RyR stabilizers against EGF-induced JB6 transformation at nontoxic concentrations (IC50 was estimated as 0.1 μM) (Fig. 6C). The β-blocker/RyR stabilizer nebivolol has been reported in a previous study to inhibit EGF-induced JB6 transformation with an IC50 of 1.09 μM (9). Thus, these RyR stabilizers that fix RyR leak via different mechanisms (14) consistently blocked EGF-induced JB6 transformation. Interestingly, the MAPK inhibitor U0126 showed the highest potency (IC50 was estimated as 0.01 μM) in inhibiting EGF-induced JB6 transformation at nontoxic concentrations (Fig. 6D). The PLC inhibitors D609 and U73122 also inhibited EGF-induced JB6 transformation (Fig. 6 E and F). Several RyR agonists were examined on the soft agar assay with or without the presence of EGF. 4-CMC at the concentration of 0.1 μM significantly increased colony formation (Fig. 6G). At this concentration, 4-CMC slightly but significantly increased the colony formation induced by EGF. Another RyR agonist suramin at 10 nM increased transformation but did not further increase the colony formation induced by 10 ng/mL EGF (Fig. 6H). Effects of BAPTA-AM and Ionomycin on Neoplastic Transformation of JB6 Cells. We next tested BAPTA-AM, a membrane-permeable chelator of intracellular Ca2+, and the results also demonstrate dose-dependent inhibition of EGF-induced JB6 transformation, although BAPTA-AM was more toxic than RyR stabilizers (Fig. 6I). Similarly, the Ca2+ ionophore ionomycin (1 μM) promoted JB6 transformation on its own, although it did not further increase the transformation induced by 10 ng/mL EGF (Fig. 6J). Effects of Gene Knockdown of RyR2/3 on Carvedilol’s Anti-transformation Activity. To directly investigate the role of the RyRs as a potential target of carvedilol, we designed lentiviral shRNA sequence directed toward mouse RyR2 (sh-RyR2). JB6 cells transiently infected with sh-RyR2 significantly suppressed gene expression at the transcriptional level by 60% compared to scramble control cells. In JB6 cells carrying sh-RyR2 knockdown, although EGF still induced colony formation, carvedilol reduced its ability to prevent EGF-mediated transformation (IC50=14.4 μM) when compared with the scramble control, in which carvedilol prevented EGF-induced transformation with an IC50 of 0.582 μM (Fig. 6K). Since JB6 cells expressed RyR3 at a higher level than RyR2 (Fig. 2A), JB6 cells were also infected with lentiviral shRNA directed for RyR3 (sh-Ryr3). In JB6 cells carrying sh-RyR3, carvedilol similarly showed reduced ability to prevent EGF-induced transformation (Fig. 6L), indicating the possible involvement of RyR2 and RyR3 as carvedilol’s targets. Effects of Carvedilol Derivatives on EGF-Mediated JB6 Transformation. The derivatives of carvedilol were synthesized and shown to have varying degrees of inhibition of RyR2-mediated store overload-induced Ca2+ release (SOICR) (10, 24, 25). The derivatives are structurally similar to carvedilol, but contain modified catechol or carbazole subunits, or altered linkers between these two groups. To identify potential association of SOICR/RyR inhibition and antitransformation activity on JB6, we examined five carvedilol derivatives with different potency in SOICR inhibition on EGF-mediated transformation of JB6 cells. Carvedilol has an IC50 of SOICR inhibition of 15.9 ± 2.5 μM (24) and inhibits EGF-mediated colony formation with an IC50 of 0.582 μM. Compound 93 showed the most potent inhibition of cell transformation (IC50 was 0.38 μM) and also the most potent one against SOICR (Table 2). Compounds 89 and 8 showed a modest degree of inhibition (IC50 0.93 μM and 1.13 μM, respectively). Compounds 77 and 88 showed the least potency on inhibition of transformation (IC50 2.85 and 6.20 μM, respectively), and also with the least anti-SOICR activity (Table 2). The IC50 values obtained from the soft agar transformation assay were positively correlated with the IC50 values of these derivatives on SOICR inhibition with the Pearson Correlation Coefficient r = 0.9526 (P = 0.003317). Effects of RyR Agonist 4-CMC on UV Radiation–Induced Skin Damage in SKH-1 Hairless Mice. UV radiation induces DNA damage and inflammation in the skin, which are early events that promote skin carcinogenesis. UV activates EGF receptor and MAPK (26), which plays an important role in leading to cutaneous inflammation and carcinogenesis (27, 28). We therefore examine our hypothesis by exposing the mice to UV radiation and applying topical 4-CMC to activate RyRs in the skin of SKH-1 mice. We detected time-dependent formation of cyclobutane pyrimidine dimers (CPD), a UV-specific DNA damage marker in the mouse skin, while topical treatments of 4-CMC significantly increased the production of CPD, with the difference more pronounced at the 6 h post–UV time point (Fig. 7A). Similarly, the plasma and skin levels of the proinflammatory marker IL-6 showed time-dependent changes, while 4-CMC greatly increased IL-6 production, with the difference more pronounced at the 24 h post–UV time point (Fig. 7 B and C). Consistently, 4-CMC also increased sunburn index, trans-epidermal water loss (TEWL, a parameter for skin barrier function), and erythema index, induced by multiple doses of UV radiations (Fig. 7 D–F). These data further support the functional cross-talk between MAPK and RyRs. Effects of RyR2-E4872Q Mutation in UV-Induced Mouse Skin Lesions. Fixing the overactivated (or leaky) RyR channel may prevent skin cancer development. Previous studies have shown that the RyR stabilizers carvedilol or R-carvedilol prevented UV-induced DNA damage, inflammation, and skin carcinogenesis in SKH-1 mice (5, 21). To further examine our hypothesis, we applied a knock-in mouse model on the SKH-1 background carrying the E4872Q mutation in the RyR2 gene which shortens the channel open times (29). The E4872Q+/− mice showed reduced intracellular Ca2+ release in RyR2-expressing tissues such as the heart (21) and neurons (30). After exposed to a single-dose UV radiation, the E4872Q+/− mice demonstrated attenuated DNA damage marker CPD, with the 6 h post–UV time point showing significant difference from the WT mice (Fig. 8A). The E4872Q+/− mice also showed attenuated plasma IL-6 levels and skin tissue mRNA level of IL-6 compared with WT control mice (Fig. 8 B and C). The E4872Q+/− mice were next exposed to multiple doses of UV. However, the E4872Q+/− mice showed the same UV-induced bifold thickness and erythema, but slightly increased sunburn index and TEWL (SI Appendix, Fig. S11). One possible explanation is that UV-induced upregulation of RyR1 and 3 in skin may compensate for the reduced RyR2 function in the E4872Q+/− mice. Discussion Although dysfunction of the ER Ca2+ store has been associated with carcinogenesis (31, 32) and tumor progression (33, 34), and proposed as a target for cancer therapy (35), the exact defects underlying aberrant Ca2+ handling are not defined. The present study provides evidence that RyRs expressed in the mouse and human epidermal keratinocytes play an important role in epidermal transformation and skin carcinogenesis. This study also indicates that carvedilol prevents cancer by targeting the skin RyRs, consistent with previous studies demonstrating the activity of carvedilol against the leaky RyR Ca2+ channels to suppress the development of heart failure and arrhythmias (10, 36). We examined the presence of red fluorescence (tdTomato) in mice carrying tdTomato inserted into the RyR2 gene. The RyR2-tdTomato signal was observed in the epidermis layer of the mouse skin isolated from homozygous or heterozygous mice. Consistent with the previous study (37), RyR2 expression was mainly found in the border layer under the stratum corneum. After exposure to UV, the RyR2-tdTomato signal was greatly increased. Interestingly, RyR2-positive cells were detected in dermis; it remains to identify cell types that express RyR2. Immunofluorescence analysis using antibodies confirmed the expression of all three subtypes which can be induced by UV radiation in skin. Such a result is expected as the three RyR genes are paralogous (arose from gene duplication), with about 65 to 70% overall amino acid sequence homology (38, 39). In vitro, RT-PCR confirmed the mRNA expression of RyRs in immortalized murine and human keratinocyte cell lines JB6 and HaCaT. These cells do not express all subtypes, but RyR2 is the subtype commonly expressed. In JB6 cells, the RyR agonist 4-CMC at concentrations of 0.5 mM or higher evoked Ca2+ release, while the sensitivity to 4-CMC was much lower in HaCaT cells, which required at least 5 mM to trigger significant Ca2+ release. Injection of EGF (the dose that transformed JB6 cells, 10 ng/mL) into the wells without 4-CMC failed to induce Ca2+ release. Interestingly, incubation of EGF with the JB6 or HaCaT cells for 30 min or longer greatly increased the cellular sensitivity to 4 CMC. Since the Ca2+ response to EGF/4-CMC was observed in a Ca2+ -containing medium, influx of Ca2+ through plasma membrane channels cannot be excluded. Future studies should conduct the Ca2+ release assay in a buffer without Ca2+ or using chelating agents to deplete Ca2+ in media to confirm whether the increased Ca2+ comes solely from the intracellular stores. Incubation with EGF at 10 ng/mL for 6 h increased the protein expression of RyR2 in JB6 cells but not the mRNA expression, suggesting that EGF may induce RyR post-transcriptional or posttranslational regulation. Since intracellular Ca2+ release causes ERK activation (40), it remains to determine whether ERK activation is the cause for RyR2 upregulation. A pilot study examined the human skin tissue arrays. Consistent with the in vitro data, the expression of RyR2 and pERK1/2 demonstrated colocalization and significant correlation in human SCC and melanoma. Similarly, a previous study reported the expression of RyR2 was upregulated in K-RAS-mutant colorectal cancer (41). The JB6 model system includes transformation-sensitive (P+) and transformation resistant (P−) cell lines. EGF induces anchorage-independent growth of the JB6 P+ cells in soft agar, but the P− cells are resistant to tumor promotion (42), attributable to a low level of MAPK/ERK activation (43). It remains to determine whether RyRs in JB6 P+ and P− cells respond to EGF differently. Although the exact mechanism is unknown, EGF- and UV-mediated RyR upregulation and functional potentiation suggest an important role of RyRs in skin cancer promotion. Treatment of JB6 and HaCaT cells with known RyR stabilizers attenuated EGF/4-CMC-evoked Ca2+ response. The RyR stabilizers include carvedilol and its enantiomers R- and S-carvedilol, dantrolene, JTV519, S107, and nebivolol. These compounds do not work like classical antagonists but stabilize the unstable conformation of the channel and reduce pathological “leakiness”. Thus, it is more precise to name these compounds as RyR stabilizers (14, 44). Dantrolene had a greater inhibitory effect in HaCaT cells possibly because it was more effective in targeting RyR1 (16). Most β-blockers we tested did not affect EGF-induced Ca2+ response, except for atenolol and nebivolol. Consistently, nebivolol has been shown to possess the ability to suppress RyR2 by stabilizing the leaky channel (19). EGF-potentiated Ca2+ release can be inhibited by an inhibitor of MAPK/ERK (U0126) and inhibitors of PLC, indicating an essential role of MAPK/PLC in mediating the effects of EGF on the RyRs. Since many extracellular and intracellular factors can activate MAPK, we determined whether other factors have the same potentiating effects. Using a genetic approach, transfection of constitutively active K-RAS and MEK1 confirmed the role of MAPK in RyR-mediated Ca2+ release. UV radiation, which can activate MAPK (45), at the dose of 15 or 50 mJ/cm2, demonstrated a similar potentiation effect. These data altogether support the hypothesis that tumor promoters-activated MAPK cross-talks with RyR-Ca2+. To examine the role of RyR-mediated Ca2+ release in EGF-mediated JB6 transformation, dose–response curves of JB6 transformation by EGF (10 ng/mL) in the presence of RyR stabilizers were plotted. These agents consistently inhibited EGF-induced soft agar colony formation in a dose-dependent manner and at nontoxic concentrations. BAPTA-AM, used to sequester intracellular Ca2+, similarly inhibited the transformation, supporting the hypothesis that increased intracellular Ca2+ release is essential for JB6 transformation. Interestingly, among all the drugs tested, U0126 showed the highest potency in inhibiting JB6 transformation, which is correlated with its effects on Ca2+ release. RyR agonists 4-CMC and suramin, and ionomycin (Ca2+ ionophore) transform JB6 cells without the need of EGF, suggesting that increased Ca2+ release alone is sufficient to transform JB6 cells. Another piece of evidence to support our hypothesis is that carvedilol’s antitransformation activity can be reduced by RyR2 or RyR3 gene knockdown using shRNA. Furthermore, the antitransformation potency of several carvedilol derivatives on JB6 is correlated with their potency of store-overload-induced Ca2+ release (SOICR)/RyR inhibition. The derivatives are structurally similar to carvedilol, but contain modified catechol or carbazole subunits to obtain various degrees of RyR inhibition (24). In transformation assay, compound 93 showed the most potent inhibition (IC50 :0.38 μM), while compound 88 showed the least potency (IC50 : 6.20 μM). The IC50 values on transformation are positively correlated with their IC50 values on RyR inhibition, indicating RyR inhibition is likely the mechanism of antitransformation for carvedilol and its derivatives. However, for compound 88, even though it does not suppress SOICR (IC50 = >1,000 μM), it still inhibits transformation (IC50 = 6.20 μM), suggesting that there are unknown mechanisms other than RyR inhibition. To prove our hypothesis in vivo, SKH-1 mice were exposed to UV to activate MAPK due to limited skin delivery of EGF. UV and the RyR agonist 4-CMC demonstrated a synergistic effect in inducing DNA damage and inflammation in mice based on single-dose and multiple-dose UV protocols. Since carvedilol and R-carvedilol attenuated UV-induced DNA damage and inflammation in hairless mice (5, 21), future studies should examine the drug effects of RyR stabilizers against skin lesions induced by UV and 4-CMC. Our hypothesis is further strengthened by the RyR2-E4872Q knock-in mouse model in which the RyR2 activity is reduced because of shortened channel open time (30). The RyR2-E4872Q mice exhibit attenuated UV-induced DNA damage and IL-6 levels. The phenotypes observed in RyR2-E4872Q mice were less pronounced than those induced by the pharmacological agonist 4-CMC in vivo, possibly because 4-CMC activates all three RyR isoforms, whereas the reduced function of the RyR2-E4872Q allele may be partially compensated by other RyR subtypes (46, 47). EGF and UV potentiate RyRs via both genomic and nongenomic pathways, reflecting their ability to activate posttranscriptional programs as well as rapid kinase signaling cascades. MAPK–RyR cross-talk may be partially mediated via PLC activation, due to the inhibitory effects of D609 and U73122, although how PLC activation potentiates RyRs needs further investigation. This study reports on the effect of EGF on RyR-Ca2+ signaling, although several previous studies suggest a cross-talk between MAPK and Ca2+ release. One report describes the effects of EGF on cholecystokinin/gastrin receptor (CCK2R)-mediated Ca2+ release, mediated by PLC/IP3 and by MAPK (48). Another study reported for 17β-estradiol (E2), which had no observable effect on basal Ca2+ but induced a rapid Ca2+ release from RyRs evoked by the Ca2+ ATPase pump (SERCA) inhibitor thapsigargin, in a human eccrine sweat gland cell line (49). Such activity of E2 was found to be mediated by estrogen receptors, PKC, PKA, and MAPK (49). Since E2 can transform nontumorous mammary gland epithelial cells (50) and activation of CCK2R can promote gastric cancer (51), we speculate that increased intracellular Ca2+ release may represent a common downstream event of activated MAPK; however, further studies are needed to test this hypothesis. Given that MAPK activates the transcriptional factor AP-1, which is required for JB6 transformation (43), future studies should evaluate whether AP-1 activity can be regulated by RyR-mediated Ca2+ release. Although many Ca2+ -binding proteins can be the downstream effectors of RyR-mediated Ca2+ release, one such effector may be NFAT, which can be activated by UV through a Ca2+ -dependent mechanism in JB6 cells (52). Future studies will identify the essential Ca2+ -binding effectors that are downstream of activated RyRs. In conclusion, the present study advances the field by demonstrating that EGF-enhanced Ca2+ release via RyR channels requires MAPK activation, which have been previously proved as an essential transducer for skin tumor promotion (43, 53). RyRs and related Ca2+ signals, potentiated by MAPK activation, play a central role in promoting neoplastic transformation and skin carcinogenesis, and therefore represent a druggable target for cancer chemoprevention. Although inhibitors of EGF receptor or MAPK pathways could have chemopreventive potential, they are not yet widely used for this purpose due to side effects that limit their long-term effectiveness. Findings from this study suggest that RyR stabilizers such as FDA-approved drugs carvedilol and dantrolene can be repurposed as preventive agents for skin cancer. Materials and Methods Cells, Reagents, Kits, and Equipment. The JB6 CI 41–5a cell line (named JB6 throughout this report) was purchased from American Type Culture Collection (ATCC, Manassas, VA). The detailed methods of cell culture condition are described in SI Appendix. Animals. All animal studies were approved by the Institutional Animal Care and Use Committees (IACUC) at the Western University of Health Sciences and were performed in accordance with US NIH guidelines. The mouse strains and animal study protocols are described in SI Appendix. Statistical Analysis. Data are expressed as the mean ± SEM unless described differently in the figure legends. Statistical significance was set at P < 0.05. Data, Materials, and Software Availability. All study data are included in the article and/or SI Appendix. Supplementary Material This article contains supporting information online at https://www.pnas.org/lookup/suppl/doi:10.1073/pnas.2505055122/-/DCSupplemental.