TRPM7’s Channel Activity Promotes the Pathogenesis of Abdominal Aortic Aneurysm

Abstract

Abdominal aortic aneurysm (AAA) is seen in 1–2% of the elderly. Rupture of AAA usually causes uncontrollable lethal hemorrhage, and its risk increases with AAA size. However, there is no effective pharmacological therapy for hindering AAA growth. Here we show global or VSMC-specific TRPM7 knockout in mice prevented AAA formation, featured by inhibited VSMC reprogramming, reduced inflammatory infiltration, and suppressed matrix degradation. Mechanistically, we demonstrated TRPM7-mediated Ca²⁺ signaling promotes KLF4 activation, driving VSMC reprogramming and accelerating AAA growth. By generating channel-dead and kinase-inactive knock-in mice, we found it is the channel function, rather than kinase activity, that is required for TRPM7-mediated AAA pathogenesis. Importantly, TRPM7 inhibitor NS8593 suppressed VSMCs reprogramming, and protected mice against AAA formation. Our data suggest that TRPM7 is a promising therapeutic target for developing effective prophylactic medications to limit AAA progression. Additionally, the channel-dead TRPM7 knock-in mice will serve as a valuable tool for elucidating the roles of TRPM7 in other pathophysiological conditions.

Introduction

Abdominal aortic aneurysm (AAA) is an irreversible and localized dilation of abdominal aorta distal to the renal arteries¹, with a high prevalence in aged population (1–2% of the elderly > 65-year-old)². Once ruptured, uncontrollable major hemorrhage leads to a mortality as high as 85%¹. Unfortunately, there is no effective pharmacological therapy to inhibit the growth of AAAs. A key pathological feature of AAA is reprogramming of vascular smooth muscle cells (VSMCs) from contractile to proliferative/synthetic³, causing profound inflammatory responses in the aortic walls⁴. Reprogramming of VSMCs during aneurysm is primarily controlled by the transcriptional factor Kruppel-like factor 4 (KLF4)⁵. KLF4 shuts off the expression of contractile phenotype related genes, while enhancing the expression of pro-inflammatory genes⁶. However, it remains unclear how KLF4 is abnormally activated in the development and progression of AAA⁶.

TRPM7 is a Ca²⁺-permeable channel-kinase. Recently, a small G-protein ADP-ribosylation factor-like protein 15 (ARL15) has been shown to be an endogenous negative regulator of TRPM7. Yet, the pathophysiological significance of ARL15 regulation on TRPM7 is unknown^7,8. Ubiquitously expressed in different cells, TRPM7 is involved in various cellular functions⁹. The pathophysiological functions of TRPM7 in vivo, however, remain poorly understood. This is because as a bi-functional channel-kinase, it is hard to separately study the contribution of TRPM7’s channel or kinase function in many conditions. Although the kinase function has been defined in several studies using the kinase-dead TRPM7 knock-in mice¹⁰, the pathophysiological roles of TRPM7’s channel function remain poorly understood due to the lack of TRPM7 channel-dead knock-in mice. In this study, by generating channel-dead TRPM7 knock-in mice and using kinase-dead TRPM7 knock-in mice, we sought to comprehensively understand the contribution of TRPM7 in AAA pathogenesis.

Results

Global TRPM7 knockout inhibits AAA formation and growth

To examine the role of TRPM7 in AAA, we crossed Trpm7^loxp/lox -Cre^ERT2 mice with Apoe^−/− mice and induced AAA formation using subcutaneous angiotensin II (Ang II) infusion¹¹. Before Ang II treatment, Trpm7 deletion was induced by tamoxifen for daily injection for 1 week (Extended Data Fig. 1a), and successful knockout was confirmed by genotyping (Extended Data Fig. 1b), current recording (Extended Data Fig. 1c) and Western blotting (Extended Data Fig. 1d). As shown in Fig. 1a, Trpm7^−/−Apoe^−/− mice exhibited smaller AAA diameter compared to Trpm7^+/+Apoe^−/− littermates. Moreover, the survival rate was markedly higher (Fig. 1b and 1c) whereas the prevalence of AAA (Fig. 1d) was lower in Trpm7^−/−Apoe^−/− mice. Post-mortem autopsy further confirmed that Trpm7^−/−Apoe^−/− mice have lower incidence of AAA formation and smaller AAA size (Fig. 1e). A key feature of AAA progression is elastin degradation, primarily caused MMPs³. We found that there was reduced elastin degradation in Trpm7^−/−Apoe^−/− mice (Fig. 1f and Extended Data Fig. 1e). Moreover, in Trpm7^−/−Apoe^−/− mice, there was lower MMP2 expression (Fig. 1g) and reduced MMP activity (Fig. 1h) in the media of abdominal aorta segments adjacent to the aneurysm lesions. One critical cause of elastin degradation is aortic inflammation caused by immune cell invasion. By reanalyzing the single cell sequencing data of control and aneurysmal human aortic samples (GSE155468)¹², we found that except B cells and plasma cells, almost all the infiltrated immune cells have a high expression of TRPM7 (Extended Data Fig. 1g and 1h). In the aneurysmal lesions of Trpm7^+/+Apoe^−/− mice, we found a marked increase in IL-1β expression, accompanied by a massive infiltration of CD45⁺ immune cells into the adventitia, both of which were attenuated in Trpm7^−/−Apoe^−/− mice (Fig. 1i). Meanwhile, the enhanced expression of the chemokine monocyte chemoattractant protein-1 (MCP1) in the adventitia was inhibited in Trpm7^−/−Apoe^−/− mice (Fig. 1j), which could explain the reduced immune cell invasion. We further evaluated the protective effect of TRPM7 knockout against AAA induced by CaCl₂ (Extended Data Fig. 1a). We found that, Trpm7^−/− mice exhibited smaller AAA size two weeks after applying CaCl₂, as shown by both abdominal ultrasound (Fig. 1k) and post-mortem autopsy (Fig. 1l), as well as attenuated elastin degradation (Fig. 1m). These data collectively suggest that TRPM7 activity is an important promoter in driving AAA pathogenesis.

Fig. 1: TRPM7 Knockout Prevents Aneurysm Formation.

(a), AAA evaluation by abdominal ultrasound (animals n = 6, 9, 6, 12). (b, c), Survival curves and mortality. (d), Incidence of aneurysm formation revealed by autopsy. (e), Representative macroscopic features of AAA (animals n = 6, 9, 6, 12). (f), Elastin degradation evaluation (animals n = 5, 5). (g), MMP2 (green) expression / distribution in the media (animals n = 5, 5) (Red: α-SMA). (h), in situ zymography using DQ-gelatin showing MMP activity (green) in the media (animals n = 5, 5) (Red: α-SMA). (i), CD45⁺ immune cells (red colocalized with DAPI) and IL-1β (green) expression / distribution in the aorta (animals n = 5, 5). (j), MCP1 (green) expression / distribution in the aorta (animals n = 5, 5) (Red: α-SMA). (k), AAA evaluation by abdominal ultrasound after PBS or CaCl₂ application on infrarenal abdominal aorta (animals n = 5, 10, 5, 10). (l), Representative macroscopic features of AAA (animals n = 5, 10, 5, 10). (m), Elastin degradation evaluation (animals n = 5, 5). (Statistical significance was determined using a two-tailed Student’s t-test, or two-sided ANOVA was used for comparing multiple groups with one variable).

Global TRPM7 knockout inhibits immune cell invasion

To better characterize the detailed features of infiltrated immune cells and determine the potential contribution of TRPM7, we performed flow cytometry analysis of single-cell suspensions collected after digestion of aneurysmal lesions. Compared to Trpm7^+/+Apoe^−/− mice, TRPM7 knockout inhibited the infiltration of all major immune cell populations, including myeloid cells, neutrophils, Ly6C^high and Ly6C^low macrophages, lymphocytes, B cells, and T cells (Fig. 2a to 2c), suggesting that TRPM7 knockout does not selectively influence the infiltration of specific immune cell populations. Additionally, TRPM7 knockout reduced the proliferation of CD29⁺ fibroblasts in aneurysm lesions (Fig. 2a, 2b and 2d)¹³. Interestingly, we did not observe a significant difference between the endothelial cell counts between Trpm7^+/+Apoe^−/− and Trpm7^−/−Apoe^−/− mice (Fig. 2a, 2b and 2e). The generically reduced immune cell infiltration could potentially result from inhibiting hematopoiesis by TRPM7 knockout, as TRPM7 is critical for early development¹⁴. Thus, we analyzed the peripheral leukocytes population in the same mice we used for identifying the immune cell population in aneurysm lesions. The results showed that there was no difference of major immune cell populations between Trpm7^+/+Apoe^−/− and Trpm7^−/−Apoe^−/− mice (Fig. 2f to 2h), indicating that TRPM7 does not influences immune cell generation in our animal model.

Fig. 2: TRPM7 Knockout Reduces Immune Cell Invasion in Aneurysm.

(a, b), Calcein blue was used to gate alive cells, which was further divided into immune cells and non-immune cells based on CD45 gate. Based on light scatter (cell size), CD45⁺ immune cells were divided into myeloid cells and lymphocytes. Neutrophils were separated using Ly6G. Macrophages were characterized using CD11b and Ly6C. B cells and T cells were gated from lymphocytes using CD19 and CD3, respectively. In the non-immune cell population, CD31 was used to separate endothelial cells, and CD29⁺ fibroblast population was separated from CD31⁻ cells. (c-e), Quantification of immune cell, fibroblast and endothelial cell counts (animals n = 6, 6). (f, g), Immune cell population in the peripheral blood. (h), Quantification of peripheral immune cell counts (animals n = 6, 6). (Statistical significance was determined using a two-tailed Student’s t-test, or two-sided ANOVA was used for comparing multiple groups with one variable).

TRPM7 knockout inhibits VSMC reprogramming

To understand how Trpm7 deletion inhibits AAA pathogenesis and immune cell infiltration, we next examined whether VSMC reprogramming is influenced¹⁵. Our reanalysis of the single-cell sequencing data¹² revealed that, compared to control samples, aneurysmal human aortic samples exhibited an almost complete loss of a stress-related contractile VSMC population (Extended Data Fig. 2a–2c). This loss of contractile VSMCs was accompanied by a marked increase in VSMCs expressing genes related to ECM synthesis and the emergence of a population highly expressing genes commonly associated with fibroblasts (Fig. 3a; Extended Data Fig. 2b and 2c). Interestingly, this fibroblast-like population exhibits high expression of KLFs, particularly KLF2, KLF4, KLF6, and KLF9, which are not typically abundant in fibroblasts but are known to be heavily involved in VSMC reprogramming¹⁶, suggesting that this population may have originated from VSMCs. More importantly, we observed that the basal expression level of KLF4 was much lower compared to KLF2, KLF6, and KLF9 in contractile VSMCs (less than 5–10% of their levels; Extended Data Fig. 2c), but its expression substantially increased to levels comparable to those observed in reprogrammed VSMCs and the fibroblast-like population (Extended Data Fig. 2c). This indicates the critical role of KLF4 in VSMC reprogramming during AAA pathogenesis. Therefore, we chose KLF4 as a key indicator for monitoring VSMC reprogramming. We found that in Trpm7^+/+Apoe^−/− mice, there was a substantial increase of α-SMA⁺ cells in the adventitia (Fig. 3b). Compared to α-SMA⁺⁺ VSMCs in the media, those abnormal cells in the adventitia exhibited significantly lower α-SMA expression level (Fig. 3b and 3c, Extended Data Fig. 1f). Meanwhile, KLF4 expression in the nuclei of adventitial α-SMA⁺ cells was higher in Trpm7^+/+Apoe^−/− mice than that in Trpm7^−/−Apoe^−/− mice (Fig. 3b–c), suggesting a more active reprogramming process^6,15. Also, ss shown in Extended Data Fig. 2d, there is significantly a more active VSMC proliferation (α-SMA/Ki67 double positive cells) in Trpm7^+/+Apoe^−/− mice than that in Trpm7^−/−Apoe^−/− mice.

Fig. 3: TRPM7 Knockout Inhibits Reprogrammingof Vascular Smooth Muscle Cells.

(a) Analysis of the changes in KLFs and Ca²⁺ signaling molecule expression across different VSMC populations from control and aneurysmal human aortic tissues. (b, c), KLF4 (green, colocalized with DAPI) expression in the aortas (animals n = 5, 5) (Red: α-SMA). (d, e), TRPM7 current recordings (cell number = 6–10/group isolated from at least 5 mice). (f, g), Ca²⁺ imaging of VSMCs (cell number = 6, 6 isolated from at least 5 mice), and quantification of maximal Ca²⁺ changes mediated by TRPM7 at basal level and response to 50 μM naltriben (cell number = 12, 12 isolated from at least 5 mice). (h), AAA evaluation by abdominal ultrasound (animals n = 5, 9, 5, 12). (i, j), Survival curves and mortality. (k), Incidence of aneurysm formation revealed by autopsy. (l), Representative macroscopic features of AAA (animals n = 5, 9, 5, 12). (Statistical significance was determined using a two-tailed Student’s t-test, or two-sided ANOVA was used for comparing multiple groups with one variable).

To exclude other interfering factors and define the role of TRPM7 in VSMC reprogramming, we performed in vitro experiments. The purity of VSMCs was confirmed by immunostaining using α-SMA and SM22α, and collagen I to detect fibroblast contamination, CD31 to detect endothelial cell contamination, and CD45 to determine immune cell contamination (Extended Data Fig. 3a to 3d). Moreover, we also validated the purity of VSMCs across the entire culture dish by performing flow cytometry analysis for SM22α (Extended Data Fig. 3e and 3f). We found that Ang II-induced increase of nuclear KLF4 level and loss of α-SMA were inhibited in VSMCs from Trpm7^−/−Apoe^−/− mice (labeled with M7KO to distinguish in vitro data from ex vivo data) (Extended Data Fig. 3g, h). Similarly, TRPM7 knockout inhibited the enhanced proliferation activity of VSMCs (Extended Data Fig. 3g, h), and Ang II-induced MMP2 and IL-1β expression (Extended Data Fig. 3g, h). Also, we found that TRPM7 knockout preserved VSMC contractility after Ang II treatment (Extended Data Fig. 3i). We further used western blot to further confirm the protective effects of TRPM7 knockout against Ang II-induced VSMC reprogramming, including the loss of α-SMA and SM22α, the increase of MMP2 and IL-1β, as well as enhanced KLF4 nuclear translocation (Extended Data Fig. 3j).

In WT VSMCs, whole-cell current recording revealed a tiny basal current with typical features of TRPM7 current. The endogenous current was inhibited by 200 μM 2-APB but activated by 5 mM 2-APB (Fig. 3d, 3e, and Extended Data Fig. 3k and 3l), which is a unique feature of TRPM7. This 2-APB-regulated endogenous current was absent in M7KO VSMCs (Fig. 3d, 3e, and Extended Data Fig. 3k and 3l). We found that Ang II treatment potentiated TRPM7 current in WT VSMCs (Fig. 3d, 3e, and Extended Data Fig. 3k and 3l). Interestingly, Ang II did not significantly increase the response of TRPM7 current to 5 mM 2-APB (Fig. 3d, 3e, and Extended Data Fig. 3k and 3l). We also examined the very initial current recorded immediately after rupture, when the intracellular content is unaffected by the internal recording solution, better reflecting the physiological status of TRPM7. As in Extended Data Fig. 3m, the initial TRPM7 currents in VSMCs was significantly enhanced by Ang II, suggesting increased basal TRPM7 activity by Ang II. Our findings suggest that under basal conditions, there probably exists an endogenous regulator that inhibits the activation of TRPM7, and that Ang II treatment removes this inhibition.

TRPM7 is a Ca²⁺-permeable ion channel, and Ca²⁺ signaling is an important regulator of VSMC reprogramming^17,18. Our reanalysis of the single cell sequencing data also revealed the abundant expression of Ca²⁺ signaling related genes in reprogrammed VSMCs (Fig. 3a). Thus, we next evaluated whether Ang II treatment influences TRPM7-mediated Ca²⁺ influx in VSMCs following a well-established protocol¹⁹. Persistent perfusion with 10 μM nifedipine was used to inhibit the activation of L-type calcium channels, and 20 mM Ca²⁺ solution¹⁹ was used to induce the basal Ca²⁺ responses in VSMCs for 5 minutes (Fig. 3f). We also used a TRPM7 agonist naltriben to maximize TRPM7-induced Ca²⁺ responses²⁰. Ang II incubation leads to a bigger Ca²⁺ influx in VSMCs (Fig. 3f and 3g). Surprisingly, naltriben failed to induce a significantly larger Ca²⁺ response in Ang II-treated VSMCs (Fig. 3f and 3g). This finding recapitulates the results we observed in whole-cell TRPM7 current recording that 5 mM 2-APB did not further increase TRPM7 activity in VSMCs treated with Ang II (Fig. 3f and 3g). In contrast, Ang II did not induce the same enhancement of Ca²⁺ responses in M7KO VSMCs, and TRPM7 agonist naltriben did not further increase Ca²⁺ responses (Fig. 3f and 3g), suggesting that this increased Ca²⁺ response we observed in WT VSMCs is caused by enhanced TRPM7 activity. Considering the maximum response of TRPM7 currents and TRPM7-mediated Ca²⁺ influx to TRPM7 activators were not significantly affected by Ang II treatment, this further suggests that Ang II may enhance TRPM7 activity through releasing a basal intracellular inhibitor.

Moreover, we achieved inducible TRPM7 knockout in VSMCs using Myh11-cre mice (Extended Data Fig. 4a and 4b).We found that, compared to Myh11-cre⁻Trpm7^loxp/loxpApoe^−/− control littermates, Myh11-cre⁺Trpm7^loxp/loxpApoe^−/− mice exhibited substantially smaller AAA size (Fig. 3h). These mice also showed higher survival probability (Fig. 3i) and lower mortality rate (Fig. 3j). Post-mortem evaluation confirmed that Myh11-cre⁺Trpm7^loxp/loxpApoe^−/− mice had a lower incidence of AAA formation (Fig. 3k), inhibited AAA progression (Fig. 3l), and attenuated elastin degradation in aneurysmal lesions (Extended Data Fig. 4c). The similar protective effects of global Trpm7 deletion (Fig. 1a–1e) and VSMC-specific Trpm7 deletion (Fig. 3h–3l) against AngII-induced AAA strongly suggest that TRPM7 overactivation in VSMCs is responsible for the exacerbation of AAA progression, whereas the reduced immune cell infiltration and fibrosis observed in the aneurysmal lesions of TRPM7 knockout mice (Fig. 2) are likely secondary to the attenuated VSMC reprogramming and AAA pathogenesis in TRPM7 knockout mice.

ARL15-TRPM7-Calmodulin-Calcineurin signaling

To understand the mechanisms by which Ang II regulates TRPM7 activity in promoting VSMC reprogramming and inducing AAA, we investigated ARL15, a recently identified protein which inhibits TRPM7 channel activity^7,8,21,22. We found that Ang II caused a marked reduction in ARL15 expression but no significant change of TRPM7 expression in VSMCs (Fig. 4a, 4c; Extended Data Fig. 4d–4e). In aneurysm tissue from Trpm7^+/+Apoe^−/− mice, Ang II infusion resulted in a substantial increase of TRPM7 expression, accompanied by a milder but significant reduction in ARL15 (Fig. 4b, 4c; Extended Data Fig. 4f–4g). Considering our reanalysis of single cell sequencing data revealed a substantial infiltration of immune cell populations with high TRPM7 expression (Extended Data Fig. 1g and 1h), the marked increase in TRPM7 expression observed in the aneurysmal tissue from our mice may be attributed to these immune cells. Interestingly, in VSMC populations from both control and aneurysmal human aortic samples, TRPM7 mRNA expression was relatively high across many VSMCs, but its expression did not significantly increase following reprogramming (Extended Data Fig. 5a). However, TRPM7 was expressed in VSMCs characterized by high levels of MMP2, IL1B, KLF2, and KLF4, along with low levels of ACTA2 and TAGLN (Extended Data Fig. 5b). Additionally, ARL15 mRNA expression in these VSMC populations was not significantly altered, suggesting that the reduced ARL15 protein levels may result from increased degradation (Extended Data Fig. 5a). C-terminal tail of TRPM7 can be cleaved at different positions in different cell types and tissues²³. We found that there were no differences in cleaved fragments detected by western blot in between aneurysmal and control aorta (Extended Data Fig. 4d), or between AngII treated and PBS treated SMCs (Extended Data Fig. 4f). We also evaluated whether Ang II treatment triggers the dissociation of ARL15 from TRPM7, and found that Ang II did not affect the binding of ARL15 to TRPM7 (Extended Data Fig. 4h). Nonetheless, our results indicate that ARL15 is a promising candidate for explaining the basal TRPM7 inhibition observed in Fig. 3d to 3g, and ARL15 may play an important role in regulating TRPM7 activity in AAA.

Fig. 4: TRPM7 Activation Promotes Reprogramming of Vascular Smooth Muscle Cells.

(a-c), TRPM7 and ARL15 protein levels in VSMCs (dish numbers = 6, 6, 6, 6, isolated from at least 5 mice), as well as in aortas (animals n = 12, 12). (d, e), TRPM7 current recordings in WT VSMCs infected with ad-ARL15 or mutant ARL15 (Ad-ARL15-Mut) (cell number = 6–10/group isolated from at least 5 mice). (f), Ki67 (green) and as α-SMA (red) expression in primary VSMCs (cell number = 6/group isolated from at least 5 mice). (g), Collagen gel contraction assay (dish numbers = 8, 8, isolated from at least 5 mice). (h), MMP2, IL-1β, α-SMA, SM22α and KLF4 protein levels in VSMCs (dish numbers = 6/group, isolated from at least 5 mice). (i), Ca²⁺ imaging of VSMCs (cell number = 6, 6 isolated from at least 5 mice), and quantification of maximal Ca²⁺ changes mediated by TRPM7 at basal level and response to 50 μM naltriben. (cell number = 12, 12 isolated from at least 5 mice). (j), Ki67 (green, colocalized with DAPI) and α-SMA (red) expression in VSMCs 12 hours after DMSO or 10 μM Naltriben treatment (dish numbers = 6/group, isolated from at least 5 mice). (k), Graphic illustration showing the potential mechanism of TRPM7-meidated Ca²⁺ signaling in promoting VSMC reprogramming. (l), Ki67 (green) and α-SMA (red) expression in primary VSMCs isolated from WT mice 12 hours after Naltriben+DMSO, Naltriben+FK506 or Naltriben+W13 treatments (dish numbers = 6/group, isolated from at least 5 mice). (Statistical significance was determined using a two-tailed Student’s t-test, or two-sided ANOVA was used for comparing multiple groups with one variable).

We further investigated the effects of ARL15 on TRPM7 in VSMCs using adenoviral overexpression of ARL15. ARL15 R95A mutant (ARL15-Mut) that was unable to inhibit TRPM7 was used as a negative control²⁴. We found that ARL15 overexpression inhibited both the basal activation and Ang II-induced potentiation of TRPM7 (Fig. 4d, 4e, and Extended Data Fig. 5d). Similarly, there were smaller initial TRPM7 currents in VSMCs overexpressed with ARL15 (Extended Data Fig. 5e). Considering the loss of ARL15 and the enhancement of TRPM7 activity following VSMC reprogramming, along with the inhibition of Ang II-induced reprogramming by ARL15, it is highly likely that the loss of ARL15 contributes to TRPM7 overactivation and VSMC reprogramming induced by Ang II. Moreover, ARL15 overexpression inhibited Ang II-induced loss of α-SMA (Fig. 4f), increase of KLF4 (Extended Data Fig. 5F and 5G)/Ki67 (Fig. 4f)/MMP2 (Extended Data Fig. 5f and 5h)/IL1β (Extended Data Fig. 5f and 5i) expression, and compromise of contractility (Fig. 4g). Consistently, western blotting showed that ARL15 overexpression inhibited VSMC reprogramming (Fig. 4h). We next determined whether ARL15 overexpression influences Ca²⁺ responses in VSMCs. The results showed that compared to ARL15-MUT, ARL15 overexpression not only inhibited increase of the basal Ca²⁺ response after Ang II treatment, but also abolished the naltriben-induced increase of Ca²⁺ influx (Fig. 4i). Collectively, these data indicates that the enhanced TRPM7 activation caused by Ang II-induced loss of ARL15 plays a critical role in VSMCs reprogramming.

After elucidating how Ang II enhances TRPM7 activation and TRPM7-mediated Ca²⁺ influx, we sought to determine how TRPM7-mediated Ca²⁺ signaling promotes VSMC reprogramming. Since Ang II activates a vast signaling network, we selected a TRPM7 agonist, naltriben, to study the cellular effects in VSMCs following TRPM7 activation²⁰. We found that a 24-hour naltriben treatment resulted in increased KLF4 nuclear levels in WT VSMCs (Extended Data Fig. 6a). Similarly, naltriben markedly enhanced the proliferation of WT VSMCs (Fig. 4j) and increased the production of IL-1β (Extended Data Fig. 6b) and MMP2 (Extended Data Fig. 6c). Western blotting also showed that in WT VSMCs, naltriben incubation led to VSMC reprogramming (Extended Data Fig. 6d). All these changes were eliminated by TRPM7 knockout in VSMCs (Extended Data Fig. 6d). These results indicate that TRPM7 activation by naltriben drives VSMC reprogramming.

After entering cytosol, Ca²⁺ binds to various Ca²⁺ binding proteins, including calmodulin (CaM), to exert its function as a second messenger (Fig. 4k)²⁵. We found that CaM inhibitor, FK506, inhibited the naltriben-induced increased expression of KLF4 (Extended Data Fig. 6e and 6f), Ki67 (Fig. 4l), IL-1β (Extended Data Fig. 6e and 6f) and MMP2 (Extended Data Fig. 6e and 6h). An important downstream molecular of CaM is calcineurin (Fig. 4s). Calcineurin enhances the transcriptional activity of myocyte enhancer factor 2 (MEF2), and MEF2 is required for the expression of KLF4 (Fig. 4s)²⁶. Our results showed that calcineurin inhibitor W13 produced a similar inhibitory effect on naltriben-induced VSMC reprogramming as CaM inhibitor FK506 (Fig. 4l, Extended Data Fig. 6e to 6h). Similarly, Western blotting also showed that inhibition of CaM and calcineurin prevented the VSMC reprogramming induced by TRPM7 activation (Extended Data Fig. 6i). We also examined whether other signaling molecules downstream of calcineurin, such as the well-established NFAT signaling, were affected by TRPM7 activation. However, TRPM7 knockout did not affect the increase of nuclear NFAT1 by Ang II (Extended Data Fig. 6j), nor NFAT inhibition prevent naltriben-induced loss of α-SMA and SM22α (Extended Data Fig. 6k).

Besides Ca²⁺, TRPM7 is also highly permeable to Zn²⁺. As in Extended Data Fig. 7a and 7b, the Zn²⁺ chelator TPEN completely abolished the pro-reprogramming effects induced by TRPM7 activation by naltriben (Extended Data Fig. 7a and 7b), suggesting that TRPM7-mediated Zn²⁺ may play a role in VSMC reprogramming. However, TPEN can also chelate Ca²⁺ when Ca²⁺ concentration is high (KdCa = 68 μM)²⁷. Additionally, our reanalysis of single-cell sequencing data of human aneurysmal aortas revealed that there is only a slight upregulation of the Zn²⁺ binding transcription factors ZFP36 and NFE2L2, and the abundance of Zn²⁺-signaling-related molecules in VSMCs (Extended Data Fig. 7c and 7d) was much lower compared to KLFs and those associated with Ca²⁺ signaling (Fig. 3a). Therefore, although TPEN inhibits naltriben-induced VSMC reprogramming, future studies are needed to conclude whether TRPM7-mediated Zn²⁺ influx significantly contributes to VSMC reprogramming.

Our data indicate that TRPM7-mediated Ca²⁺ signaling promotes VSMC reprogramming through CaM-calcineurin-KLF4 pathway (Fig. 4k).

TRPM7’s channel function is required for VSMC reprogramming

To elucidate whether TRPM7-mediated Ca²⁺ signaling is necessary for VSMC reprogramming, we utilized kinase-dead and channel-dead TRPM7 knock-in mice. The kinase-dead TRPM7 knock-in mice generated by a point mutation eliminates kinase function but does not compromise TRPM7 channel function¹⁰ (Extended Data Fig. 8a and 8b), and were designated as Trpm7^KD (Kinase-dead) for simplicity.

The TRPM7-E1047K mutant is a dominant-negative mutant as reported previously^28,29. We confirmed that in a TRPM7-E1047K overexpressing cell line, the induction of TRPM7-E1047K expression using tetracycline eliminated the endogenous TRPM7 current (Extended Data Fig. 8c). This current could not be significantly rescued by the overexpression of wild-type TRPM7 (Extended Data Fig. 8c), suggesting that the incorporation of E1047K subunit can effectively inactivate TRPM7 channel function. We therefore generated conditional channel-dead TRPM7-E1047K knock-in mouse (Fig. 5a). Briefly, in TRPM7-E1047K mice, exon 22 is floxed and followed by a floxed inverted exon 22 containing the point mutation E1047K. Expression of Cre recombinase causes the excision of the wild-type exon 22 and the inversion of the mutated exon22 containing the E1047K substitution that inactivates the TRPM7 channel function (Extended Data Fig. 8c and 8d). We then used TRPM7^+/E1047K mice for in vivo experiments (TRPM7^CD). TRPM7^+/E1047K mice are normal, induction of channel-dead by global cre (TRPM7^CD) at adult age (8 weeks old) does not cause any abnormality. The introduction of E1047K loxp site was confirmed by genotyping PCR (Extended Data Fig. 8d). Recording showed that the channel activity of TRPM7 was absent (Extended Data Fig. 8e), while the kinase activity was intact (Extended Data Fig. 8f and 8g). Kinase-dead and channel-dead TRPM7 mutant knock-in mice were crossed with Apoe^−/− mice to generate Trpm7^KDApoe^−/− and Trpm7^CDApoe^−/−mice. For VSMCs isolated from Trpm7^KDApoe^−/− and Trpm7^CDApoe^−/−mice, we used abbreviations “M7KD” and “M7CD” for convenience (Fig. 5).

Fig. 5: TRPM7’s Channel Activity is Required for Promoting Reprogramming of Vascular Smooth Muscle Cells.

(a), Graphic illustration showing knock-in of channel-dead TRPM7. (1): Structure of the murine TRPM7 locus, targeting vector, and resulting homologous recombination locus. Between exons 21–23 of E1047K mouse is a floxed exon 22 followed by a neomycin cassette and a floxed inverted exon 22 containing the point E1047K mutation with mutant LoxP sites. (2): In the presence of Flp recombinase the neomycin cassette can be excised. (3): Expression of Cre recombinase causes the excision of the wild-type exon and the inversion of the mutated exon 22 containing the E1047K amino acid substitution that inactivates the TRPM7 channel (see details in Methods Section). (b), Ki67 (green) and α-SMA (red) expression in VSMCs (n=6/group). (c), Collagen gel contraction assay (dish numbers = 6/group, isolated from at least 5 mice). (d, e), MMP2, IL-1β, α-SMA, SM22α and KLF4 protein expression in VSMCs (dish numbers = 6/group, isolated from at least 5 mice). (f, g), TRPM7 current recordings (cell number = 6–10/group isolated from at least 5 mice). (h), Ca²⁺ imaging of VSMCs (cell number = 6/group isolated from at least 5 mice), and quantification of maximal basal Ca²⁺ changes mediated by TRPM7 at basal level and response to 50 μM naltriben (cell number = 12/group isolated from at least 5 mice). Ionomycin was used as internal control. (Statistical significance was determined using a two-tailed Student’s t-test, or two-sided ANOVA was used for comparing multiple groups with one variable).

We found that Ang II-induced loss of α-SMA (Fig. 5b), increase in nuclear KLF4 levels (Extended Data Fig. 8h and 8i), enhancement of proliferation (Fig. 5b), and increases in MMP2 and IL-1β (Extended Data Fig. 8h, 8j and 8k) expression were all inhibited in channel-dead, but not kinase-dead, VSMCs. Correspondingly, the compromised VSMC contractility in the M7KD group was preserved in the M7CD group (Fig. 5c). Similarly, M7CD, instead of M7KD, prevented the reprogramming induced by Ang II (Fig. 4d and 4e). Current recording showed that kinase-dead VSMCs still exhibited enhanced TRPM7 basal activation after Ang II treatment, which was absent in channel-dead VSMCs (Fig. 5f, 5g and Extended Data Fig. 8l and 8m). Moreover, channel-dead VSMCs, instead of kinase-dead VSMCs, exhibited resistance to the potentiation of Ca²⁺ responses after Ang II incubation (Fig. 5h). Our results indicate that the channel function of TRPM7, rather than the kinase function, plays a critical role in Ang-II induced VSMC reprogramming.

Channel-dead TRPM7 knock-in inhibits AAA formation

Trpm7^KDApoe^−/− mice did not exhibit reduced AAA diameter (Fig. 6a and 6b), improved survival curve (Fig. 6c), lowered mortality (Fig. 6d),decreased incidence of AAA formation (Fig. 6e) or reduce AAA size (Fig. 6f and 6g). Also, Trpm7^KDApoe^−/− did not reduce elastin degradation (Fig. 6h and 6i), MMP activity (Fig. 6j and 6k), MMP2 expression (Fig. 6j and 6l), and the nuclear level of KLF4 (Fig. 6j and 6m) and Ki67 (Fig. 6j and 6n). The infiltration of CD45⁺ immune cells (Fig. 6j and 6o), enhanced IL-1β expression (Fig. 6j and 6p) and increased MCP1 expression (Fig. 6j and 6q) in the adventitia of aneurysm lesions were not influenced by Trpm7^KDApoe^−/−. Moreover, the infiltration of all major immune cell populations (Fig. 6r to 6t), fibroblast proliferation (Fig. 6r, 6s and 6u) and endothelial cell counts (Fig. 6r, 6s and 6v) were not influenced in the Trpm7^KDApoe^−/− knock-in mice. We also analyzed the peripheral leukocytes population in the same mice that we used for identifying the immune cell population in aneurysm lesions, and found that there was no difference of major immune cell populations between Trpm7^+/+Apoe^−/− and Trpm7^KDApoe^−/− mice (Extended Data Fig. 9a to 9c). Although kinase function was previous shown to be important in modulating immune responses³⁰, our results indicate that TRPM7 kinase function is not required for immune cell invasion during AAA development.

Fig. 6: Genetical Inactivation of TRPM7 Kinase in Mice Does Not Prevent Aneurysm Formation.

(a, b), AAA evaluation by abdominal ultrasound (animals n = 5, 9, 5, 8). (c, d), Survival curves and mortality rate. (e), Incidence of aneurysm formation revealed by autopsy. (f, g), Representative macroscopic features of AAA (animals n = 5, 9, 5, 8). (h, i), Elastin degradation evaluation(animals n = 5, 5). (j-q), Representative images (j) and quantification of MMP activity (k) and expression of MMP2(l), KLF4(m), Ki67(n), CD45(o), IL-1β(p), and MCP1(q) in aortas (α-SMA: red) (animals n = 5, 5). (r, s), Immune cell infiltration and fibroblast proliferation in aortas as in Fig. 2a. (t-v), Quantification of immune cell, fibroblast and endothelial cell counts (animals n = 6, 6). (Statistical significance was determined using a two-tailed Student’s t-test, or two-sided ANOVA was used for comparing multiple groups with one variable).

In contrast, abdominal ultrasound detected smaller AAA diameter in Trpm7^CDApoe^−/− mice (Fig. 7a). Moreover, Trpm7^CDApoe^−/− mice exhibited higher survival rates (Fig. 7b and 7c), lower incidence of AAA formation (Fig. 7d), smaller AAA size (Fig. 7e), and reduced elastin degradation (Fig. 7f). In CaCl₂ model, Trpm7^CD mice also exhibited smaller AAA sizes (Fig. 7g and 7h) and inhibited elastin degradation (Fig. 7i). Moreover, aortas from Trpm7^CDApoe^−/− mice had reduced MMP activity and MMP2 expression (Extended Data Fig. 9d to 9f), and the inhibition of the increase of KLF4 (Fig. 7j) and Ki67 (Fig. 7k) activity, suggesting blunted VSMC reprogramming. Similarly, immune cell invasion and local inflammation, revealed by CD45 and IL-1β (Extended Data Fig. 9d and 9g), and MCP1 (Extended Data Fig. 9d and 9h) staining, were attenuated. Flow cytometry also showed a general reduction in the infiltration of all major immune cell populations (Fig. 7l to 7n). Meanwhile, Trpm7^CDApoe^−/− mice exhibited lower fibrosis in the aneurysm lesions (Fig. 7l, 7m and 7o), while endothelial cell counts (Fig. 7l, 7m and 7p) were not influenced. Meanwhile, there was no difference of major immune cell populations between Trpm7^+/+Apoe^−/− and Trpm7^CDApoe^−/− mice (Extended Data Fig. 9i to 9k).

Fig. 7: Knock-in of Channel-dead TRPM7 in Mice Prevents Aneurysm Formation.

(a), AAA evaluation by abdominal ultrasound (animals n = 6, 9, 6, 12). (b, c), Survival curves and mortality rate. (d), Incidence of aneurysm formation revealed by autopsy. (e), Representative macroscopic features of AAA (animals n = 6, 9, 6, 12). (f), Elastin degradation evaluations (animals n = 5, 5). (g), AAA evaluation by abdominal ultrasound 2 weeks after PBS or CaCl₂ application on infrarenal abdominal aorta (animals n = 5, 9, 5, 10). (h), Representative macroscopic features of AAA (animals n = 5, 9, 5, 10). (i), Elastin degradation evaluations (animals n = 5, 5). (j, k), KLF4(j) or Ki67(k) in aortas (α-SMC: red) (animals n = 5, 5). (l, m), Immune cell infiltration and fibroblast proliferation in aortas as in Fig 2a. (n-p), Quantification of immune cell, fibroblast, and endothelial cell counts (animals n = 6, 6). (Statistical significance was determined using a two-tailed Student’s t-test, or two-sided ANOVA was used for comparing multiple groups with one variable).

In summary, the above data indicate that the channel function, instead of the kinase function, drives AAA pathogenesis.

TRPM7 inhibition suppresses AAA formation

We found that that TRPM7 inhibitors FTY720 and NS8593 treatments prevented the increase of KLF4 expression (Extended Data Fig. 10a and 8b), proliferation activity (Fig. 8a), IL-1β and MMP2 expression (Extended Data Fig. 10a, 10c and 10d), and loss of contractility (Fig. 8b). Importantly, TRPM7 inhibitors produced a similar protective effect as chelating extracellular Ca²⁺ using EGTA (Fig. 8a, Extended Data Fig. 10a to 10d), strongly supporting our mechanistic notion that TRPM7 mediated Ca²⁺ signaling is critical in VSMC phenotype switching. Western blotting showed that NS8593 and FTY720 prevented the reprogramming induced by Ang II (Fig. 8c).

Fig. 8: TRPM7 Inhibition Prevents Aneurysm Formation in Mice.

(a), Ki67 and α-SMA expression in VSMCs 24 hours after Ang II (10 ng/mL) treatment in the presence of DMSO, FTY720, NS8593 or EGTA (dish numbers = 6/group, isolated from at least 5 mice). (b), Collagen gel contraction assay (dish numbers = 6/group, isolated from at least 5 mice). (c), MMP2, IL-1β, α-SMA, SM22α and KLF4 protein expression in WT VSMCs (dish numbers = 6/group, isolated from at least 5 mice). (d-i) 4-weeks daily NS8593 administration immediately after Ang II perfusion. (d), AAA evaluation by abdominal ultrasound (animals n = 4, 10, 4, 12). (e, f), Survival curves and mortality rate. (g), Incidence of aneurysm formation revealed by autopsy. (h), Representative macroscopic features of AAA (animals n = 4, 10, 4, 12). (i), Elastin degradation evaluation (animals n = 5, 5). (j-o) Daily NS8593 administration 2 weeks after Ang II perfusion. (j), AAA evaluation by abdominal ultrasound (animals n = 5, 10, 5, 12). (k, l), Survival curves and mortality rate. (m), Incidence of aneurysm formation revealed by autopsy. (n), Representative macroscopic features of AAA (animals n = 5, 10, 5, 12). (o), Elastin degradation evaluation (animals n = 5, 5). (Statistical significance was determined using a two-tailed Student’s t-test, or two-sided ANOVA was used for comparing multiple groups with one variable).

Given that immune cell infiltration plays a critical role in AAA, the immunosuppressant FTY720 could complicate the interpretation of results. Therefore, we chose NS8593 to more evaluate the role of TRPM7 inhibition in AAA. We found that NS8593 inhibited endogenous TRPM7 currents in VSMCs with an IC₅₀ of 58.0 nM and began to suppress the loss of α-SMA and SM22α at a concentration as low as 0.3 μM, and at 3 μM, it almost completely abolished the effects of Ang II (Extended Data Fig. 10f). Additionally, we confirmed that NS8593 did not have any further effect on M7KO VSMCs (Extended Data Figure 10g), nor did it provide additional protection against VSMC reprogramming in M7KO VSMCs (Extended Data Figure 10h). We also determined the plasma concentration of NS8593 in mice 12 hours after daily intraperitoneal administration at a dose of 5 mg/kg. We found that the plasma concentrations of NS8593 were 349.0±41.38 nM, 714.1±24.34 nM, and 921.3±113.8 nM on days 2, 3 and 5 (Extended Data Figure 10).

Based on the above data, we designed two administration strategies to evaluate the potential protection of NS8593 against AAA (Extended Data Fig. 10j). Firstly, we began NS8593 treatment immediately after Ang II infusion. We found that intraperitoneal injection of NS8593 during the 4-week Ang II infusion led to reduced AAA diameter (Fig. 8d), increased survival probability (Fig. 8e and 8f), lowered AAA incidence (Fig. 8g and 8h), and prevented elastin degradation (Fig. 8i and Extended Data Fig. 10k). Secondly, we sought to determine whether intervention by TRPM7 inhibition in the middle of AAA development could also produce a similar effect (Extended Data Fig. 10j). We found that NS8593 application 2 weeks after Ang II infusion still led to reduced AAA diameter (Fig. 8j), increased survival probability (Fig. 8k and 8l), lowered AAA incidence (Fig. 8m and 8n), and attenuated elastin degradation (Fig. 8o). Our results indicate that, similar to TRPM7 global knockout and TRPM7 channel-dead knock-in, the TRPM7 inhibitor NS8593 produces a potent protective effect against AAA in mice.

In conclusion, we demonstrated that ARL15-TRPM7-Ca²⁺-CaM-calcineurin signaling plays a crucial role in promoting KLF4 activation and inducing VSMC phenotype switching, thereby accelerating the development of AAA. By employing the Trpm7 deletion, TRPM7 channel-dead TRPM7 mutant knock-in, and kinase-inactive knock-in mouse models, we found that it is the channel function, rather than kinase activity, which is essential for VSMC reprogramming and AAA development. Importantly, we found that TRPM7 inhibitor suppresses VSMC reprogramming in vitro and slows AAA progression in vivo. Our findings suggest that TRPM7 is a promising therapeutic target for developing effective prophylactic pharmacological medications for limiting AAA growth.

Discussion

The lack of channel-dead mutant mice made it impossible to conclusively distinguish the role of channel and kinase functions in various physiological and pathological functions of TRPM7 at the animal level. In this study, we paired our channel-dead TRPM7 knock-in mice (Trpm7^CD) with kinase-inactive mice generated by knocking in of point-mutation of TRPM7¹⁰ (TRPM7^KD) for in vivo studies. The key advantage of this kinase-dead mouse strain is that the channel function of TRPM7 is intact¹⁰, where kinase-inactive mice generated by truncation of kinase domain (TRPM7-Δkinase) significantly impaired TRPM7 channel function, causing early embryonic lethality in homozygous mice and abnormalities such as impaired Mg²⁺ absorption even in heterozygous mice³¹. Using our Trpm7^CD and TRPM7^KD mice side by side in the in vivo AAA experiments, we demonstrated that it is the channel function that is necessary and sufficient for TRPM7-mediated pathogenesis in Ang II-induced AAA, whereas TRPM7 kinase function appears not involved in Ang II-induced VSMC reprogramming and AAA growth. Importantly, our Trpm7^CD mice provide an invaluable tool for investigating the role of channel and kinase functions of TRPM7 in other physiological and pathological functions.

Although VSMC reprogramming is known to play a central role in aneurysm formation and progression, contradicting results have been reported about the role of VSMC proliferation in aneurysm³², with some studies showing that VSMC proliferation inhibits diseases progression ³³, whereas others demonstrating limited role of VSMC proliferation in aneurysm formation and progression. The discrepancy results could be partially attributable to the studies using different stages of AAA³⁴. We found that in WT mice, there was a massive increase in the number of α-SMA^low cells in the adventitia from aneurysmal lesions, which was significantly inhibited in TRPM7 knockout mice, suggesting that TRPM7 may play a role in the generation of abnormally reprogrammed, proliferative, and degradative VSMCs³⁵. However, since we did not use lineage-tracking models, the origin of these adventitial α-SMA^low cells remains uncertain and may potentially derive from the expansion of fibroblast populations during AAA pathogenesis. A future comprehensive evaluation using lineage tracking will be necessary to confirm whether these cells are indeed derived from abnormal VSMC proliferation.

An exciting finding in this study is the inhibition of ARL15 in VSMC reprogramming cascade. Our group and others have recently discovered that ARL15 is an endogenous negative regulator of TRPM7 activity^7,22,24. The biological significance of TRPM7 regulation by ARL15, however, is completely unknown^7,24. In the current study, we found that ARL15 expression was reduced in aneurysm lesions and in synthetic/proliferative VSMCs, which was accompanied by an increase of TRPM7 activity. Moreover, ARL15 overexpression inhibited VSMC reprogramming induced by Ang II. Aligned with our discoveries, ARL15 has been shown to promote the activation of transforming growth factor β (TGFβ)-Smad4 signaling³⁶, and activation of TGFβ-Smad4 in VSMCs has been demonstrated to inhibit VSMC reprogramming and AAA development³⁷, all of which supports our hypothesis that inhibiting ARL15 in VSMCs releases this restraint of TRPM7, thereby promoting AAA development.

Methods

Animals

All the experimental mice bred and hosted in the animal facility building of University of Connecticut School of Medicine (UCONN Health) were fed with standard chow diet (minerals supplementation: calcium: 1%, phosphorus 0.7%, non-phytate phosphorus: 0.4%, sodium: 0.2%, potassium: 0.6%, chloride: 0.4%, magnesium: 0.2%, zinc: 70 mg/kg, manganese: 100 mg/kg, copper: 15 mg/kg, iodine: 6 mg/kg, iron: 200 mg/kg, selenium: 0.23 mg/kg) and water ad libitum. Standard housing conditions were maintained at a controlled temperature with a 12-h light/dark cycle. All experimental procedures and protocols were approved by the Institutional Animal Care and Use Committee (IACUC) of University of Connecticut School of Medicine (animal protocol: AP-200135–0723) and were conducted in accordance with the U.S. National Institutes of Health Guidelines for the Care and Use of Laboratory Animals. To avoid variations among different sex of animals, male mice at 8 to 12 weeks old were used in this study.

Trpm7^loxp/loxp mice were kindly provided by Dr. Clapham, and the kinase-dead TRPM7 knock-in mice were generated by Dr. Matsushita. The channel-dead TRPM7 knock-in mice were created by Dr. Runnels in collaboration with Ingenious Targeting Laboratory. The graphic illustration of generating inducible channel-dead TRPM7 knock-in mice is shown in Fig. 5a. Briefly, the exon 22 containing E1047 was floxed, followed by a neomycin cassette and a floxed inverted exon 22 containing the point E1047K mutation with mutant LoxP sites. Neomycin cassette was excised by FLP1 recombinase, generating heterozygous Trpm7^+/F-E1047K mice (or Trpm7^+/E1047K for simplicity). The wild-type exon 22 can be excised by Cre recombinase, accompanied by the inversion of the mutated exon 22 containing the E1047K substitution that inactivates the channel function of TRPM7 (Fig. 5a). Compared with wild-type mice, we did not observe noticeable differences in growth, survival, fertility, or physiological appearance in heterozygous Trpm7^+/F-E1047K mice.

To generate inducible global knockout of TRPM7 or knock-in of TRPM7-E1047K, Trpm7^loxp/loxp mice and Trpm7^+/F-E1047K mice were crossed with Rosa26-Cre^ERT2 mice. Tamoxifen was used to induce TRPM7 knockout or TRPM7-E1047K knockin, and Cre⁻ littermates were used as WT control mice. Both Trpm7^loxp/loxp and Trpm7^+/F-E1047K mice, as well as kinase-dead TRPM7 knock-in mice were crossed with Apoe^−/− mice (JAX laboratory) for induction of AAA. To generate inducible VSMC specific knockout of TRPM7, Trpm7^loxp/loxp mice were crossed with Myh11-cre mice (JAX laboratory), which were then crossed with Apoe^−/− mice for induction of AAA. All the mice used in the experiments were backcrossed with C57BL/6 mice for more than 10 generations to ensure that they all have identical C57BL/6 strain background.

Induction of TRPM7 knockout and TRPM7-E1047K knock-in

As previously reported, TRPM7 is pivotal for embryonic development and organogenesis, but is dispensable in adult mice. Therefore, we only treat the mice with Tamoxifen to induce TRPM7 knockout or TRPM7-E1047K mutant knockin when mice at adult age, 7~ 8 weeks old. After TRPM7 knockout or TRPM7-E1047K knockin at adult age, mice appear healthy and normal. Experiments were conducted after confirming that TRPM7 knockout or TRPM7-E1047K knockin by genotyping, western-blot, and functional current recordings.

Genotyping and confirmation of knockout or knock-in

Trpm7^loxp/loxp and Rosa26-Cre^ERT2 genotyping and confirmation of Trpm7 deletion were performed as previously reported. Homozygous kinase-inactive knock-in mice were conformed following the protocol from Dr. Dr. Matsushita¹⁰. Floxed allele of Trpm7^+/F-E1047K mice were confirmed by genotyping with primers 5’-GGTGTCTTGAGCATTGTTCAACCGTGTGC-3’ and 5’-ACCACCATTACCATCATTCCTTGAAGTGGC-3’. Inversion of exon 22 containing the E1047K mutation was confirmed after tamoxifen treatment by PCR using primers 5’- GCCACTTCAAGGAATGATGGTAATGGTGGT-3’ and 5’- TCACTACAATGTAGAACATATTGGCCACCTG-3’.

Trpm7 deletion and channel-dead knock-in were further confirmed by functional detection using western blot and whole-cell TRPM7 current recordings by patch-clamp.

Tamoxien was used for inducible knockout or knockin. Trpm7^loxp/loxp-cre^ERT2+Apoe^−/−, Trpm7^loxp/loxp-cre^ERT2-Apoe^−/−, Trpm7^+/f-E1047-cre^ERT2+Apoe^−/−, Trpm7^loxp/loxp-Myh11-cre-Apoe^−/− and Trpm7^+/f-E1047-cre^ERT2-Apoe^−/− mice were treated by tamoxifen (1 mg/kg, by gavage every day, for 7 times). Knockout of TRPM7, knock-in of channel-dead TRPM7 mutant kinase-inactive mutant were further confirmed by genotyping, and functional detection using patch-clamp current recording for channel function, WB for expression levels, and in vitro phosphorylation of kinase activity.

Primary vascular smooth muscle cell isolation, culture, and treatments

Mice were euthanized based on our animal protocol, and abdominal aortas were harvested into ice-cold DMEM (with pyruvate) medium (Thermal Fisher Scientific, 11-995-040) supplemented with 1% Pen-Strep (Thermal Fisher Scientific, 15-140-122). One isolation requires at least 6 adult mice. After being rinsed using PBS to thoroughly wash off remanent blood, aortas were transferred into appropriate amount of adventitia digestion solution (DMEM + 1% Pen-Strep + 166 U/mL Collagenase Type 2 (Worthington, S8N10850)). After an incubation at 37 °C for 10 to 20 minutes, adventitia of aortas was completely peeled off as taking off stocks. Then aortas were cut off longitudinally and intima was scratched using forceps to remove endothelium. Prepared media were cut into small pieces around 1 mm² size, which were transferred into appropriate amount of media digestion solution (DMEM + 1% Pen-Strep + 10% FBS (Thermal Fisher Scientific, A4766801) + 366 U/mL Collagenase Type 2 (Worthington, S8N10850)). After incubation at 37 °C for 3 to 4 hours, the tissue chunks were pipetted up and down for 6 to 16 time to release the VSMCs. The suspension was centrifuged at 600 g for 6 minutes, and the cell pellets were resuspended DMEM medium to washing off the remaining collagenase. Then the suspension was centrifuged at 600 g for 6 minutes again, and the VSMCs pellet were resuspended in appropriate amount of VSMCs culture medium (DMEM + 1% Pen-Strep + 10% FBS) and plated into cell culture dish. After 3 days of culture, medium was changed for the 1^st time, then medium was changed every 2 days. The cell confluence usually reaches 70 to 80% at the 4^th to 5^th day after the isolation. A purity of 97–99% can be confirmed by flow cytometry, as well as by immunostaining of α-smooth muscle actin, smooth muscle protein 22-alpha, collagen 1 (to detect fibroblast contamination), CD31 (to detect endothelial cell contamination) and CD45 (to detect immune cell contamination).

VSMCs at P2 to P3 generations were used for all in vitro experiments in this study. For VSMCs reprogramming induced by angiotensin II, angiotensin II (Cayman Chemical, 17150) was diluted in VSMCs culture medium to achieve a working concentration of 10 ng/mL. DMSO (0.1 %) or Chemicals were also added to achieve the corresponding working concentration (FTY720 (10 μM), NS8593 (10 μM) or EGTA (0.5 mM)). Then original culture medium in the dish was replaced by the treatment solutions. For VSMCs reprogramming induced by TRPM7 activation, naltriben mesylate (TOCRIS, 0812) was diluted in VSMCs culture medium to achieve a working concentration of 10 μM, in combination with DMSO (0.1 %), W13 (10 μM) or FK506 (10 μM).

Antibodies, chemicals, and reagents

MMP-2 (D4M2N) Rabbit mAb (Cell Signaling Technology, 40994, 1:100 in 5% bovine calf serum (BCS)/10% goat serum (GS) for IF; 1:1000 in TBST for WB); CD45 (D3F8Q) Rabbit mAb (Alexa Fluor^® 594 Conjugate) (Cell Signaling Technology, 35154, 1:100 in 5% BCS/10% GS for IF); IL-1β (3A6) Mouse mAb (Cell Signaling Technology, 12242, 1:100 in 5% BCS/10% GS for IF, 1:1000 in TBST for WB); MCP-1 (E8Y7P) Rabbit mAb (Cell Signaling Technology, 81559, 1:100 in 5% BCS/10% GS for IF); KLF4 Rabbit mAb (Abclonal, A13673, 1:100 in 5% BCS/10% GS for IF; 1:1000 in TBST for WB); Ki-67 (D3B5) Rabbit mAb (Alexa Fluor^® 488 Conjugate) (Cell Signaling Technology, 11882, 1:200 in 5% BCS/10% GS for IF); α-Smooth Muscle Actin (D4K9N) XP^® Rabbit mAb (Alexa Fluor^® 594 Conjugate) (Cell Signaling Technology, 36110, 1:200 in 5% BCS/10% GS for IF; 1:1000 in TBST for WB); Transgelin/TAGLN Antibody (Cell Signaling Technology, 40471, 1:200 in 5% BCS/10% GS for IF; 1:1000 in TBST for WB); COL1A1 (E8F4L) XP^® Rabbit mAb (Cell Signaling Technology, 72026, 1:200 in 5% BCS/10% GS for IF); NFAT1 Antibody (Cell Signaling Technology, 4389S; 1:1000 in TBST for WB); Lamin B1 (D9V6H) Rabbit mAb (Cell Signaling Technology, 13435S; 1:1000 in TBST for WB); TRPM7 antibody (Novus, NB500–243, 1:1000 in TBST for WB); ARL15 Polyclonal Antibody (Thermal Fisher Scientific, 11934–1-AP, 1:1000 in TBST for WB); Anti-mouse IgG (H+L), F(ab’)2 Fragment (Alexa Fluor^® 488 Conjugate) (Cell Signaling Technology, 4408, 1:600 in 5% BCS/10% GS for IF); Anti-rabbit IgG (H+L), F(ab’)2 Fragment (Alexa Fluor^® 488 Conjugate) (Cell Signaling Technology, 4412, 1:600 in 5% BCS/10% GS for IF); DAPI Fluoromount-G^® (Southern Biotech, 0100–20); GAPDH Rabbit pAb (Abclonal, AC027, 1:5000 in TBST for WB); β-Tubulin Rabbit pAb (Abclonal, AC008, 1:5000 in TBST for WB); Calcein Blue AM (BD Biosciences, 564060, 1 μM in 2% FBS/PBS for FC); BV605 Hamster Anti-Mouse CD29 (BD Biosciences, 740365, 1:200 in 2% FBS/PBS for FC); BV711 Rat Anti-Mouse CD31(BD Biosciences, 740680, 1:200 in 2% FBS/PBS for FC); Alexa Fluor^® 700 anti-mouse Ly-6C Antibody (Biolegend, 128024, 1:200 in 2% FBS/PBS for FC); APC anti-mouse CD45.2 Antibody (Biolegend, 109814, 1:200 in 2% FBS/PBS for FC); FITC anti-mouse Ly-6G Antibody (Biolegend, 127606, 1:200 in 2% FBS/PBS for FC); PE anti-mouse CD3ε Antibody (Biolegend, 100308, 1:200 in 2% FBS/PBS for FC); CD11b Monoclonal Antibody (M1/70), APC-eFluor^™ 780, eBioscience^™ (Thermal Fisher Scientific, 47-0112-82, 1:200 in 2% FBS/PBS for FC); CD19 Rat anti-Mouse, PerCP-Cy5.5, Clone: 1D3, BD (Thermal Fisher Scientific, BDB551001, 1:200 in 2% FBS/PBS for FC); Naltriben mesylate (TOCRIS, 0812); W-13 (hydrochloride) (Cayman chemical, 14277); FK-506 (Cayman chemical, 10007965); NS8593 (hydrochloride) (Cayman chemical, 29774); FTY720 Phosphate (Cayman chemical, 10008639); EGTA (Cayman chemical, 11706); TPEN (Cayman chemical, 13340); NFAT Inhibitor (Cayman chemical, 13855); NP40 (Thermal Fisher Scientific, 28324); Triton^™ X-100 (T-9284), Bovine Serum Albumin (Sigma-Aldrich, 9048-46-8); Fura-2-AM (Thermal Fisher Scientific, F1201); Pluronic^™ F-127 (Thermal Fisher Scientific, P6866); Invitrogen^™ MitoSOX^™ (Thermal Fisher Scientific, M36008); Collagenase from Clostridium histolyticum Type XI (Sigma, C7657); Hyaluronidase from bovine testes Type Is (Sigma, H3506); Deoxyribonuclease I from bovine pancreas (Sigma, DN25). All chemicals for making solutions for live cell imaging (see below) and recording solution (see below) were purchased from Sigma-Aldrich.

Abdominal Aortic Aneurysm (AAA) Mouse Model

In brief, eight-to nine-weeks old male mice (~25 g) were used in this study. The reason we only used male animals is that the Ang II model is typically used in male mice to induce AAA, and AAA is much more commonly seen in male human individuals. ALZET mini-osmotic pumps filled with either PBS or angiotensin II (Cayman Chemical, 17150, 1 μg/kg/min) were implanted subcutaneously in the nuchal area³⁸. AAA formation was evaluated after 4 weeks of AngII infusion. In the experiment for evaluating the effects of inhibiting TRPM7 on AAA, Trpm7^+/+Apoe^{−/ −}and Trpm7^−/−Apoe^−/− mice were used for evaluating the effects of TRPM7 inhibitor NS8593 (5mg/kg, i.p., daily after Ang II infusion) on AAA. Death of mice was recorded for preparing survival curve and quantifying mortality. Death rate was calculated as the % of mice within each group that died from a verified rupture (formation of blood clots in their abdominal aortas that indicates the abdominal aortic rupture event). AAA incidence was defined as a at least 50% increase in maximal diameter of suprarenal abdominal aorta at day 28 of Ang II infusion and included mice that died from aneurysm rupture. After 4 weeks, AAA diameter was measured by abdominal ultrasound using VisualSonics vevo3100, and incidence of AAA formation was confirmed by dissection after euthanasia.

CaCl₂ model: In brief, mice were anesthetized, and their abdominal cavity was accessed with a 2 cm incision. The infrarenal abdominal aorta was carefully exposed, and a saline-soaked rubber strip was placed beneath the aorta. A piece of cotton soaked with 0.5 M CaCl₂ was carefully applied onto the abdominal aorta for 15 minutes, followed by a PBS rinse and suture of the incision. After 14 days, the AAA diameter was measured by abdominal ultrasound using VisualSonics Vevo3100, and the incidence of AAA formation was confirmed by dissection after euthanasia.

Aneurysm aortas were frozen in Scigen Tissue-Plus^™ O.C.T. Compound (Thermal Fisher Scientific, 23-730-571) for tissue slicing, or fast frozen in liquid nitrogen and kept in −80 °C refrigerator for protein extraction. Cryo-section was performed to generate aneurysm tissue sections at the thickness of 6 μm.

Quantification of elastin degradation

Elastin breaks were quantified as number of elastin breaks per representative 20X magnification image. Elastin break and degradation scoring was done based on the following criteria: Score of 0 indicates normal tissue which shows no elastin degradation region. Score of 1 indicates that elastin degradation is < 25% of the entire area. Score of 2 indicates between 25 and 50% elastin degradation. Score 3 indicates a 50 to 75% elastin degradation and grade 4 indicates > 75% elastin degradation³⁹.

Immunofluorescence staining

Full-length aortas collected from mice were mounted in OCT. Prior to staining, slides were left at room temperature for 30 minutes for the dehydration of OCT residue. Samples were fixed in 4% paraformaldehyde for 20 minutes and washed in PBS for 3 times. Samples were then permeabilized using 0.3% Triton X for 15 minutes and incubated in blocking buffer (PBS with 10% BCS, 5% goat serum, and 0.1% Triton X) for 2 hours at room temperature. Primary antibodies were diluted in antibody dilution buffer (PBS with 5% BCS, 10% goat serum, and 0.1% Triton X), and samples were incubated for 3 hours at room temperature and then washed in PBS for 3 times. Secondary antibodies were diluted at appropriate concentrations in antibody dilution buffer (1:600 for the goat anti-mouse antibody, and 1:400 for the goat anti-rabbit antibody), and samples were incubated for 1 hour at room temperature. Samples were then washed in PBS for 3 times. Samples were then blocked for 1 hour at room temperature using above-mentioned blocking buffer. 594-conjugated α-smooth muscle actin antibody was diluted (1:200) in antibody dilution buffer and incubated on samples overnight at 4 °C. The next day, samples were washed in PBS for 3 times and mounted using mounting solution supplemented with DAPI.

For in vitro immunofluorescence staining, primary VSMCs were trypsinized and plated onto 25 mm² square coverslips precoated with gelatin at an appropriate density. After chemical treatments for 24 hours (angiotensin II) or 12 hours (naltriben), coverslips were rinsed using ice-cold PBS for 3 times, which is followed by a fixation in 4% paraformaldehyde. After 15 minutes, coverslips were rinsed using PBS for 3 times, following a permeabilization using 0.3% Triton X in PBS. After 10 minutes, coverslips were incubated in blocking solution (PBS with 10% BCS, 5% goat serum, and 0.1% Triton X). After 1 hour, coverslips were incubated with primary antibodies at appropriate dilutions at room temperature for 2 hours. Then, coverslips were rinsed using PBS for 3 times, and incubated with secondary antibodies at appropriate dilutions at room temperature for 1 hour, following by a wash using PBS for 3 times. Next, coverslips were incubated with 594-conjugated α-smooth muscle actin at room temperature for 2 hours. After washing using PBS for 3 times, coverslips were mounted onto cover slides using mounting solution supplemented with DAPI.

Images were taken using Olympus APX 100 imaging system. For tissue sections, CD45 quantifications were analyzed as a density of number of CD45⁺ cells per image area. IL-1β and MCP1 quantifications were analyzed as a ratio of area of IL-1β or MCP1 per total aneurysm area. DQ Gelatin zymography analyses and MMP2 quantifications were done based on a ratio of area of DQ gelatin or MMP2 fluorescence to elastin fiber-enclosed area. Ki67 and KLF4 quantifications were calculated as a ratio of Ki67⁺⁺ or KLF4⁺⁺ cells per total aneurysm area. All quantifications were performed using ImageJ.

In situ zymography

Aneurysm tissue slices were obtained as described in immunofluorescence staining protocol. Slides were left at room temperature for 30 minutes for the dehydration of OCT residue, then were washed in PBS for 3 times. Next, slides were incubated in DQ^™ Gelatin From Pig Skin, Fluorescein Conjugate (Thermal Fisher Scientific, D12054) diluted (1:50) in zymography buffer (50 mM Tris-HCl, 150 mM NaCl, 5 mM CaCl₂, 0.2 mM sodium azid, adjusted to pH 7.4). Samples were covered with plastic film to prevent evaporation and placed in 37 °C incubator for 2 hours. Slides were then washed in PBS for 3 times. Samples were then fixed in 4% paraformaldehyde for 20 minutes. Slides were then washed in PBS for 3 times. To co-stain with α-smooth muscle actin, samples were permeabilized with 0.1% Triton X for 15 minutes, blocked in blocking buffer for 2 hours, and incubated in 594-conjugated α-smooth muscle actin antibody for 2 hours at room temperature. Slides were then washed in PBS for 3 times and mounted using mounting solution supplemented with DAPI. Images were taken using Olympus APX 100 imaging system. Quantifications were performed using ImageJ.

Flow cytometry analysis following aorta digestion

Mice were euthanized based on animal protocol. Abdominal aortas were obtained, rinsed thoroughly using ice-cold PBS for 3 times, cut into around 1mm² small pieces, and incubated in aorta digestion solution (HBSS, 2 mM Ca²⁺, 3 mg/mL Collagenase Type I, 120 U/mL Collagenase Type XI, 60 U/mL DNAse and 60 U/mL hyaluronidase 1a). After incubation at 37 °C for 45 minutes, tissue pellets were resuspended to release cells inside, and the suspension was filtered using a 70-μm cell strainer, following a centrifuge at 600 g for 6 minutes. The cell pellets collected were washed using digestion washing buffer (10 % FBS in PBS) for 2 times, and resuspended in flow cytometry washing buffer (2 % FBS in PBS). Antibodies were diluted in flow cytometry washing buffer appropriately, and incubated with isolated aortic cells and peripheral leukocytes at 37 °C for 36 minutes. Stained cells were washed using flow cytometry washing buffer for 2 times, and subjected to flow cytometry analysis using BIO-RAD ZE5 Cell Analyzer. Collected data was analyzed using Flowjo 10.1.2.

Adenovirus preparation and infection

Recombinant adenoviruses expressing human wild type (WT) ARL15 and the R95A mutant were made by first subcloning the expression cassettes from the human ARL15 (NM_019087.3) and R95A mutant in the pcDNA3.1+/C-(K)-DYK vector (GenScript) vector²² by directional cloning into the pENTR^™/D-TOPO^™ Gateway vector. The following primers were employed to introduce the V5-tag after the COOH-terminus of ARL15 protein sequence: 5’- CCACCATGTCTGATCTCCGAATAACTGAGG-3’ and 5’- CTAATCCAAACCCAACAGCGGATTCGGGATCATTCTTACAGCTTCATGGTCTTTTTCTTC-3’. Next, pENTR-hARL15-WT and pENTR-hARL15-R95A were used to introduce the open reading frames with the V5 tag into pAd/CMV/V5-DEST using Gateway LR clonase, following manufacturer’s instruction (Thermofisher). pAd/CMV/V5-hARL15 and pAd/CMV/V5-hARL15-R95A were then transfected into the 293A cell line (Thermofisher) to produce recombinant AD-hARL15 and AD-hARL15-R95A, following the manufacturer’s instructions for the ViraPower Adenoviral Expression System (Thermofisher).

Collagen gel contraction assay

VSMCs after angiotensin II and / or chemical treatments were used for this assay. 100,000 VSMCs were resuspended in 450 μL of culture medium, which was mixed with 150 μL of type 1 collagen (4 mg/mL) following 3 μL of 1M NaOH. 500 μL of mixture was transferred into a well of 24-wells plate. After collagen gel was solidified (30 minutes), 500 μL of culture medium was added into the well and a 200 μL pipette tip was used to make the gel floating. Images were taken 48 hours after the plating. Gel sizes were measured using ImageJ.

Ca²⁺ imaging

Fura2-AM imaging was performed as described previously. In brief, pre-warmed VSMCs culture medium was used to dilute Fura-2 AM to a working concentration at 2.5 μM, and 0.02% Pluronic^™ F-127 (Thermal Fisher Scientific, P3000MP) was added to facilitate loading of Fura-2 AM. VSMCs on 25 mm² glass coverslips were washed using HBSS for 3 times, and then incubated with 0.1 ml of Fura-2 AM working solution. After incubation at 37 °C for 20 min, coverslips were rinsed using HBSS for 3 times. Fluorescence intensities at 510 nm with 340 nm and 380 nm excitation were collected at a rate of 1 Hz using CoolSNAP HQ2 (Photometrics) and data were analyzed using NIS-Elements (Nikon).

Western blotting

NP-40/Triton lysis buffer (10% NP40, 1% Triton^™ X-100, 150 mM NaCl, 1 mM EDTA, 50 mM Tris, pH=8.0) containing proteinase inhibitors and phosphatase inhibitors was used to lyse both cultured VSMCs and frozen aneurysm tissue. Cell and tissue lysate were further lysed by ultrasound using an ultrasonic cleaner filled with ice-cold water for 30 min. After incubated on ice for 1 h, lysate was centrifuged at 13000 g for 30 min and supernatant was collected. Protein concentration was measured using Pierce^™ Rapid Gold BCA Protein Assay Kit. 30–50 μg of total protein was loaded and separated proteins were transferred to Nitrocellulose membranes. Membranes were blocked with 5% BSA and 2.5% goat serum in Tris buffered saline (TBS, pH=7.4) at room temperature for 2 h, and incubated with primary antibodies in TBS with 0.05% Tween (TBS-T) at room temperature for 2 h. Then membranes were incubated with secondary antibodies in TBS-T for 1 h at room temperature for 1 h for detection. Blots were developed with ImageQuant LAS 4000 imaging system. Band intensity was quantified using ImageJ software and normalized with appropriate loading controls.

Immunoprecipitation

300 μg of protein was taken and diluted using NP-40 lysis buffer to a total volume of 500 μl. 0.1 μg of ARL15 antibody was added to the diluted protein sample. After the protein-antibody mixture was incubated on ice for 2 hours, 25 μl of pre-washed Protein A/G PLUS-Agarose (Santa Cruz Biotechnology, sc-2003) was added, and the whole mixture was incubated at 4 °C overnight. The mixture was then centrifuged at 2500g for 1 minute to collect the agarose beads. The beads were washed with NP-40 lysis buffer seven times, mixed with an equal amount of 2x Laemmli Sample Buffer (BIO-RAD, 1610737), and boiled at 95 °C for 5 minutes. The samples were then ready for Western blotting analysis.

Electrophysiology

Whole cell currents were recorded using an Axopatch 200B amplifier as we previously reported^40–42. Data were digitized at 10 or 20 kHz and digitally filtered offline at 1 kHz. Patch electrodes were pulled from borosilicate glass and fire-polished to a resistance of ~3 MΩ when filled with internal solutions. Series resistance (R_s) was compensated up to 90% to reduce series resistance errors to <5 mV. Cells in which R_s was >10 MΩ were discarded ⁴³. For TRPM7 current recordings, voltage stimuli lasting 250 ms were delivered at 1-second intervals, with voltage ramps ranging from −100 to +100 mV at a holding potential of 0 mV to elicit currents. A fast perfusion system was used to exchange extracellular solutions and to deliver agonists and antagonists to the cells, achieving a complete solution exchange in about 1–3 seconds. Normal Tyrode solution was used as the extracellular solution, containing (in mM): 145 NaCl, 5 KCl, 2 CaCl₂, 10 HEPES, and 10 glucose, with an osmolarity of 290–320 mOsm/kg and pH adjusted to 7.4 with NaOH. The internal pipette solution for whole-cell current recordings of TRPM7 in VSMCs contained (in mM): 145 Cs-methanesulfonate, 8 NaCl, 10 EGTA, and 10 HEPES, with pH adjusted to 7.2 with CsOH. For current recording in VSMCs, 1 μM tetrodotoxin, 10 mM tetraethylammonium (TEA), and 10 μM nifedipine were included in the external perfusion solution to block voltage-gated Na⁺ currents, voltage-gated K⁺ currents, and voltage-gated Ca²⁺ currents, respectively. Patch-clamp data were analyzed using Clampfit 10.0.

Measurement of NS8593 plasma concentration using LC-MS/MS system

Plasma concentration of NS8593 was measured using LC-MS/MS system, similar to the method as we previously reported. In brief, NS8593 was injected intraperitoneally at 5 mg/kg daily. Blood samples were collected 12 hours after injection from the heart of mice at 2, 3, and 5 days after daily intraperitoneal injections of NS8593 (corresponding to two, three, and five injections, respectively). Heparin was added to the collected blood, and plasma was isolated by centrifugation at 600 g for 10 minutes. Samples with significant hemolysis were excluded from the study. The collected plasma was stored at −80°C prior to extraction. After all samples from all groups were collected, 800 μL of methanol was added to the diluted plasma (100 μL plasma + 100 μL PBS). The mixture was vortexed for 30 seconds, incubated for 1 hour at −20°C, and centrifuged at 16,000 g at 4°C for 15 minutes. The supernatant was transferred to a new tube, and the solvent was evaporated using a vacuum concentrator. The extracts were reconstituted in methanol/water (1:1) solution, sonicated for 1 minute in an ice-cold bath, and centrifuged at 16,000 g at 4°C for 15 minutes to remove debris. The methanol extracts were then diluted with Optima grade pure water and frozen for lyophilization in pre-weighed tubes. Once completely dried, the dry mass was calculated, and the extracts were reconstituted at a concentration of 10 mg/mL. 10 μL of these reconstituted samples were used for analysis using the LC-MS/MS system at the UConn Proteomics & Metabolomics Facility and the Center for Open Research Resources & Equipment (COR²E). The concentration of the plasma NS8593 was calculated based on the standard curve constructed using six different concentrations of NS8593.

Reanalysis of single cell sequencing data

The single-cell sequencing dataset (SRP274475) was obtained from the NCBI database and filtered using a Python Jupyter Notebook. The filtering criteria included a gene count per cell ranging from 200 to 4000 (or 5000), with mitochondrial gene content less than 10%. After filtering, the data underwent normalization, followed by scaling, dimensionality reduction, cluster identification, and visualization using the Leiden method at a resolution of 0.5. To facilitate cluster identification across multiple merged datasets, additional integration was conducted after the normalization step but before scaling. Conserved marker genes for each cluster were determined using the Leiden algorithm. For reclustering, UMAP coordinates of the relevant cells were extracted and analyzed in the same manner as clusters from the combined datasets.

Quantification and statistical analysis

All data are expressed as mean ± standard error of the mean (SEM), and “n” indicates the number of biological replicates. Experiments were technically repeated at least three times for the in vitro experiments, and the results of technical replicates were averaged, and considered as 1 biological sample. Sample size was estimated by Power analysis using the web tool (https://clincalc.com/stats/samplesize.aspx) or G*Power software (http://www.gpower.hhu.de/), utilizing independent study groups, a type I error α=0.05, and power =80%. Randomization was used to divide animals or experimental samples into separate groups for different treatments to avoid bias in all our experiments. Subjects were randomly assigned to different treatment groups using GraphPad’s “Random Number Calculators.” Statistical analyses were performed using GraphPad Prism v9 (GraphPad Software, Inc.). Data quantification and analysis were performed blindly.

Prior to statistical analysis, data were assessed for normality using the Kolmogorov-Smirnov test or Shapiro-Wilk test, justifying the use of Student’s t-test or analysis of variance (ANOVA). Equal variance was analyzed using the Bartlett test and F-test for multiple comparison groups and two-sample t-test, respectively. If normality and equal variance assumptions were not met, the Mann-Whitney test was used for statistical analysis for two-group comparisons, and the Kruskal-Wallis non-parametric test with Dunn’s multiple comparison was used for comparison between multiple groups.

In GraphPad Prism 9.0, Welch’s ANOVA and Brown-Forsythe ANOVA automatically adjust the calculations of the F ratio and degrees of freedom to account for heterogeneity of within-group variances. The P value is interpreted similarly to that in the analysis of variance. For two-group comparisons, statistical significance was determined using a two-tailed Student’s t-test. For multiple groups, statistical significance was determined using one-way or two-way ANOVA, followed by Bonferroni post-test with correction. One-way ANOVA was used for comparing multiple groups with one variable, and two-way ANOVA was used when there were two independent variables. A two-sided P < 0.05 was considered significant.

The sample size for Western blot experiments consisted of six dishes of cells from at least five mice. The sample size for live cell imaging experiments consisted of 20–40 cells from at least five dishes of cells, collected from more than five mice (see Figure legend). For animal experiments, flow cytometry analysis, and aortic section staining, “n” indicates the number of mice in each group. For in vitro experiments, “n” indicates the number of dishes of cells isolated from the same number of mice.

The IC₅₀ of NS8593 for inhibiting TRPM7 activity in VSMCs was calculated by fitting the concentration–response curves using Boltzmann equation of the form: E = E_max{1/[1 + (EC₅₀/C)ⁿ]}, where E is the effect at concentration C, E_max is maximal effect, EC₅₀ is the concentration for half-maximal effects and n is the Hill coefficient. EC₅₀ is replaced by IC₅₀ when the effect is an inhibitory effect.

Extended Data

Extended Data Fig. 1: TRPM7 Knockout in mice.

(a), Graphic illustration timeline for inducing TRPM7 knockout and subsequent AAA modeling. 8-week-old male mice were used in this study. After 7 days of daily tamoxifen intraperitoneal injections, either Ang II or CaCl2 AAA models were performed. AAA evaluation was conducted 4 weeks after Ang II perfusion or 2 weeks after CaCl2 application. (b), Representative genotyping PCR after tamoxifen treatment. The presence of the 200 bp band, instead of the 1000 bp band, indicates successful TRPM2 knockout. (c), Quantification and representative traces of TRPM7 whole-cell current recording from aortic VSMCs isolated from Trpm7+/+Apoe−/− and Trpm7−/−Apoe−/− mice (animals n = 13, 16). (d), Western blotting confirmation of TRPM7 knockout in the aorta, as well as validation of antibody. (e), Quantification of elastin score in the PBS control group of Fig. 1f (animals n = 5/group). (f), Quantification of α-SMA fluorescence intensity in Fig. 3a. (g), UMAP plot of 22 major cell populations (Leiden clustering algorithm, resolution = 0.4) in control and aneurysmal human aortic tissues after re-analysis of the GSE155468 raw sequencing data downloaded from the GEO repository. (h), Violin plot of TRPM7 expression pattern in different immune cell populations (from a pool of 48,128 cells after filtration from three control samples and eight aneurysmal samples. Box plots indicate the median (center line), the first and third quartiles (bounds of the box), and the minima and maxima within 1.5 times the interquartile range (whiskers). Outliers are represented as individual points beyond the whiskers.). (Statistical significance was determined using a two-tailed Student’s t-test).

Extended Data Fig. 2: The substantial activation of KLF4 signaling during VSMC reprogramming in human aneurysmal lesions.

(a), UMAP plot of 5 major VSMCs and fibroblast populations (Leiden clustering algorithm, resolution = 0.4) in control and aneurysmal human aortic tissues after re-analysis of the GSE155468 raw sequencing data downloaded from the GEO repository. (b), Bubble plot characterization of the molecular features of the 5 populations. (c), Violin plot of KLF2, KLF4, KLF6, and KLF9 expression pattern in VSMCs and fibroblast populations (from a pool of 48,128 cells after filtration from three control samples and eight aneurysmal samples. Box plots indicate the median (center line), the first and third quartiles (bounds of the box), and the minima and maxima within 1.5 times the interquartile range (whiskers). Outliers are represented as individual points beyond the whiskers.). (d), Representative images and quantification of IF staining showing Ki67 (green, colocalized with DAPI) expression in the aortas from Trpm7+/+Apoe−/− and Trpm7−/−Apoe−/− mice 4 weeks after Ang II infusion (animals n = 5, 5) (Red: α-SMA). (Statistical significance was determined using a two-tailed Student’s t-test).

Extended Data Fig. 3: TRPM7 knockout inhibits VSMC reprogramming.

(a-d), Representative images of IF staining showing α-SMA / Collagen I (a), SM22α / Collagen I (b), α-SMA / CD31 (c) and α-SMA / CD45 (d) in isolated primary VSMCs to confirm purity. (e, f), Confirmation of VSMC purity using flow cytometry analysis of SM22α expression in cells from the entire culture dish (4 dishes from 4 mice) after a 1-week culture. (g-h), Quantification and representative images of IF staining showing KLF4 and Ki67 (colocalized with DAPI), MMP2 and IL-1β, as well as α-SMA expression in WT and M7KO VSMCs 24 hours after Ang II treatment (10 ng/mL) (dish numbers = 6/group, isolated from at least 5 mice). (i), Quantification and representative images of collagen gel contraction assay using VSMCs from WT and M7KO mice pretreated with Ang II for 24 hours (dish numbers = 8/group, isolated from at least 5 mice). (j), Western blotting analysis of the expression levels of MMP2, IL-1β, α-SMA, SM22α and KLF4 in WT and M7KO VSMCs with or without Ang II treatment (dish numbers = 6/group, isolated from at least 5 mice). (k), Representative traces of whole-cell TRPM7 current recordings in VSMCs from M7KO (paired with Fig. 3d) mice pretreated with PBS or Ang II (10 ng/mL) (5 mM 2-APB was used to activate and 200 μM 2-APB was used to inhibit TRPM7) (cell numbers = 6–10/group, isolated from at least 5 mice). (l), Additional statistical analysis for Fig. 3d. (m), Quantification of and representative whole-cell TRPM7 current recordings from the very initial stable traces immediately after membrane breaking in VSMCs from WT and M7KO mice pretreated with PBS or Ang II (10 ng/mL) (cell numbers = 6–10/group, isolated from at least 5 mice). (Statistical significance was determined using a two-tailed Student’s t-test, or one-way or two-way ANOVA was used to compare multiple groups with one or more variables).

Extended Data Fig. 4: Conditional TRPM7 knockout in VSMCs.

(a), Graphic illustration showing how VSMC specific TRPM7 knockout was achieved. (b), Confirmation of VSMC specific TRPM7 knockout by TRPM7 current recording in isolated primary VSMCs. (c), Representative images of elastin morphology, and quantification of elastin degradation score and number of elastin breaks for Fig. 3m (animals n = 5, 5). (d-g), Full size blotting of TRPM7 and ARL15 for Fig. 4a & 4b showing the cleavage products of TRPM7 primarily at about 100 and 40 kDa. (h), Immunoprecipitation of TRPM7 by anti-ARL15 antibody and the and the quantification of the pulling ability of ARL15 for TRPM7. (Statistical significance was determined using a two-tailed Student’s t-test, or one-way or two-way ANOVA was used to compare multiple groups with one or more variables).

Extended Data Fig. 5: ARL15 overexpression inhibits VSMC reprogramming.

(a), Violin plot of TRPM7 and ARL15 expression pattern in VSMCs and fibroblast populations from control and aneurysmal human aortic tissues as in Extended Data Fig. 3b (from a pool of 48,128 cells after filtration from three control samples and eight aneurysmal samples. Box plots indicate the median (center line), the first and third quartiles (bounds of the box), and the minima and maxima within 1.5 times the interquartile range (whiskers). Outliers are represented as individual points beyond the whiskers.). (b-c), UMAP plot of the expression pattern of TRPM7, ACTA2 (α-SMA), TAGLN (SM22α), MMP2, IL1B, KLF2 and KLF4 in VSMCs and fibroblast populations. (d), Additional statistical analysis with WT VSMCs without virus infection for Fig. 4d. (e), Quantification of and representative whole-cell TRPM7 current recordings from the very initial stable traces immediately after membrane breaking (cell numbers = 6–10/group, isolated from at least 5 mice). (f-i), Representative images and quantification of IF staining showing KLF4(i), MMP2(j), and IL-1β(k) (green), as well as α-SMA (red) expression in primary VSMCs isolated from WT mice infected with ad-ARL15 or ad-ARL15-Mut for 24 hours before Ang II treatment (dish numbers = 6/group, isolated from at least 5 mice). (Statistical significance was determined using a two-tailed Student’s t-test, or one-way or two-way ANOVA was used to compare multiple groups with one or more variables).

Extended Data Fig. 6: Activation of TRPM7 by naltriben induces VSMC reprogramming.

(a-c), Quantification and representative images of IF staining showing α-SMA, KLF4 (a), MMP2 (b) and IL-1β (c) expression in Trpm7+/+Apoe−/− VSMCs 24 hours after naltriben treatment (10 μM) (dish numbers = 6/group, isolated from at least 5 mice). (d), Western blotting analysis of the expression levels of MMP2, IL-1β, α-SMA, SM22α and KLF4 in WT and M7KO VSMCs with or without naltriben treatment for 24 hours (dish numbers = 6/group, isolated from at least 5 mice). (e-h), Quantification and representative images of IF staining showing α-SMA, KLF4 (f), MMP2 (g) and IL-1β (h) expression in Trpm7+/+Apoe−/− VSMCs 24 hours after naltriben treatment (10 μM) with the treatments of DMSO, FK506 and W13 (dish numbers = 6/group, isolated from at least 5 mice). (i), Western blotting analysis of the expression levels of MMP2, IL-1β, α-SMA, SM22α and KLF4 in WT and M7KO VSMCs with or without naltriben treatment for 24 hours, with the treatments of DMSO, FK506 (10 μM) and W13 (10 μM) (dish numbers = 6/group, isolated from at least 5 mice). (j), Western blotting analysis of the nuclear NFAT1 levels in WT and M7KO VSMCs after a 24-hour Ang II treatment (dish numbers = 6/group, isolated from at least 5 mice). (k), Western blotting analysis of the expression levels of α-SMA and SM22α in WT VSMCs after a 24-hour naltriben treatment in the presence of NFAT inhibitory peptide VIVIT (dish numbers = 6/group, isolated from at least 5 mice). (Statistical significance was determined using a two-tailed Student’s t-test, or one-way or two-way ANOVA was used to compare multiple groups with one or more variables).

Extended Data Fig. 7: Limited changes of Zn2+ signaling in VSMCs from human aneurysmal tissues.

(a, b), Western blotting analysis of the expression levels of MMP2, IL-1β, α-SMA, SM22α and KLF4 in WT VSMCs with or without naltriben treatment for 24 hours, with or without the treatment of 10 μM TPEN (dish numbers = 6/group, isolated from at least 5 mice). (c), UMAP plot of 5 major VSMCs and fibroblast populations (Leiden clustering algorithm, resolution = 0.4) in control and aneurysmal human aortic tissues as in Extended Data Fig. 3b. (d), Heatmap characterization of the changes of Zn2+ signaling pathway molecules. (Statistical significance was determined using a two-tailed Student’s t-test, or one-way or two-way ANOVA was used to compare multiple groups with one or more variables).

Extended Data Fig. 8: Side-by-side comparison of kinase-dead and channel-dead TRPM7 knock-in.

(a), Representative genotyping PCR of kinase-dead loxp sites. The presence of the 211 bp band and disappearance of the 129 bp band indicate the presence of kinase-dead loxp sites. (b), Quantification and representative traces of TRPM7 whole-cell current recording from VSMCs isolated from Trpm7+/+Apoe−/− and Trpm7KDApoe−/− mice (cell number = 15, 15, isolated from at least 5 mice). (c), Quantification and representative traces of TRPM7 whole-cell current recording from TRPM7E1047K overexpression HEK293T cell line transfected with GFP control plasmid or TRPM7, with or without the treatment with tetracycline for inducing TRPM7E1047K expression. (d), Representative genotyping PCR of channel-dead loxp site. The presence of the 234 bp band indicates the presence of E1047K loxp site. (e), Quantification and representative traces of TRPM7 whole-cell current recording from VSMCs isolated from Trpm7+/+Apoe−/− and Trpm7CDApoe−/− mice (cell number = 15, 16, isolated from at least 5 mice). (f, g), Representative Western blot analysis of myelin basic protein (MBP) in vitro phosphorylation by TRPM7 kinase following co-immunoprecipitation enrichment of TRPM7 using lysates from transfected HEK293T cells (f) or hearts (g). (h-k), Representative images and quantification of IF staining showing KLF4(i), MMP2(j), and IL-1β(k) expression (green), as well as α-SMA (red) expression in primary VSMCs isolated from Trpm7+/+Apoe−/− (WT), Trpm7KDApoe−/− (Kinase-dead) and Trpm7CDApoe−/− (Channel-dead) mice 24 hours after Ang II treatment (10 ng/mL) (dish numbers = 6/group, isolated from at least 5 mice). (l), Additional statistical analysis for Fig. 5f. (m), Quantification of and representative whole-cell TRPM7 current recordings from the very initial stable traces immediately after membrane breaking in VSMCs from WT, M7KD, and M7CD mice pretreated with PBS or Ang II (10 ng/mL) (cell number = 6–10/group, isolated from at least 5 mice). (Statistical significance was determined using a two-tailed Student’s t-test, or one-way or two-way ANOVA was used to compare multiple groups with one or more variables).

Extended Data Fig. 9: Kinase-dead or channel-dead TRPM7 knock-in do does not affect peripheral leukocyte populations.

(a, b), Representative images of flow cytometry analysis of immune cell population in the peripheral blood from Trpm7+/+Apoe−/− and Trpm7KDApoe−/− mice 4 weeks after angiotensin II infusion. (c), Quantification of immune cell counts (animals n = 6, 6). (d-h), Representative images (d) and quantification of IF staining showing MMP activity (in situ zymography using DQ-gelatin) (e), and expression of MMP2(f), CD45(g), IL-1β(g), and MCP1(h) (see color labels in the images) in aorta sections from Trpm7+/+Apoe−/− and Trpm7CDApoe−/− mice 4 weeks after Ang II infusion (α-SMC: red) (animals n = 5, 5). (i, j), Representative images of flow cytometry analysis of immune cell population in the peripheral blood from Trpm7+/+Apoe−/− and Trpm7CDApoe−/− mice 4 weeks after angiotensin II infusion. (k), Quantification of immune cell counts (animals n = 6, 6). (Statistical significance was determined using a two-tailed Student’s t-test, or one-way or two-way ANOVA was used to compare multiple groups with one or more variables).

Extended Data Fig. 10: TRPM7 inhibition suppresses VSMC reprogramming.

(a-d), Representative images (a) and quantification of IF staining showing KLF4(b), MMP2(c), IL-1β(d) expression, as well as α-SMA expression in primary VSMCs from Trpm7+/+Apoe−/− (WT) mice at 24 hours after Ang II (10 ng/mL) treatment in the presence of DMSO, FTY720, NS8593 or EGTA (dish number = 6/group, isolated from at least 5 mice). (e), Representative TRPM7 current recordings in VSMCs perfused with NS8393 at various concentrations, as well as a concentration-response curve with an estimated IC50 obtained by Boltzmann equation fitting (cell number = 5, isolated from at least 5 mice). (f), Western blotting analysis of the expression levels of α-SMA and SM22α in WT VSMCs with or without Ang II treatment for 24 hours, with the treatment of NS8593 at different concentrations (dish numbers = 6/group, isolated from at least 5 mice). (g), Whole-cell current recording in M7KO VSMCs to confirm if NS8593 produces an additional effect without the presence of TRPM7. (h), Western blotting analysis of the expression levels of α-SMA and SM22α in WT and M7KO VSMCs with or without Ang II treatment for 24 hours, with or without the treatment of NS8593. (i), LC-MS/MS Measurement of plasma concentration of NS8593 in mice 12 hours after injection (5 mg/kg), following a series of daily injections (2 days, 3 days, and 5 days). Data points were connected using smooth fitting (animals n = 2/group). (j), Graphic illustration showing the administration strategy of NS8593. (k), Quantification of elastin score in the PBS control group of Fig. 8i (animals n = 5/group). (Statistical significance was determined using a two-tailed Student’s t-test, or one-way or two-way ANOVA was used to compare multiple groups with one or more variables).

Data Availability

All original data that support the plots within this paper and other findings of this study is included in the source data files published with this study.