Targeted renal knockdown of Na⁺/H⁺ exchanger regulatory factor Sip1 produces uric acid nephrolithiasis in Drosophila

Abstract

Nephrolithiasis is one of the most common kidney diseases, with poorly understood pathophysiology, but experimental study has been hindered by lack of experimentally tractable models. Drosophila melanogaster is a useful model organism for renal diseases because of genetic and functional similarities of Malpighian (renal) tubules with the human kidney. Here, we demonstrated function of the sex-determining region Y protein-interacting protein-1 (Sip1) gene, an ortholog of human Na⁺/H⁺ exchanger regulatory factor (NHERF1), in Drosophila Malpighian tubules and its impact on nephrolithiasis. Abundant birefringent calculi were observed in Sip1 mutant flies, and the phenotype was also observed in renal stellate cell-specific RNA interference Sip1 knockdown in otherwise normal flies, confirming a renal etiology. This phenotype was abolished in rosy mutant flies (which model human xanthinuria) and by the xanthine oxidase inhibitor allopurinol, suggesting that the calculi were of uric acid. This was confirmed by direct biochemical assay for urate. Stones rapidly dissolved when the tubule was bathed in alkaline media, suggesting that Sip1 knockdown was acidifying the tubule. SIP1 was shown to collocate with Na⁺/H⁺ exchanger isoform 2 (NHE2) and with moesin in stellate cells. Knockdown of NHE2 specifically to the stellate cells also increased renal uric acid stone formation, and so a model was developed in which SIP1 normally regulates NHE2 activity and luminal pH, ultimately leading to uric acid stone formation. Drosophila renal tubules may thus offer a useful model for urate nephrolithiasis.

INTRODUCTION

Nephrolithiasis is a common renal disease, with a high and increasing prevalence rate (5% in women and 12% in men) (13) but with poorly understood etiology. Despite a large amount of investment in treatment, research, and medication of the disease worldwide (more than $5.3 billion/yr in the United States alone) (30), limited progress in the medical treatment of nephrolithiasis has been achieved in the last few decades (40, 54). The incidence of nephrolithiasis has been increasing in parallel with other epidemics such as cardiovascular disease and hypertension (21), depression (28), diabetes mellitus (71), and metabolic diseases (54). For example, the prevalence rate of uric acid is increasing globally, i.e., in the United States by >1%, in northern Europe by between 0.4 and 0.7%, and in southern Europe by >3% (60).

Although all the underlying causes behind the formation of kidney stones are not fully known, the literature suggests genetics as a crucial factor in susceptibility to some types of nephrolithiasis (4, 19) along with environmental and dietary factors (40, 43). There are two main models describing the role of genetics in kidney stone formation: the monogenic codominant model (26), in which a single gene actively accelerates stone accumulation (24), and the polygenic or heterogeneous coinheritance model (52, 63), in which two or more genes act coherently to each other to accelerate or inhibit stone accumulation (25, 51). Thus, basic research to determine the role of genes in stone formation with new animal models to investigate the pathophysiology of the disease may play a vital role in the advancement of the field, leading to new therapeutic agents for the management of kidney disease (36).

Approximately 70% of Drosophila melanogaster genes have human homologs, many of which are associated with kidney diseases (19). With its transparent renal system and powerful genetic technologies, Drosophila is an ideal system to study several different types of nephrolithiasis (2, 3, 10, 12, 19, 30, 40, 61, 72, 73). The Drosophila renal system comprises two pairs of Malpighian tubules (MTs), one anterior and one posterior (69), with critical roles in excretion and osmoregulation, functionally analogous to mammalian kidneys. MTs are composed of two major cell types, principal cells (PCs) and stellate cells (SCs), which are responsible for ion, water, and organic solute transport (5, 18, 40). MTs regulate body calcium, magnesium, potassium, phosphate, and carbonate levels, thereby influencing the formation of intraluminal stones (19).

Here, we demonstrated a novel role of the sex-determining region Y protein-interacting protein-1 (Sip1 or CG10939) gene, an ortholog of human Na⁺/H⁺ exchanger regulatory factor 1 (NHERF1) (32), in renal uric acid stone formation by selective knockdown of Sip1 in SCs. To identify the intraluminally accumulated stones, we performed physiological, chemical, pharmacological, and genetic analyses including the development of a chemical approach to quantify uric acid accumulation in MTs. Sip1, Moesin, and Na⁺/H⁺ exchanger isoform 2 (NHE2) were colocalized in wild-type, Sip1 mutant, and Moesin RNA interference (RNAi) flies, suggesting a model in which Sip1 regulates NHE2 to regulate luminal H⁺, resulting in a favorable environment for uric acid stone formation.

MATERIALS AND METHODS

Drosophila stocks.

D. melanogaster strains were reared at 22°C and 55% humidity with a 12:12-h light-dark photoperiod and standard cornmeal diet. The following strains were used: Canton-S (CS) as wild type, upstream activating sequence (UAS)-CG10939 RNAi (UAS-CG10939 RNAi, BDSC no. 65156), UAS-Moe RNAi (BDSC no. 31135), and rosy¹ (ry¹) mutant (BDSC no. 584) (42) from Bloomington Drosophila Stock Center (Bloomington, IN), and UAS-NHE2 RNAi (VDRC no. 106053) from the Vienna Drosophila Research Center (Vienna, Austria). UAS-Drosophila aquaporin labeled with enhanced yellow fluorescent protein (UAS-DRIP-eYFP) was as previously described by Cabrero et al. (8). UAS lines were driven by either capability receptor-galactose-responsive transcription factor GAL4 (CapaR-Gal4), specific to tubule PCs (59), or Cl⁻ channel-a-Gal4 (ClC-a-Gal4), specific to tubule SCs (8). The mutant Sip1^5a/CyO (49) line was a kind gift from Dr. Cédric Polesello (Toulouse University, Toulouse, France). Fly crosses were performed at 26°C to increase the efficiency of the GAL4/UAS binary system.

Dietary allopurinol assay.

Allopurinol [4-hydroxypyrazolo (3,4-d)pyrimidine, Sigma] was dissolved in standard Drosophila diet to make a final concentration of 250 ng/ml (75) and kept in vials at room temperature for 1 day. Five-to-seven-day-old adult flies were transferred in drug-containing vials and kept for 2 days before dissection and imaging steps. The following lines were fed with allopurinol: wild type, Sip1^(−/−), and ry^(−/−).

RNA preparation and quantitative RT-PCR.

Knockdown efficiency of the targeted gene relative to parental lines was assessed by quantitative RT-PCR. Tubules were dissected from 50 flies of the specified genotype, and RNA was isolated using RNeasy Mini Kit (Qiagen) following the manufacturer’s recommendations. cDNA was generated using the protocol as previously described elsewhere (7). Quantitative RT-PCR was performed using Opticon DNA Engine 4 (Bio-Rad) using Brilliant III Ultra-Fast SYBR Green QPCR Master Mix (Agilent) using the following primer sequences: Sip1 (CG10939), forward 5′-GCTGTTCGCTTTCGTTTCGTTTAG-3′ and reverse 5′-TGTCCTGGTTTCACCTTCTCCG-3′; NHE2 (CG9256), forward 5′-CACAATGTCCTGGCTGACCTTTC-3′ and reverse 5′-CTCCACCACCGAGAGATAAAACC-3′; and ribosomal protein L32 (Rpl32; CG7939), forward 5′-TGACCATCCGCCCAGCATAC-3′ and reverse 5′-ATCTCGCCGCAGTAAACGC-3′. The specificity of amplicons was verified with melting curve analysis, the messenger level was normalized using Rpl32 as an internal control gene, and the expression level was calculated using the ∆∆C_t method (where C_t is threshold cycle) (37).

Imaging and pH sensitivity of renal stones.

Adult flies (5–7 days old) were dissected in PBS (pH 5). Intact MTs were mounted on glass slides in PBS adjusted to pH 5–10, and MTs were immediately imaged using a microscope (Axioskop 2, Zeiss) under polarized light. As the visualization of the birefringent crystals is transitory, intact tubule samples, from wild-type flies and from the specified genotypes, were imaged immediately after dissection. Images were taken every minute for 30 min and were quantified once the timeframe was completed. Imaging conditions were maintained as previously described (11). Total stones present within the tubule at 0 min were considered 100%, and the stones that accumulated after 1, 10, 20, and 30 min were quantified with respect to the initial quantity.

Quantification of renal stones.

Quantification of the stones was achieved using ImageJ software as per a previously described protocol (11). Briefly, the tubular area of interest was outlined, and the pixel intensity was obtained. Any tubular pixel intensity above the threshold was considered as stones. The total area of stones in the lumen was calculated by subtracting background intensity.

Immunohistochemistry.

Immunostaining procedures were performed as previously described (8). Adult MTs were dissected in PBS and fixed with 4% (wt/vol) paraformaldehyde for 30 min at room temperature. The following primary antibodies were used: rabbit anti-NHE2 (short isoform, 1:300), rabbit anti-NHE2 (long isoform, 1:300) (15), and rabbit anti-moesin phosphorylated at Thr⁵⁵⁹ (anti-Moe-P) and rabbit anti-SIP1 (1:200) (53). Alexa Fluor 488/546 goat anti-rabbit antibody (ThermoFisher Scientific) was used in a dilution of 1:1,000 for visualization of the primary antiserum. Incubations in primary and secondary antibodies were performed overnight. Tubules were incubated with markers such as 4′,6-diamidino-2-phenylindole (DAPI; 1 µg/ml, Sigma-Aldrich) and/or rhodamine-Alexa Fluor 633-coupled phalloidin (1:100, ThermoFisher Scientific). All samples were mounted in Vectashield (Vector Laboratories), and images were taken using a confocal microscope (LSM 800, Zeiss) and processed with Zen software and Adobe Photoshop/Illustrator CS 5.1.

Uric acid colorimetric assay.

The total quantity of uric acid stones that accumulated in whole tubule homogenates of wild-type, Sip1 mutant, and Sip1/NHE2 knockdown flies was quantified using the QuantiChrom colorimetric uric acid assay kit (DIUA-250, BioAssay Systems) according to the manufacturer’s instructions. Six adult fly MTs per sample were homogenized in 12 μl of Tween 20 (Sigma-Aldrich), and 200 µl of working reagent were added to 5 μl of each tubule sample in 96-well plates (3 replicates for each sample). Samples were incubated for 30 min at room temperature, and optical density was measured at 590 nm using a Mithras LB940 automated 96-well plate reader (Berthold Technologies). Data were analyzed using MikroWin software.

Statistical analysis.

Data are presented as means ± SE. The significance of differences was assessed with a Student’s t-test (2-tailed) for unpaired samples or one-way ANOVA followed by Dunnett’s test, with significance taken as P < 0.05.

RESULTS

Mutation of Sip1 induces stone accumulation.

A previous mammalian study has shown that NHERF1 may play an important role in renal stone formation (35), so we determined the role of the Drosophila ortholog of NHERF1, Sip1, in mediating stone formation in Drosophila MTs. Immunocytochemical experiments using anti-SIP1 showed that SIP1 was indeed expressed in the wild-type fly kidney, but specifically in SCs, which are easily recognizable by their stellar shape (Fig. 1A). No immunostaining was detected in MTs of homozygous Sip1 mutants (Fig. 1A), thus confirming that the signal observed in wild type corresponds to SIP1 protein and that SIP1 expression is abolished in Sip1 homozygous mutants.

Fig. 1.

Mutation of sex-determining region Y protein-interacting protein-1 (Sip1) causes accumulation of stones intraluminally. A: SIP1 protein was specifically expressed in Malpighian tubule (MT) stellate cells (SCs) of wild-type (WT) flies, whereas no expression was detected in Sip1 mutant flies. Blue, DAPI; green, SIP1. PC, principal cell. Scale bars = 10 μm. B–G: representative polarized microscopy images of wild-type (WT) and Sip1^−/− mutant flies immediately after dissection (at time 0). Sip1^−/− MTs showed intraluminal accumulation of birefringent stones. Scale bars = 500 μm. H and I: bars represent the percentage of total stones in the anterior and posterior MTs of male (H) and female (I) Sip1 mutant flies compared with WT MTs. Bar diagrams were constructed by considering the accumulated stones at time 0 as 100%. Data are presented as means ± SE; n = 5 MTs. *P < 0.05 by one-way ANOVA followed by Dunnett’s test.

We next investigated the stone phenotype of MTs of Sip1 mutant flies. Mutation of Sip1 results in the formation of a very high number of small birefringent stones in the lumen of both male and female MTs compared with wild-type tubules (Fig. 1, B–G). Quantification of the mineralized area covered between 70% and 80% of both anterior and posterior tubule areas of male and female flies (Fig. 1, H and I). The anterior tubules have an enlarged initial segment (58, 69), which handles most of the organism’s excess calcium (20); however, this region did not develop birefringent stones in Sip1 mutants, and the stone burden was similar in anterior and posterior tubules (Fig. 1), suggesting that these calculi were not calcium based.

We next investigated whether renal, cell-specific knockdown of Sip1 resulted in the same phenotype. The UAS-Sip1 RNAi line produced a significant knockdown (>70%) in overall tubule expression of Sip1 when driven in SCs (ClC-a-Gal4>UAS-Sip1 RNAi; Fig. 2A). Specific silencing of the Sip1 gene in SCs showed a marked increase of birefringent stones compared with parental control lines (ClC-a-Gal4/+ and UAS-Sip1 RNAi/+; Fig. 2, B and C). However, no knockdown was observed when Sip1 RNAi was driven in PCs (Fig. 2D), suggesting that Sip1 is expressed uniquely in SCs. Accordingly, specific knockdown of the Sip1 gene in tubule PCs using the CapaR-Gal4 driver line resulted in unchanged stone quantity compared with controls (Fig. 2E), indicating a novel role of Sip1 in tubule SCs in mediating stone formation. Taken together, these results suggest that mutation of Sip1, and specific knockdown of Sip1 in SCs, promotes lithiasis.

Fig. 2.

Quantification of stones that accumulated in the lumen of sex-determining region Y protein-interacting protein-1 (Sip1) knockdown Malpighian tubules (MTs). A: expression of Sip1 was significantly decreased in Clc-a-Gal4>UAS-Sip1 MTs compared with parental lines ClC-a-Gal4/+ and UAS-Sip1 RNAi/+. B: representative polarized microscopy images of Clc-a-Gal4>UAS-Sip1 RNAi knockdown flies compared with parental controls. C: quantification of stones that accumulated in MTs under knockdown (stellate cell specific) and control conditions. D: expression of Sip1 showed no downregulation when specifically knocked down in principal cells (PCs). E: representative polarized images of MTs of PC-specific Sip1 knockdown flies (CapaR-Gal4>UAS-Sip1 RNAi) compared with parental lines CapaR-Gal4/+ and UAS-Sip1 RNAi/+. Data are presented as means ± SE; n = 5 MTs. Scale bars = 500 μm in B and E. CapaR, capability receptor; ClC-a, Cl⁻ channel-a; Gal4, galactose-responsive transcription factor GAL4; RNAi, RNA interference; UAS, upstream activating sequence; NS, nonsignificant. *P < 0.05 by one-way ANOVA followed by Dunnett’s test.

Modulation of pH affects stone solubility.

To determine the chemical nature of the intraluminally accumulated stones, Sip1 mutant tubules were incubated under acid or alkaline load by altering bathing pH between 5 and 10. At pH 5 and pH 6, no change in the quantity of stones after 30 min was noted. However, at pH 7, the total accumulated stones started to dissolve significantly within 20 min, and this process occurred faster with increased pH of the bathing solution, where 90% of the stones were dissolved within 10 min at pH 10 (Fig. 3A).

Fig. 3.

pH modulates solubility of Malpighian tubule (MT) stones. A: percentage of undissolved stones corresponding to the pH change of the bathing solution. B and C: pH (pH 6.6 and 6.7, respectively) at which stones start dissolving over a 30-min period. Data are expressed as means ± SE; n = 5 MTs. NS, nonsignificant. *P < 0.05 by one-way ANOVA followed by Dunnett’s test.

To precisely determine at which pH stones start dissolving, the pH of the bathing solution was altered by 0.1 pH unit ranging between pH 6 and pH 7. We showed that intraluminal stones start dissolving significantly at pH 6.7 and above (Fig. 3, B and C). Uric acid is a weak acid [pK_a 5.5 (38, 55)] that is relatively insoluble compared with its sodium salt (33, 70). Therefore, the stones that accumulated in Sip1 mutant MTs share similar chemical behavior with uric acid stones.

Inhibition of the function of xanthine oxidase leads to the disappearance of stones in Sip1 mutant tubules.

Uric acid is a product of purine metabolism (Fig. 4A). The pathway includes xanthine oxidase (XO), which is responsible for converting hypoxanthine to xanthine and converting xanthine to uric acid. Allopurinol inhibits the function of XO, thereby blocking the biosynthesis of uric acid (48) and causing a concomitant increase in hypoxanthine and xanthine concentration (47). Rosy (ry) is the second mutation discovered in D. melanogaster (19) and encodes the enzyme xanthine dehydrogenase/XO. Ry mutants closely recapitulate the symptoms of human xanthinuria type I (16, 65). In particular, both ry mutants and allopurinol-treated flies show elevated levels of hypoxanthine and xanthine and extremely low levels of urate and allantoin, as shown by metabolomic analysis (1, 34). Therefore, we studied the formation of calculi in both Sip1 and ry mutants under feeding treatment with allopurinol. On a standard diet, wild-type and ry flies (with no XO enzyme activity) did not produce uric acid stones (Fig. 4, B, D, and F). Furthermore, wild-type flies and Sip1 and ry mutants were fed an allopurinol-containing diet, leading to the disappearance of birefringent crystals in MTs (Fig. 4, C, E, and G), phenocopying the xanthine stone (ry flies; Fig. 4F). Thus, pharmacological inhibition of XO by dietary exposure to allopurinol led to the disappearance of stones, confirming that the intraluminally accumulated stones are uric acid stones.

Fig. 4.

Biochemical pathway for uric acid formation and blockade by allopurinol. A: uric acid biosynthesis pathway. Uric acid is the end product of purine metabolism catalyzed by different enzymes, including xanthine oxidase. B–G: representative images of Malpighian tubules from wild-type (WT) flies and sex-determining region Y protein-interacting protein-1 (Sip1) and rosy mutants on a normal or allopurinol diet. In all cases, flies fed with allopurinol did not accumulate stones. Arrowheads indicate the location of the lumen in the tubule. Scale bars = 500 μm in B–G.

Uric acid quantification in Sip1 knockdown tubules.

We next quantified the concentration of uric acid in MTs of wild-type, Sip1 mutant, and SC-specific Sip1 knockdown flies. The total concentration of uric acid in MTs of Sip1 mutant flies was 8.5-fold higher compared with wild-type flies (Fig. 5A). Similarly, in lines in which Sip1 RNAi is targeted to MT SCs (ClC-a-Gal4>UAS-Sip1 RNAi), a threefold increase in the quantity of uric acid compared with the parental controls was observed (ClC-a-Gal4/+ and UAS-Sip1 RNAi/+; Fig. 5B).

Fig. 5.

Concentration of uric acid in sex-determining region Y protein-interacting protein-1 (Sip1) mutant and knockdown flies. A: solubilized levels of uric acid in Sip1 mutant Malpighian tubules (MTs) were significantly higher compared with control [wild type (WT)] tubules. B: uric acid concentration was significantly higher in MTs of Sip1 knockdown flies (Clc-a-Gal4>UAS-Sip1 RNAi) compared with parental lines. Data are presented as means ± SE; n = 5 MTs. ClC-a, Cl⁻ channel-a; Gal4, galactose-responsive transcription factor GAL4; RNAi, RNA interference; UAS, upstream activating sequence. *P < 0.05 by Student’s t-test in A and one-way ANOVA followed by Dunnett’s test in B.

Taken together, these results unambiguously demonstrate the presence of uric acid stones within MTs of Sip1 mutant and Sip1 knockdown flies and that Sip1 gene expression in SCs mediates proper tubular lumen acidification.

Sip1 and Moesin localize to the apical membrane of tubule SCs.

Sip1 encodes a protein that functions as a scaffold linking the plasma membrane and cytoskeletal linker proteins encoded by Moesin (32), where SIP1 and moesin interact with each other to maintain epithelial integrity via phosphorylation (32, 50). We tested the colocalization of these proteins in MTs using polyclonal antibodies raised against both SIP1 and moesin. SIP1 immunostaining was detected exclusively in SCs of ClC-a-Gal4>UAS-DRIP-Venus MTs expressing DRIP-eYFP, a marker of apical membranes in SCs (Fig. 6A). An optical section made through one of the SCs clearly emphasizes that SIP1 and DRIP-eYFP colocalized to the luminal side of the nucleus (Fig. 6, B–D).

Fig. 6.

Sex-determining region Y protein-interacting protein-1 (SIP1) protein is expressed in the apical membrane of stellate cells (SCs). A: immunostaining of adult Malpighian tubules (MTs) using anti-SIP1 antibody in ClC-a-Gal4>UAS-DRIP-Venus expressing Drosophila aquaporin labeled with enhanced yellow fluorescent protein (DRIP-eYFP), a marker of apical membranes in SCs. B–D: cross section of a single SC showing colocalization between SIP1 and DRIP-eYFP in the apical membrane. E: expression of moesin protein in MT SCs. F: moesin was specifically expressed in the apical membrane of MT SCs of wild-type flies. G and H: no expression was seen in Moesin knockdown (KD) flies. Blue, DAPI; red, SIP1; green, moesin. Scale bars = 100 μm in A and E and 10 μm in B–D and F–H.

Moesin is known to participate with Crumbs in the development of apical basal polarity and to mark the apical domain of epithelia (39, 41). Immunostaining using anti-moesin antibody also showed specific labeling of moesin only in SCs (Fig. 6E), and a z-stack image revealed that the subcellular location was on the apical side of the plasma membrane (Fig. 6F). As expected, no immunostaining was observed in tubules from Moesin knockdown flies (Fig. 6, G and H), confirming the specificity of the antibody. These results confirm that SIP1 and moesin are both localized to the apical membrane in polarized epithelial SCs, which suggests a potential functional relationship.

SIP1 colocalizes with NHE2 and moesin in SCs.

The function of NHEs was first characterized in isolated cortical brush-border membrane vesicles showing Na⁺-driven H⁺ movement and H⁺-driven Na⁺ movement across the membrane (44). Furthermore, computational modeling of the hydrophobic-hydrophilic nature and predicted structure of NHEs has also shown interaction between NHEs and NHERF1/SIP1 (46). It is known that SIP1 is a scaffold protein required for the regulation of several transmembrane receptors and ion transporters (32, 62), so we hypothesized that SIP1 could regulate the activity of the alkali-metal/proton exchanger (NHE) protein family in Drosophila tubules.

NHEs play an important role in the transport of Na⁺ and H⁺ across the membrane (22) as well as in the maintenance of cellular and epithelial integrity. D. melanogaster has three NHE genes, NHE1, NHE2, and NHE3, which are expressed in multiple tissues (Supplemental Fig. S1; Supplemental Data for this article is available online at https://doi.org/10.6084/m9.figshare.8429165.v1) (22), but the NHE2 long isoform is SC specific (15). Interestingly, we found that the NHE2 long isoform also had clear localization in SCs of wild-type MTs, whereas NHE2 short isoform antibody labeled the apical membrane of tubule PCs (Supplemental Fig. S2). Furthermore, when moesin was specifically knocked down in MT SCs, no birefringent crystals were observed in Moesin RNAi lines (ClC-a-Gal4>UAS-Moe RNAi; data not shown) compared with parental controls (ClC-a-Gal4/+ and UAS-Moe RNAi/+), suggesting the absence of the role of moesin alone in the formation of uric acid stones. However, specific knockdown of NHE2 in SCs (∼70%) resulted in higher intraluminal accumulation of birefringent stones (∼30% of the tubule area; Fig. 7, B and C), and the solubilized uric acid levels quantified using the colorimetric assay were significantly increased compared with parental lines (Fig. 7D). Interestingly, no such phenotype was observed in PC-specific NHE2 knockdown (Fig. 7E). Taken together, these results demonstrate the roles of Sip1 and NHE2 in renal urate nephrolithiasis.

Fig. 7.

Silencing of Na⁺/H⁺ exchanger isoform 2 (NHE2) in stellate cells (SCs) of Malpighian tubules (MTs) causes accumulation of stones. A: expression of NHE2 was significantly decreased in knockdown flies (ClC-a-Gal4>UAS-NHE2 RNAi) compared with parental controls ClC-a-Gal4/+ and UAS-NHE2 RNAi/+. B: representative polarized microscopic images of NHE2 knockdown flies and parental controls. Scale bars = 500 μm. C: quantification of stones that accumulated in NHE2 knockdown MTs. D and E: quantification of the uric acid concentration in SC- and principal cell-specific NHE2 knockdown MTs, respectively. Data are presented as means ± SE; n = 5 MTs. CapaR, capability receptor; ClC-a, Cl⁻ channel-a; Gal4, galactose-responsive transcription factor GAL4; RNAi, RNA interference; UAS, upstream activating sequence; NS, nonsignificant. *P < 0.05 by one-way ANOVA followed by Dunnett’s test.

Consistent with the colocalization of SIP1, Moesin, and NHE2 proteins, we investigated a putative functional relationship between these proteins. To achieve this, we used an immunocytochemical approach using anti-NHE2 long isoform and anti-NHE2 short isoform rabbit polyclonal antibodies to stain tubules of Sip1 and Moesin mutant flies. Interestingly, no immunostaining using both NHE antibodies was observed in tubules from Sip1 and Moesin mutant flies, suggesting that SIP1, moesin, and NHE proteins are part of a scaffold linking the plasma membrane and cytoskeleton of tubule SCs (Fig. 8 and Supplemental Fig. S2).

Fig. 8.

Expression of the Na⁺/H⁺ exchanger 2 (NHE2) long isoform in wild-type (WT), sex-determining region Y protein-interacting protein-1 (Sip1) mutant, and Moesin knockdown (KD) flies. The NHE2 long isoform showed clear localization in the apical membrane of Malpighian tubule stellate cells (SCs). Thin bright lines represent nonspecific staining of the trachea, which is known to be sticky to antibodies. Blue, DAPI; green, NHE2 long isoform. PC, principal cell. Scale bars = 20 μm.

DISCUSSION

Mammalian NHERF1 was first characterized in the rabbit border membrane as an essential cofactor for cAMP inhibition of NHE (45, 66). Here, the role of the Drosophila ortholog of NHERF1, Sip1, in mediating uric acid stone formation in Drosophila MTs was characterized by biochemical, pharmacological, and genetic assays. Insects, like birds, are considered to have uricotelic excretory systems, in which waste nitrogen is dumped as uric acid, to conserve water, and so uric acid calculi are constitutive in most terrestrial insects (17). However, adult Drosophila tubules express very high levels of urate oxidase (uricase) (64), and so urate crystals are not normally abundant. In this context, the extreme accumulations observed here in Sip1 mutants are remarkable.

What mediates precipitation of uric acid stones in the tubule? In mammals, interaction between SIP1 and urate transporters has been suggested (14); our results suggest that SIP1 connects plasma membrane proteins such as NHE2 with members of the ezrin-radixin-moesin (ERM) family, thereby regulating lumen acidification (32, 62). In mammals, the ERM protein complex interacts with the plasma membrane and actin cytoskeleton (29, 67) within specific domains to systematize the plasma membrane (27) and thereby provide a regulated linkage between the plasma membrane and actin cytoskeleton. Recent genetic and biochemical studies have shown that NHERF1/Sip1 plays an essential role in the activation of ERM proteins in mammals (6) and also in D. melanogaster (32). Intriguingly, targeted deletion of NHERF1 in mice elevates intestinal deposition of calcium and also triggers calcium oxalate and uric acid crystal formation (57). However, loss of ERM proteins results in mislocalization of NHERF1 in mice (56). In D. melanogaster, Moesin is the sole representative of the ERM family (53). Sip1 promotes Moesin function by affecting interaction with Sterile20-like kinase (Slik) kinase; genetic and functional interactions between Sip1, Moesin, and Slik kinase have been shown in Drosophila pupae and cultured S2 cells (32). We demonstrated expression of SIP1 and moesin in MT SCs, potentially suggesting an interaction in SCs.

NHEs are integral membrane proteins that comprise multiple transmembrane domains and a large cytosolic COOH-terminal domain (74). A previous study in the mammalian model has shown that NHERF1 phosphorylates NHEs, thereby affecting their activity (9). Interestingly, rabbit NHERF1 is involved in the regulation of the renal brush-border NHEs (23). Also, computational modeling of the hydrophobic-hydrophilic nature and predicted structure of NHEs has shown an interaction between NHEs and NHERF1 (46). Supporting these previous findings, our immunocytochemical experiments revealed that NHE2 (long isoform) is localized to MT SCs but is not expressed in Sip1 and Moesin mutant MTs. Thus, all three proteins, SIP1, moesin, and NHEs, are localized specifically in SCs with potential functional interactions.

Collectively, our experimental results allow a model for the formation of uric acid stones in MTs of D. melanogaster (Fig. 9). Although our model does not allow us to distinguish uric acid stones from hyperuricosuria alone (rare), aciduria (very common), or both, we demonstrated that a common class of kidney stones can usefully be studied in the Drosophila renal system, where we can benefit from the uniquely powerful genetic interventions characteristic of this organism.

Fig. 9.

Model illustrating the role of sex-determining region Y protein-interacting protein-1 (SIP1) protein in uric acid stone formation in Drosophila Malpighian tubules (MTs). MTs contain two main cell types, principal cells (gray) and stellate cells (SCs, yellow), and the transport processes of these cells have been described elsewhere (5). In principal cells, apically localized V-type H⁺-ATPase energizes transepithelial secretion, providing electrogenic transport of H⁺ into the lumen, coupled with a cation/H⁺ antiporter. In SCs, Cl⁻ moves down an electrochemical gradient through Cl⁻ channels in SCs, and water follows by osmosis through water channels in SCs (7a). The apically located SIP1 interacts with Na⁺/H⁺ exchanger isofrom 2 (NHE2) and activates the efflux of Na⁺ and influx of H⁺. Mutation of Sip1 and NHE2 leads to accumulation of H⁺ intraluminally and tubular lumen acidification, mediating uric acid stone formation. NHA, Na⁺/H⁺ antiporter.

GRANTS

This work was supported by the European Union’s Horizon 2020 research and innovation program under Marie Skłodowska-Curie Grant Agreement 64293 RENALTRACT (to J. A. T. Dow and S. A. Davies), by UK Biotechnology and Biological Sciences Research Council Grant BB/L002647/1 (to S. A. Davies, J. A. T. Dow, and S. Terhzaz), and by National Institute of Diabetes and Digestive and Kidney Diseases Grants R01-DK-092408 and U54-DK-100227 (to J. A. T. Dow and M. F. Romero).

DISCLOSURES

No conflicts of interest, financial or otherwise, are declared by the authors.

AUTHOR CONTRIBUTIONS

S.G., S.T., P.C., M.F.R., S.A.D., and J.A.T.D. conceived and designed research; S.G., S.T., and P.C. performed experiments; S.G. and P.C. analyzed data; S.G., S.T., P.C., M.F.R., and J.A.T.D. interpreted results of experiments; S.G. prepared figures; S.G. and S.T. drafted manuscript; S.G., S.T., P.C., M.F.R., S.A.D., and J.A.T.D. edited and revised manuscript; S.G., S.T., P.C., M.F.R., S.A.D., and J.A.T.D. approved final version of manuscript.