Influence of glycoprotein MUC1 on trafficking of the Ca2+-selective ion channels, TRPV5 and TRPV6, and on in vivo calcium homeostasis

Abstract

Polymorphism of the gene encoding mucin 1 (MUC1) is associated with skeletal and dental phenotypes in human genomic studies. Animals lacking MUC1 exhibit mild reduction in bone density. These phenotypes could be a consequence of modulation of bodily Ca homeostasis by MUC1, as suggested by the previous observation that MUC1 enhances cell surface expression of the Ca²⁺-selective channel, TRPV5, in cultured unpolarized cells. Using biotinylation of cell surface proteins, we asked whether MUC1 influences endocytosis of TRPV5 and another Ca²⁺-selective TRP channel, TRPV6, in cultured polarized epithelial cells. Our results indicate that MUC1 reduces endocytosis of both channels, enhancing cell surface expression. Further, we found that mice lacking MUC1 lose apical localization of TRPV5 and TRPV6 in the renal tubular and duodenal epithelium. Females, but not males, lacking MUC1 exhibit reduced blood Ca²⁺. However, mice lacking MUC1 exhibited no differences in basal urinary Ca excretion or Ca retention in response to PTH receptor signaling, suggesting compensation by transport mechanisms independent of TRPV5 and TRPV6. Finally, humans with autosomal dominant tubulointerstitial kidney disease due to frame-shift mutation of MUC1 (ADTKD-MUC1) exhibit reduced plasma Ca concentrations compared to control individuals with mutations in the gene encoding uromodulin (ADTKD-UMOD), consistent with MUC1 haploinsufficiency causing reduced bodily Ca²⁺. In summary, our results provide further insight into the role of MUC1 in Ca²⁺-selective TRP channel endocytosis and the overall effects on Ca concentrations.

Keywords: Ca homeostasis; MUC1; TRPV5; TRPV6; mucin 1.

Figure 1

MUC1 increases cell surface expression and decreases rates of endocytosis selectively for TRPV5 and TRPV6 in polarized epithelial cells.A, cell surface expression and endocytosis of TRPV5 were examined. MDCK cells with, or without, stable expression of MUC1 were transiently transfected with TRPV5-GFP. Cell surface proteins were labeled with membrane-impermeant sulfo-NHS-SS-biotin on ice and moved to a circulating water bath at 37 °C for 0, 10, or 20 min. Cells were returned to ice, and surface biotin was stripped with MESNA. Cell lysates were incubated overnight at 4 °C with neutravidin-conjugated beads. Beads were washed and proteins eluted in sample buffer with β-mercaptoethanol at 60 °C for 5 min before SDS-PAGE and immunoblotting for either TRPV5 with anti-GFP antibodies or for MUC1 with anti-cytoplasmic tail (CT2) antibodies. Differences in the oligomerization state of TRPV5-GFP were the result of the overnight incubation at 4 °C and not indicative of changes in situ (see Fig. S1). Therefore, all oligomerization states were included for the quantification of TRPV5-GFP. Time 0 represents basal cell surface expression of TRPV5. MUC1 increased TRPV5 cell surface expression (28%, ∗p < 0.05 by Student’s t test). Endocytosis time-courses (N = 3) show that MUC1 reduces TRPV5 endocytosis (∗∗p < 0.01 by two-way ANOVA). Endocytosis of TRPV5 in the presence of MUC1 did not differ significantly from that of MUC1 itself. Representative blots are shown for each analysis. B, cell surface expression and endocytosis time-course of TRPV6 with, and without, coexpression with MUC1 were examined as in (A). Blots were developed with rabbit anti-TRPV6 or anti-MUC1 cytoplasmic tail antibodies. Time 0–biotinylated TRPV6 demonstrated increased basal cell surface TRPV6 in the presence of MUC1 (63%, ∗p < 0.05 by Student’s t test). MUC1 coexpression significantly reduced TRPV6 endocytosis (∗∗∗p = 0.001 by two-way ANOVA). Endocytosis of TRPV6 in the presence of MUC1 did not differ significantly from that of MUC1 itself. Representative blots are shown for each analysis. C, endocytosis of an endogenous protein, podocalyxin, in MDCK cells is no different in the presence or absence of MUC1 expression (p = NS by two-way ANOVA). Representative blots are shown for each analysis. Error bars represent SD of the mean. MDCK, Madin-Darby canine kidney.

Figure 2

MUC1 increases TRPV6 whole cell current density.A, whole cell currents were evoked by linear ramps from −100 to 70 mV from a holding potential of −10 mV as described in Experimental procedures. B, typical current traces are shown after maximal activation for HEK293 cells transfected with TRPV6 or cells cotransfected with TRPV6 and MUC1. The dashed line represents zero current. C, summary of the effect of MUC1 on TRPV6-mediated whole cell currents measured at −80 mV. Cotransfection of MUC1 resulted in increased TRPV6 current density (−63 ± 13 pA/pF, n = 6, compared with controls expressing TRPV6 alone (−24 ± 4 pA/pF, n = 7)). Data are presented as means ± SEM (∗p < 0.05 by Student’s t test).

Figure 3

Absence of MUC1 in vivo shifts localization of TRPV5 and TRPV6 away from the cell apex.A, representative immunofluorescence images show redistribution of TRPV5 from the renal tubule cell apex toward cytoplasm in kidneys from Muc1^−/− mice. Red represents TRPV5 staining. Green represents parvalbumin staining, indicating early distal convoluted tubule. Blue represents TO-PRO-3 staining. Panel below shows the mean of line scans of TRPV5 staining intensity from the cell surface, extending into the cytosol from Muc1^−/− mice and from Muc1^+/+ controls (one line scan per TRPV5-expressing cell). Y-axis represents mean normalized intensity in early DCT epithelial cells. X-axis indicates distance from apparent cell surface. Shading behind black line represents 95% confidence interval of line scans (N = 3 animals per genotype; n = 251 cells for Muc1^+/+ mice; n = 332 cells for Muc1^−/− mice). Note, divergence of line scans within the cytoplasmic region, indicating more TRPV5 staining in the cytosol in Muc1^−/− mice. Second panel represents mean cytoplasm/apical staining intensity, as calculated from arbitrarily chosen cytosolic and apical cell regions. Cytoplasm/cell apex staining of TRPV5 is greater in Muc1^−/− mice (N = 3 animals per genotype, n = 147 cells for Muc1^+/+ mice; n = 192 cells for Muc1^−/− mice; ∗∗∗∗p < 0.0001 by Student’s t test). B, TRPV6 subcellular localization in mouse kidney. Red represents TRPV6 staining. Panel below shows the mean of line scans of TRPV6 staining intensity. TRPV6 staining intensity is shifted away from the cell apex, toward cytoplasm (N = 3 animals per genotype, n = 180 cells for Muc1^+/+ mice; n = 380 cells for Muc1^−/− mice). Adjacent panel shows cytoplasm/apical TRPV6 staining in tubular epithelial cells. Cytoplasm/apical TRPV6 staining is greater in Muc1^−/− mice (N = 3 animals per genotype, n = 184 cells for Muc1^+/+ mice; n = 270 cells for Muc1^−/− mice; ∗∗∗∗p < 0.0001 by Student’s t test). C, duodenal epithelium TRPV6 staining is also shifted toward cytoplasm in Muc1^−/− mice. Line scans and cytoplasm/apical staining were examined from 5 to 10 cells per villus in 9 to 10 individual villi per duodenum from a total of three mice from each genotype (n = 45 cells for Muc1^+/+ mice; n = 48 for Muc1^−/− mice). Line scans and cytoplasm/apical staining both show a shift of TRPV6 staining away from cell apex, toward the cytoplasm in Muc1^−/− mice (∗∗∗∗p < 0.0001 by Student’s t test). In all images, gray scale bars represent 10 μm. Error bars represent SD of the mean. DCT, distal convoluted tubule.

Figure 4

Muc1^−/−mice exhibit lower blood Ca²⁺, but no differences inCaexcretion.A, blood Ca²⁺ levels are lower in female but not male Muc1^−/− mice. Female Muc1^−/− mice exhibited lower blood Ca²⁺ than Muc1^+/+ controls (∗p < 0.05 and ∗∗p < 0.01 by Student’s t test). Male mice had similar blood Ca²⁺ in Muc1^+/+ controls compared to Muc1^−/− animals (p = NS). There was no genotype-specific difference when male and female mice were pooled (p = NS). B, urine Ca excretion rate (U_CaV̇) was not different in Muc1^−/− mice compared to littermate controls. Urine was collected over 3.5 h in metabolic cages, and urinary Ca excretion (U_CaV̇) was measured. No difference in U_CaV̇ was observed overall (p = NS by Student’s t test) or when mice were stratified on the basis of sex (p = NS in pairwise comparison by Student’s t test). C, plasma levels of 1,25(OH)₂ Vitamin D did not differ in Muc1^+/+versus Muc1^−/− mice, either overall or after stratification by sex (p = NS in pairwise comparison by Student’s t test). D, PTH levels were not different in Muc1^+/+versus Muc1^−/− mice either overall or after stratification by sex (p = NS in pairwise comparison by Student’s t test). Error bars represent SD of the mean.

Figure 5

Urinary electrolyte excretion in response to the stable PTH analog, teriparatide, did not differ between Muc1^+/+and Muc1^−/−mice.A, urine Na excretion (U_NaV̇), phosphorus excretion (U_PhosV̇), and Ca excretion (U_CaV̇) were compared following injection of 5% volume/body weight of NSS. Two days later, the same animals received 5% volume/body weight of NSS with 150 μg/kg TPT. N = 8 Muc1^−/− mice and 10 Muc1^+/+ mice. Open or closed symbols represent excretion following injection with NSS or TPT, respectively. Blue or green lines and symbols represent excretion from Muc1^+/+ or Muc1^−/− mice, respectively. TPT influenced U_NaV̇ (p < 0.01), U_PhosV̇ (p < 0.0001), and U_CaV̇ (p < 0.0001), however, the genotype-treatment interaction term was not significant by mixed effects modeling (p = NS). Error bars represent standard error. B, TPT induced a similar degree of hypercalcemia in Muc1^−/− and Muc1^+/+ animals. Blood Ca²⁺ was similar in TPT-treated Muc1^+/+ and Muc1^−/− mice (p = NS). Error bars represent SD of the mean. NSS, normal saline solution; TPT, teriparatide.

Figure 6

Plasma Ca is lower in individuals with autosomal dominant tubulointerstitial kidney disease due to MUC1 mutation (ADTKD-MUC1) than in control individuals with autosomal dominant tubulointerstitial kidney disease due to UMOD mutation (ADTKD-UMOD). Violin plots representing plasma Ca extend from minimum to maximum values. Solid lines represent the median; dashed lines represent quartiles. Plasma Ca levels were lower in ADTKD-MUC1 patients than in controls (∗∗p < 0.01 by Student’s t test). After stratification by sex, this difference persisted in women (∗p < 0.05 by Student’s t test) but not in men (p = NS).