Does the composition of urinary extracellular vesicles reflect the abundance of renal Na+/phosphate transporters?

Abstract

Studies addressing homeostasis of inorganic phosphate (Pi) are mostly restricted to murine models. Data provided by genetically modified mice suggest that renal Pi reabsorption is primarily mediated by the Na⁺/Pi cotransporter NaPi-IIa/Slc34a1, whereas the contribution of NaPi-IIc/Slc34a3 in adult animals seems negligible. However, mutations in both cotransporters associate with hypophosphatemic syndromes in humans, suggesting major inter-species heterogeneity. Urinary extracellular vesicles (UEV) have been proposed as an alternative source to analyse the intrinsic expression of renal proteins in vivo. Here, we analyse in rats whether the protein abundance of renal Pi transporters in UEV correlates with their renal content. For that, we compared the abundance of NaPi-IIa and NaPi-IIc in paired samples from kidneys and UEV from rats fed acutely and chronically on diets with low or high Pi. In renal brush border membranes (BBM) NaPi-IIa was detected as two fragments corresponding to the full-length protein and to a proteolytic product, whereas NaPi-IIc migrated as a single full-length band. The expression of NaPi-IIa (both fragments) in BBM adapted to acute as well to chronic changes of dietary Pi, whereas adaptation of NaPi-IIc was only detected in response to chronic administration. Both transporters were detected in UEV as well. UEV reflected the renal adaptation of the NaPi-IIa proteolytic fragment (but not the full-length protein) upon chronic but not acute dietary changes, while also reproducing the chronic regulation of NaPi-IIc. Thus, the composition of UEV reflects only partially changes in the expression of NaPi-IIa and NaPi-IIc at the BBM triggered by dietary Pi.

Keywords: Phosphate; Slc34a1; Slc34a3; Urinary extracellular vesicles.

Fig. 1

Urinary and plasma levels of Pi and Ca²⁺ correlate with dietary Pi content. (a) Plasma concentration of Pi, (b) urinary excretion of Pi, (c) fractional excretion of Pi, (d) plasma concentration of Ca²⁺, (e) urinary excretion of Ca²⁺, (f) fractional excretion of Ca²⁺, (g) plasma creatinine, (h) urinary excretion of creatinine and (i) urinary volume in samples from rats fed chronically (5 days) with low Pi (LL), acutely changed (12 h) from low to high Pi (LH), chronically fed with high Pi (HH) and acutely changed from high to low Pi (HL). Statistical significances were calculated with one-way ANOVA with Bonferroni’s multiple comparison test. n = 5 for each group, * P < 0.05, **P < 0.01, *** P < 0.001, **** P < 0.0001

Fig. 2

Plasma levels of Pi-regulating hormones differentially adapt to changes in dietary Pi. Plasma levels of (a) intact FGF-23, (b) intact PTH and (c) 1,25(OH)₂ vitamin D₃ in samples from rats fed chronically (5 days) with low Pi (LL), acutely changed (12 h) from low to high Pi (LH), chronically fed with high Pi (HH) and acutely changed from high to low Pi (HL). Statistical significances were calculated with one-way ANOVA with Bonferroni’s multiple comparison test. n = 5 for each group, **** P < 0.0001

Fig. 3

The abundance of the proteolytic fragment of NaPi-IIa in UEV partially correlates with its renal expression. Western blots for NaPi-IIa in renal BBM from rats fed (a) chronically (5 days) low Pi (LL) and acutely changed (12 h) from low to high Pi (LH), (b) chronically fed high Pi (HH) and acutely changed from high to low Pi (HL), (c) chronically fed low (LL) or high Pi (HH), and (d) UEV isolated from all 4 groups. Graphs show the quantifications of the full-length and proteolytic fragment normalized to (a–c) LiCor total protein stain (suppl Fig. 1), (d) urinary creatinine and TSG101. (a’–c’) real-time PCRs on renal RNA samples from the same groups. Statistical significances were calculated by t-test (a–c and a’–c’) or with one-way ANOVA with Bonferroni’s multiple comparison test (d). n = 5 for each group, *P < 0.05, **P < 0.01, *** P < 0.001, **** P < 0.0001. Red asterisks indicate changes in UEV similar to those described in renal BBM

Fig. 4

The abundance of NaPi-IIc in UEV partially correlates with its renal expression. Western blots for NaPi-IIc in renal BBM from rats fed (a) chronically (5 days) low Pi (LL) and acutely changed (12 h) from low to high Pi (LH), (b) chronically fed high Pi (HH) and acutely changed from high to low Pi (HL), (c) chronically fed low (LL) or high Pi (HH), and (d) UEV isolated from all groups. Graphs show the quantifications normalized to (a–c) LiCor total protein stain (supplementary Fig. 1), (d) urinary creatinine and TSG101. (a’–c’) real time PCRs on renal RNA samples from the same groups. Statistical significances were calculated by t-test (a–c and a’–c’) or with one-way ANOVA with Bonferroni’s multiple comparison test (d). n = 5 for each group, *P < 0.05, ** P < 0.01. Asterisks indicate changes in UEV similar to those described in renal BBM

Fig. 5

The abundance of AQP2 in UEV correlates with its renal expression. Western blots for AQP2 in renal BBM from rats fed (a) chronically (5 days) low Pi (LL) and acutely changed (12 h) from low to high Pi (LH), (b) chronically fed high Pi (HH) and acutely changed from high to low Pi (HL), (c) chronically fed low (LL) or high Pi (HH), and (d) UEV isolated from all 4 groups. Graphs show the quantifications of the unglycosylated and glycosylated channel normalized to (a–c) LiCor total protein stain (supplementary Fig. 2), (d) urinary creatinine and TSG101. Statistical significances were calculated by t-test (A–C) or with one-way ANOVA with Bonferroni’s multiple comparison test (D). n = 5 for each group, *P < 0.05