Somatic Mutations in MCOLN3 Are Associated with Aldosterone-Producing Adenomas

Abstract

Background:

Primary aldosteronism is a common but underdiagnosed cause of endocrine hypertension that contributes to global cardiovascular morbidity and mortality. It is characterized by renin-independent hyperaldosteronism that originates from adrenal lesions — the majority of which are found to harbor aldosterone-driver somatic mutations in genes encoding ion-transporting proteins. These mutations disrupt intracellular calcium homeostasis, facilitating pathologic increase in aldosterone synthase expression and aldosterone production. Elucidating the exact mechanisms causing aldosterone excess in primary aldosteronism would further the development of targeted treatments and alleviate the global hypertension burden.

Methods:

Next-generation sequencing analysis of formalin-fixed paraffin-embedded aldosterone-producing adenomas identified novel somatic variants in MCOLN3 (encoding the cation-permeable channel, TRPML3). Electrophysiologic, fura-2 calcium measurements, gene expression and steroid quantification studies were performed in adrenal HAC15 cells to characterize the functional effects of the novel MCOLN3 mutations.

Results:

Three somatic MCOLN3 variants (p.Y391D, p.F415I and p.N411_V412delinsI) were identified in aldosterone-producing adenomas from four male primary aldosteronism patients. Mutated MCOLN3 expressed in HAC15 cells resulted in a gain-of-function phenotype, which induced cell membrane depolarization and calcium influx and, in turn, triggered a significant increase in aldosterone synthase expression and aldosterone production.

Conclusions:

This is the first report of disease-causing MCOLN3 mutations in humans and first to implicate mutated MCOLN3 as a driver of dysregulated aldosterone production in primary aldosteronism.

Graphical Abstract

Introduction

Primary aldosteronism (PA) is the most prevalent cause of endocrine hypertension, accounting for 5–15 %^{1, 2} and 14–21 %³ of general and resistant hypertension cases, respectively. PA is characterized by renin-independent production of aldosterone from one (unilateral) or both (bilateral) adrenals that results in chronic hypertension.

Unilateral PA is most often attributed to aldosterone-producing adenomas (APA) or to smaller lesions called aldosterone-producing nodules (APN).⁴ Autonomous aldosterone biosynthesis in these lesions is primarily caused by aldosterone-driver somatic mutations in genes that encode ion pumps, transporters, and channels. These mutations induce cytosolic calcium influx via cell membrane depolarization and, in turn, upregulate aldosterone synthase (CYP11B2) expression, which facilitates aldosterone production. Next-generation sequencing (NGS) has identified aldosterone-driver mutations in genes including KCNJ5 (potassium channel),⁵ ATP1A1 (sodium/potassium ATPase), ATP2B3 (plasma membrane calcium ATPase),^{6, 7} CACNA1D (voltage-gated L-type calcium channel),^{6, 8} CACNA1H (voltage-gated T-type calcium channel),⁹ CLCN2 (chloride channel),¹⁰ and more recently SLC30A1 (zinc transporter).¹¹ In rare cases, mutations have also been found in genes encoding other proteins including CADM1,¹² CTNNB1,¹³ PRKACA,¹⁴ and GNAQ/11.¹⁵

The advances in CYP11B2 immunohistochemistry (IHC)-guided DNA sequencing using formalin-fixed paraffin-embedded (FFPE) tissue have boosted the detection of somatic mutations in aldosterone-producing lesions to 89–98 %.⁹ This, in turn, has provided important clues into the pathogenesis of PA. Herein, we describe the discovery and characterization of novel aldosterone-driver variants in MCOLN3 that encodes mucolipin-3/TRPML3 (transient receptor potential cation channel 3). Importantly, MCOLN3 mutations have neither been previously reported in APA/APN nor has this gene been associated with any disease in humans. MCOLN3 has primarily been studied in context of endosome trafficking and autophagy,^{16, 17} but its functional role in the adrenal gland is unknown. In this study, we have demonstrated that novel MCOLN3 mutations in APA lead to increased calcium influx that causes elevated aldosterone production, thereby resulting in PA.

Methods

The data that support the findings of this study are available from the corresponding author upon reasonable request. Extended details on the methodology are provided in the Supplemental Material. Please see the Major Resources Table in the Supplemental Materials.

Ethics and Study Approval

Consented patients with unilateral PA (based on the institutional consensus or Endocrine Society’s clinical practice guidelines) from the University of Michigan, University of Pennsylvania, Vanderbilt University, and Charité-Universitätsmedizin Berlin were included in this study. The inclusion criteria further considered the availability of FFPE blocks of resected adrenals. Clinical information was collected from each patient at the time of presentation. This study was conducted with the approval of the institutional review boards of the University of Michigan (HUM00083056 and HUM00069665), University of Pennsylvania (IRB#831990), Vanderbilt University Medical Center (IRB#150995), and the ethics committee of the Charité-Universitätsmedizin Berlin (#EA1/097/21). Informed consent was not required for retrospective analysis of archived material including UM102, ELA25, and AA30. For CUB-APA79, informed consent was obtained retrospectively. Deceased renal donor adrenal tissue for gene expression analysis was obtained from the University of Michigan kidney transplant donor program.

Statistical Analysis

Results are expressed as mean ± SEM unless otherwise stated. Statistical comparisons of parameters between groups were performed using the Kruskal-Wallis test or an ordinary one-way and two-way ANOVA followed by the appropriate post hoc tests. Statistical analyses were performed using GraphPad Prism version 10 (GraphPad Software, Inc., San Diego, CA). Extended details are provided in the Supplemental Material.

Results

In this study, whole exome sequencing (WES) analysis was performed on DNA from two FFPE APA without known aldosterone-driver mutations based on our aforementioned CYP11B2 IHC-guided targeted NGS approach. WES identified a novel missense somatic mutation, p.Y391D in MCOLN3 (MCOLN3^Y391D) in two APA DNA samples (UM102 and ELA25) (Table S1–S2), which was confirmed using an updated targeted NGS panel (Table S3). The panel also identified another novel in-frame mutation p.N411_V412delinsI (MCOLN3^{N411_V412delinsI}) in one APA (AA30) that was previously found to be devoid of known aldosterone-driver mutations. Both MCOLN3 mutations were validated by Sanger sequencing (Figure 1A–B). Subsequent WES studies identified an additional missense mutation p.F415I (MCOLN3^F415I) in one APA (CUB-APA79), which was also confirmed with our targeted NGS approach. Collectively, three MCOLN3 mutations were identified in four APA, all derived from men with PA across different racial groups, between ages 57–76 years (Table 1). Cross-sectional imaging and adrenal vein sampling confirmed that three patients (ELA25, AA30, and CUB-APA79) presented with a solitary APA. The fourth patient (UM102) had an aldosterone-producing lesion on the left adrenal gland and a non-functional adenoma on the contralateral gland. All four patients exhibited hypokalemia and severely elevated aldosterone-to-renin ratios. The patients harboring MCOLN3^{N411_V412delinsI} and MCOLN3^F415I presented with more severe hypertension compared to patients with MCOLN3^Y391D. However, these patients were on fewer antihypertensive medications. Notably, the patient harboring MCOLN3^{N411_V412delinsI}, who is of African ancestry, presented with more severe hypokalemia, which might suggest a higher sensitivity to sodium and aldosterone.^{18, 19} Analysis of the overall frequency of somatic MCOLN3 mutations per participating institute pointed to mutated MCOLN3 as a rare cause of PA (0.98 % prevalence; Table S4). Post-surgical follow-up of these patients indicated complete resolution of biochemical PA.

Figure 1. MCOLN3 mutations in APA. Sanger sequence chromatograms of tumor and adjacent adrenal tissue DNA for validation of the MCOLN3^Y391D and MCOLN3^{N411_V412delinsI} variants.

Sanger chromatogram sequences for the (A) MCOLN3^Y391D variant (UM102) resulting in a single residue change (blue) and (B) MCOLN3^{N411_V412delinsI} mutation (AA30) presenting with an indel variant (red, deleted; yellow, inserted). Mutations are absent in the tumor adjacent adrenal tissue (wild-type (WT); top sequences). (C) Simplified schematic representation of the mucolipin-3 transmembrane. Location of MCOLN3^Y391D (cyan), MCOLN3^{N411_V412delinsI} (yellow), and MCOLN3^F415I (magenta) in segment 4 (S4) and 5 (S5) is shown in close proximity to the ion selectivity filter. (D) Alignment of orthologs among various species shows the conservation of Tyr391, Val412, and F415 in the mucolipin-3 protein. Affected amino acid residue associated with the novel variants are highlighted in blue (MCOLN3^Y391D), red and green (MCOLN3^{N411_V412delinsI}), and purple (MCOLN3^F415I). In grey are residues conserved in orthologs. Figure 1C was created in BioRender. Van Rooyen, D. (2025) https://BioRender.com/yt2ztz0

Table 1.

Clinical characteristics of patients with mutant MCOLN3-harboring APA.

Sample ID	UM102	ELA25	AA30	CUB-APA79
Center	University of Michigan, USA	University of Pennsylvania, USA	Vanderbilt University, USA	Charité-Universitätsmedizin Berlin, Germany
Age at surgery (years)	63	76	57	74
Sex	Male	Male	Male	Male
Ancestry	European	Asian	African	European
Baseline characteristics
MCOLN3 somatic variant	p.Y391D	p.Y391D	p.N411_V412delinsI	p.F415I
Blood pressure (mmHg)	130/66	156/70	193/109	197/89
# of antihypertensive medications	3	6	5	2
Lowest serum potassium (mEq/L)	3.3	3.6	2.5	3.5
Potassium supplementation	No	Yes	Yes	Yes
PAC (ng/dL)	100.2	25	65.9	35.5
PRA (ng/mL/hr) or *ng/L	<0.1	0.14	0.3	1.1*
ARR [(ng/dL)/(ng/mL/h)]	high	179	220	322
Adrenal vein sampling
Lateralization ratio	11.6	10.4	13.2	Unsuccessful
Side of aldosterone excess	Left	Left	Right	Right
Imaging study
Side of adrenal tumor	Both	Left	Right	Right
Side of adrenalectomy	Left	Left	Right	Right
Resected tumor diameter (mm)	17	12	16	20
Post-surgical follow-up data
Follow-up period	0.5 months	242 months	4 months	46 months
Blood pressure (mmHg)	148/86	126/78	168/96	130/80
# of antihypertensive medications	1	2	2	1
PAC (ng/dL)	8.7	Unavailable	1.6	Unavailable
PRA (ng/mL/hr)	0.4	Unavailable	0.2	Unavailable
ARR [(ng/dL)/(ng/mL/h)]	21.75	Unavailable	8	Unavailable
Clinical outcome	Partial	Partial	Partial	Partial
Biochemical outcome	Complete	Unavailable	Complete	Unavailable

N/A, not available; PAC, plasma aldosterone concentration; PRA, Plasma renin activity; ARR, aldosterone:renin ratio

MCOLN3 encodes a transient receptor potential cation channel. All three MCOLN3 mutations are in close proximity to the TRPML3 ion-conducting pore and ion selectivity filter (Figure 1C). Furthermore, Tyr391, Val412, and Phe415 are highly conserved in orthologs of various species (Figure 1D) including the mouse. In mice, disease-causing MCOLN3 germline mutations have been reported close to the ion selectivity filter.^{20, 21} The role of MCOLN3 in the adrenal gland is unclear. Herein, we investigated the differential expression of MCOLN3 mRNA in multiple human tissues and found highest expression in the adrenal cortex (Figure S1A). MCOLN3 levels were significantly higher in APA (2-fold; P<0.05) vs normal adrenal, but no difference was observed in cortisol-producing adenomas (CPA) (Figure S1B). CYP11B2 mRNA expression was significantly higher in APA (102-fold; P<0.001) compared to both normal adrenals and CPA (Figure S1B). Single-nucleus RNA transcriptome analysis using an adrenal gland from a 21-year-old male deceased renal donor yielded distinct adrenal cortex zonal cell population clusters, including two clusters for zona glomerulosa (ZG) (clusters ZG1 and ZG2) (Figure S2A). The aldosterone-producing cells (with high CYP11B2 expression) were confined to the ZG1 cluster, whereas a strong enrichment of MCOLN3 was observed in both ZG1 and ZG2.

Previous investigations of morphologically normal adrenals reported an age-related change in CYP11B2 expression pattern, highlighting that not all ZG cells are aldosterone-producing (or CYP11B2-expressing).²² We investigated the relationship between MCOLN3 expression with that of CYP11B2-expressing ZG cells by immunohistochemical analysis in adrenals with continuous as well as fragmented ZG CYP11B2 expression (Figure S2B). We observed that the immunoreactivity for MCOLN3 was localized to the ZG but not restricted to the CYP11B2-expressing ZG cells.

Histologic assessment of mutated MCOLN3-harboring tumors (Figure 2) showed increased immunoreactivity for MCOLN3 and CYP11B2 in tumors with the missense mutation as compared to the deletion-insertion variant. While MCOLN3 expression in the APA with the MCOLN3^{N411_V412delinsI} variant was sparse as compared to the ZG of adjacent adrenal tissue, CYP11B2 immunoreactivity in the tumor was heterogenous. We also investigated the immunoreactivity of 17α-hydroxylase/17,20-lyase (CYP17A1), a steroidogenic enzyme that is expressed in APA harboring KCNJ5²³ and SLC30A1¹¹ mutations. The CYP17A1 expression was observed to vary from scant up to high in the MCOLN3 mutation-harboring APA.

Figure 2. Histopathological characteristics of APA with somatic MCOLN3 mutations.

Hematoxylin and eosin (H&E) staining, and MCOLN3/TRPML3, CYP11B2 and CYP17A1 immunohistochemical staining of the resected APA harboring the MCOLN3 variants. Scale bar, 2.5 mm.

We also performed targeted RNA sequencing on mutant MCOLN3-harboring tissue (APA_UM102, APA_ELA25, and APA_AA30) and compared the gene expression profiles to the adrenocortical zones (ZG, zona fasciculata (ZF) and zona reticularis (ZR)), and other adrenocortical tumor types (non-MCOLN3-mutated APA, CPA, and adrenocortical carcinomas (ACC)). Principal component analysis (PCA) demonstrated clustering of APA bearing MCOLN3 mutations with the aldosterone-biosynthesizing ZG (Figure S3A). In addition, we observed that mutated MCOLN3-harboring APA (n=3) grouped with APA of other genotypes, and distinctly separated from both ACC and CPA (Figure S3B).

To characterize the functional effects of the MCOLN3^Y391D and MCOLN3^{N411_V412delinsI} variants, we transfected HAC15 adrenocortical cells with plasmid DNA encoding mutant MCOLN3 variants (HAC15-MCOLN3^Y391D and HAC15-MCOLN3^{N411_V412delinsI}) and assessed the changes in MCOLN3, CYP11B2, NR4A2 (NURR1, an upstream regulator of CYP11B2 transcription),²⁴ CYP11B1 (11β-hydroxylase) and CYP17A1 transcript levels, as well as aldosterone production after 48 h. Data was compared to cells transfected with the empty pIRES-CD8²⁵ plasmid vector (HAC15-empty) and the wild-type MCOLN3 variant (HAC15-MCOLN3^WT). Transfection was determined to be successful based on elevated MCOLN3 mRNA levels detected in cells transfected with MCOLN3^WT, MCOLN3^Y391D, and MCOLN3^{N411_V412delinsI} (13.1-, 9.4-, and 10.3-fold, respectively; P<0.0001) compared to empty control (Figure 3A). No change in transcript levels of NR4A2, CYP11B2, CYP11B1, or CYP17A1 was observed in HAC15-MCOLN3^WT vs the empty control. Additionally, aldosterone production was unaffected in HAC15-MCOLN3^WT (Figure 3 and Figure S4).

Figure 3. Impact of the MCOLN3^Y391D and MCOLN3^{N411_V412delinsI} variant on NR4A2 and CYP11B2 mRNA levels and aldosterone production.

HAC15 cells were transfected with pIRES-CD8 plasmid (empty), or plasmid containing MCOLN3^WT (WT), MCOLN3^Y391D (Y391D), or MCOLN3^{N411-V412delinsI} (∆N411_V412delinsI) via electroporation. 48 h post-transfection, relative mRNA levels of (A) MCOLN3, (B) NR4A2, and (C) CYP11B2, were quantified by qRT-PCR. (D) Aldosterone content in cell medium was quantified and normalized to total well protein content. Data are presented as the mean ±SEM of four independent experiments performed in triplicate (n=12). Statistical analyses were performed using a one-way ANOVA with Dunnett’s multiple comparison test relative to empty (A) and WT group (B-D). * P<0.05; *** P<0.001; ****P<0.0001. Volcano plot (E) and heatmap (F) representation of whole transcriptomic analyses of HAC15-MCOLN3^WT and HAC15-MCOLN3^Y391D after 48 h show increased expression of NR4A2 and CYP11B2. Figures represent data from four independent experiments (n=4), each consisting of an equivalent amount of pooled sample from experimental triplicates.

In contrast, the role of MCOLN3^Y391D and MCOLN3^{N411_V412delinsI} in driving hyperaldosteronism was highlighted by the significant increase in both NR4A2 (17.5 and 10.7-fold, respectively; P<0.0001) (Figure 3B) and CYP11B2 mRNA (4.7 and 4.2-fold, respectively; P<0.0001) (Figure 3C) vs MCOLN3^WT. CYP11B1 expression also differed significantly in HAC15-MCOLN3^Y391D and HAC15-MCOLN3^{N411_V412delinsI} vs HAC15-MCOLN3^WT (4-fold; P<0.0001) (Figure S4). While immunoreactivity of CYP17A1 co-localized with MCOLN3 expression in the tumors (Figure 2), HAC15 adrenocortical cells expressing wild-type or mutant MCOLN3 showed no change in CYP17A1 mRNA levels (Figure S4). Importantly, increased aldosterone production was observed in adrenocortical cells expressing MCOLN3^Y391D and MCOLN3^{N411_V412delinsI} (1.9- and 1.8-fold, respectively; P<0.0001) compared to HAC15-MCOLN3^WT. (Figure 3D). The biosynthesis of hybrid steroids, 18-hydroxycortisol and 18-oxocortisol, which requires both CYP11B2 and CYP17A1^{26, 27} — was also found to be elevated in HAC15-MCOLN3^Y391D vs HAC15-MCOLN3^WT, whereas cortisol remained unchanged (Figure S5). Bulk transcriptomic analysis of HAC15-empty, HAC15-MCOLN3^WT and HAC15-MCOLN3^Y391D further corroborated the augmentative effect of MCOLN3^Y391D on NR4A2 and CYP11B2 expression compared to MCOLN3^WT (Figure 3E–F).

PA-related mutations have been found to cause elevated CYP1B2 expression and dysregulated aldosterone via multiple molecular mechanisms. Some PA-associated mutations cause an influx of calcium directly though the mutated protein.⁸ Other PA-related mutated ion channels/transporters cause an indirect increase in intracellular calcium levels via opening of the voltage-gated L-type calcium channels owing to pathological depolarization of membrane potential as a result of abnormal sodium permeability.^{5, 11, 28, 29} Lastly, some mutations also function by concurrently adopting both the aforementioned mechanisms.²⁹ Since MCOLN3 is reportedly permeable to both calcium and sodium,³⁰ we tested the ability of MCOLN3^Y391D and MCOLN3^{N411_V412delinsI} to cause membrane depolarization and disruption in the regulation of cytosolic calcium levels via direct calcium influx and/or abnormal sodium conductance.

We investigated the basal cell membrane potential of HAC15-MCOLN3^Y391D and HAC15-MCOLN3^{N411_V412delinsI} by whole-cell patch-clamp recordings and compared it with the HAC15-empty and HAC15-MCOLN3^WT. The basal membrane potential of HAC15-MCOLN3^Y391D (−13.6 ±2.4 mV) and HAC15-MCOLN3^{N411_V412delinsI} (−18.5 ±2.0 mV) was strongly depolarized compared to the HAC15-MCOLN3^WT (−64.6 ±1.9 mV) and HAC15-empty (−67.4 ±1.9 mV) (Figure 4A). The depolarization of membrane potential seen in the HAC15-MCOLN3^Y391D and HAC15-MCOLN3^{N411_V412delinsI} cells was accompanied by significant elevation in basal intracellular calcium compared to HAC15-empty and HAC15-MCOLN3^WT (P<0.05) (Figure 4B). Cytosolic calcium activity was continuously monitored over time and upon exposure to extracellular solutions with varying calcium content (Figure 4C). Cytosolic calcium levels in HAC15-MCOLN3^Y391D and HAC15-MCOLN3^{N411_V412delinsI} showed a greater dependency on extracellular calcium levels compared to HAC15-MCOLN3^WT. This was indicated by a reduction in intracellular calcium levels in mutant cells following exposure to calcium-free extracellular solution, and a surge in cytosolic calcium under high extracellular calcium conditions. Cytosolic calcium levels in HAC15-MCOLN3^WT and HAC15-empty cells were unaffected by altered calcium content in extracellular solutions (Figure 4C). We also tested the ability of mutated MCOLN3 to cause disruptions in the regulation of cytosolic calcium levels under sodium-free conditions caused by replacing extracellular sodium with N-methyl-d-glucamine⁺ (NMDG⁺). Under sodium-free conditions, cytosolic calcium levels increased again in HAC15-MCOLN3^Y391D and HAC15-MCOLN3^{N411_V412delinsI}, while no change was observed in control cells. At the end of calcium measurements, we tested the effects of increased extracellular potassium (K⁺, 18 mM), which leads to membrane depolarization under physiological conditions. High potassium induced a significant cytosolic calcium increase in HAC15-MCOLN3^WT and HAC15-empty. Intracellular elevation in calcium levels in mutant MCOLN3 cells was minimal under high potassium, likely owing to the increased basal depolarization of these cells (Figure 4C).

Figure 4. Impact of mutant MCOLN3^Y391D and MCOLN3^{N411_V412delinsI} on basal membrane potential and cytosolic calcium (Ca²⁺) activity in adrenal HAC15 cells.

Transfected HAC15 cells expressing pIRES-CD8 plasmid (empty), MCOLN3^WT (WT), MCOLN3^Y391D (Y391D) or MCOLN3^{N411_V412delinsI} (∆N411_V412delinsI) were used for whole-cell patch-clamp experiments. (A) Membrane potential measurements of HAC15-MCOLN3^Y391D (n=17) and -MCOLN3^{N411_V412delinsI} (n=12) were compared to HAC15-empty (n=7) or HAC15-MCOLN3^WT (n=8). (B) Basal Ca²⁺ activities of HAC15-MCOLN3^Y391D (n=10) and HAC15-MCOLN3^{N411_V412delinsI} (n=9) were compared to HAC15-empty (n=10) or HAC15-MCOLN3^WT (n=11). (C) Changes in fura-2 ratio over time in cells overexpressing empty vector, MCOLN3^WT and mutated MCOLN3 variants upon application of the indicated extracellular solutions. Ca²⁺ in control was 1.3 mM compared to high-Ca²⁺ with 5 mM. K⁺ was 3.6 mM in control and 18 mM in high-K⁺. Statistical significance was tested using one-way ANOVA analysis with Tukey’s multiple comparison (HAC15-MCOLN3^Y391D and HAC15-MCOLN3^N411_V412I vs HAC15-MCOLN3^WT (*P<0.05,) or HAC15-empty control (^#P<0.05)).

We further investigated whether increased sodium conductance could play a role in the pathogenesis of mutant MCOLN3, as seen previously in APA with mutated KCNJ5 and SLC30A1.^{5, 11} The strong depolarization of HAC15-MCOLN3^Y391D and HAC15-MCOLN3^{N411_V412delinsI} was characterized by measuring the whole-cell currents at different voltage-clamp steps (Figure S6 A–D). As opposed to HAC15-empty and HAC15-MCOLN3^WT, HAC15-MCOLN3^Y391D and HAC15-MCOLN3^{N411_V412delinsI} exhibited an additional large inward current at voltages between −100 and −40 mV. This current was significantly reduced under sodium-free conditions. In addition, removing extracellular sodium resulted in significant reduced depolarization of the cell membrane in HAC15-MCOLN3^Y391D (−34.7 ±2.9 mV) and HAC15-MCOLN3^{N411_V412delinsI} (−32.4 ±3.2 mV) (Figure S6E). The mutated cells, however, did not hyperpolarize to the level of HAC15-MCOLN3^WT and HAC15-empty. Interestingly, HAC15-MCOLN3^WT were also hyperpolarized to some extent by sodium removal (Figure S6E). Overall, the patch-clamp data suggested an abnormal basal sodium conductance of mutant MCOLN3, which caused cell membrane depolarization.

We further examined the impact of mutant MCOLN3-induced membrane depolarization on the activation of voltage-dependent L-type Ca²⁺ channels by treating the WT and mutant MCOLN3-expressing HAC15 cells with varying concentrations (0.1, 1 and 10 µM) of verapamil (an L-type calcium channel blocker) (Figure S7). Both NR4A2 and CYP11B2 demonstrated a dose-dependent abrogation in HAC15-MCOLN3^Y391D. Similarly, HAC15-MCOLN3^{N411_V412delinsI} exhibited a decreasing trend for CYP11B2 levels with increasing verapamil concentrations, with significance reached at 10 µM. For the HAC15-empty and HAC15-MCOLN3^WT control cells, an insignificant, albeit discernible reduction in CYP11B2 transcript levels was observed at 10 µM verapamil, which may be explained by the non-specific actions of verapamil at high concentrations.

Collectively, the fura-2 calcium measurements revealed an abnormal calcium influx and impaired calcium-handling capacity which is likely caused by the abnormal basal ion conducting activity of MCOLN3^Y391D and MCOLN3^{N411_V412delinsI}. This mechanism underpins the autonomous aldosterone production in the mutant MCOLN3-expressing cells.

Discussion

Adrenal CYP11B2 expression and aldosterone production is tightly regulated by intracellular calcium levels.^{31, 32} Somatic aldosterone-driver mutations in aldosterone-producing adrenal lesions affect intracellular calcium buffering in a direct⁸ or indirect^{5, 7, 11} manner. In this study, we report the first set of disease-causing mutations in MCOLN3 identified in APA from four men. In vitro investigations of the MCOLN3 mutations (MCOLN3^Y391D and MCOLN3^{N411_V412delinsI}) showed increased aldosterone production by altering intracellular calcium flux.

Research related to the regulatory role of MCOLN3 in endocytic and autophagic pathways has increased the last few decades.³³ Human MCOLN3 is located on chromosome 1 (1p22.3) and encodes TRPML3,³⁴ which consists of a conserved ion-conducting pore domain formed by four subunits. Each subunit has six transmembrane segments and a pore loop located between segments 5 and 6. The pore loop (L444_I455 and M460_K466) is located on either side of the ion selectivity motif (Figure 1).³⁵ In varitint-waddler mice, the A419P located close to the ion transport region and leads to gain-of-function of MCOLN3.³⁴ As a result, the constitutively active inward current carried by calcium and sodium ions causes auditory disease, circling behavior and abnormal pigmentation^{21, 30} (heterozygous) or death (homozygous).^{20, 21, 33} The mouse and human MCOLN3 share ~91 % sequence homology.³⁶ MCOLN3 mutations have not yet been associated with disease in humans. Notably, the residues altered by MCOLN3^Y391D, MCOLN3^F415I, and MCOLN3^{N411_V412delinsI} identified in the APA were found to be in close proximity to the conserved ion transport motif³⁵ suggesting a possible gain or loss of function of the channel.

Our study demonstrated that MCOLN3 is prominently expressed in the adrenal cortex, as was shown previously^{37, 38} as well as aldosterone-producing tumors. Our single nucleas-RNAseq analysis corroborated previous findings by Altieri, et al., which demonstrated the ZG localization of MCOLN3 in the human adrenal cortex⁴¹. Interestingly, while MCOLN3 was abundantly expressed throughout the adrenal ZG, its expression did not necessarily correlate with that of CYP11B2, suggesting that the expression/function of the two genes is independent each other. This is also corroborated by our in vitro studies which demonstrate that overexpression of MCOLN3^WT does not alter either CYP11B2 transcription or aldosterone production in adrenal cells. Thus, the physiological function of MCOLN3^WT in the adrenal cortex remains to be elucidated.

Co-expression of CYP11B2 and CYP17A1 is an immunohistochemical characteristic of adrenal tumors harboring KCNJ5²⁷ and SLC30A1¹¹ mutations. Consequently, elevated biosynthesis of the hybrid steroids, including 18-hydroxycortisol and 18-oxocortisol, is detected in these patients.^{39, 40} In our study, CYP17A1 tumor immunoreactivity accompanied that of CYP11B2 and MCOLN3. While we were unable to determine the increase in hybrid steroids in patient serum owing to lack of sample availability, elevated hybrid steroid levels in our in vitro experimental adrenal cell media suggest that these steroids could be considered for serum steroid biomarker studies for PA patients with MCOLN3 mutation-harboring adrenal lesions. Our in vitro studies showed the selective elevation of the key steroidogenic enzymes and steroids in the aldosterone biosynthesis pathway.

The mutant MCOLN3 cells suggest that mutated MCOLN3 specifically targeted aldosterone biosynthesis and was not responsible for an increase in cortisol production. Interestingly, the immunoreactivity of MCOLN3 and CYP11B2 was more intense in the tumors harboring MCOLN3^Y391D than the tumor harboring MCOLN3^{N411_V412delinsI}. The MCOLN3^{N411_V412delinsI} variant was identified in an African American patient who was the youngest of the four patients in this study and presented with more severe hypertension than the patients with APA harboring the missense MCOLN3 mutations. Previous studies have found that patients of African ancestry have a positive correlation between the severity of hyperaldosteronism and blood pressure^{41, 42} and a higher susceptibility to sodium-sensitive hypertension and hyperaldosteronism.^{18, 19} While only 4 patients with APA MCOLN3 mutations have thus far been identified, there appears to be a series of clinical implications for this rare cause of PA. These include a male dominance, older age of onset, and greater disease severity as illustrated by hypokalemia and a high ARR. Furthermore, MCOLN3 mutations are associated with larger adenomas, similar to the APA harboring mutant KCNJ5 and ATP2B3.⁴³ In addition, surgical treatment appears to be highly effective for resolution of biochemical PA for patients with the mutant MCOLN3-bearing APA.

To our knowledge, this study is the first report of disease-causing MCOLN3 mutations in humans and, importantly, the first to implicate mutant MCOLN3 as a cause of PA. Electrophysiologic studies demonstrated that overexpression of MCOLN3^Y391D and MCOLN3^{N411_V412delinsI} in HAC15 adrenal cells induces calcium influx via both direct and indirect mechanisms (Figure 5). First, the mutations cause a gain-of-function phenotype leading to a constitutively active calcium-permeable channel that causes a direct calcium influx in the mutant cells. This mechanism is characteristic of the pathologic machinery causing aldosterone excess in CACNA1D mutation-harboring APA.⁸ Second, the mutant MCOLN3 causes abnormal sodium conductance as seen in other aldosterone-driver mutations including KCNJ5, SLC30A1 and ATP1A1.^{5, 11, 28, 29} This could lead to depolarization of the cell membrane and an increase in cytosolic calcium activity via opening of voltage-gated L-type calcium channels. This hypothesis was supported by the abrogation of mutated MCOLN3-induced adrenal cell CYP11B2 expression by low concentrations of verapamil (1 µM) which can be explained by the inhibition of L-type calcium channels. If membrane depolarization and activation of L-type calcium channels were the sole mediators of aldosterone excess in the mutant MCOLN3 cells, then treatment with verapamil would be expected to reduce CYP11B2 transcript levels in these cells to the basal levels observed in the control (empty vector and wild-type) cells. However, verapamil-treated mutant cells exhibited increased CYP11B2 transcript levels vs control cells, suggesting that additional mechanisms beyond L-type calcium channel activation contribute to aldosterone overproduction. In this case, direct calcium influx through mutant MCOLN3 would be a modulator. Additionally, it is likely that the increased sodium conductance and depolarization could impair calcium export via the sodium/calcium-exchanger (NCX) as previously reported with the mutant KCNJ5.⁴⁴ The elevated cytosolic calcium observed under sodium-free conditions (Figure 4C) could, thereby, be a result of impaired calcium export function of NCX or even possibly the calcium import reverse-mode of NCX triggered under pathological conditions.⁴⁵ Collectively, as seen in the case of CACNA1H mutations,²⁹ both the direct and indirect (depolarization-dependent) mechanisms of increased calcium activity trigger increased CYP11B2 expression and cause dysregulated aldosterone production.

Figure 5. Mechanisms of cytosolic calcium (Ca²⁺) influx caused by mutant MCOLN3 variants.

Mutant MCOLN3 resulted in a gain-of-function Ca²⁺ and/or sodium (Na⁺)-permeable channel which alters calcium activity directly or indirectly. (A) MCOLN3 causes direct Ca²⁺ influx via the mutant channel. (B) Increased sodium conductance by mutated MCOLN3 induces Na⁺-dependent membrane depolarization which indirectly causes Ca²⁺ influx via the activation of voltage-gated calcium channels. Additionally, increased Na⁺ conductance may lead to higher cytosolic Na⁺ concentration and membrane depolarization that would impair Ca²⁺ export via the Na⁺/Ca²⁺-exchanger (NCX), thus contributing to increased cytosolic Ca²⁺ (not shown here). Elevated cytosolic Ca²⁺ levels induce the transcription of aldosterone synthase (CYP11B2) and hyperaldosteronism. Created in BioRender. Van rooyen, D. (2025) https://BioRender.com/nehq7dj

In conclusion, our data implicate mutated MCOLN3 as direct and indirect facilitator of cytosolic calcium influx, which, consequently, drives autonomous aldosterone production in PA.

Perspectives

Recurrent somatic mutations in MCOLN3, located near its highly conserved ion transport pore and selectivity filter, were discovered in aldosterone-producing adenomas from four male patients. This study is the first to demonstrate mutations in MCOLN3 as a cause of disease, specifically PA, in humans. In vitro studies demonstrated that the newly identified MCOLN3 mutations lead to increased adrenal cell calcium influx that causes elevated adrenal cell aldosterone production. Although MCOLN3 mutations encompass a rare genetic cause of sporadic primary aldosteronism, this study provides a better understanding of the molecular mechanisms underlying elevated aldosterone production in aldosterone-producing adenomas.

Novelty and Relevance.

What is New?

• Somatic mutations in MCOLN3, which encodes mucolipin-3 (calcium-permeable cation channel), were identified in aldosterone-producing adenomas from men.

• In vitro studies demonstrate mutated MCOLN3 as a direct and indirect facilitator of cytosolic calcium influx, which drives autonomous aldosterone production.

• This is the first report of disease-causing MCOLN3 mutations in humans and to implicate mutations in MCOLN3 as a cause of primary aldosteronism.

What is Relevant?

• In vitro analyses suggest that mutant MCOLN3 induces membrane depolarization and calcium influx resulting in increased CYP11B2 expression and aldosterone production.

• Mutations in MCOLN3 constitute a rare genetic cause of sporadic primary aldosteronism.

Clinical/Pathophysiological Implications

• The study broadens our knowledge of the pathomechanisms causing hyperldosteronism in aldosterone-producing adenomas.

Non-standard Abbreviations and Acronyms:

PA

primary aldosteronism

APA

aldosterone-producing adenomas

APN

aldosterone-producing nodules

Y391D

MCOLN3 ^Y391D

∆V411_V412delinsI

MCOLN3 ^{N411_V412delinsI}

WT

MCOLN3 ^WT

NGS

Next-generation sequencing

FFPE

formalin-fixed paraffin-embedded

CYP11B2

aldosterone synthase