Human Molecular Genetics, 2021, Vol. 30, No. 15 1413–1428 https://doi.org/10.1093/hmg/ddab131 Advance Access Publication Date: 13 May 2021 General Article GENERAL ARTICLE Novel Dent disease 1 cellular models reveal biological processes underlying ClC-5 loss-of-function Mónica Durán1,† , Carla Burballa1,2,† , Gerard Cantero-Recasens1 , Cristian M. Butnaru2 , Vivek Malhotra2,3 , Gema Ariceta1,4 , Eduard Sarró1,‡,* and Anna Meseguer1,5,6,‡,* 1 Renal Physiopathology Group, Vall d’Hebron Research Institute (VHIR)-CIBBIM Nanomedicine, Barcelona, Spain, 2 Centre for Genomic Regulation, The Barcelona Institute of Science and Technology, Barcelona, Spain, 3 Institució Catalana de Recerca i Estudis Avançats, Barcelona, Spain, 4 Pediatric Nephrology Department, Vall d’Hebron University Hospital, Universitat Autònoma de Barcelona, Barcelona, Spain, 5 Departament de Bioquímica i Biologia Molecular, Unitat de Bioquímica de Medicina, Universitat Autònoma de Barcelona (UAB), Bellaterra, Spain and 6 Red de Investigación Renal (REDINREN), Instituto de Salud Carlos III-FEDER, Madrid, Spain *To whom correspondence should be addressed at: Passeig Vall d’Hebron 119-129, 08035, Barcelona, Spain. Tel: +34 934894061; Email: eduard.sarro@vhir.org (Eduard Sarró) and ana.meseguer@vhir.org (Anna Meseguer) Abstract Dent disease 1 (DD1) is a rare X-linked renal proximal tubulopathy characterized by low molecular weight proteinuria and variable degree of hypercalciuria, nephrocalcinosis and/or nephrolithiasis, progressing to chronic kidney disease. Although mutations in the electrogenic Cl− /H+ antiporter ClC-5, which impair endocytic uptake in proximal tubule cells, cause the disease, there is poor genotype–phenotype correlation and their contribution to proximal tubule dysfunction remains unclear. To further discover the mechanisms linking ClC-5 loss-of-function to proximal tubule dysfunction, we have generated novel DD1 cellular models depleted of ClC-5 and carrying ClC-5 mutants p.(Val523del), p.(Glu527Asp) and p.(Ile524Lys) using the human proximal tubule-derived RPTEC/TERT1 cell line. Our DD1 cellular models exhibit impaired albumin endocytosis, increased substrate adhesion and decreased collective migration, correlating with a less differentiated epithelial phenotype. Despite sharing functional features, these DD1 cell models exhibit different gene expression profiles, being p.(Val523del) ClC-5 the mutation showing the largest differences. Gene set enrichment analysis pointed to kidney development, anion homeostasis, organic acid transport, extracellular matrix organization and cell-migration biological processes as the most likely involved in DD1 pathophysiology. In conclusion, our results revealed the pathways linking ClC-5 mutations with tubular dysfunction and, importantly, provide new cellular models to further study DD1 pathophysiology. Introduction Dent disease 1 (DD1; OMIM #300009) is a rare X-linked renal variable degree and occurrence of hypercalciuria, nephrocalci- tubulopathy affecting about 330 families world-wide (1) and nosis, nephrolithiasis and hypophosphatemic rickets (2,3). DD1 characterized by low molecular weight proteinuria (LMWP), and progresses to renal failure between the third and fifth decades of † The first two authors should be regarded as joint first authors. ‡ Eduard Sarró and Anna Meseguer should be considered as joint senior authors. Received: February 8, 2021. Revised: April 23, 2021. Accepted: April 26, 2021 © The Author(s) 2021. Published by Oxford University Press. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted reuse, distribution, and reproduction in any medium, provided the original work is properly cited. 1413 1414 Human Molecular Genetics, 2021, Vol. 30, No. 15 life in 30–80% of affected males, while female carriers are usually generate a protein that reaches the plasma membrane and early asymptomatic (4). Progression to CKD in DD1 is associated with endosomes correctly, but shows reduced or abolished currents. decreasing calcium excretion in urine, which normalizes in 40% Yet, very little is known regarding how these mutations lead of affected adults < 30 years, and in 85% of those >30 years of age to specific disease manifestations. In this sense, the consider- (5). Lack of hypercalciuria, non-specific manifestations, overlap able intra-familial variability in disease severity and the lack of with glomerular disorders (6) and other genetic forms of renal genotype–phenotype correlation suggest that unknown mech- Fanconi syndrome with overt LMWP (7) may explain common anisms might be involved in PTCs dysfunction leading to DD1 underdiagnosis of DD1 which accurate prevalence is unknown. progression. In order to identify these ClC-5 mutation-associated There is no current curative treatment for DD1, nor clinical trials pathways, we have knock down the CLCN5 gene or introduced ongoing, and patient’s care is supportive, focusing on the treat- the ClC-5 mutations p.(Val523del) (not classified), p.(Glu527Asp) ment of hypercalciuria and the prevention of nephrolithiasis (8). (class 2) or p.(Ile524Lys) (class 1) in RPTEC/TERT1 cells, pro- DD1 is caused by loss-of-function mutations in the CLCN5 gene ducing specific cellular models carrying exogenous wild-type encoding the electrogenic 2 Cl− /H+ antiporter ClC-5, which is (WT) or mutant ClC-5 proteins. The parental cell line represents abundantly expressed in the epithelia of kidney and intestine, one of the most well-differentiated and stable human proximal though it is also expressed in brain, lung and, to a lesser extent, tubular cell line currently available, retaining sodium-dependent liver (9). In the human kidney, ClC-5 is mainly expressed in phosphate uptake and an intact functionality of the megalin– proximal tubule cells (PTCs), where it is predominantly located cubilin transport system (23,24). Gene expression profiling and in intracellular subapical endosomes and participates in endo- functional analysis in these newly created DD1 cells revealed the somal acidification (2,10). A small fraction of ClC-5 is also found biological processes (BP) related to proximal tubule dysfunction on the plasma membrane of PTCs, where it is proposed to in DD1, likely explaining phenotype variability of the disease and mediate plasma membrane chloride currents (10) or participate the progression to renal failure. in the macromolecular complexes responsible for low molecular weight (LMW) protein and albumin endocytosis (11). PTCs reabsorb ∼65% of filtered load and most of filtered LMW Results proteins mainly via receptor-mediated endocytosis (12). The Expression and subcellular localization of ClC-5 main actor in LMW protein reabsorption is the endocytic com- plex, which is comprised by the multiligand tandem receptors mutants p.(Val523del), p.(Glu527Asp) and p.(Ile524Lys) megalin and cubilin. Receptor-mediated endocytosis requires a in RPTEC/TERT1 cells continuous cycling of megalin and cubilin between the apical To explore the molecular mechanisms underlying PTCs dysfunc- plasma membrane, where they specifically bind ultrafiltrated tion in DD1, we first generated stable RPTEC/TERT1 cell lines LMW proteins and other ligands, and the early endosome, silenced for CLCN5 gene or carrying the pathogenic ClC-5 muta- where the receptors dissociate from their bound ligands (13). tions p.(Val523del), p.(Glu527Asp) or p.(Ile524Lys) (described in This process requires vesicular acidification for dissociating the DD1 patients (25–27)). A scheme summarizing the generation of ligand-receptor complex, recycling of receptors to the apical the RPTEC/TERT1 DD1 cell model and the localization of shRNA membrane and progression of ligands into the lysosomes. sequences and mutations in the corresponding ClC-5 protein Endosomal acidification is achieved by ATP-driven transport of is provided in Figures S1A and B, respectively. First, to fully cytosolic H+ through the vacuolar H+ -ATPase (14). Inactivating characterize these cell lines, RNA was extracted from 10-day mutations of CLCN5 in Dent disease patients (15) as well as the differentiated control (Ctrl shRNA), knockdown (CLCN5 shRNA), deletion of CLCN5 in knock-out (KO) mice (16,17) lead to severe and knockdown cells carrying the HA-tagged recombinant WT LMWP due to a defective endocytic uptake in PTCs, which has (rClC-5 WT), or mutant ClC-5 forms p.(Val523del), p.(Glu527Asp) been associated with the disappearance of megalin and cubilin and p.(Ile524Lys), to assess the levels of endogenous, exogenous at the brush border of PTCs. ClC-5 was initially postulated to (HA-tagged) and total CLCN5 levels by means of quantitative real- provide a Cl− shunt into the lumen of endosomes to dissipate V- time PCR (RT-qPCR). Our results showed that endogenous mRNA ATPase-mediated H+ accumulation, thereby enabling efficient levels of CLCN5 were strongly reduced in all cell lines transduced endosomal acidification (14,18). However, mutations in ClC-5 with the shRNA against CLCN5, compared with control cells (Ctrl causing Dent disease do not necessarily lead to a defective endo- shRNA): CLCN5 shRNA (6.2%, P < 0.0001 vs Ctrl shRNA), rClC-5 somal acidification (19), suggesting that the disease may result WT (16.1%, P < 0.0001 vs Ctrl shRNA), rClC-5 p.(Val523del) (4.4%, from an impaired exchange activity, namely uncoupling Cl− /H+ P < 0.0001 vs Ctrl shRNA), rClC-5 p.(Glu527Asp) (8.2%, P < 0.0001 co-transport and altered Cl− accumulation at early endosomes vs Ctrl shRNA) and rClC-5 p.(Ile524Lys) (13%, P < 0.0001 vs Ctrl (20). Thus, the precise molecular role of ClC-5 in endosomal shRNA) (Fig. 1A). Re-introduction of HA-tagged WT (rClC-5 WT) physiology and endocytosis, as well as several aspects of its ion and mutant (rClC-5 p.(Val523del), rClC-5 p.(Glu527Asp) and rClC- transport properties remain to be fully elucidated. 5 p.(Ile524Lys)) forms in former silenced cells restored CLCN5 To date, a total of 266 pathogenic variants of CLCN5 have been mRNA levels, as assessed with RT-qPCR probes against the HA reported consisting of nonsense, missense, splice site, insertion tag (Fig. 1B) and total CLCN5 (Supplementary Material, Fig. S1C), and deletion mutations (1,21). According to the latest reports, although mutants rClC-5 p.(Val523del) (33.8%, P < 0.0001 vs WT) CLCN5 mutations are grouped into three classes on the basis of and rClC-5 p.(Glu527Asp) (68.7%, P = 0.0372 vs rClC-5 WT), but functional data (19,21,22): class 1 mutations result in defective not rClC-5 p.(Ile524Lys) (87.9%, P = 0.7362 vs rClC-5 WT) exhibited protein processing and folding, thereby inducing retention of the lower exogenous CLCN5 levels than those of rClC-5 WT cells. At mutant protein in the endoplasmic reticulum (ER), where they the protein level, ClC-5 was detected as a lower band running at are early degraded by quality control systems; class 2 mutations 80–90 kDa and a higher diffuse band running as a smear at about impair protein processing and stability, leading to a function- 100 kDa, which was consistent with previous reports (28,29) ally defective protein lacking electric currents; these mutants (Fig. 1C). Loading equivalent amount of cell extract revealed that show reduced expression in the plasma membrane, but a nor- the protein levels of all the three ClC-5 mutants were strongly mal distribution in the early endosomes; and class 3 mutations reduced in comparison with rClC-5 WT (Fig. 1C). Human Molecular Genetics, 2021, Vol. 30, No. 15 1415 techniques. Our results show that rClC-5 WT localized at the plasma membrane (P.M.), early endosomes (E.E.) and endoplas- mic reticulum (E.R.), as shown by co-localization with the spe- cific subcellular compartment markers N-cadherin (P.M.), Rab-5 (E.E.) and KDEL (E.R.) (Fig. 2A, B and C). Quantification of the co- localization of ClC-5 forms with KDEL using the Manders’ over- lap coefficient (MOC) demonstrated that rClC-5 p.(Glu527Asp) (MOC = 0.29, P = 0.0256 vs rClC-5 WT) and rClC-5 p.(Ile524Lys) (MOC = 0.39, P = 0.0024 vs rClC-5 WT), but not rClC-5 p.(Val523del) (MOC = 0.17, P = 0.6005 vs rClC-5 WT) accumulated at the ER to a greater extent than rClC-5 WT (MOC = 0.10). Moreover, all three mutants showed a reduced co-localization with Rab5 (MOC rClC- 5 WT = 0.29, rClC-5 p.(Val523del) = 0.09 (P = 0.0002 vs rClC-5 WT), rClC-5 p.(Glu527Asp) = 0.16 (P = 0.0012 vs rClC-5 WT) and rClC- 5 p.(Ile524Lys) = 0.07 (P < 0.0001 vs rClC-5 WT)) and N-cadherin (MOC rClC-5 WT = 0.33, rClC-5 p.(Val523del) = 0.03 (P < 0.0001 vs rClC-5 WT), rClC-5 p.(Glu527Asp) = 0.02 (P < 0.0001 vs rClC-5 WT) and rClC-5 p.(Ile524Lys) = 0.01 (P < 0.0001 vs rClC-5 WT)) in com- parison with rClC-5 WT, indicating that their presence in E.E. and P.M. was reduced. These results confirmed, in the case of rClC-5 p.(Glu527Asp) and rClC-5 p.(Ile524Lys) subcellular localizations, previous results in HEK-MSR cells (19). ClC-5 p.(Ile524Lys) mutant, but not p.(Val523del) or p.(Glu527Asp) presents an altered glycosylation pattern It has been previously described that ClC-5 undergoes several post-translational modifications, including glycosylation (30). Moreover, mutations on ClC-5 N-glycosylation sites trigger poly-ubiquitination and proteasomal degradation (29,30). To gain more insight on whether the differences in protein levels and subcellular localization between rClC-5 WT and ClC-5 mutants could be related to impaired glycosylation processing, cell lysates from each of the conditions were treated with Endoglycosidase H (Endo H), which cleaves asparagine-linked mannose rich oligosaccharides, but not highly processed complex oligosaccharides, and Peptide:N-glycosidase F (PNGase F), which cleaves between the innermost GlcNAc and asparagine residues of high mannose, hybrid, and complex oligosaccha- rides. Our results show that rClC-5 WT and all mutant forms of ClC-5 were sensitive to PNGase F digestion, confirming that all of them were N-glycosylated (Fig. 3). On the other hand, only the lower migrating band of p.(Ile524Lys) mutant was sensible to EndoH digestion, as observed by a reduction in its molecular weight. These results suggest that this faster migrating band of Figure 1. Expression levels of ClC-5 mutants p.(Val523del), p.(Glu527Asp) and the p.(Ile524Lys) mutant might correspond to a core-glycosylated p.(Ile524Lys) in RPTEC/TERT1 cells. (A) mRNA levels of endogenous CLCN5 EndoH-sensitive form of the protein, correlating with the higher from control cells (Ctrl), CLCN5-silenced cells (CLCN5 shRNA), and CLCN5- degree of ER retention observed for this mutant. silenced cells carrying WT (rClC-5 WT) and mutant (rClC-5 p.(Val523del), rClC-5 p.(Glu527Asp) and rClC-5 p.(Ile524Lys)) ClC-5 were measured by RT-qPCR using To investigate whether the expression of ClC-5 mutants, and specific probes targeting the intact shRNA target sequence. (B) The expression more specifically, p.(Ile524Lys), could be inducing the unfolded levels of re-introduced WT (rClC-5 WT) and mutant (rClC-5 p.(Val523del), rClC- protein response (UPR) and ER stress as a result of their accumu- 5 p.(Glu527Asp) and rClC-5 p.(Ile524Lys)) ClC-5 were assessed by RT-qPCR using lation in the ER, we checked the phosphorylation of the ER stress specific probes against the HA-tag, which is only present in the exogenous ClC- marker PERK (31) (Supplementary Material, Fig. S2A) and the 5. (C) Protein levels of re-introduced WT and mutant ClC-5 were analyzed by cleavage of XBP-1 mRNA (31) (Supplementary Material, Fig. S2B) western blot using an antibody against the HA-tag. Tubulin was used to ensure that equal amounts of total cell extract were loaded in each lane. For all exper- in cells expressing WT or mutant ClC-5. Our results show that iments, Ctrl shRNA corresponds to cells transduced with both shRNA empty neither expression of rClC-5 WT nor any of the ClC-5 mutants vector and re-expression empty vector. Dots represent individual values and studied induced detectable levels of ER stress. columns indicate the mean ± SEM, n > 3. Statistical significance was determined using One-way ANOVA followed by Tukey’s multiple comparisons post hoc test. ∗ , P < 0.05; ∗∗ , P < 0.01. # P < 0.01 compared with Ctrl shRNA; & P < 0.01 compared CLCN5 silencing or re-introduction of p.(Val523del), with CLCN5 shRNA. p.(Glu527Asp) and p.(Ile524Lys) ClC-5 mutants impairs albumin endocytosis We next analyzed the subcellular localization of WT and To determine the effect of CLCN5 silencing and the selected ClC- mutant ClC-5 in RPTEC/TERT1 cells by using immunostaining 5 mutations on the endocytic capacity of RPTEC/TERT1 cells, we 1416 Human Molecular Genetics, 2021, Vol. 30, No. 15 Figure 2. p.(Val523del), p.(Glu527Asp) and p.(Ile524Lys) mutations alter the subcellular localization of ClC-5 in RPTEC/TERT1 cells. Subcellular localization of WT (rClC-5 WT) and mutant (rClC-5 p.(Val523del), rClC-5 p.(Glu527Asp) and rClC-5 p.(Ile524Lys)) ClC-5 was analyzed in RPTEC/TERT1 cells seeded on glass coverslips by determining their co-localization with the ER marker KDEL (A), early endosomes (EE) marker Rab-5 (B) and plasma membrane (PM) marker N-cadherin (C), using the corresponding antibodies. Cell nuclei were stained with Hoescht. Scale bars correspond to 5 μm. Quantification of the co-localization was performed using the Manders’ overlap coefficient (MOC). Dots represent individual values and columns indicate the mean ± SEM, n = 3 (minimum five cells/field). Statistical significance was determined using One-way ANOVA followed by Tukey’s multiple comparisons post hoc test. ∗ , P < 0.05; ∗∗ , P < 0.01. Abbreviations: E.R., endoplasmic reticulum; P.M., plasma membrane. analyzed Alexa Fluor 488-labelled albumin uptake. Detection of albumin, whereas the endocytic capacity was re-established labeled albumin within the cell boundaries, both in orthogonal when the WT ClC-5 was re-introduced into the silenced cells views and single planes, demonstrated that the endocytic (number of albumin Particles/cell Ctrl shRNA = 7.23, CLCN5 machinery functions properly in the RPTEC/TERT1 cell line shRNA = 2.27 (P = 0.0065 vs Ctrl shRNA) and rClC-5 WT = 6.57 (Fig. 4A). CLCN5 silencing strongly reduced the uptake of labelled (P = 0.6942 vs Ctrl shRNA) (Fig. 4A and B). By contrast, none of Human Molecular Genetics, 2021, Vol. 30, No. 15 1417 Figure 3. ClC-5 mutant p.(Ile524Lys), but not p.(Val523del) or p.(Glu527Asp), presents an altered glycosylation pattern. To explore the effects of selected ClC-5 mutations on the glycosylation pattern of ClC-5, total cell lysates from RPTEC/TERT1 cells carrying each of the mutations were treated with Endoglycosidase H (E), which cleaves asparagine-linked mannose rich oligosaccharides, but not highly processed complex oligosaccharides, and Peptide:N-glycosidase F (PNGase F) (P), which cleaves between the innermost GlcNAc and asparagine residues of high mannose, hybrid, and complex oligosaccharides. After glycosidase reactions, samples were analyzed by western blot using an anti HA-tag antibody. the ClC-5 mutants were able to restore this activity (number of from the 1563 genes altered by CLCN5 silencing, up to 452 albumin particles/cell rClC-5 p.(Val523del) = 3.15 (P = 0.0304 vs (from a total of 466 commonly regulated genes) were regulated Ctrl shRNA), rClC-5 p.(Glu527Asp) = 3.7 (P = 0.0350 vs Ctrl shRNA) in the opposite direction by CLCN5 re-introduction (Fig. 5C). and rClC-5 p.(Ile524Lys) = 2.58 (P = 0.0093 vs Ctrl shRNA)), what Among these, 237 (52%) genes were down-regulated and 215 indicates that these residues are essential for ClC-5-mediated (48%) were up-regulated by the effect of CLCN5 silencing. Only endocytosis (Fig. 4A and B). the genes that were commonly regulated by CLCN5 silencing and rClC-5 WT re-introduction (452 genes) were considered for subsequent analysis. To study in which BP were involved CLCN5 silencing or re-introduction of ClC-5 mutations these genes, we performed an analysis of over-represented gene p.(Val523del), p.(Glu527Asp) and p.(Ile524Lys) alter the ontology (GO) terms. A list of the BP GO terms significantly global gene expression profile of RPTEC/TERT1 cells enriched is shown in Figure 5D. We identified GO terms related In order to discover the putative mechanisms involved in prox- to (intersection size/term size): response to wounding (36/372), imal tubule dysfunction secondary to the loss of ClC-5, the matrix organization (27/397), anion homeostasis (5/65), cell gene expression profile of DD1 cell models was analyzed using adhesion (16/275), cell migration (29/569), positive regulation of DNA microarrays. To make the data comparable, as well as to reactive oxygen species (ROS) (12/107) and nephron development remove technical biases, microarray data were first normalized (13/146), among others. In anion homeostasis, significantly and batch effect corrected. The principal component analysis changed transcripts were SLC34A2, NR1H4, TFAP2B, SFRP4 and (PCA) shown in Figure 5A, indicated that samples from the same SLC7A11, and in nephron development were ADAMTS16, TFAP2B, experimental condition were aggregated. In addition, and to NOG, FMN1, LGR4, DLL1, COL4A4, KIF26B, SULF1, NID1, TACSTD2, assess the reliability of the microarray data, the expression levels PROM1 and BMP4. of EMX2, PTPRD, STEAP1, ZPLD1, CDH1, NR1H4 and EHF genes In order to explore the effects of the selected ClC-5 muta- were analyzed by qRT-PCR (Supplementary Material, Fig. S3). Val- tions on the transcriptome of RPTEC/TERT1 cells, gene expres- idation genes were selected among those that met the following sion profiles of p.(Val523del), p.(Glu527Asp) and p.(Ile524Lys) requirements: (i) altered expression between the WT condition mutants were compared to that of WT ClC-5. As shown in and any of the mutations studied, and (ii) modified expression by the heatmap in Figure 6A, p.(Val523del) and p.(Ile524Lys) were CLCN5 silencing and total or partial recovery upon ClC-5 WT re- the conditions that exhibited the greatest and the smallest introduction. Our results show that the validation gene selected differences in the gene expression profile, respectively, when panel presented an expression pattern consistent to that found compared to WT ClC-5. Moreover, only five genes (CHCHD7, in the DNA microarray (Supplementary Material, Fig. S3), thereby PREX1, PLAG1, LIN54 and ZRANB3) were commonly affected by confirming the trustworthiness of the microarray data. For com- all three mutations (Fig. 6B). Re-introduction of p.(Val523del) parative analysis, genes were defined as differentially expressed mutant altered the expression of 831 genes compared to ClC- genes (DEGs) if they presented an adjusted P value lower than 5 WT cells, from which 231 were down-regulated and 600 were 0.05 and a log2 Fold Change (logFC) higher or equal to 0.5 in any up-regulated. Roughly 20% (47/231) of the down- and 5% (27/600) of the comparisons studied. Top up- and down-regulated DEGs in of the up-regulated genes by Val523del expression were also each of the comparisons are shown in Supplementary Material, found down- or up-regulated, respectively, in CLCN5-silenced Tables S1–S8. cells, and might represent genes whose expression levels cannot The heatmap in Figure 5B graphically illustrates the differ- be restored by p.(Val523del) mutant to the levels achieved by ences in the gene expression profile of CLCN5-silenced cells WT ClC-5. On the other hand, genes altered by p.(Val523del) that in comparison with control cells and CLCN5-silenced cells were not affected in CLCN5-silenced cells could be indicating a with re-introduced WT ClC-5. These results show that CLCN5 gain of functionality of this mutant with regard to CLCN5 silenc- silencing elicited a marked effect on the gene expression ing. An analysis of the significantly enriched BPs (GO terms) profile of RPTEC/TERT1 cells, and that this effect was partially showed that p.(Val523del) altered the expression of genes mainly reversed by re-introducing the rClC-5 WT (Fig. 5B). Notably, related to DNA replication, but also related to carboxylic acid and 1418 Human Molecular Genetics, 2021, Vol. 30, No. 15 Figure 4. CLCN5 silencing or p.(Val523del), p.(Glu527Asp) and p.(Ile524Lys) ClC-5 mutations impair Albumin endocytosis. (A) Control, CLCN5 shRNA, rClC-5 WT, rClC-5 p.(Val523del), rClC-5 p.(Glu527Asp) and rClC-5 p.(Ile524Lys) cells were seeded on glass coverslips and incubated with 50 μg/mL Alexa Fluor 488-conjugated Albumin for 60 min. In the orthogonal view, dotted lines across the images demarcate the top surface of the cell. Scale bars correspond to 5 μm. (B) Quantification of albumin uptake was performed by measuring the number of albumin particles/cell. Dots represent individual values and columns indicate the mean ± SEM, N (number of experiments) > 3, n (number of cells) > 30. Outliers were removed using the Grubb’s test (alpha = 0.05). Statistical significance was determined using One-way ANOVA followed by Tukey’s multiple comparisons post hoc test. ∗ , P < 0.05; ∗∗ , P < 0.01. anion transport and renal system development, among others significantly enriched GO term in the over-representation anal- (Fig. 6C). ClC-5 mutant p.(Glu527Asp) altered the expression of ysis. Moreover, it might also account for the low number of 510 genes (204 down-regulated and 306 up-regulated). Among common genes found at the intersection of the three mutant them, 24% of the down- (49/204) and 16% (48/306) of the up- models. regulated genes were also found down- or up-regulated, respec- tively, by effect of CLCN5 silencing. Genes whose expression was CLCN5 silencing and ClC-5 mutations downregulate significantly altered by p.(Glu527Asp) in comparison to WT ClC- 5 were specially enriched in processes related to tissue remod- epithelial markers expression, increase cell eling and morphogenesis, but also in other processes including proliferation and impair cell-to-substrate adhesion and (intersection size/term size): sialic acid transport (3/7), sodium collective cell migration ion transmembrane transport (8/152) and also renal system We next explored whether the changes observed in the gene development (19/315), for example (Fig. 6D). Finally, p.(Ile524Lys) expression profiles by CLCN5 silencing or loss-of-function ClC-5 mutation altered the expression of only 32 genes com- ClC-5 mutations correlated with changes in the epithelial pared to ClC-5 WT, with 6 of them down-regulated and 26 up- characteristics (e.g. substrate adhesion or cell migration). For regulated. This low number of genes did not allow to obtain any this purpose, we analyzed epithelial markers’ levels (e.g. CDH1, Human Molecular Genetics, 2021, Vol. 30, No. 15 1419 Figure 5. CLCN5 silencing alters the global gene expression profile of RPTEC/TERT1 cells. To identify the potential mechanisms involved in the proximal tubule dysfunction secondary to the loss of ClC-5, we analyzed the gene expression profile of RPTEC/TERT1 cells carrying CLCN5 silencing or rClC-5 WT using DNA microarrays. (A) PCA obtained after normalization and batch effect corrections of the DNA microarray data. (B) Heatmap graphically illustrating the differences in the expression levels of the 1754 genes that were up- or down-regulated with a logFC > |0.5| and an adj.p.val < 0.05 in CLCN5-silenced cells (CLCN5 shRNA) in comparison to control cells (control shRNA) and CLCN5-silenced cells where WT ClC-5 was re-introduced (rClC-5 WT). (C) Venn diagrams depicting the genes commonly regulated by the effect of CLCN5 silencing and rClC-5 WT re-introduction. The table indicates the number of common genes that are up- or down-regulated in each comparison. (D) Analysis of over-represented GO terms BPs related to the 452 genes that were commonly and oppositely up- or down-regulated by the effect of CLCN5 silencing and rClC-5 WT re-introduction with a logFC > |0.5| and an adj.p.val < 0.05, using the GProfiler server. occludin, and Keratin-7 and -18), cell proliferation, substrate E-cadherin and keratin-7 expression (Fig. 7A). Our results also adhesion and collective cell migration. Results in Figure 7A show that, in comparison with control cells, CLCN5 silencing show that CLCN5 silencing strongly reduced E-Cadherin and increased cell-to-substrate adhesion (P < 0.0001) (Fig. 7B) and Keratin-7 levels, while it had no effect on occludin and Keratin- cell proliferation (P = 0.0193) (Fig. 7C), and reduced collective 18 levels. Re-introduction of rClC-5 WT, but not ClC-5 mutants cell migration (P = 0.0005) (Fig. 7D). On the other hand, all p.(Val523del), p.(Glu527Asp) or p.(Ile524Lys), totally rescued three ClC-5 mutants increased cell-to-substrate adhesion when 1420 Human Molecular Genetics, 2021, Vol. 30, No. 15 Figure 6. ClC-5 mutations p.(Val523del), p.(Glu527Asp) and p.(Ile524Lys) exert different effects on RPTEC/TERT1 gene expression profile. In order to explore the effects of the selected ClC-5 mutations on the transcriptome of RPTEC/TERT1 cells, gene expression profiles of p.(Val523del), p.(Glu527Asp) and p.(Ile524Lys) mutants were compared to that of WT ClC-5. (A) Heatmap showing the expression levels of the 1195 genes that were up-or down-regulated with a logFC > |0.5| and an adj.p.val < 0.05 in CLCN5-silenced cells where p.(Val523del), p.(Glu527Asp) or p.(Ile524Lys) ClC-5 was re-introduced in comparison to CLCN5-silenced cells where WT ClC-5 was re- introduced (rClC-5 WT). This results show that p.(Val523del) and p.(Ile524Lys) were the conditions that exhibited the greatest and the smallest differences in the gene expression profile, respectively, when compared to WT ClC-5. (B) Venn diagrams depicting the number of commonly regulated genes by ClC-5 mutations. Only five genes were shown to be commonly affected by all three mutations. GProfiler analysis of the significantly enriched BPs associated with the 831 genes up- or down-regulated with a logFC > |0.5| and adj.p.val < 0.05 by p.(Val523del) mutation (C), or associated with the 510 genes up- or down-regulated with a logFC > |0.5| and adj.p.val < 0.05 by p.(Glu527Asp) mutation (D), in comparison to ClC-5 WT. compared with the respective control rClC-5 WT (P < 0.0001 for (P = 0.0073 vs rClC-5 WT), but not p.(Val523del) (P = 0.9254 vs p.(Val523del) and p.(Glu527Asp) vs rClC-5 WT and P = 0.0018 for rClC-5 WT), induced a reduction in collective cell migration p.(Ile524Lys) vs rClC-5 WT) (Fig. 7E). In addition, ClC-5 mutations (Fig. 7G). Taken together, these results suggest that both CLCN5 p.(Val523del) and p.(Glu527Asp) increased cell proliferation silencing and expression of ClC-5 mutants could lead cells into a (P = 0.025 and P = 0.0277 vs rClC-5 WT, respectively), while dedifferentiated state, in agreement with some of the enriched ClC-5 mutation p.(Ile524Lys) did not (P = 0.8762 vs rClC-5 WT) GO BPs found in the previous section, including renal system (Fig. 7F). Finally, p.(Glu527Asp) (P = 0.0104 vs WT) and p.(Ile524Lys) development and tissue remodeling and morphogenesis. Human Molecular Genetics, 2021, Vol. 30, No. 15 1421 Figure 7. CLCN5 silencing and mutations p.(Val523del), p.(Glu527Asp) and p.(Ile524Lys) impair cell-to-substrate adhesion and collective cell migration. To explore whether the changes observed in the gene expression profiles by CLCN5 silencing or loss-of-function ClC-5 mutations correlated with changes in the epithelial characteristics, we analyzed substrate adhesion, proliferation and collective cell migration in RPTEC/TERT1 cells. (A) Epithelial markers CDH1, occludin, and Keratin- 7 and -18 were analyzed by western blot. (B and E) Cell-to-substrate adhesion as assessed by the ability of RPTEC/TERT1 cells to bind to tissue culture substrate. N (number of experiments) = 4, n (number of replicates) = 42. Statistical significance was determined using Mann-Whitney test, rank comparison (P < 0.01) for Ctrl shRNA vs CLCN5 shRNA comparison and non-parametric one-way ANOVA (Kruskal–Wallis test) followed by Dunn’s multiple comparison post hoc test for WT vs mutant ClC-5 comparisons. (C and F) Cell proliferation was measured by staining cells with CFSE and quantifying the fluorescence of the cells at the onset of the experiment and 4 days later. Statistical significance was analyzed using an unpaired t test for Ctrl shRNA vs CLCN5 shRNA comparison, and one-way ANOVA followed by Fisher’s LSD test for WT vs mutant ClC-5 comparisons. (D and G) collective cell migration, which depends on the integrity of cell–cell contacts, was determined with the wound healing migration assay as indicated in methods. Statistical significance was determined using an unpaired t test with equal SD (two-tail) for Ctrl vs CLCN5 shRNA (N = 4) comparison and a one-way ANOVA followed by Tukey’s multiple comparisons post hoc test for rClC-5 WT vs mutant (N ≥ 3) comparisons. For all experiments, dots represent individual values and columns indicate the mean ± SEM. ∗ , P < 0.05; ∗∗ , P < 0.01. 1422 Human Molecular Genetics, 2021, Vol. 30, No. 15 Discussion PTCs taken from CLCN5 KO mice had increased expression of proliferation markers and oxidative scavengers, suggesting that Besides reports describing the effects of ClC-5 mutations on PT dysfunction in CLCN5 KO mice was associated with oxidative chloride currents and endosomal acidification, there is a paucity stress, dedifferentiation and increased cell proliferation, which of studies addressing the impact of these mutations on the are in agreement with the results found in this study. Moreover, phenotype and expression profile of PTCs. Consequently, little the urinary proteome of patients with Dent disease has been is known about the putative mechanisms involved in proximal shown to be enriched with proteins actively participating in tubule dysfunction secondary to the loss of ClC-5. Moreover, interstitial matrix remodeling (34). Thereby, our results are evidences indicate that ClC-5 mutations causing defective aligned with the hypothesis raised by Devuyst and Luciani proximal tubular endocytosis and endosomal trafficking may (20) explaining the potential mechanism by which the loss not necessarily impair endosomal acidification, suggesting that of ClC-5 may cause proximal tubule dysfunction. According these processes may not be coupled and that tubular dysfunc- to the authors, and in addition to the impaired trafficking tion in DD1 may not result from reduced endosomal acidification and recycling of apical receptors and the defective receptor- (19,32). In this sense, one of the most striking conclusions of this mediated endocytosis, loss of ClC-5 would be also associated study is that despite the fact that lack of ClC-5 or the presence with altered lysosomal function. This might compromise the of ClC-5 mutations p.(Val523del), p.(Glu527Asp) and p.(Ile524Lys) lysosomal mediated-degradation and clearance of autophago- compromised the endocytic capacity of RPTEC/TERT1 in a somes containing ubiquitinated proteins and dysfunctional similar manner, they do not exert equivalent effects on the mitochondria, leading to excessive production of ROS. The gene expression profile of these cells, and only five genes were increase in ROS might alter the integrity of the junctional commonly modulated by these mutations. This suggests that complex proteins, releasing transcription factors, which will mutant forms of ClC-5 may alter different cellular processes translocate to the nucleus and promote proliferation. It is beyond the endocytic pathway. Accordingly, in this work we noteworthy that we did not detect a significant enrichment of aimed to discover the mechanisms that underlie tubule dysfunc- GO terms related to phagocytosis in neither CLCN5-silenced cells tion in DD1 by characterizing the phenotypical consequences of nor cells carrying p.(Val523del), p.(Glu527Asp) nor p.(Ile524Lys) ClC-5 mutations on PTCs. ClC-5 mutations, thereby suggesting the existence of other For this purpose, we have generated novel DD1 cell models molecular mechanisms converging on PTCs dedifferentia- by stably transfecting CLCN5 shRNA and pathogenic ClC-5 tion and dysfunction. In this sense, and in addition to the mutations into the RPTEC/TERT1 cell line, which maintains abovementioned GO terms, CLCN5 silencing also altered the many differentiation hallmarks (23). The use of established expression of genes related to BPs such as anion homeostasis, and well characterized cell lines takes advantage of a uniform chemotaxis or response to growth factor. Moreover, terms genetic background thus avoiding compensatory effects due such as morphogenesis of an epithelial bud and nephron to currently unknown polymorphisms or mutations in other development point at a role of ClC-5 in kidney development. genes that might be present in other primary cell models. It is worth mentioning that GO terms found in CLCN5-silenced In addition, use of cell models is of special interest in DD1, RPTEC/TERT1 cells including organ development, ion transport, where renal biopsy is not routinely indicated since (i) laboratory response to external stimulus, response to wounding, regulation findings and genetic testing can be sufficient for diagnosis of cell differentiation, chemotaxis and taxis were also found in and (ii) most DD1 patients are children and it is not ethical the gene microarray analysis from proximal S1 and S2 tubules to apply an invasive procedure if it cannot provide relevant of CLCN5 KO mice kidneys of the Guggino group (35), indicating information when only performed at one specific time point. that our human cell model mimics the PTCs of the mouse In this study, we have characterized the ClC-5 mutant proteins model. By contrast, we did not found terms related to lipid and the functionally of the cell models that we have produced, metabolism, which was the class with the greatest number of to further identify genes and BPs specifically regulated by CLCN5 changes in gene transcript level in the CLCN5 KO mice (35). That silencing or mutant forms of ClC-5. We chose to study these result was surprising because overall changes in lipids have not mutations because they are closely located within the ClC-5 P been reported in Dent disease, likely suggesting species-specific helix, which is involved in dimer interface formation, and exhibit related processes between mouse and human DD1 models. lack of ClC-5 activity; albeit that, they differentially affect ClC-5 When we analyzed the list of genes altered by CLCN5 silencing, functionality (19,25,26). p.(Ile524Lys) is a class I mutation and we found that among the most down-regulated genes there were abolished currents have been related to its retention in the ER genes that could be relevant to understand the pathophysiology (19). p.(Glu527Asp) is a type 2 ClC-5 mutant, and it lacks currents of DD1, at the proximal tubule level. Such an examples are despite its normal presence in the endosome compartment and secretory leukocyte peptidase inhibitor (SLPI; logFC -5.01), which partially (30%) reaching the plasma membrane when transfected has been related to PTCs regeneration (36), MUC1 (Mucin1; logFC in HEK-MSR cells (19). Expression of both p.(Glu527Asp) and -4.72), whose mutation causes a rare form of tubulointerstitial p.(Ile524Lys) mutants in HEK-293 cells also resulted in impaired fibrosis (37), SLC34A2 (sodium-dependent phosphate transport endosomal acidification and altered protein stability (19). protein 2B; logFC –4.01), which may contribute to the diminution p.(Val523del) effects on subcellular localization and endosomal in the uptake of both sodium and phosphate in the proximal acidification have not yet been described. In this regard, we tubules in Dent disease patients, the Rab GTPase RAB27B (logFC found that RPTEC/TERT1 cells lacking CLCN5 showed a marked –1.50), which is involved in exosome secretion, COL4A4 (Collagen loss of epithelial markers, increased cell to substrate adhesion, Type IV Alpha 4 Chain; logFC –0.92), which is mutated in reduced collective cell migration and a trend suggesting an patients with Alport syndrome, the kidney-specific Cadherin increase in cell proliferation, all of them characteristics of CDH16 (logFC –2.25), which is involved in cell–cell adhesions epithelial dedifferentiation. This correlated with an altered or KLF4 (Kruppel-like factor 4; logFC –0.99), which has been expression of genes related to ROS, cell–cell adhesion, cell identified as a renal linage master regulatory transcription migration, extracellular matrix organization or cell motility factor (38). Taken together, these results suggest that lack of among others. In the same direction, Gally et al. (33) found that ClC-5 widely affects the phenotype of RPTEC/TERT1 cells, but it Human Molecular Genetics, 2021, Vol. 30, No. 15 1423 remains to be known whether lack of ClC-5 impacts on these as the two most up-regulated genes in this condition, as well processes through its effect on endosomal acidification, altered as SLC27A2 (Fatty Acid Transporter FATP2), SLC16A4 (Monocar- chloride transport, protein endocytosis, or its participation in boxylate Transporter 4 MCT4) and SLC4A4 (Sodium Bicarbonate macromolecular complexes. For instance, endocytic trafficking Cotransporter NBC1), all of them being involved in renal diseases contributes to cell adhesion and migration in different ways (42). To cite some of them related to proximal tubule dysfunction, (39). First, internalization of chemokines by scavenger receptors SLC17A1 mutations cause dysregulation of urate handling by the is essential for sensing the chemotactic gradients, whereas proximal tubule cells (43) and SLC4A4 mutations cause defec- endocytosis and subsequent recycling of chemokine receptors is tive reabsorption of HCO3 − , leading to proximal tubule acidosis key for sustaining the responsiveness of migrating cells. Second, (44). By contrast, neither CLCN5-silenced cells nor any mutant endosomal pathways modulate adhesion by delivering integrins condition showed an altered expression of the sodium-bile acid to their site of action and supplying factors for focal adhesion cotransporter SLC10A2, which was one of the gene transcripts disassembly. Finally, endosomal transport also contributes to most increased in transcript number (17 fold) in the CLCN5 cell migration by delivering membrane type 1 matrix metallo- knockout mice proximal tubules of the Guggino group (35). How- protease to the leading edge facilitating proteolysis-dependent ever, we found that p.(Val523del) cells had a significant enrich- chemotaxis. ment of the BP ‘bile acid synthesis and transport,’, and genes One of the most striking results of the present work is the contained in this GO term, such as the bile acid receptor NR1H4, reduced number of BPs commonly altered by CLCN5 silenc- and the nuclear receptor NR1D1 or the very long-chain acyl-CoA ing and re-introduction of ClC-5 mutations, and also between synthetase SLC27A2, were also among the most up-regulated them, even though all conditions impaired albumin endocyto- genes in p.(Val523del) cells. Thus, the fact that p.(Val523del) ClC- sis and cell differentiation compared to control cells. Only the 5 up-regulates so many genes codifying for apical and baso- GO terms ‘extracellular matrix organization’ and ‘extracellu- lateral membrane co-transporters may suggest the existence lar structure organization’ were shared between CLCN5 silenc- of compensatory pathways to overcome the defective receptor- ing and p.(Val523del) and p.(Glu527Asp) conditions. Moreover, mediated endocytosis caused by ClC-5 loss-of-function. Besides although up to 154 genes were modulated by both p.(Val523del) that, and as mentioned above, renal development was another and p.(Glu527Asp), only the GO terms ‘urogenital system devel- BP altered in p.(Val523del) cells. Amid the genes belonging to this opment’ and ‘renal system development’ were jointly enriched BP, we found highly up-regulated (logFC 2.31) the transcription in these conditions. Interestingly, the BP ‘nephron development,’ factor HES1 (hairy and enhancer of split-1), previously identified which is a child term of the just mentioned GO terms, appeared as a renal linage master regulatory transcription factor, playing significantly overrepresented in the CLCN5 silencing condition. an important role in the Notch signaling pathway (38). As for This lack of complete functional equivalence between absence of p.(Val523del) down-regulated genes, it is remarkable to note that ClC-5 and presence of ClC-5 mutants in relation to the affected some of the top down-regulated genes, such as CDH1, MFAP5 or BPs, points to a gain of functionality of the mutated forms of ClC- LUM, participate in extracellular matrix organization and cell– 5. Thus, we hypothesize that, beyond endocytosis impairment, cell adhesion. This effect could in part explain the increased mutant proteins still exhibit a certain degree of functionality cell-to-substrate adhesion of the cells carrying the p.(Val523del) that is fully lost in CLCN5-silenced cells. In this direction, each mutation. Finally, it is also worth mentioning that cells carrying of the studied mutations might modify ClC-5 function in dis- p.(Val523del) mutation downregulate SLC3A1 gene, which cod- tinct ways, yielding to a different gene expression profile in the ifies for the amino acid transporter ATR1 that is mutated in corresponding cells. That is despite the close location of the patients affected of cystinuria (42). mutated residues in ClC-5’s helix P and the fact that amino As mentioned earlier, Glu527 is one of the most conserved acids Val523 and Glu527 are highly conserved in all known amino acids and is present in all the known ClCs, including ClCs (19,25,26). those from plants, yeast, Escherichia coli, cyanobacteria, fish and Interestingly, p.(Val523del) was the condition, among the dif- mammals (19,26). In addition, it has been previously described ferent mutants studied, with the largest differences in gene that the p.(Glu527Asp) ClC-5 mutant has a dominant negative profile when compared to the WT form. This could explain, effect on endosomal acidification (19), and mutation of the cor- in part, that p.(Val523del) ClC-5 mutation has been found in a responding residue in ClC-0 results in a reversion of voltage pediatric patient with a severe clinical phenotype (40), although dependence, i.e. currents were activated by hyperpolarization the lack of more individuals carrying the same mutation makes instead of depolarization (45). It is striking that, in our cell model, impossible to establish such a correlation. Genes altered by a large part of the BPs altered by the p.(Glu527Asp) mutation were p.(Val523del) ClC-5 mutation were mainly involved in cell cycle related to tissue remodeling, morphogenesis, differentiation and and proliferation, which are processes that have been linked to development. Interestingly, the BPs ‘Organ development’ and a dedifferentiation state, but also in carboxylic acid and anion ‘organ morphogenesis’ were the second and eighth GO terms, transport, and renal system development BPs. In this sense, cells respectively, with the greatest number of significantly changed carrying p.(Val523del) ClC-5 showed a non-significant increase transcripts in the CLCN5 KO mice of the Guggino group (35). Other in cell proliferation compared to rClC-5 WT cells, thereby sug- BPs, such as sialic acid transport, sodium ion transmembrane gesting that the p.(Val523del)-modulated genes included in pro- transport and, cell substrate adhesion, BMP signaling pathway liferation GO terms could indeed be mediating dedifferentia- or T cell activation appeared altered in RPTEC/TERT1 cells car- tion. On the other hand, we found altered an elevated num- rying p.(Glu527Asp) ClC-5. In this sense, genes related to T cell ber of genes from the solute carrier (SLC) group of membrane activation, such as genes of the human leukocyte antigen (HLA) transport proteins. SLC transporters show high expression lev- system (HLA-DRB1, logFC 1.79; HLA-DMA, logFC 1.55; and HLA- els in metabolically active organs such as the kidney, liver or DPA1, logFC 1.44) or the lymphocyte cytosolic protein 1 (LCP-1, brain (41), and the kidney has been identified as one of the logFC -1.46) and Thy-1 cell surface antigen (THY1, logFC -1.44) are target organs for most high expression of SLCs-mediated dis- among the most up or down regulated genes by p.(Glu527Asp) eases (42). For instance, we found the type I sodium-dependent re-introduction. These results are consistent with microarray phosphate transporters SLC17A1 (NPT1) and SLC17A3 (NPT4), data obtained from intestine of CLCN5 KO mice showing altered 1424 Human Molecular Genetics, 2021, Vol. 30, No. 15 expression of genes implicated in the immune system (46), and likely explaining the phenotypic variability of DD1 patients. In with the proposed role for ClC-5 in the immunopathogenesis of this sense, we found that BPs related to kidney development, ulcerative colitis (47). In addition, we also found biological path- anion homeostasis, organic acid transport, extracellular matrix ways related to bone remodeling. In this sense, it is well known organization and cell migration, were among the pathways that that the kidney is the major organ involved in the regulation more likely could explain the pathophysiology of DD1. of calcium and phosphate homeostasis, which is essential for bone mineralization and development (48). Thus, an enrichment Materials and Methods of such BPs could be indeed reflecting an altered phosphate regulation in PTCs. Moreover, alteration of these processes in Cell culture PTCs could be related to the increased bone turnover previously Renal proximal tubule epithelial cells RPTEC/TERT1 were described in the CLCN5 KO mouse model of Dent’s disease, likely obtained from the American Type Culture Collection (#CRL-4031, explaining the propension to altered bone homeostasis in young ATCC® ). RPTEC/TERT1 were cultured in Dulbecco’s Modified Dent’s patients (49). Eagle Medium: Nutrient Mixture F-12 (1:1, v/v) (#31331093, Curiously, introduction of the ClC-5 mutant p.(Ile524Lys) in Thermo Fisher Scientific) supplemented with 20 mM HEPES RPTEC/TERT1 cells only altered the expression of a reduced num- (#15630–080, Gibco), 60 nM sodium selenite (#S9133, Sigma- ber of genes, yielding, among the different mutants studied, the Aldrich), 5 μg/ml transferring (#T1428, Sigma-Aldrich), 50 nM gene expression profile that more closely resembled that of the dexamethasone (#D8893, Sigma-Aldrich), 100 U/ml penicillin rClC-5 WT condition. These results were unexpected considering and 100 μg/ml streptomycin (#15240–062, Gibco), 2% fetal bovine that (i) the p.(Ile524Lys) mutant presented the highest degree of serum (#10270, Gibco), 5 μg/ml insulin (#I9278, Sigma-Aldrich), ER localization among the different ClC-5 mutants studied, and 10 ng/ml epidermal growth factor (#E4127, Sigma-Aldrich) (ii) the lack of a correspondence between p.(Ile524Lys) mRNA and and 3 nM triiodothyronine (#T5516, Sigma-Aldrich). Cultures protein levels might indicate that this mutant exhibits reduced were maintained at 37◦ C in a 5% CO2 atmosphere. Unless protein stability or impaired post-translational processing. The otherwise indicated, cells were cultured for 10 days to allow increased ER retention of p.(Ile524Lys), however, does not trans- cell differentiation. Cells were routinely tested for mycoplasma late to an induction of the UPR, suggesting that p.(Ile524Lys) contamination. would be rapidly targeted to proteasomal degradation without displaying and ER stress gene signature. Proteasomal degrada- tion, in turn, could explain the reduced levels of p.(Ile524Lys) Gene silencing protein that we detected by western blot. A possible explanation For CLCN5 silencing, the MISSION® TRC shRNA transfer vector for the reduced phenotypic effects of p.(Ile524Lys) would be that, containing the ClC-5 shRNA target sequence CACCGAGA- although most of the proteins would be retained in the ER, a GATTACCAATAA (#TRCN0000043904, Sigma-Aldrich) was co- small but sufficient amount of the p.(Ile524Lys) proteins would transfected with the third generation vectors VSVG, RTR2 manage to escape from the ER and reach its functional localiza- and PKGPIR, which provide the envelope, packaging and tion. This hypothesis would imply that p.(Ile524Lys) is, beyond reverse-expressing proteins, respectively, into HEK-293T cells. its retention in the ER, a functional protein able to produce Supernatants containing viral particles were then harvested, chloride currents. The reduced number of genes modulated by supplemented with 10% FBS, 1% non-essential amino acids p.(Ile524Lys) did not allow to find statistically enriched GO terms. and 8 μg/mL polybrene (#TR-1003, Sigma-Aldrich) and added However, an analysis of the DEGs in p.(Ile524Lys) cells shed some to RPTEC/TERT1 cells, followed by antibiotic-mediated selection light about the potential processes altered by this mutation. In (8 μg/mL puromycin, #ant-pr, Invivogen). this sense, the top-down regulated genes, CDH1 (E-cadherin) and KRT7 (keratin-7), correspond to well-established epithelial mark- ers. This expression profile is aligned with the reduced collective Vectors and site-directed mutagenesis cell migration observed in rClC-5 p.(Ile524Lys) cells. Moreover, For shRNA rescue experiments, the canonical form (NM_000084.4; the third most down-regulated gene, CATSPER1, corresponds to mRNA variant 3) of WT human CLCN5 was cloned into a voltage-gated calcium channel, and the fifth, SLC38A8, to a pDONR vectors (pDONR™221, #12536–017, Invitrogen) using the putative sodium-dependent amino-acid/proton antiporter. On Gateway cloning system (Invitrogen). In order to escape from the other hand, PREX1, a guanine nucleotide exchange factor for degradation by the RISC complex, silent mutations (c.[99C > T; RAC1, and the ferroxidase enzyme Ceruloplasmin (CP) were the 100C > A; 102A > G; 105G > A; 108T > C; 111C > A; 114T > C]) most up-regulated genes by p.(Ile524Lys). PREX1, which is one of were introduced in the shRNA targeting sequence of human the few genes altered by all the three ClC-5 mutations studied, CLCN5. Over the shRNA-rescuing ClC-5 vector, we introduced has been identified as an important factor in tumor cell invasion the following mutations in the CLCN5 gene: p.(Val523del) and metastasis in a number of cancer models (50). As for CP, it (c.1566-1568del), p.(Glu527Asp) (c.1581A > T) and p.(Ile524Lys) has been described that it plays an important role in cellular iron (c.1571T > A). In addition, a HA-tag was also added in the homeostasis and could protect kidney against a damage from C-terminus of each cDNA. Subsequently, all inserts were iron excess (51). Interestingly, CP was also found up-regulated in sub-cloned to an expression vector containing hygromycin KO mice of the Guggino group, and the molecular function ‘iron resistance (pLenti CMV Hygro DEST 117–1, Addgene) using ion binding’ appeared in sixth position in the GO miner analysis the Gateway recombination system. All constructs generated of DEGs (35). were stably transduced into previously CLCN5-silenced cells In conclusion, in this work we have generated new cell mod- using lentiviral particles produced in HEK-293T cells and were els of Dent disease that accurately reproduce ClC-5 defects pre- subsequently selected with 400 μg/ml hygromycin (#ant-hg- viously described in CLCN5 KO mice and DD1 patients and we 5, Invivogen). Site-directed mutagenesis was performed with have demonstrated that, besides the established critical function the QuikChange® II XL Site-Directed Mutagenesis Kit (#200522, of ClC-5 in endocytosis, there are other mutation-associated Agilent Technologies) and primers were designed using the pathways that could be relevant for the etiopathogenesis of DD1, QuikChange Primer Design tool (Agilent Technologies). Human Molecular Genetics, 2021, Vol. 30, No. 15 1425 RNA extraction and qPCR at room temperature. Aldehyde groups were quenched in ® 50 mM NH4 Cl/PBS for 30 min and non-specific binding sites Total RNA was isolated from cells using TRIzol Reagent were blocked with 5% BSA in PBS for 60 min. Coverslips (#15596–026, Life Technologies) following the manufacturer’s were incubated overnight at 4◦ C with a 1:100 dilution with protocol, and 1000 ng of RNA in a final volume of 20 μl one of the following primary antibodies: HA (#867423001, was retro-transcribed using the High-Capacity cDNA Reverse Roche), KDEL (#Ab10C3, Abcam), Rab5 (#3547S, Cell Signaling) Transcription Kit (#4387406, Applied Biosystems). For each probe and N-Cadherin (#610920, BD Transduction Labs), followed to analyze, 20 ng of cDNA was amplified in a 10 μl final volume by incubation with the corresponding fluorescent-conjugated of reaction mix containing SYBR Green MasterMix (#A25742, secondary antibodies (1:500 dilution, #A11004, #A28175, #A11011, Applied Biosystems) or TaqMan MasterMix (#4369016, Applied #A27034, #A21247, and #A27012, Thermo Fisher Scientific) for 1 h Biosystems), according to the manufacturer’s instructions. at room temperature. Finally, cells were incubated with Hoechst Endogenous levels of CLCN5 mRNA were measured using 33342 (1:2000 dilution, #H1399, Invitrogen) for 5 min to stain cell specific probes targeting the intact shRNA target sequence nuclei. Coverslips were then mounted on slides with Prolong (Endogenous CLCN5 primers: 5 -GGGATAGGCACCGAGAGAT- diamond mounting medium (#P36961, Thermo Fisher Scientific) 3 and 5 -GGTTAAACCAGAATCCCCCTGT-3 ). Levels of re- and fluorescence labeling was visualized in a confocal spectral introduced (exogenous) forms of CLCN5 were measured using Zeiss LSM 980 microscope. Acquired images were processed specific probes against the C-terminal HA-tag, which is only using ImageJ software (52), and the co-localization analysis was present in re-introduced forms (Exogenous CLCN5 primers: 5 - performed using the ImageJ JACoP plugin (53). Co-localization GGTTACACACAACGGGCGAT-3 , and 5 -CGTAATCTGGAACATC was measured using Pearson’s and Manders co-localization GTA-3 ). Total levels of CLCN5 were measured with a commercial coefficients. TaqMan probe (#Hs00163986_m1, Applied Biosystems). Data were normalized against TBP using the following primers: 5 - CGGCTGTTTAACTTCGCTTC-3 and 5 -CAGACGCCAAGAAACAGT Glycosylation assay GA-3 , or the commercial TaqMan probe TBP Hs00427620_m1 RPTEC/TERT1 cells were lysed using RIPA buffer supplemented (Applied Biosystems). In order to validate the microarray with protease inhibitor cocktail (#P8340, Sigma-Aldrich) and analysis, the following TaqMan probes was used: (Applied equal amounts of whole cell extracts were digested for 18 h Biosystems): STEAP1 (Hs00185180_m1), ZPLD1 (Hs00604192_m1), with endoglycosidase H (Endo H, #P0702S, New England Biolabs) PTPRD (Hs00369913_m1), CDH1 (Hs01023895_m1), EMX2 (Hs00244 or peptide N-glycosidase F (PNGase F, #P0704S, New England 574_m1), NR1H4 (Hs01026590_m1), and EHF (Hs00171917_m1). Biolabs) enzymes following the manufacturer’s instructions. Analysis was performed using the 7900HT Sequence Detection System (Applied Biosystems). Relative expression fold change was determined by the comparative 2(−CT) method after Albumin uptake normalizing to TBP. For the analysis of XBP-1 splicing, we used Albumin uptake was measured to investigate receptor-mediated the following primers: 5 -AAACAGAGTAGCAGCGCAGACTGC-3 endocytosis. RPTEC/TERT1 cells were seeded on glass coverslips and 5 -TCCTTCTGGGTAGACCTCTGGGAG-3 . (Marlenfeld GmbH & Co. KG) and grown for 10 days. To measure albumin uptake, cells were exposed to 50 μg/mL Alexa Fluor Western blot 488-conjugated Albumin (#A13100, Thermo Fisher Scientific) for Cells were lysed in SET buffer (10 mM Tris–HCl pH 7.4, 150 mM 60 min. At the end of the incubation period, cells were washed NaCl, 1 mM EDTA and 1% SDS) and the protein concentration was six times with ice-cold PBS and fixed in −20◦ C methanol for quantified by the BCA assay (#23225, Thermo Fisher Scientific). 5 min. Instead of stripping the cells, non-specifically bound Equal amounts of whole cell extracts (ranging from 20 to albumin was excluded by using the confocal orthogonal view. 80 μg of protein depending on the antibody) were resolved To reveal cellular perimeter, slides were incubated 1 h at room in 10% SDS-PAGE gels and transferred to PVDF membranes temperature with a 1:5000 dilution of Phalloidin conjugated (#ISEQ00010, Millipore). Membranes were blocked with 5% non- to tetramethyl-rhodamine-isothiocyanate (TRITC, excitation/e- fat dry milk diluted in PBS-T (1x PBS, 0.1% Tween-20) for 1 h and mission: 540–545/570–573 nm; Sigma-Aldrich). Cell nuclei were incubated overnight at 4◦ C with the appropriated antibodies: stained with Hoechst 33342 (1:2000 dilution, #H1399, Invitrogen) HA (1:1000 dilution, #11867423001, Roche), β-tubulin (1:5000 for 5 min at room temperature and slides were mounted with dilution, #T4026, Sigma-Aldrich), E-Cadherin (1:1000 dilution, Prolong diamond mounting medium (#P36961, Thermo Fisher #610181, BD Transduction Labs), Occludin (1:1000 dilution, Scientific). Images were acquired with a 63X objective on a con- #71–1500, Invitrogen), PERK (1:1000 dilution, #5683 T, Cell focal laser scanning microscope (Zeiss LSM 980) and processed Signaling), Cytokeratin-7 (undiluted, #790–4462, Ventana Medical using ImageJ software. Co-localization analysis was performed Systems) and Cytokeratin-18 (1:1000 dilution, #51582, Santa using the ImageJ JACoP plugin. Cruz). Membranes were then incubated with the corresponding secondary antibodies (rabbit anti-mouse IgG/HRP, #P0260, Dako, DNA microarray and goat anti-rat IgG/HRP, #A9037, Sigma-Aldrich) at a 1:5000 dilution. Membranes were visualized using chemiluminescence Total RNA for DNA microarray was isolated as indicated before. reagent (#WBLUF0500, Millipore) and exposed on Odyssey Fc RNA quality was checked using Bioanalyzer nano assay (Agilent Imaging System (Li-Cor). Technologies). RNA samples representing five separate exper- iments from each of the conditions were used. Ten indepen- dent microarrays were performed using the Clariom D arrays Immunocytochemistry (#902922, Affymetrix-Genechip array) according the manufac- RPTEC/TERT1 cells were cultured on glass coverslips (#0111450, turer’s protocol. Briefly, 200 ng of total RNA was retro-transcribed Marlenfeld GmbH & Co. KG) for 10 days. Cells were then using WT pico HT Kit (#902622, Thermo Fisher Scientific). The washed in cold PBS and fixed in −20◦ C methanol for 5 min resulting cDNA was fragmented, labeled and hybridized using 1426 Human Molecular Genetics, 2021, Vol. 30, No. 15 the GeneAtlas hybridization, wash, and stain kit for WT array was removed leaving a defined 500 μm (according to the strips (#901667, Thermo Fisher), and finally, the Clariom D array manufacturer’s information) cell-free gap between the two com- was placed in the GeneTitan MC System for scanning (#00–0373, partments previously seeded with cells. Cells were then gently Affymetrix). washed twice with medium to remove cell debris, and gap clo- sure (wound healing) was monitored by obtaining digital images Over representation analysis of the cell-covered area at the onset of the experiments and 21 hours later with a Thunder microscope (Leica). Area measure- GO terms over-representation analysis was performed using ments at each time point were performed using ImageJ software. the webserver g:Profiler (https://biit.cs.ut.ee/gprofiler/gost) as described in (54). Statistical analysis XBP1 splicing assay All data sets were first tested for normality using the D’Agostino XBP-1 mRNA splicing was analyzed by PCR in untreated cells and Pearson omnibus normality test. Statistical significance or control cells treated with ER stress inducers Brefeldin-A was determined by Student’s t test (continuous data, two (1 μM, #B6542, Sigma-Aldrich) and Tunicamycin (5 μM, #T7765, groups) or one-way analysis of variance (ANOVA) (for the Sigma-Aldrich). RNA was then isolated and transcribed to comparison of more than two groups) followed by Turkey’s post cDNA as described previously. For each condition, one μl of hoc test. If data were not normally distributed, we used Mann– cDNA was amplified in a 25 μl final volume of reaction mix Whitney’s unpaired t test (two groups) or the non-parametric containing 2.5 mM MgCl2 , 0.2 mM dNTPs, 0.2 μM forward ANOVA (Kruskal–Wallis) (more than two groups) followed by primer (5 -AAACAGAGTAGCAGCGCAGACTGC-3 ), 0.2 μM reverse Dunn’s post hoc test. All statistical analyses were performed primer (5 -TCCTTCTGGGTAGACCTCTGGGAG-3 ), and 1 unit of using GraphPad Prism 6 software (RRID:SCR_002798). Values are Taq polymerase (Bioline, Meridian Bioscience). PCR conditions expressed as mean ± SEM. Criteria for a significant statistically were as follow: an initial melting step of 94◦ C for 2 min; then significant difference were: ∗ P < 0.05; ∗∗ P < 0.01. Each specific 30 cycles of 94◦ C for 30 s, an annealing step of 60◦ C for 30 s, test is indicated in figure legends. and 72◦ C for 30 s; followed by a final elongation step of 72◦ C for 5 min. Cyclophilin A was used as a housekeeping control gene (forward primer 5 -ATGGTCAACCCCACCGTGTTC-3 ; reverse Supplementary Material primer 5 -TTCGAGTTGTCCACAGTCAGCAAT-3 ). PCR products Supplementary Material is available at HMG online. were separated in 3% agarose gels. Cell proliferation Acknowledgements We thank the patient advocacy group Asdent (Asociación de la Cell proliferation was performed as previously described (55). Enfermedad de Dent, https://www.asdent.es) for its continuous Briefly, cells were incubated with 5 μM carboxyfluorescein suc- support. We thank all members of the Meseguer Lab for valuable cinimidyl ester (CFSE, #21888, Sigma-Aldrich) for 10 min at 37◦ C. discussions. Cell cytometry was carried out at the High Technol- The unbound CFSE was quenched by washing cells twice in ogy Unit at Vall d’Hebron Research Institute (VHIR). Fluorescence complete medium. An aliquot of cells was used to measure microscopy was performed at the Advanced Light Microscopy cell fluorescence at the onset of the experiment. The rest of Unit at the CRG, Barcelona; and at the High Technology Unit labeled cells were seeded on tissue plates and incubated at 37◦ C at VHIR. DNA microarrays were analyzed at the Statistics and until reaching a confluency of 40–60% (3–4 days). At the end of Bioinformatics Unit (UEB) at VHIR. V. Malhotra is an Institució this period, fluorescence of daughter cells was measured. Cell Catalana de Recerca i Estudis Avançats (ICREA) professor at fluorescence was measured on a FACS calibur flow cytometer the Centre for Genomic Regulation. This work reflects only the (Becton Dickinson) and proliferation indices were determined authors’ views, and the EU Community is not liable for any use using the Cell Quest software (Becton Dickenson). that may be made of the information contained therein. E.S. and M.D. were supported by the generous contribution of Asdent Cell adhesion Patients Association. C.B. was a recipient form the PhD4MD RPTEC/TERT1 cells cultured for 10 days were trypsinized, washed program from CRG and VHIR. twice with culture medium to eliminate trypsin and counted. Conflict of Interest statement. The authors have declared that no Fifty thousand cells/well were then seeded onto two duplicated conflict of interest exists. 96-well plates for 60 min at 37◦ C. After this period, unattached cells from one of the plates were removed by washing cells twice with PBS, followed by an additional incubation period of 60 min in medium to facilitate cell recovery. The amount Funding of attached cells (from the washed plate) and the total cells This work was supported in part by Asdent Patients Asso- (from the unwashed plate) was determined using XTT assay ciation and grants from Ministerio de Ciencia e Innovación (# 11465015001, Sigma-Aldrich) following the manufacturer’s (SAF201459945-R and SAF201789989-R to A.M.), the Fundación instructions. Senefro (SEN2019 to A.M.) and Red de Investigación Renal REDinREN (12/0021/0013). A.M. group holds the Quality Mention Wound healing migration assay from the Generalitat de Catalunya (2017 SGR). For wound healing migration assay (also termed scratch assay), 2.25 × 104 cells were seeded on each of the two compartments of silicone culture inserts (#81176, Ibidi) and grown for 10 days. Ethics Approval At the onset of the experiment, the silicone culture insert Compliance with ethical standards. Human Molecular Genetics, 2021, Vol. 30, No. 15 1427 Author Contributions the proton pump in endocytotically active kidney cells. Proc. Natl. Acad. Sci. U. S. A., 95, 8075–8080. E.S. and A.M. conceived and designed the research studies. M.D., 19. Smith, A.J., Reed, A.A.C., Loh, N.Y. et al. (2009) Charac- C.B1 ., G.C.R., C.B2 . and E.S. conducted the experiments. M.D., terization of dent’s disease mutations of CLC-5 reveals a C.B1 ., G.C.R., E.S. and A.M. analyzed the data. M.D., C.B1 ., G.C.R., correlation between functional and cell biological conse- V.M., G.A., E.S. and A.M. draft the manuscript. All authors read quences and protein structure. Am. J. Physiol. Ren. Physiol., and approved the final manuscript. 296, 390–397. 20. Devuyst, O. and Luciani, A. (2015) Chloride transporters and receptor-mediated endocytosis in the renal proximal tubule. Chloride transporters and receptor-mediated endocytosis in References the renal proximal tubule. J. Physiol. (2015), 593, 4151–4164. 1. Mansour-Hendili, L., Blanchard, A., Le Pottier, N. et al. (2015) 21. Gianesello, L., Ceol, M., Bertoldi, L. et al. (2020) Genetic analy- Mutation update of the CLCN5 gene responsible for Dent ses in dent disease and characterization of CLCN5 mutations disease 1. Hum. Mutat., 36, 743–752. in kidney biopsies. Int. J. Mol. Sci., 21, 1–20. 2. Gianesello, L., Del Prete, D., Ceol, M. et al. (2020) From protein 22. Lourdel, S., Grand, T., Burgos, J. et al. (2012) ClC-5 mutations uptake to Dent disease: an overview of the CLCN5 gene. Gene, associated with Dent’s disease: a major role of the dimer 747, 144662. interface. ClC-5 mutations associated with Dent’s disease: a 3. Shipman, K.E. and Weisz, O.A. (2020) Making a Dent in Dent major role of the dimer interface. Pflugers Arch. Eur. J. Physiol. disease. Function, 1, 1–9. (2012), 463, 247–256. 4. Scheinman, S.J. (1998) X-linked hypercalciuric nephrolithi- 23. Wieser, M., Stadler, G., Jennings, P. et al. (2008) hTERT alone asis: clinical syndromes and chloride channel mutations. immortalizes epithelial cells of renal proximal tubules with- Kidney Int., 53, 3–17. out changing their functional characteristics. Am. J. Physiol. 5. Blanchard, A., Curis, E., Guyon-Roger, T. et al. (2016) Observa- Physiol., 295, F1365–F1375. tions of a large Dent disease cohort. Kidney Int., 90, 430–439. 24. Aschauer, L., Carta, G., Vogelsang, N. et al. (2015) Expression of 6. Norden, A.G.W., Scheinman, S.J., Deschodt-Lanckman, M.M. xenobiotic transporters in the human renal proximal tubule et al. (2000) Tubular proteinuria defined by a study of Dent’s cell line RPTEC/TERT1. Toxicol. Vitr., 30, 95–105. (CLCN5 mutation) and other tubular diseases. Kidney Int., 57, 25. Wu, F., Reed, A.A.C., Williams, S.E. et al. (2009) Mutational 240–249. analysis of CLC-5, cofilin and CLC-4 in patients with dent’s 7. Lemaire, M. (2020) Novel Fanconi renotubular syndromes disease. Nephron - Physiol., 112, 53–62. provide insights in proximal tubule pathophysiology. Am. J. 26. Lloyd, S.E., Günther, W., Pearce, S.H.S. et al. (1997) Characteri- Physiol. Physiol., 2507, ajprenal., 00214, 2020. sation of renal chloride channel, CLCN5, mutations in hyper- 8. Devuyst, O. and Thakker, R.V. (2010) Dent’s disease. Orphanet calciuric nephrolithiasis (kidney stones) disorders. Hum. Mol. J. Rare Dis., 5, 28. Genet., 6, 1233–1239. 9. Steinmeyer, K., Schwappach, B., Bens, M. et al. (1995) Cloning 27. Takemura, T., Hino, S., Ikeda, M. et al. (2001) Identification and functional expression of rat CLC-5, a chloride channel of two novel mutations in the CLCN5 gene in Japanese related to kidney disease. J. Biol. Chem., 270, 31172–31177. patients with familial idiopathic low molecular weight pro- 10. Devuyst, O., Christie, P.T., Courtoy, P.J. et al. (1999) Intra-renal teinuria (Japanese Dent’s Disease). Am. J. Kidney Dis., 37, and subcellular distribution of the human chloride channel, 138–143. CLC-5, reveals a pathophysiological basis for Dent’s disease. 28. Grand, T., Mordasini, D., L’Hoste, S. et al. (2009) Novel CLCN5 Hum. Mol. Genet., 8, 247–257. mutations in patients with Dent’s disease result in altered 11. Hryciw, D. H., Ekberg, J., Ferguson, C. et al. (2006) Regulation ion currents or impaired exchanger processing. Kidney Int., of albumin endocytosis by PSD95/Dlg/ZO-1 (PDZ) scaffolds: 76, 999–1005. Interaction of Na+-H+ exchange regulatory factor-2 with 29. D’Antonio, C., Molinski, S., Ahmadi, S. et al. (2013) Con- ClC-5. J. Biol. Chem., 281, 16068–16077. formational defects underlie proteasomal degradation of 12. Zhuo, J.L. and Li, X.C. (2013) Proximal nephron. Compr. Phys- Dent’s disease-causing mutants of ClC-5. Biochem. J., 452, iol., 3, 1079–1123. 391–400. 13. Eshbach, M.L. and Weisz, O.A. (2017) Receptor-mediated 30. Schmieder, S., Bogliolo, S. and Ehrenfeld, J. (2007) N- endocytosis in the proximal tubule. Annu. Rev. Physiol., 79, glycosylation of the Xenopus laevis ClC-5 protein plays a role 425–448. in cell surface expression, affecting transport activity at the 14. Satoh, N., Suzuki, M., Nakamura, M. et al. (2017) Functional plasma membrane. J. Cell. Physiol., 210, 479–488. coupling of V-ATPase and CLC-5. World. J. Nephrol., 6, 14. 31. Oslowski, C.M. and Urano, F. (2011) Measuring ER stress 15. Scheinman, S.J., Guay-Woodford, L.M., Thakker, R.V. et al. and the unfolded protein response using mammalian tissue (1999) Genetic disorders of renal electrolyte transport. N. culture system. Methods Enzymol., 490, 71–92. Engl. J. Med., 340, 1177–1187. 32. Gorvin, C.M., Wilmer, M.J., Piret, S.E. et al. (2013) Receptor- 16. Piwon, N., Gunther, W., Schwake, M. et al. (2000) ClC-5 Cl– mediated endocytosis and endosomal acidification is channel disruption impairs endocytosis in a mouse model impaired in proximal tubule epithelial cells of Dent disease for Dent’s disease. Nature, 408, 369–373. patients. Proc. Natl. Acad. Sci. U. S. A., 110, 7014–7019. 17. Wang, S.S., Devuyst, O., Courtoy, P.J. et al. (2000) Mice lack- 33. Gailly, P., Jouret, F., Martin, D. et al. (2008) A novel renal ing renal chloride channel, CLC-5, are a model for Dent’s carbonic anhydrase type III plays a role in proximal tubule disease, a nephrolithiasis disorder associated with defec- dysfunction. Kidney Int., 74, 52–61. tive receptor-mediated endocytosis. Hum. Mol. Genet., 9, 34. Santucci, L., Candiano, G., Anglani, F. et al. (2016) Urine 2937–2945. proteome analysis in Dent’s disease shows high selective 18. Günther, W., Lüchow, A., Cluzeaud, F. et al. (1998) CIC-5, the changes potentially involved in chronic renal damage. J. chloride channel mutated in Dent’s disease, colocalizes with Proteome, 130, 26–32. 1428 Human Molecular Genetics, 2021, Vol. 30, No. 15 35. Wright, J., Morales, M.M., Sousa-Menzes, J. et al. (2008) mutations in several protein regions. J. Gen. Physiol., 110, Transcriptional adaptation to Clcn5 knockout in proximal 165–171. tubules of mouse kidney. Physiol. Genomics, 33, 341–354. 46. Maritzen, T., Rickheit, G., Schmitt, A. et al. (2006) Kidney- 36. Ochi, A., Chen, D., Schulte, W. et al. (2017) MIF-2/D-DT specific upregulation of vitamin D3 target genes in ClC-5 KO enhances proximal tubular cell regeneration through SLPI- mice. Kidney Int., 70, 79–87. and ATF4-dependent mechanisms. Am. J. Physiol. Ren. Physiol., 47. Alex, P., Ye, M., Zachos, N.C. et al. (2010) Clcn5 knockout 313, F767–F780. mice exhibit novel immunomodulatory effects and are more 37. LaFavers, K.A. and El-Achkar, T.M. (2020) Autosomal domi- susceptible to dextran sulfate sodium-induced colitis. J. nant tubulointerstitial kidney disease: a new tool to guide Immunol., 184, 3988–3996. genetic testing. Kidney Int., 98, 549–552. 48. Wei, K., Yin, Z. and Xie, Y. (2016) Roles of the kidney in the 38. Hiratsuka, K., Monkawa, T., Akiyama, T. et al. (2019) Induction formation, remodeling and repair of bone. J. Nephrol., 29, of human pluripotent stem cells into kidney tissues by 349–357. synthetic mRNAs encoding transcription factors. Sci. Rep., 9, 49. Silva, I.V., Cebotaru, V., Wang, H. et al. (2003) The ClC-5 1–13. knockout mouse model of Dent’s disease has renal hyper- 39. Maritzen, T., Schachtner, H. and Legler, D.F. (2015) On the calciuria and increased bone turnover. J. Bone Miner. Res., 18, move: endocytic trafficking in cell migration. Cell. Mol. Life 615–623. Sci., 72, 2119–2134. 50. Cook, D.R., Rossman, K.L. and Der, C.J. (2014) Rho guanine 40. Ramos-Trujillo, E., Claverie-Martin, F., Garcia-Nieto, V. et al. nucleotide exchange factors: regulators of Rho GTPase (2013) Dent’s disease: identification of seven new pathogenic activity in development and disease. Oncogene, 33, mutations in the CLCN5 gene. J. Pediatr. Genet., 2, 133–140. 4021–4035. 41. Schumann, T., König, J., Henke, C. et al. (2020) Solute car- 51. Jiang, B., Liu, G., Zheng, J. et al. (2016) Hephaestin and cerulo- rier transporters as potential targets for the treatment of plasmin facilitate iron metabolism in the mouse kidney. Sci. metabolic disease. Pharmacol. Rev., 72, 343–379. Rep., 6, 1–11. 42. Zhang, Y., Zhang, Y., Sun, K. et al. (2018) The SLC transporter 52. Schindelin, J., Arganda-Carrera, I., Frise, E. et al. (2009) Fiji - an in nutrient and metabolic sensing, regulation, and drug open platform for biological image analysis. Nat. Methods, 9. development. J. Mol. Cell Biol., 11, 1–13. 53. Bolte, S. and Cordelières, F.P. (2006) A guided tour into subcel- 43. Lederer, E. and Miyamoto, K.i. (2012) Clinical consequences lular colocalization analysis in light microscopy. J. Microsc., of mutations in sodium phosphate cotransporters. Clin. J. 224, 213–232. Am. Soc. Nephrol., 7, 1179–1187. 54. Reimand, J., Isserlin, R., Voisin, V. et al. (2019) Pathway 44. Suzuki, M., Vaisbich, M.H., Yamada, H. et al. (2008) Functional enrichment analysis and visualization of omics data using analysis of a novel missense NBC1 mutation and of other g:Profiler, GSEA, cytoscape and enrichment map. Nat. Protoc., mutations causing proximal renal tubular acidosis. Pflugers 14, 482–517. Arch. Eur. J. Physiol., 455, 583–593. 55. Sarró, E., Durán, M., Rico, A. et al. (2020) Cyclophilins A and B 45. Ludewig, U., Jentsch, T.J. and Pusch, M. (1997) Inward oppositely regulate renal tubular epithelial cell phenotype. rectification in ClC-0 chloride channels caused by J. Mol. Cell Biol., 34.