The Renal Micro-RNA Expression Profile of Pkd1^RC/RC Mice Changes Longitudinally with the Progression of the Disease

Visual Abstract

Abstract

Key Points

• Experimental autosomal dominant polycystic kidney disease is associated with longitudinal changes in the kidney micro-RNA and mRNA profiles.

• Alterations in cellular transport, a feature of CKD, and a common biological process within polycystic kidney disease models and human autosomal dominant polycystic kidney disease, is present from early stages.

• Early versus late micro-RNA expression changes in autosomal dominant polycystic kidney disease could be targeted for a more tailored therapeutic intervention in the disease.

Background

Autosomal dominant polycystic kidney disease (ADPKD) is characterized by the progressive development and enlargement of bilateral kidney cysts, which often leads to kidney failure. This study comprehensively characterized the expression profile of micro-RNAs (miRNAs) and their target genes in Pkd1^RC/RC mice and further compared them with other murine models of polycystic kidney disease (PKD) and a model of CKD and individuals with ADPKD.

Methods

Pkd1^RC/RC and wild-type (WT) mice (n=10 each, five males and five females) were studied at 1, 6, and 12 months. At each time point, kidney volume was determined in vivo (magnetic resonance imaging), followed by ex vivo histomorphometric analyses. In randomly selected Pkd1^RC/RC and WT mouse kidneys (n=5/genotype, at 1 and 12 months), miRNA-sequencing (seq) was performed, followed by mRNA-seq, integrated (miRNA-seq/mRNA-seq analysis), and functional analysis of miRNA target genes. Venn diagrams were constructed to identify overlapping and novel differentially expressed (DE) miRNAs in Pkd1^RC/RC and other commonly used murine models of PKD, diabetic kidney disease, and individuals with ADPKD.

Results

miRNA-seq analysis identified 41 and 181 miRNAs DE in Pkd1^RC/RC versus WT kidneys at 1 and 12 months, respectively, which were confirmed by quantitative PCR. Target genes of miRNAs DE in Pkd1^RC/RC at early stages encoded for proteins mainly implicated in cell proliferation and α-ketoglutarate (α-KG) transport, whereas those DE at late stages encoded for transport proteins involved in proinflammatory and metabolic processes. Urine α-KG concentration (¹H nuclear magnetic resonance spectroscopy), its fractional excretion, and tissue levels were higher in Pkd1^RC/RC during the entire course of the disease and associated with decreased protein expression of α-KG transporters sodium-coupled dicarboxylate transporter 3 and organic anion transporter 1. We further identified common DE miRNAs among murine models of PKD, diabetic kidney disease, and previous reports in individuals with ADPKD, as well as several novel miRNAs DE in early and late Pkd1^RC/RC kidneys, which have not been previously reported in either other murine models of PKD or patients with ADPKD.

Conclusions

Our study demonstrates that the renal miRNA expression profile changes longitudinally with the progression of the disease and might suggest that the post-transcriptional regulation of α-KG transport could represent a novel early feature of the disease.

Introduction

Autosomal dominant polycystic kidney disease (ADPKD) is the fourth leading cause of kidney failure (KF) among adults.¹ Nevertheless, substantial challenges remain to be addressed in several aspects of the disease. In 2018, the vasopressin receptor antagonist tolvaptan was approved for adults at risk of rapidly progressing ADPKD.² However, although tolvaptan slows kidney function decline, it does not stop disease progression, and treated patients may experience aquaresis-related adverse events and, in rare cases, hepatotoxicity.^3–6 Therefore, additional interventions are needed to halt or slow disease progression with acceptable risk and side-effect profiles.

Another critical problem is assessing ADPKD severity, follow-up, and evaluation of treatment response at early stages because of the extensive phenotypic variability, the natural course of the disease, and limitations in currently available biomarkers. In ADPKD, total kidney volume (TKV) increases exponentially and directly correlates with GFR decline,⁷ but it takes time to observe changes and does not provide mechanistic disease insights. Therefore, early, real-time biomarkers related to the underlying molecular mechanisms might detect renal injury before permanent anatomical damage occurs.

ADPKD is characterized by the progressive development and enlargement of bilateral renal cysts, which gradually compress and injure neighboring structures, triggering inflammation and interstitial fibrosis, leading to KF in most cases.⁸ However, the mechanisms regulating these processes, their contribution to disease progression, and the timing of transitions remain largely unclear. Micro-RNAs (miRNAs) regulate several biological processes implicated in the pathogenesis of ADPKD. For example, the miR-17 miRNA family has been reported to promote cyst epithelial cell proliferation and metabolic reprogramming,⁹ whereas miR-21 has been shown to promote disease progression by suppressing cyst epithelial apoptosis and facilitating cyst growth.¹⁰ Recently, miR-214 has been implicated in modulating the cyst microenvironment by attenuating inflammation and disease progression.¹¹ While these studies suggest an important role of miRNAs in the pathogenesis of ADPKD, the timing of the miRNA expression changes across the course of the disease remains to be clarified.

Circulating levels of several miRNAs differ between patients with ADPKD and non-CKD controls and among patients on the basis of their CKD stage.^12,13 Studies from human urinary exosomes and kidney tissue specimens identified a few miRNAs that correlated with eGFR and kidney length, further suggesting their potential to track disease progression.¹⁴ However, these studies have been mainly cross-sectional and limited to adult patients with more advanced disease. Changes that occur early on are more likely to be disease-specific and result from the primary insult rather than a consequence of CKD and sustained damage to the kidney. Therefore, miRNAs that change early on have the potential to provide ideal disease-specific biomarkers that could be used longitudinally from the early stages of ADPKD.

In this study, we (1) determined the longitudinal changes in miRNA expression patterns accompanying the dynamics of the kidney phenotype of Pkd1^RC/RC mice, (2) singled out early and late miRNA expression changes and determined the biological functions of their differentially expressed (DE) targeted genes, and (3) compared the DE miRNA profiles of early and late stages of Pkd1^RC/RC mice with other murine models of polycystic kidney disease (PKD) and a model of CKD (diabetic kidney disease [DKD]), as well as with the profile of individuals with ADPKD.

Methods

Detailed materials and methods are provided in the Supplemental Material.

Results

Animal Characteristics and Laboratory Parameters

The study protocol involved the inclusion of male and female wild-type (WT) mice inbred in a mixed background (F1 progeny of 129S6/SvEvTac [129] WT and C57BL/6J [B6] WT) and male and female Pkd1^RC/RC mice (a mouse model mimicking the human PKD1 variant, p.Arg3277Cys [RC] that develops slowly progressive disease and has preserved renal function for at least 1 year^15,16) in the same mixed background. We chose to perform our studies in a mixed background because this combination balances moderate disease progression with minimal variability between animals, in contrast to using homogeneous 129 or B6 backgrounds.¹⁶

The main characteristics of WT and Pkd1^RC/RC mice are shown in Supplemental Figure 1 and Supplemental Table 1. Body weight was similar between WT and Pkd1^RC/RC at all time points. Percent total kidney weight to body weight (TKW/BW), body length-adjusted TKV (Supplemental Figure 1A), and cystic area (Supplemental Figure 1B) were higher in Pkd1^RC/RC versus WT since month 1. Fibrotic area (Supplemental Figure 1C), plasma creatinine, and BUN did not differ between Pkd1^RC/RC and WT until 12 months. Urine output was similar between Pkd1^RC/RC and WT at all time points. In Pkd1^RC/RC, TKW/BW, body length-adjusted TKV, and cystic area increased from 1 to 6 months and remained higher at 12 months. However, TKW/BW did not increase between 6 and 12 months, consistent with previous reports in Pkd1^RC/RC mice.¹⁶ Fibrotic area and BUN did not change from 1 to 6 months, but increased from 6 to 12 months, consistent with the progression of the disease in patients with ADPKD.^7,8,17

Differential miRNA Expression Analysis of Pkd1^RC/RC versus WT at Early and Late Disease Stages

To visualize early and late miRNA expression changes in Pkd1^RC/RC versus WT kidneys, we performed DE analysis at 1 and 12 months. At 1 month, we identified 28 miRNAs upregulated and 13 downregulated in Pkd1^RC/RC versus WT and 119 miRNAs upregulated and 62 downregulated at 12 months (Figure 1A and Supplemental Figure 2).

Figure 1.

Early and advanced changes in the kidney miRNA profile of Pkd1^RC/RC mice versus WT. (A) Volcano plots showing 28 miRNAs upregulated and 13 downregulated in Pkd1^RC/RC compared with WT kidneys at 1 month and 119 miRNAs upregulated and 62 downregulated in Pkd1^RC/RC compared with WT kidneys at 12 months (n=5 each). The vertical axis (y axis) corresponds to the −log₂ (P value), and the horizontal axis (x axis) displays the log two-fold change (Pkd1^RC/RC/WT) value. Upregulated and downregulated miRNAs are shown as red and blue dots, respectively. By contrast, nonsignificant genes are shown as gray dots (P value ≤ 0.05 and fold changes >2 are indicated by black dashed lines). (B) Venn diagrams showing miRNAs upregulated (left), and downregulated (right) in Pkd1^RC/RC compared with WT kidneys at 1 and 12 months, and miRNAs commonly upregulated (left) and downregulated (right) at 1 and 12 months. miRNA, micro-RNA; WT, wild-type.

To gain insights into the specific changes associated with early and late stages of the disease, we performed Venn diagram analysis of DE miRNAs at 1 and 12 months. We identified that 9 and 100 miRNAs were exclusively upregulated in Pkd1^RC/RC versus WT at 1 and 12 months, respectively, and 19 miRNAs were commonly upregulated in Pkd1^RC/RC at 1 and 12 months (Figure 1B). Contrarily, eight and 57 miRNAs were exclusively downregulated in Pkd1^RC/RC at 1 and 12 months, respectively, and five miRNAs were commonly downregulated in Pkd1^RC/RC at 1 and 12 months. Unlike previous studies,^9,18 expression of miR-17∼92 miRNA cluster members did not differ between Pkd1^RC/RC and WT (Supplemental Figure 3), likely because of different animal models and backgrounds, time points, and criteria to identify DE miRNAs. Expression of miR-21a-5p, which was similar between the groups at 1 month, was higher in Pkd1^RC/RC versus WT at 12 months (Supplemental Figure 4).

To validate our miRNA-sequencing (seq) analysis, we performed quantitative PCR on randomly selected top upregulated and downregulated miRNAs. Expression of candidate miRNAs followed the same patterns as the miRNA-seq findings (Supplemental Figure 5).

Integrated miRNA-Seq/mRNA-Seq Analysis of Pkd1^RC/RC versus WT Mice at Early and Late Disease Stages

To single out early versus late miRNA expression changes and determine the biological processes of their DE-targeted genes, we classified miRNAs DE at 1 or 12 months or both based on their expression patterns, which rendered six different groups: group 1, including miRNAs upregulated only at 1 month (n=9); group 2 (miRNAs downregulated only at 1 month, n=8); group 3 (miRNAs upregulated at 1 and 12 months, n=19); group 4 (miRNAs downregulated at 1 and 12 months, n=5), group 5 (miRNAs upregulated only at 12 months, n=100); and group 6 (miRNAs downregulated only at 12 months, n=57; Figure 2). None of the DE miRNAs had an opposite direction between 1 and 12 months.

Figure 2.

Groups of DE miRNAs at early and advanced stages of the disease. Expression pattern of miRNAs DE in Pkd1^RC/RC compared with WT kidneys at 1 and 12 months. Upregulated and downregulated miRNAs at each time point are shown in red and blue, respectively, while nonsignificant miRNAs are shown in gray. miRNAs were classified into six groups: miRNAs upregulated only at 1 month (group 1), miRNAs downregulated only at 1 month (group 2), miRNAs upregulated at 1 and 12 months (group 3), miRNAs downregulated at 1 and 12 months (group 4), miRNAs upregulated only at 12 months (group 5), and miRNAs downregulated only at 12 months (group 6). *miRNAs not mapped in our miRNA-seq at that specific time point. DE, differentially expressed; miRNA-seq, micro-RNA sequencing.

Next, we performed integrated miRNA-seq/mRNA-seq analysis to investigate whether target genes of DE miRNAs in each group followed opposite expression patterns compared with their regulatory miRNAs. We found that miRNAs in group 1 are capable of targeting 6687 genes (16,104 miRNA-mRNA interactions, Supplemental Data File 5), of which 18 genes (Supplemental Table 2) followed the opposite direction of their miRNAs (downregulated in Pkd1^RC/RC at 1 month; Supplemental Figure 6). miRNAs in group 2 can target 9005 genes (26,284 interactions Supplemental Data File 6), of which 66 genes (Supplemental Table 3) were upregulated in Pkd1^RC/RC only at 1 month (Supplemental Figure 7). miRNAs in group 3 can target 11,872 genes (55,115 interactions Supplemental Data File 7), of which 11 genes (Supplemental Table 4) were downregulated in Pkd1^RC/RC at 1 and 12 months (Supplemental Figure 8). miRNAs in group 4 can target 4261 genes (8290 interactions Supplemental Data File 8), of which 164 genes (Supplemental Table 5) were upregulated in Pkd1^RC/RC at 1 and 12 months (Supplemental Figure 9). miRNAs in group 5 can target 16,716 genes (274,214 interactions Supplemental Data File 9), of which 135 genes (Supplemental Table 6) were downregulated in Pkd1^RC/RC only at 12 months (Supplemental Figure 10). miRNAs in group 6 can target 15,954 genes (192,888 interactions Supplemental Data File 10), of which 1039 genes (Supplemental Table 7) were upregulated in Pkd1^RC/RC only at 12 months (Supplemental Figure 11).

Functional Analysis Characterizes the Top Biological Processes of Target Genes of miRNAs DE in Pkd1^RC/RC Kidneys at Different Stages of the Disease

Target genes of miRNAs downregulated in Pkd1^RC/RC from early stages of the disease (groups 2 and 4) encoded for cell surface, chromosome, and cytoskeletal proteins (Supplemental Figure 12A) with binding, transcription factor, and ATP-dependent activity (Supplemental Figure 13A), mainly implicated in cell cycle, adhesion, and division (Figure 3A). By contrast, target genes of miRNAs upregulated in Pkd1^RC/RC from early stages of the disease (groups 1 and 3) encoded for plasma membrane proteins (Supplemental Figure 12B) with transporter activity (Supplemental Figure 13B), implicated in α-ketoglutarate (α-KG) transport (Figure 3B).

Figure 3.

Functional analysis of target genes of miRNAs DE in Pkd1^RC/RC at early and advanced disease stages. (A) Functional annotation clustering analysis showing the top 20 biological processes associated with target genes of miRNAs downregulated (groups 2 and 4) and (B) upregulated (groups 1 and 3) in Pkd1^RC/RC compared with WT kidneys at early disease stage. (C) Functional annotation clustering analysis showing the top 20 biological processes associated with target genes of miRNAs downregulated (group 5) and (D) upregulated (group 6) in Pkd1^RC/RC compared with WT kidneys at the advanced disease stage. k is the number of genes in the intersection of the query set with a set from MSigDB. GSEA, gene set enrichment analysis; MSigDB, Molecular Signatures Database.

Target genes of miRNAs downregulated in Pkd1^RC/RC at late stages of the disease (group 6) encoded for cell surface and side membrane proteins (Supplemental Figure 12C) with signaling receptor binding and regulator activity (Supplemental Figure 13C), primarily involved in several immune cell processes, including proinflammatory cytokine production and lymphocyte activation (Figure 3C). Contrarily, target genes of miRNAs upregulated in Pkd1^RC/RC at late stages of the disease (group 5) encoded for endoplasmic reticulum and plasma membrane proteins (Supplemental Figure 12D) with transporter activity (Supplemental Figure 13D), linked to transmembrane transport and lipid and amino acid metabolic processes (Figure 3D).

Altogether, our analyses suggest that genes likely regulated by miRNAs from the early stages of the disease encode proteins implicated in cell proliferation and α-KG transport, whereas those DE at late stages of the disease encode cell surface and transport proteins involved in proinflammatory, immune-mediated, and metabolic processes.

Urine and Kidney Tissue Analyses Confirm Alterations in α-KG Levels Associated with Decreased Sodium-Coupled Dicarboxylate Transporter 3 and Organic Anion Transporter 1 Expression in Pkd1^RC/RC from Early Disease Stages

Our functional analyses identified α-KG transport as a prominent biological process dysregulated from the early stages. This was associated with downregulation of Solute Carrier Family 13 Member 3 (slc13a3), which encodes sodium-coupled dicarboxylate transporter 3 (NaDC3), and Solute Carrier Family 22 Member 6 (slc22a6), which encodes organic anion transporter 1 (Oat1), both genes being targets of miRNAs upregulated early on. To validate this, we determined protein expression of NaDC3 and Oat1 at 1 and 12 months. We found that expression of both was downregulated (Supplemental Figure 14) from early stages.

NaDC3 and Oat1, a sodium-coupled dicarboxylate transporter and an organic anion (OA)/dicarboxylate exchanger,^19–22 are both expressed on the basolateral membrane of proximal tubular cells and act in concert with the apical OA transporter to regulate renal secretion of OA and therapeutic drugs by taking up α-KG from the blood (Figure 4A).^23–25 Overall, along with the apical sodium-coupled dicarboxylate transporter 1, which uptakes α-KG from urine,²⁶ these regulate α-KG levels in plasma, urine, and tissue, important for maintaining proper cellular metabolism, as well as OA and drug detoxification.^26,27 Interestingly, diseases associated with SLC13A3 variants present with increased urinary excretion of α-KG.^28,29

Figure 4.

Urine, plasma, and kidney tissue α-KG concentration and its FE in Pkd1^RC/RC and WT mouse through the disease. (A) Diagram of α-KG metabolism. (B) Quantification of α-KG concentration in urine, (C) plasma, (D) its FE, and (E) concentration in kidney tissue by ¹H NMR in Pkd1^RC/RC (red dots) and WT (blue dots) at 1, 6, and 12 months (n=10/genotype/time point). Urine α-KG concentration is adjusted by mole of creatinine, and kidney tissue concentration is adjusted by g of wet tissue. ***P < 0.001; ****P < 0.0001. α-KG, α-ketoglutarate; FE, fractional excretion; Gln, glutamine; Glu, glutamate; GLS, glutaminase; GS, glutamine synthetase; GDH, glutamate dehydrogenase; GTs, glutamine transaminases; IDH3, isocitrate dehydrogenase 3; NaDC3, sodium-coupled dicarboxylate transporter 3; NMR, nuclear magnetic resonance; OA⁻, organic anion; OAT1, organic anion/dicarboxylate exchanger 1; TCA, tricarboxylic acid.

To determine whether the functional predictions and decreased expression in these transporters were associated with alterations in α-KG levels, we identified and quantified urine, plasma, and tissue α-KG concentration at different stages of the disease. We found that α-KG concentration was higher in Pkd1^RC/RC urine versus WT since the early stages of the disease and remained higher through its entire course (Figure 4B). Higher urine levels of α-KG may result from increased filtration, decreased reabsorption, or increased production by the kidney. Glomerular hyperfiltration is a common early feature of children and adolescents with ADPKD, although the GFR eventually declines as the disease progresses.^30,31 In this model, plasma creatinine and BUN levels were not different until the 12-month timepoint (Supplemental Table 1). We then calculated plasma α-KG concentration, followed by its fractional excretion (FE α-KG), with the equation: (Urine [α-KG]×Plasma [creatinine]/Plasma [α-KG]×Urine [creatinine]×100). Interestingly, plasma α-KG concentration was not different at any point, but the FE α-KG was higher throughout the disease (Figure 4, C and D). Finally, we determined kidney tissue α-KG concentration, which was higher in Pkd1^RC/RC versus WT kidneys from the early stages of the disease (Figure 4E). Taken together, these data suggest that the increased urine levels of α-KG are independent of the filtration function and systemic production of the metabolite and are likely because of increased kidney production, associated with alterations in its transport.

Comparison of DE miRNAs in Pkd1^RC/RC with Other Murine Models of PKD Identifies Common and Novel DE miRNAs in Pkd1^RC/RC at Different Stages of the Disease

To determine whether different murine models of PKD may share similar regulatory mechanisms downstream of the basic pathogenic defect directly caused by the gene mutation and its protein product defect, we compared the DE miRNA profile of the Pkd1 hypomorphic (Pkd1^RC/RC) model with that of other models that were publicly available. Ten studies reporting DE miRNAs in the kidney tissue of PKD murine models fulfilled the entry criteria (Table 1). Venn diagram analysis of miRNAs upregulated at the early stages of the disease in Pkd1^RC/RC identified three, two, and one miRNAs commonly upregulated with Pkd1^−/− and Pkd2^−/− early onset models, and PKD/Mhm (cy/+) rats, respectively (Figure 5A), as well as two commonly upregulated between Pkd1^RC/RC and Pkd1^−/− and Pkd2^−/− models. There were no commonly upregulated miRNAs among the four models at the early stages of the disease. By contrast, we identified zero, two, and zero miRNAs commonly downregulated with Pkd1^−/− and Pkd2^−/− early onset models, and PKD/Mhm (cy/+) rats, respectively (Figure 5B), and one commonly downregulated between Pkd1^RC/RC and Pkd1^−/− and Pkd2^−/− models. There were no commonly downregulated miRNAs among the four models at the early stages of the disease.

Table 1.

Studies reporting differentially expressed micro-RNAs in kidney tissue of other murine models of polycystic kidney disease

First Author (Year)	Species	Model	Age	Disease Stage	Method	PubMed (PMID)
Pandey (2008)76	Rat	PKD/Mhm (cy/+)	36 d	Early	Microarray	19102782
Dweep et al. (2013)73	Rat	PKD/Mhm (cy/+)	36 d	Early	Microarray	23326503
Woo (2017)77	Mouse	HoxB7/Cre; Pkd1fl/fl	1 and 3 d	Early	miRNA-seq	29074972
Woo (2017)77	Mouse	HoxB7/Cre; Pkd2fl/fl	1 and 3 d	Early	miRNA-seq	29074972
Shin (2018)78	Mouse	HoxB7/Cre; Pkd2fl/fl	1 and 3 d	Early	miRNA-seq	29475398
Lakhia et al. (2016)10	Mouse	Pkhd1/Cre; Pkd2fl/fl	14 d	Early	qPCR	26677864
Magayr et al. (2020)14	Mouse	Pkd1 nl/nl	14 d	Early	qPCR	32622528
Sun (2015)79	Rat	PKD/Mhm (cy/+)	16 wk	Advanced	qPCR	25588980
Woo (2017)77	Mouse	HoxB7/Cre; Pkd1fl/fl	7 d	Advanced	miRNA-seq	29074972
Lakhia et al. (2016)10	Mouse	Ksp/Cre; Pkd1fl/fl	10 d	Advanced	qPCR	26677864
Hajarnis et al. (2017)9	Mouse	Ksp/Cre; Pkd1fl/fl	10 d	Advanced	Microarray	28205547
Kim (2019)80	Mouse	HoxB7/Cre; Pkd1fl/fl	10 and 13 d	Advanced	qPCR	30332302
Woo (2017)77	Mouse	HoxB7/Cre; Pkd2fl/fl	7 d	Advanced	miRNA-seq	29074972
Shin (2018)78	Mouse	HoxB7/Cre; Pkd2fl/fl	7 d	Advanced	miRNA-seq	29475398
Lakhia et al. (2016)10	Mouse	Pkhd1/Cre; Pkd2fl/fl	21 and 28 d	Advanced	qPCR	26677864
Hajarnis et al. (2017)9	Mouse	Pkhd1/Cre; Pkd2fl/fl	21 d	Advanced	Microarray	28205547
Patel et al. (2013)18	Mouse	Ksp/cre; Kif3afl/fl	28 d	Advanced	Microarray	23759744
Magayr et al. (2020)14	Mouse	Pkd1 nl/nl	10 wk	Advanced	qPCR	32622528

miRNA-seq, micro-RNA sequencing; PKD, polycystic kidney disease; PMID, PubMed identifier; qPCR, quantitative PCR.

Figure 5.

Comparison of DE miRNAs in Pkd1^RC/RC with other murine models of PKD. Venn diagrams showing novel and commonly upregulated (A–C) and downregulated (B–D) miRNAs in Pkd1^RC/RC kidneys (blue) and kidney tissue from other PKD models, including Pkd1^−/− (yellow) and Pkd2^−/− (green) early onset mouse models, and PKD/Mhm (cy/+) rats (red) or Kif3a^−/− mice (red) from the early stages left and advanced (right) stages of the disease. *miR-199a-5p is also reported as upregulated in PKD/Mhm (cy/+) rats at advanced stages. PKD, polycystic kidney disease.

At the late stages of the disease, our analyses identified five, five, and five miRNAs commonly upregulated with Pkd1^−/− and Pkd2^−/− early onset models, and Kif3a^−/−, respectively (Figure 5C), as well as four commonly upregulated between Pkd1^RC/RC and Pkd1^−/− and Pkd2^−/− models. In addition, we identified four commonly upregulated miRNAs among the four models. Contrarily, we identified four, eight, and one miRNAs commonly downregulated with Pkd1^−/− and Pkd2^−/− early onset models, and Kif3a^−/−, respectively (Figure 5D), and one miRNA was commonly downregulated between Pkd1^RC/RC and Pkd1^−/− and Pkd2^−/− models. However, there were no commonly downregulated miRNAs among the four models at the late stages of the disease.

Overall, our analysis suggests that there is a modest overlap in the miRNA profile among these models at the early stages of the disease, likely because of specific gene/mutation signatures affecting the miRNA profile and the differences in disease severity/course among models. On the other hand, the higher overlap at advanced stages of the disease likely reflects secondary signatures related to CKD influencing the miRNA profile. Finally, we also identified 20 and ten novel, upregulated and downregulated, respectively, miRNAs from the early stages of the disease that may constitute other important regulatory mechanisms in early ADPKD.

Comparison of DE miRNAs in Pkd1^RC/RC with Reported DE miRNAs in Individuals with ADPKD Reveals Their Timing in the Disease Course

Previous studies in individuals with ADPKD have identified several miRNAs as potential candidate biomarkers for the disease. However, whether these miRNAs may reflect early or more advanced stages of the disease is still unclear. To address this gap in knowledge, we compared the miRNA profile in the kidneys of the slowly progressive Pkd1^RC/RC with previously reported DE miRNAs in individuals with ADPKD. Ten studies reporting DE miRNAs in individuals with ADPKD fulfilled our entry criteria (Table 2).

Table 2.

Studies reporting differentially expressed micro-RNAs in patients with autosomal dominant polycystic kidney disease

First Author (Year)	Source	GFR (Mean)	Disease Stage	Method	PubMed (PMID)
Małachowska (2019)81	Serum	GFR >60	Early	qPCR	34900055
Kocyigit et al. (2017)12	Serum	GFR >60	Early	qPCR	28558802
Ghasemi (2023)82	Urine	GFR >60	Early	qPCR	37060340
Ali (2024)83	Urine EVs	GFR >60	Early	miRNA-seq	38404150
Magayr et al. (2020)14	Urine EVs	GFR >60	Early	miRNA-seq	32622528
Ali (2024)83	Urine EVs	GFR <60	Advanced	miRNA-seq	38404150
Streets (2017)84	Urine EVs	GFR <60	Advanced	qPCR	28077374
Magayr et al. (2020)14	Urine EVs	GFR <60	Advanced	miRNA-seq	32622528
Sun (2015)79	Renal cystic tissue	ESKD	Advanced	qPCR	25588980
de Stephanis (2018)85	Renal cystic tissue	ESKD	Advanced	qPCR	29323708
Kim (2019)80	Renal cystic tissue	ESKD	Advanced	Microarray	30332302
Lakhia et al. (2020)11	Renal cystic tissue	ESKD	Advanced	qPCR	32182218
Magayr et al. (2020)14	Renal cystic tissue	N/A*	Advanced	miRNA-seq	32622528

EV, extracellular vesicle; miRNA-seq, micro-RNA sequencing; PMID, PubMed identifier; qPCR, quantitative PCR.

^*Data abstracted from cystic samples.

First, we compared DE miRNAs in the kidney of Pkd1^RC/RC versus DE miRNAs in individuals with ADPKD, independent of disease stage and analytical platform. We identified 19 miRNAs upregulated in Pkd1^RC/RC previously reported as upregulated in individuals with ADPKD (Supplemental Figure 15A). Similarly, we identified 14 commonly downregulated miRNAs between Pkd1^RC/RC and individuals with ADPKD (Supplemental Figure 15B). When separating this analysis by utilized platforms, most of the commonly upregulated and downregulated miRNAs have been previously reported from kidney tissue specimens from individuals with ADPKD, as expected (Supplemental Figure 15, C and D).

To identify which DE miRNAs previously reported in individuals with ADPKD may better reflect early versus common mechanisms implicated in CKD, we repeated this analysis by disease stage. Our analysis showed zero and one miRNA upregulated in Pkd1^RC/RC previously reported as upregulated in individuals with eGFR >60 ml/min per 1.73 m² in urine/urine extracellular vesicles (EVs) or plasma/serum, respectively (Figure 6A). No miRNAs were commonly upregulated among the three groups. Similarly, we found zero and one miRNA downregulated in Pkd1^RC/RC previously reported as downregulated in individuals with eGFR >60 ml/min per 1.73 m² in urine/urine EVs or plasma/serum, respectively (Figure 6B). There were no miRNAs commonly downregulated among the three groups.

Figure 6.

Comparison of DE miRNAs in Pkd1^RC/RC with reported DE miRNAs studies in individuals with ADPKD. Venn diagrams showing novels and commonly upregulated (A–C) and downregulated (B–D) miRNAs in Pkd1^RC/RC kidneys (blue) at early (left) and advanced (right) disease stages and those previously reported in the urine/urine EVs (yellow), and plasma/serum (green), or renal cystic tissue (green), of individuals with ADPKD. ADPKD, autosomal dominant polycystic kidney disease; EV, extracellular vesicle.

At the late stages of the disease, we identified 0 and 16 miRNAs upregulated in Pkd1^RC/RC previously reported as upregulated in patients with eGFR <60 ml/min per 1.73 m² in urine/urine EVs or cystic kidney tissue at KF, respectively (Figure 6C). No miRNAs were commonly upregulated among the three groups. By contrast, we found 1 and 10 miRNAs downregulated in Pkd1^RC/RC previously reported as downregulated in patients with eGFR <60 ml/min per 1.73 m² in urine/urine EVs or cystic kidney tissue at KF, respectively (Figure 6D). There was 1 miRNA commonly downregulated among the three groups.

Together, our analysis of DE miRNAs in Pkd1^RC/RC mice compared with DE miRNAs previously reported in individuals with ADPKD indicates that most of the overlapping DE miRNAs correspond with advanced disease in Pkd1^RC/RC (n=28 versus 2 overlapping DE miRNAs at late or early stages, respectively).

Functional Analysis Distinguishes the Top Biological Processes of Target Genes of miRNAs DE in PKD at Different Stages of the Disease and Compared with Other Etiology of CKD

To discern the extent to which dysregulated miRNAs/mRNAs observed at late stages in Pkd1^RC/RC reflect CKD progression, we compared the miRNA profile in Pkd1^RC/RC with that of another murine model of CKD (DKD). Venn diagram analysis was generated to compare DE miRNAs between our study and those reported by Opazo-Ríos et al., which identified 99 miRNAs upregulated in advanced diabetic nephropathy (ob/ob) compared with WT mouse kidneys.³² Interestingly, we found that 13 of these miRNAs were also upregulated in Pkd1^RC/RC versus WT kidneys at 12 months (Supplemental Figure 16A).

Finally, we performed functional analysis of target genes of overlapping DE miRNAs between Pkd1^RC/RC and at least one other murine model of PKD (Figure 5 and Supplemental Figure 17), the murine model of DKD (Supplemental Figure 16B), individuals with ADPKD (Figure 6 and Supplemental Figure 18), and a combination of DE miRNAs from Pkd1^RC/RC, other murine models of PKD and individuals with ADPKD (Supplemental Figure 19). We found that target genes of overlapping DE miRNAs in these comparisons were implicated in similar functions of those DE in Pkd1^RC/RC compared with WT kidneys (Figure 3). Specifically, while alterations in cellular transport are a feature of CKD, and a common biological process within PKD models and DKD at late stages, it is present from early stages of PKD models and human ADPKD.

Discussion

This study shows that experimental ADPKD is associated with longitudinal changes in the kidney miRNA and mRNA profiles. We found that genes targeted by miRNAs dysregulated from the early stages of the disease encode for proteins mainly implicated in cell proliferation and α-KG transport, whereas those targeted by miRNAs dysregulated at late stages of the disease encode for transport proteins involved in proinflammatory and metabolic processes. Furthermore, urine (but no plasma) α-KG concentration, its FE, and tissue levels were higher in Pkd1^RC/RC during the entire course of the disease, and associated with decreased protein expression of α-KG transporters NaDC3 and Oat1, suggesting an increased kidney production associated with alterations in its transport. Our comparison with other murine models of PKD, another model of CKD (DKD), and individuals with ADPKD suggests that, at the early stages of the disease, gene/mutation-specific signatures may be affecting the miRNA profile, whereas, at advanced stages of the disease, changes in the miRNA profile likely reflect CKD. We further identified several novel miRNAs DE in early and late Pkd1^RC/RC kidneys, which have not been previously reported in either other murine models of PKD or patients with ADPKD. Therefore, our observations may serve as a springboard for future studies to establish the role of these novel miRNAs as regulatory mechanisms, as well as the novel and commonly DE miRNAs early on in the disease as potential biomarkers or therapeutic targets in ADPKD.

While several DE miRNAs have been reported in ADPKD,^9–11 whether they are present from early stages or are associated mainly with late stages of the disease has not been resolved. Here, we identified several miRNAs dysregulated in Pkd1^RC/RC (groups 1–4) from the early stages of the disease. These miRNAs are likely associated with the primary defect and might be involved in regulating mechanisms related to cyst formation and enlargement and might be more specific for the disease. We also singled out several miRNAs dysregulated at late stages of the disease (groups 5 and 6), likely to result from and further regulate mechanisms associated with CKD progression and not specific to ADPKD.

Our integrated analysis identified genes targeted by miRNAs dysregulated from early stages and those targeted by miRNAs dysregulated at late stages of the disease. Functional analysis revealed that upregulated target genes of miRNAs downregulated from early stages in Pkd1^RC/RC are mainly implicated in cell cycle, division, and proliferation, including cyclin E2 (ccne2), which has been implicated in cystic epithelial cell proliferation.³³ Contrarily, downregulated target genes of miRNAs upregulated in Pkd1^RC/RC from early stages encoded for proteins implicated in α-KG transport. Specifically, we found that the expression of slc13a3 and slc22a6, and their protein products, NaDC3 and Oat1, respectively, were lower in Pkd1^RC/RC kidneys from early stages of the disease. Interestingly, in humans, diseases associated with SLC13A3 variants have been accompanied by increased urine excretion of α-KG that persists over time.^28,29 This is in line with our observation of higher urine α-KG concentration in Pkd1^RC/RC versus WT throughout the course of the disease.

In the kidney, NaDC3 is located in the basolateral membrane (blood-facing side) of the renal tubular cells and is involved in OA secretion by taking up α-KG from the blood²³ (Figure 4). The process, which involves Na⁺/K⁺-ATPase, NaDC3, and Oat1, relies on a gradient of α-KG within the tubular cells and is referred to as tertiary active transport.³⁴ The outwardly directed gradient for α-KG is maintained through a Na⁺-coupled secondary active uptake of α-KG, and by its intracellular metabolism, such that differences in metabolic state and the available exchangeable intracellular pool of α-KG produced by metabolism may influence its transport.^22,35 Overall, along with the apical sodium-coupled dicarboxylate transporter 1, which uptakes most of the filtered α-KG,²⁶ these transporters regulate α-KG levels in plasma, urine, and tissue, important for maintaining proper cellular metabolism and OA and drug detoxification. Therefore, alterations in these transporters may result in dysregulations in cellular α-KG metabolism³⁶ and disturbances in OA and drug elimination, which could have important implications for ADPKD. Conversely, intracellular abnormalities in α-KG metabolism driven by increased utilization and/or synthesis may influence its transport and, more broadly, OA and drug bioavailability.

At normal plasma α-KG concentration, the majority is actively reabsorbed mainly in the proximal tubule, with low amounts being excreted in the urine.²⁶ Increased urine α-KG concentration may result from increased filtration, decreased reabsorption, or increased kidney production. Our observations of normal plasma α-KG levels, normal renal function until late stages, and increased FE and tissue concentration suggest an increased kidney production associated with alterations in its transport. Kidney α-KG can be generated through three major metabolic pathways: (1) oxidative decarboxylation of isocitrate in the tricarboxylic acid cycle catalyzed by NAD-dependent isocitrate dehydrogenase 3 in the mitochondria, (2) glutaminolysis, in which deamination of glutamine (Gln) by glutaminase results in glutamate and subsequent deamination of this latter by glutamate dehydrogenase produces α-KG and ammonia, and (3) through Gln transamination, ultimately releasing α-KG and ammonia (Figure 4). Therefore, an increased kidney α-KG production may reflect underlying changes in Gln metabolism, consistent with previous studies demonstrating abnormal Gln utilization in PKD^37–41 and amelioration of the disease with Gln inhibitors^37,38 or interference with glutaminolysis by targeting asparagine synthetase.^42,43 Future studies are needed to determine whether modulation of α-KG metabolism may represent an additional therapeutic target in ADPKD.

At late stages of the disease, we found that upregulated target genes of downregulated miRNAs are primarily involved in proinflammatory cytokine production and lymphocyte activation. Tumor necrosis factor-α progressively increases with age and disease progression in cystic kidneys of cpk mice⁴⁴ and has been found in the cyst fluids of individuals with ADPKD.⁴⁵ This proinflammatory cytokine downregulates endothelial nitric oxide synthase and contributes to vascular damage, in line with our previous observation of progressive endothelial dysfunction and capillary loss in the kidneys of PCK rats.⁴⁶ Similarly, a few upregulated members of the complement pathway were targets of miRNAs downregulated in Pkd1^RC/RC kidneys. Excessive activation of the complement pathway has been reported in ADPKD^47,48 and might imply disease progression rather than PKD-specific mechanisms.

In contrast, we found that downregulated target genes of miRNAs upregulated at late stages of the disease encode proteins primarily implicated in transmembrane transport and lipid and amino acid metabolic processes. This is in line with studies showing differences in gene expression of fatty acid metabolism in mice cystic kidneys and patients with ADPKD^9,40,49–51 and alterations in tryptophan metabolism.^52–54 Thus, post-transcriptional modulation of these transporters and metabolic enzymes could represent an important regulatory mechanism and contribute to disease progression in Pkd1^RC/RC mice. Further mechanistic studies are needed to validate these findings and confirm whether the identified DE miRNAs and their target genes could meaningfully impact disease development and progression.

Identifying regulatory mechanisms downstream of the basic pathogenic defect directly caused by the gene mutation and its protein product dysfunction could aid in developing additional therapeutic strategies.⁵⁵ Several PKD rodent models with overlapping phenotypes have contributed significantly to understanding genetic mechanisms, interactions, and disease pathogenesis and suggested some pathogenic mechanisms are likely conserved.^56–64 In this study, we identified a few miRNAs commonly upregulated between Pkd1^RC/RC, Pkd1^−/−, and Pkd2^−/−, or both early onset models, at early stages of the disease. Among them is miRNA 206-3p, previously reported to upregulate under a hypoxic microenvironment in PTEN-KO cells.⁶⁵ Studies suggested that polycystin deficiency may mimic a low-oxygen environment, proposing a new role for the PCs in oxygen sensing and cell metabolism.⁶⁶ Therefore, the upregulation of miRNA 206-3p may represent a common regulatory mechanism in the context of polycystin deficiency that may contribute to cyst development and enlargement. We also identified a few miRNAs commonly downregulated between Pkd1^RC/RC, Pkd1^−/−, and Pkd2^−/−, or both early onset models, at the early stages of the disease. Specifically, the common downregulation of miR-3470b observed in the three models warrants further mechanistic studies to determine its early role in the disease. However, we did not identify commonly up- or downregulated miRNAs among the four models Pkd1^RC/RC, Pkd1^−/−, Pkd2^−/− mice and PKD/Mhm (cy/+) rats at the early stages. The PKD/Mhm (cy/+) rat is a substrain of the Han:SPRD polycystic rat, and although it has been used as an ADPKD model, it is not an orthologous model for ADPKD.^67–69 Altogether, these data position miR-206-3p and miR-3470b as potential common regulatory pathways directly downstream of PCs dysfunction. Contrarily, the finding of four commonly upregulated miRNAs among all models at late stages of the disease, and 13 commonly upregulated miRNAs with a model of DKD, suggests these regulatory mechanisms may be downstream of secondary signatures related to CKD. Importantly, while alterations in cellular transport are a feature of CKD, and a common biological process within PKD models and DKD at late stages, it is present from early stages of PKD models and human ADPKD.

Our analyses also identified 20 and ten novel, upregulated and downregulated, respectively, miRNAs from early stages of the disease. Among them is the upregulation of miR-147-3p, which regulates inflammatory responses⁷⁰ and is a key mediator in tricarboxylic acid cycle dysfunction and pseudohypoxia in cancer,⁷¹ and therefore, may constitute an important early regulatory mechanism in ADPKD. Similarly, downregulation of miR-5126 and miR-6538 has been reported in the cargo of small EVs of ciliary mutant Bbs4^−/− cells,⁷² further suggesting a role of these miRNAs in EV-dependent ciliary signaling mechanisms and possibly ADPKD.

Studies in individuals with ADPKD have identified several miRNAs as potential candidate biomarkers for the disease,^12–14 but whether they reflect early or more advanced stages and their changes through the disease is still unclear. In this study, analysis of DE miRNAs previously reported in individuals with ADPKD compared with DE miRNAs in Pkd1^RC/RC mice indicated that most of the overlapping DE miRNAs correspond with advanced disease in Pkd1^RC/RC. Therefore, the DE miRNAs identified early on in this study may yield earlier disease biomarkers related to the underlying molecular mechanisms and warrant further investigation in a young cohort of individuals with ADPKD.

Our study is limited by the small number of samples per time point for miRNA-seq and bulk mRNA-seq analyses, which preclude any definitive conclusions concerning the cell-type involvement or differences between the sexes. miRNA binding site predictions may detect false positives, and different search algorithms could detect different targets. To mitigate this, we used miRWalk,⁷³ which identifies both predicted and experimentally validated miRNA-target interaction pairs by combining results with search algorithms from eight other platforms for miRNA target prediction. Because miRNAs generally regulate translation, differences in mRNA-seq levels could be modest, and pathway analysis might be associated with false positives. Nonetheless, our study provides a longitudinal analysis and comparison with other murine PKD models and individuals with ADPKD and hypothesis-generating for future mechanistic and preclinical studies in mice, as well as biomarker discovery in humans. Owing to the lack of human tissue material during the disease's non-KF phases, our study cannot demonstrate miRNA overlap between animals and humans. Studies have shown that although gene expression is similar between mice and humans, there are still considerable differences in the regulatory networks controlling the activity of the immune system, metabolic functions, and responses to stress, all of which have important implications in ADPKD.^74,75 Although our results cannot predict human response, they provide valuable information on candidate pathways that should be further investigated in humans at different stages of the disease.

In summary, our comprehensive characterization allows a better understanding of early versus late miRNA expression changes in ADPKD that could be targeted for a more tailored therapeutic intervention and may serve as real-time specific biomarkers for the disease.

Disclosures

Disclosure forms, as provided by each author, are available with the online version of the article at http://links.lww.com/KN9/B472.

Author Contributions

Conceptualization: Maria V. Irazabal.

Data curation: Maria V. Irazabal.

Formal analysis: Yahya Alsawaf, Alfonso Eirin, Christian Hanna, Marie C. Hogan, Maria V. Irazabal, Slobodan Macura, Nalin Sharma, Santu K. Singha, Ivan Vuckovic, Youwen Zhang, Jamie Zheng.

Funding acquisition: Maria V. Irazabal.

Investigation: Yahya Alsawaf, Alfonso Eirin, Christian Hanna, Marie C. Hogan, Maria V. Irazabal, Slobodan Macura, Nalin Sharma, Santu K. Singha, Ivan Vuckovic, Xiaohong Xu, Youwen Zhang, Jamie Zheng.

Methodology: Yahya Alsawaf, Alfonso Eirin, Christian Hanna, Marie C. Hogan, Maria V. Irazabal, Slobodan Macura, Nalin Sharma, Santu K. Singha, Ivan Vuckovic, Youwen Zhang, Jamie Zheng.

Project administration: Maria V. Irazabal.

Resources: Maria V. Irazabal.

Software: Yahya Alsawaf, Alfonso Eirin, Christian Hanna, Marie C. Hogan, Maria V. Irazabal, Slobodan Macura, Nalin Sharma, Santu K. Singha, Ivan Vuckovic, Youwen Zhang, Jamie Zheng.

Supervision: Maria V. Irazabal.

Validation: Maria V. Irazabal.

Visualization: Maria V. Irazabal.

Writing – original draft: Maria V. Irazabal.

Writing – review & editing: Alfonso Eirin, Christian Hanna, Marie C. Hogan, Maria V. Irazabal, Slobodan Macura, Santu K. Singha, Ivan Vuckovic, Youwen Zhang.

Funding

M.V. Irazabal: National Institute of Diabetes and Digestive and Kidney Diseases (DK128017 and DK118391) and Congressionally Directed Medical Research Programs (PR221810). A. Eirin: National Institute of Diabetes and Digestive and Kidney Diseases (DK129240).

Declarative Statements

All animal experiments were conducted in accordance with the NIH Guide for the Care and Use of Laboratory Animals or an equivalent standard that meets or exceeds the ethical and welfare requirements outlined in the NIH Guide. All protocols were approved by the appropriate institutional animal care and use committee.

Data Availability Statements

Original data generated for the study are or will be made available in a public access repository upon publication. Data Type: Raw Data/Source Data. Repository Name: Figshare. Linkable Citation: https://doi.org/10.6084/m9.figshare.30086419. The raw and processed data obtained from miRNA-seq and mRNA-seq analysis of Pkd1^RC/RC and WT mice at 1 and 12 months (Supplemental Data Files 1–4), as well as miRNA-mRNA interactions for groups 1–6 (Supplemental Data Files 5–10) were deposited in the figshare database (https://figshare.com/): Supplemental Data File 1: https://doi.org/10.6084/m9.figshare.28131632, Supplemental Data File 2: https://doi.org/10.6084/m9.figshare.28131656, Supplemental Data File 3: https://doi.org/10.6084/m9.figshare.28131668, Supplemental Data File 4: https://doi.org/10.6084/m9.figshare.28131710, Supplemental Data Files 5–10: https://doi.org/10.6084/m9.figshare.28131725.

Supplemental Material

This article contains the following supplemental material online at http://links.lww.com/KN9/B473.

Supplemental Materials and Methods

Supplemental Table 1. Animal characteristics by timepoint.

Supplemental Table 2. Targets of miRNAs in group 1.

Supplemental Table 3. Targets of miRNAs in group 2.

Supplemental Table 4. Targets of miRNAs in group 3.

Supplemental Table 5. Targets of miRNAs in group 4.

Supplemental Table 6. Targets of miRNAs in group 5.

Supplemental Table 7. Targets of miRNAs in group 6.

Supplemental Figure 1. Cross-sectional analyses of the kidney phenotype in Pkd1^RC/RC and WT animals.

Supplemental Figure 2. Early and advanced changes in the kidney miRNA profile of Pkd1^RC/RC mice versus WT.

Supplemental Figure 3. miR-17∼92 miRNA cluster in Pkd1^RC/RC versus WT kidneys at early and advanced stages of the disease.

Supplemental Figure 4. miR-21a-5p in Pkd1^RC/RC versus WT kidneys at early and advanced stages of the disease.

Supplemental Figure 5. Validation of miRNA-seq analysis.

Supplemental Figure 6. Targets of miRNAs in group 1.

Supplemental Figure 7. Targets of miRNAs in group 2.

Supplemental Figure 8. Targets of miRNAs in group 3.

Supplemental Figure 9. Targets of miRNAs in group 4.

Supplemental Figure 10. Targets of miRNAs in group 5.

Supplemental Figure 11. Targets of miRNAs in group 6.

Supplemental Figure 12. Functional analysis (cellular component) of target genes of miRNAs DE in Pkd1^RC/RC at early and advanced disease stages.

Supplemental Figure 13. Functional analysis (molecular function) of target genes of miRNAs DE in Pkd1^RC/RC at early and advanced disease stages.

Supplemental Figure 14. Validation of targeted genes at the protein level.

Supplemental Figure 15. Comparison of DE miRNAs in Pkd1^RC/RC with reported DE miRNAs studies in individuals with ADPKD, independent of disease stage.

Supplemental Figure 16. Comparison with murine models of other forms of CKD and top biological processes.

Supplemental Figure 17. Functional clustering analysis of target genes of overlapping upregulated and downregulated miRNAs in murine PKD.

Supplemental Figure 18. Functional clustering analysis of target genes of overlapping upregulated and downregulated miRNAs between Pkd1^RC/RC and human PKD.

Supplemental Figure 19. Comparison among murine models of PKD and individuals with ADPKD.