ift140-Deficient Zebrafish as a Model for Kidney Cystogenesis and an F0-Based Screen for Genetic Modifiers of Kidney Cysts

Visual Abstract

Abstract

Key Points

• ift140-deficient zebrafish exhibited phenotypes observed in mammals, including skeletal defects and kidney cysts.

• Beyond ciliogenesis, ift140 also regulated cell polarity and intracellular microtubule stability, potentially contributing to cystogenesis.

• Microhomology-mediated end joining–induced mosaic ift140 crispants enabled evaluation of cystogenesis in adult fish and F0-based screening for modifiers of kidney cysts.

Background

Genetic modifiers are believed to play an important role in the onset and severity of polycystic kidney disease, but identifying these modifiers has been challenging due to the lack of effective methodologies.

Methods

We generated zebrafish mutants of IFT140, a skeletal ciliopathy gene and newly identified autosomal dominant polycystic kidney disease gene, to examine skeletal development and kidney cyst formation in larval and juvenile mutants. In addition, we used ift140 crispants, generated through efficient microhomology-mediated end joining–based genome editing, to compare phenotypes with mutants and conduct a pilot genetic modifier screen.

Results

ift140 mutants developed kidney cysts and bone defects similar to those seen in mammalian models. ift140 crispants recapitulated mutant phenotypes while bypassing the early lethality of the mutants, enabling the analysis of kidney cyst formation in adult fish. In addition to cilia defects, we identified nonciliary phenotypes, including disrupted cell polarity and aberrant cytoplasmic microtubule stabilization in kidney epithelial cells, as potential contributors to ift140-associated cystogenesis. The ability to detect ift140-associated kidney cysts with ease allowed us to develop an F0-based genetic screen to identify potential protective modifiers. A pilot screen of 16 genes previously implicated in dysregulated signaling pathways in autosomal dominant polycystic kidney disease revealed both known and novel modifiers, including mtor and ulk1a. We further found that inhibition of mtor and ulk1a reversed both cilia-related and non–cilia-related abnormalities in the kidney.

Conclusions

By establishing a zebrafish model of ift140-associated cystic kidney disease, we recapitulated ift140’s ciliary role and uncovered a nonciliary function in kidney cystogenesis. Importantly, we demonstrated the feasibility of using ift140 mosaic crispants to evaluate cystogenesis in adult fish and to perform F0-based screening for identifying genetic modifiers of kidney cysts.

Introduction

IFT140 encodes a core component of the intraflagellar transport complex A. Biallelic mutation of IFT140 is associated with skeletal ciliopathies, such as the short rib thoracic dysplasia and Mainzer-Saldino syndrome, with extraskeletal phenotypes including kidney cysts.^1,2 In experimental models, Ift140 ablation causes skeletal ciliopathies and cystic kidneys.^3,4 Recently, IFT140 was identified as a new autosomal dominant polycystic kidney disease (ADPKD) gene.⁵ It is potentially the third most mutated gene in ADPKD, following PKD1 and PKD2.⁵ As a retrograde intraflagellar transport, IFT140 coordinates the removal of cargos from cilia, and it also regulates the entry of proteins into cilia.^6–8 However, little is known about the mechanisms underlying its functions in kidney cyst development.

ADPKD is one of the most prevalent and potentially lethal genetic diseases.^9–12 Patients with ADPKD exhibit highly variable phenotypes, of which genetic modifiers are thought to play an important role.^13–16 Genome-wide association studies and quantitative trait locus analysis are commonly used to identify genetic modifiers. However, these statistical approaches have significant limitations: (1) establishing a precise genotype–phenotype relationship is challenging, and (2) experimental validation of candidate modifiers in rodent models is low-throughput, requiring the generation of mutants and multigenerational crosses. As a result, only a limited number of modifier genes, such as HNF1β, TSC2, and DKK3, have been reported,^17–19 which significantly impedes the development of effective therapeutic interventions. Thus, an alternative animal model with more efficient genetic tools and a novel method enabling rapid discovery of genetic modifiers of ADPKD are highly desirable.

Zebrafish have been used to study cystogenesis, with mutations in genes associated with human cystic diseases, such as pkd1 and hnf1b, leading to kidney cyst formation in zebrafish embryos.^20–22 Large-scale compound screens in zebrafish embryos have identified potential therapeutic drugs for PKD, some of which have showed promising effects in mammalian models.^23,24 These screens have relied on the characteristic body curvature of pkd2 mutants as a readout rather than kidney cyst formation, as pkd2 mutants do not develop pronephric cysts.^21,25 While pkd1 mutants form pronephric cysts with full penetrance by 3 days postfertilization (dpf), the small cyst size and complications from edema make visual detection challenging. Thus, hematoxylin and eosin staining of tissue sections remains the most reliable detection method but is impractical for large-scale screening.²² Consequently, a kidney cyst-based screen, particularly a genetic modifier screen, has yet to be conducted in any zebrafish cystic models.

Recently, clustered regularly interspaced short palindromic repeats (CRISPR)–induced mosaic knockout (KO) has been used to identify genes involved in morphogenesis, regeneration, and behavior phenotypes in F0 zebrafish embryos.^26–29 However, multilocus targeting of a single gene is needed to achieve functional disruption due to the nonhomologous recombination mechanism of CRISPR KO.²⁹ On the other hand, microhomology-mediated end joining (MMEJ)–based CRISPR can achieve highly efficient KO through single-locus targeting because frame-shifting genetic lesions can be predicted and preselected.³⁰ Using the MMEJ technology, genetic modifiers of cardiomyopathy have been successfully identified through F0-based screens both in zebrafish embryos and adults.^31–33

Here, we aim to determine whether IFT140 has conserved roles in kidney cyst formation and skeletal development in zebrafish and to elucidate the underlying mechanisms of cystogenesis. We also aim to assess the feasibility of combining the zebrafish ift140 model with efficient MMEJ-mediated genome editing to identify potential modifiers of kidney cysts through an F0-based genetic screen.

Methods

Zebrafish Strains

The ift140 genomic lesion was generated through MMEJ-mediated genome editing technology.^30,33 Genotyping was performed using derived cleaved amplified polymorphic sequences primers (forward: 5′-CTC​CTG​ACA​GCA​GTG​GAG​TGC​CGG​TCC​CCA T-3′; reverse: 5′-GTT​GCT​CTT​ATG​TCA​GGT​AGT​G-3′) to introduce a C-to-T substitution, creating an NcoI restriction enzyme site in the PCR-amplified mutant allele.³⁴ To minimize potential off-target effects, F1 fish were outcrossed, and subsequent experiments were conducted using F3 or F4 generations. All protocols and procedures were approved by the Mayo Clinic Institutional Animal Care and Use Committee (Rochester, MN) and performed in accordance with the NIH Guide for the Care and Use of Laboratory Animals.

In addition, mtor and ulk1a mutants were generated previously.^33,35 To ensure sufficient sample size, approximately 100 embryos from mtor or ulk1a out-crosses were injected with ift140 sgRNA, yielding approximately 50 mtor^+/− or ulk1a^+/− and approximately 50 wild-type (wt) embryos per experiment. Similarly, approximately 150–200 embryos from ift140;mtor or ift140;ulk1a in-crosses were analyzed to obtain approximately nine or more desired double mutants per experiment. This was based on our power analysis indicating that a two-tailed t test (α=0.05, power=0.8) requires at least 6–8 samples. mtor genotyping followed published methods,³⁵ while ulk1a genotyping used derived cleaved amplified polymorphic sequences primers (forward: 5′-AGC​CCC​AAA​ACA​TCC​TTC​TCT​CAT​AGA​G-3′; reverse: 5′-CCA​CAT​AAC​ACC​CAA​GAG​C-3′) to introduce a Bsp1286I restriction site in the PCR-amplified wildtype allele.

MMEJ-Based F0 Assay in Zebrafish Embryos and Adult Fish

MMEJ-inducing sgRNAs were designed using the online tool microhomology-mediated end joining knockout target heuristic utility (http://genesculpt.org/menthu/) and synthesized from Synthego Corporation.^31,32 Each sgRNA was mixed with Alt-R Cas9 protein (Integrated DNA Technologies, 1081058) to achieve a final concentration of 5 µM and 300 ng/µl, respectively, and injected into one-cell stage embryos. The resulting embryos were assessed for pronephric cysts at 4–6 dpf. In some cases, injected embryos were raised to adulthood to evaluate kidney phenotypes in adult fish. To determine sgRNA efficiency, genomic DNA was extracted from embryos or the tail fin of adult fish, followed by PCR amplification of the targeted region and mutation rate calculation using the Inference of CRISPR Edits Tool (https://www.synthego.com/products/bioinformatics/crispr-analysis). Sequences of sgRNAs and PCR primers are presented in Supplemental Table 1.

Alcian Blue and Alizarin Red Staining

Cartilage staining and mineralized bone staining of zebrafish larvae were adapted from published protocols.^36,37 In brief, larvae were euthanized with MS222 (0.02%) and fixed in 4% paraformaldehyde overnight at 4^°C. After three washes with PBS with 0.1% Tween-20, samples were bleached in 1.5% H₂O₂ and 1% KOH for 20 minutes. For cartilage staining, samples were immersed in 0.1% Alcian blue (w/v; Sigma–Aldrich) in 70% ethanol overnight. For skeletal staining, specimens were incubated with 30% saturated borax for 5 hours. This was followed by incubation with 0.01% Alizarin red (w/v; Sigma-Aldrich) in 1% KOH overnight. The specimens from both stainings were then cleared using successive changes of 20% glycerol (Thermo Fisher) in 1% KOH, 50% glycerol in 1% KOH, and 80% glycerol in 1% KOH. Imaging was conducted using a Leika dissecting microscope with digital camera attached.

Western Blotting, Histologic Analysis, and Immunofluorescence Labeling

Western blotting was performed as described previously.²² Histologic analysis of zebrafish pronephros was performed on JB4 sections as described previously,²² while adult zebrafish kidney histology was conducted on paraffin sections following established protocols.³⁸ Immunofluorescence analysis of the zebrafish pronephros was conducted using whole-mount immunostaining of embryos or cryosectioned samples, as described previously.^22,38,39 For adult zebrafish kidneys, immunofluorescence was performed on paraffin sections of either 5 µm thin or 20 µm thick.³⁸ Tissue sections were labeled with renal tubular segment markers, including rhodamine-labeled Dolichos biflorus agglutinin and fluorescein-labeled lotus tetragonolobus lectin (LTL), along with the nuclear marker DAPI (Vector Laboratories). Primary antibodies used were anti-Ift140 (Proteintech, 17460-I-AP), anti-Actin (Sigma-Aldrich, A3854), antiacetylated α-tubulin (Sigma–Aldrich, T7451), anti–γ-tubulin (Sigma-Aldrich, T5326), anti-PKC (Santa Cruz Biotechnology, sc-216), anti-proliferating cell nuclear antigen (Sigma-Aldrich, p8825), and anti-Na⁺/K⁺ ATPase α-1 subunit (α6F, Developmental Studies Hybridoma Bank). Secondary antibodies were Alexa Fluor conjugated (Life Technologies). Images were captured using a Zeiss Axioplan II microscope equipped with ApoTome and AxioVision software (Carl Zeiss Microscopy).

Optical Clearing and Whole-Mount Immunostaining of Juvenile Zebrafish

Juvenile zebrafish at 20–30 dpf were fixed in 4% paraformaldehyde overnight and treated with a bleaching solution (1.5% H₂O₂ in 1% KOH) to eliminate melanocyte pigmentation. The abdominal internal organs were removed while preserving the kidney attached to the dorsal body wall. Optical clearing, whole-mount immunostaining, and imaging were then performed as previously described.³⁸

Pharmacologic Treatment of Zebrafish Embryos

Embryos were treated with rapamycin (300 nM; LC Laboratories) or the ULK1 inhibitor SBI-0206965 (6 µM; MedChemExpress) from approximately 40 hpf to 5 dpf, with daily replacement of fresh solutions. Pronephric cysts were assessed at 5 dpf.

Statistics

Data were analyzed using GraphPad Prism software. To compare differences between two groups, an unpaired two-tailed Student t test was used; to compare differences among multiple groups, one-way ANOVA followed by the Tukey post hoc test was performed. A P value of < 0.05 was considered as statistically significant. All data collection and analysis were conducted in a blinded fashion to offset individual bias.

Results

ift140^e2/e2 Fish Recapitulated Kidney and Skeletal Phenotypes in Mammalian Models

The zebrafish orthologue of mammalian IFT140 is located on chromosome 24 (zfin.org) and is expressed in the kidney as well as other regions, including the notochord (Supplemental Figure 1). We generated a stable ift140 mutant (Figure 1A) that lacked Ift140 protein expression (Figure 1B), suggesting a null mutant allele. All ift140^e2/e2 embryos developed distinctive pronephric cysts, detectable as early as 2 dpf (Figure 1C). These cysts progressively enlarged; however, the mutants did not exhibit abnormal body curvature or signs of cardiac or whole-body edema (Figure 1D and Supplemental Figure 2A). By approximately 2 weeks of age, mutant larvae were significantly smaller and began to die, with none surviving past 30 dpf (Supplemental Figure 2, B and C).

Figure 1.

Inactivation of ift140 resulted in the formation of kidney cysts in both the pronephros and the newly formed mesonephros. (A) Schematic representation of MMEJ-mediated genetic lesions in ift140. Underlined: sgRNA target sequence; red: Pam sequence; shaded: microhomology sequence. Nucleotide deletions (indicated by dashes) caused a coding frameshift that presumably resulted in a premature stop codon (*). The corresponding amino acid sequence is shown below the DNA sequence. (B) Ift140 protein expression was abolished in homozygous ift140 mutants. Four wild-type and four mutant zebrafish were collected for Western blot analysis at 14 dpf, and representative images are shown. (C and D) Inactivation of ift140 led to the formation of pronephric cysts. The dashed box indicates the glomerular neck region of the pronephros. Images of embryos at 2 dpf (C) and 4 dpf (D) are shown. (E) Inactivation of ift140 caused mesonephric tubular dilation. Whole-mount immunostaining was performed on optically cleared juvenile fish using an anti-PKC antibody. The open triangle, white arrow, blue arrow, and filled triangle indicate the glomerulus, dilated mesonephric tubule, undilated mesonephric distal tubule, and pronephric tubule, respectively. Representative images are shown. Scale bar: 50 µm. dpf, days post fertilization; MMEJ, microhomology-mediated end joining; WT, wild-type.

To assess mesonephros development, which begins around 12–14 dpf,⁴⁰ we performed optical clearing-whole mount imaging on surviving ift140^e2/e2 larvae. At 26 dpf, mutants exhibited significantly fewer nephrons than age-matched siblings, indicating delayed mesonephric kidney development (Figure 1E). By contrast, pronephric kidney development appeared unaffected, with expanded nephrin and slc13a1 expression reflecting dilation in the glomerular neck and proximal tubular regions (Supplemental Figure 3). Importantly, newly formed mesonephric tubules in mutants displayed pronounced dilation (arrow), particularly in the proximal region adjacent to the glomerulus, while the distal region fused to the pronephros remained unaffected (blue arrow) at 26 dpf (Figure 1E). By 30 dpf, extensive tubule dilation obscured the assessment of distal tubules (Figure 1E).

We also examined craniofacial and axial skeletal development. At 6 dpf, craniofacial cartilage and calcified bone structures showed no significant morphologic differences between wild-type and mutant embryos (Figure 2, A and B). By 26 dpf, nonmineralized craniofacial cartilage was present (Figure 2C1). Although the Meckel (mk) and ceratohyal (ch) structures were shorter in mutants than in wild-type larvae, this may be attributed to their overall smaller body size (Figure 2C2). However, the dorsal and anal fins were barely detectable (Figure 2C1), and only the hypural (hy) cartilage structure was visible in the caudal fin complex (Figure 2C3). At this developmental stage, defects in ossified bone structures were most striking in mutants (Figure 2D1). Craniofacial bones were nearly absent, the number of vertebrae was significantly reduced, with only a few present in the rostral part of the axial skeleton (Figure 2D2), and no caudal fin structures were observed (Figure 2D3). These findings suggest that ift140 is essential for skeletal development, although developmental delay and growth arrest may contribute to the absence of caudal fin cartilage.

Figure 2.

Inactivation of ift140 resulted in a lack of ossification of larval bones. (A) Alcian blue staining of embryos at 6 dpf showed no differences in craniofacial cartilage structures between sibling and mutant zebrafish. (B) Alizarin red staining of embryos at 6 dpf indicated no differences in craniofacial calcified bone structures between sibling and mutant. (C1–C3) Alcian blue staining of larval fish at 26 dpf revealed the presence of craniofacial cartilage structures and the absence of caudal fin cartilage structures, except for the hy. (D1–D3) Alizarin red staining of larval fish at 26 dpf showed that most calcified bones were absent, with only a few vertebrae present in the rostral part of the axial skeleton. Shown are lateral views of whole larvae (C1 and D1), ventral views of craniofacial bones (C2 and D2), and lateral views of the caudal fin complex (C3 and D3). Scale bar: 1 mm. Label abbreviations: af, anal fin; cf, caudal fin; ch, ceratohyal; df, dorsal fin; ep, epurals; hs, haemal spines; hy, hypural; mk, Meckel's cartilage; oc, opisthural cartilage; ot, otolith; ust, urostyle; wb, Weberian apparatus.

Figure 3.

MMEJ-mediated mosaic KO of ift140 enabled the analysis of kidney cysts in adult fish in the F0 generation. (A) Gross morphology of adult ift140^MJ fish. Embryos injected with MMEJ-inducing sgRNA targeting ift140 were allowed to grow to adulthood. At 3 months of age, ift140^MJ fish were generally shorter in length and often exhibited a distorted body shape compared with wild-type fish. (B and C) The kidneys in adult ift140^MJ fish lost their saddle-like shape (B) and became enlarged (C). The kidney area was normalized to body length. Eight wild-type fish and eight ift140^MJ fish with a KO efficiency (% KO) approximately 80% were analyzed. % KO was calculated using genomic DNA from the tail fin. (D and E) Kidney cyst severity correlated with % KO of ift140. Representative H&E images are shown (D), with glomerular cysts indicated by asterisks. A significant correlation between % KO of ift140 and cystic burden was demonstrated by Pearson correlation coefficient (r=0.83, P < 0.001; E). The cystic index (cyst area/total kidney area) was analyzed using three sections per kidney. (F) The segment identity of cystic tubules was examined by co-staining kidney sections with LTL (green; labels proximal tubules and major collecting ducts), DBA (red; labels distal tubules), and DAPI (blue; labels nuclei). Unstained tubule is indicated by #, and the glomerulus cyst is marked with asterisk. (G and H) Cyst-lining epithelial cells exhibited hyperproliferation. Kidney sections were immunostained using a PCNA antibody (G), and the percentage of PCNA(+) LTL-labeled cells (arrows) was quantified (H). Four fish per group and three sections per kidney were used for analysis (F–H), with approximately 200 LTL-labeled cells assessed per section (H). Scale bars: 1 cm (A), 1 mm (B),100 µm (D), and 50 µm (F and G). BL, body length; DAPI, 4′,6-diamidino-2-phenylindole; DBA, Dolichos biflorus agglutinin; H&E, hematoxylin and eosin; KO, knockout; LTL, lotus tetragonolobus lectin; PCNA, proliferating cell nuclear antigen.

ift140^MJ Fish Enabled the Analysis of Kidney Phenotypes in Adult Zebrafish in the F0 Generation

The early lethality of ift140^e2/e2 mutants prevented the assessment of kidney phenotypes in adult fish, whereas ift140^e2/+ fish remained generally normal for at least 1 year (data not shown). To overcome this limitation, we used a F0-based mosaic KO approach using MMEJ-mediated genome editing. sgRNA-injected fish, designated as ift140^MJ, formed pronephric cysts and some were viable. Consistent with MMEJ editing, >90% of ift140 genetic lesions were the predicted 5-nt deletion, suggesting relatively low genetic heterogeneity. To investigate the relationship between KO efficiency and disease severity, we titrated the sgRNA dose. At approximately 70%–80% KO efficiency, many adult ift140^MJ fish exhibited significant body shortening and tail distortion (Figure 3A). Their kidneys lost their characteristic saddle-like shape (Figure 3B), were significantly enlarged (Figure 3C), and displayed moderate cystic burden (Figure 3D). At KO efficiencies exceeding 90%, most fish either died or were severely undersized, with kidneys filled with numerous cysts (Figure 3D). However, at approximately 50% KO efficiency, fish appeared largely normal, displaying only minor kidney tubular dilation (Figure 3D). Pearson correlation analysis confirmed a strong positive relationship between ift140 KO efficiency and kidney cystic burden (r=0.83, P < 0.001; Figure 3E).

Notably, kidney cysts in ift140^MJ fish included both glomerular and tubular cysts (Figure 3D). Large tubular cysts were either unstained or partially stained by LTL, while medium-sized cysts were LTL(+), and no Dolichos biflorus agglutinin(+) cysts were detected (Figure 3F). These findings suggest that tubular cysts primarily originate from the proximal tubules, where epithelial cells undergo dedifferentiation after cyst expansion. However, we cannot rule out a collecting duct origin for some tubular cysts, as LTL labels both proximal tubules and major collecting ducts in zebrafish.⁴¹

Hyperproliferation of cyst-lining epithelial cells is a hallmark of ADPKD.^11,42 Consistently, LTL(+) cystic tubules in the ift140^MJ fish exhibited significantly more proliferating cell nuclear antigen-positive epithelial cells compared with those in wild-type fish (Figure 3, G and H). In addition, LTL(+) noncystic tubules in the F0 animals also showed a tendency for increased cell proliferation (Figure 3H). Kidney fibrosis, another hallmark of ADPKD, has been observed in Ift140-depleted mouse kidneys.^3,43 In line with this, the ift140^MJ kidneys showed some evidence for collagen deposition (Supplemental Figure 4).

Inactivation of ift140 Resulted in the Loss of Distal Single Cilia and Disrupted Epithelial Polarity and Cytoplasmic Microtubule Hyperacetylation in the Proximal Tubules

To elucidate the mechanisms underlying cystogenesis in the kidney, we examined ciliogenesis in ift140 mutants. The zebrafish kidney contains two types of cilia: single cilia, primarily distributed in the distal tubule and proximal region near the glomerulus, and multicilia bundles, localized in the middle segment of the tubules (Figure 4A).^38,44–46 In ift140^e2/e2 embryos, distal single cilia were shorter, and multicilia bundles lacked the well-aligned orientation seen in wild-type embryos at 2 dpf (Supplemental Figure 5). At 4 dpf, distal single cilia were significantly shortened and reduced in number (Figure 4, B and B1), while cilia bundle misorientation became more pronounced (Figure 4C). By 6 dpf, distal single cilia were nearly absent (data not shown). Interestingly, intracellular microtubules were hyperacetylated, as indicated by acetylated α-tubulin (Ac-tub) staining at the cell periphery and within the cytoplasm of certain cells in the anterior multiciliated cell (MCC) region of mutant embryos (Figure 4D), suggesting increased cytoplasmic microtubule stabilization.⁴⁷ Expression analysis of histone acetyltransferases (atat1) and histone deacetylases (hdac6 and sirt2) revealed reduced sirt2 expression in mutant embryos (Supplemental Figure 6), which may contribute, at least in part, to cytoplasmic microtubule hyperacetylation.⁴⁸

Figure 4.

The effect of ift140 inhibition on kidney cilia and intracellular microtubule acetylation. (A) Schematic representation of single cilia and multicilia bundle distribution in wild-type fish. Cilia were visualized by whole-mount immunostaining with an antibody against acetylated α-tubulin (Ac-tub, green), and the pronephros were labeled using an antibody against Na⁺/K⁺ ATPase (α6F, red). Boxed regions are enlarged in panels (B–D). (B and B1) ift140^e2/e2 embryos exhibited shorter and fewer distal single cilia. Representative images are shown in (B), with quantifications of cilia length and number provided in (B1). Single cilia counts were normalized to the measured area length. (C) ift140^e2/e2 embryos displayed misorientation of multicilia bundles. (D) ift140^e2/e2 embryos showed Ac-tub staining around the cell periphery and within the cytoplasm (arrows). (E) ift140^e2/e2 embryos exhibited abnormal basal body docking (arrow). Cryosections were immunostained with anti-atypical PKC (red) and anti–γ-tubulin (green). Representative cross-section images from the glomerular-neck region (asterisk) are shown. g: glomerulus. (F and G) Adult ift140^MJ kidneys contained fewer single ciliated cells. Representative cilia staining (red) is shown in (F), with quantification of ciliated tubular epithelial cells in (G). (H) Adult ift140^MJ kidneys exhibited increased Ac-tub staining around the cell periphery and within the cytoplasm of LTL(+) tubules. Arrowhead indicates multicilia bundle. (I) Adult ift140^MJ kidneys showed displacement of basal bodies (arrow). Cystic tubules are marked with #. Fish were analyzed at 4 dpf (A–C and E), 6 dpf (D), and 3 months (F–I). Ten embryos per genotype were examined in (B–D), five embryos in (E), and 4–10 fish in (F–I), with approximately 1000 tubular epithelial cells counted (G). Scale bars: 10 µm (B and E), and 20 µm (C, D, F, H, and I).

Both the disorganization of multicilia bundles and cytoplasmic microtubule stabilization suggest potential defects in epithelial polarity.^49–51 In polarized kidney epithelial cells, the basal body is typically localized at the apical side of the cell. In ift140^e2/e2 embryos, basal bodies largely remained at the apical surface of the distal tubules (Supplemental Figure 7A). Consistently, the apical expression of atypical PKC and the basolateral localization of Na⁺/K⁺-ATPase appeared largely unaffected (Supplemental Figure 7B). However, in the region adjacent to the glomerulus, basal bodies were mislocalized to lateral junctions, the center of cells, or even the basal surface, accompanied by a loss of apical PKC expression (Figure 4E). Given that these ciliary and nonciliary defects could affect cilia beating and subsequent kidney fluid output, we performed a dye excretion assay and observed reduced fluid flow in ift140^e2/e2 embryos (Supplemental Figure 8), suggesting impaired cilia motility and/or disorganized cilia beating.^22,44

Cilia phenotypes in other organs were also examined. ift140 mutants displayed comparable cilia length in the Kupffer's vesicle and spinal canal to controls but showed reduced multicilia density in the anterior rim of the nasal pit, as previously reported (Supplemental Figure 9).⁵² Cytoplasmic microtubule hyperacetylation was not detected (data not shown). These data are consistent with normal cardiac asymmetry and absence of hydrocephalus in the mutants (data not shown) but suggest impaired mucus clearance.

In the kidneys of adult ift140^MJ fish, similar defects to those in embryonic kidneys were observed, including a severe depletion of single cilia (Figure 4, F and G), cytoplasmic microtubule hyperacetylation in LTL(+) epithelial cells (Figure 4H), and randomization of basal body positioning in cystic tubules (Figure 4I). Collectively, these findings suggest that ift140 inactivation leads to cilia loss, aberrant stabilization of intracellular microtubule, and defective apical docking of the basal body, with the latter two abnormalities being particularly prominent in the dilated proximal tubules.

ift140^MJ Embryos Could Serve as a Platform for Screening Protective Genetic Modifiers of Kidney Cysts

Since approximately 90% of ift140^MJ embryos, similar to ift140 mutants, developed pronephric cysts without apparent edema complications, and these cysts can be easily detected under a light microscope at 4–6 dpf (Figure 5A), ift140^MJ embryos presented a promising platform for cyst-based screening of genetic modifiers. To test this hypothesis, we conducted a pilot screen targeting 16 genes implicated in dysregulated signaling pathways in ADPKD using the MMEJ-mediated genetic assay (Figure 5B).^53,54 sgRNAs for each gene were coinjected with ift140-sgRNA into one-cell stage embryos, and pronephric cyst formation was assessed at 5 dpf. As expected, inactivation of genes highlighted in blue, which were previously suggested to have protective effects, such as mtor, reduced the number of ift140^MJ embryos with pronephric cysts.^55–57 Conversely, inactivation of genes marked in red, which were previously associated with detrimental effects, failed to suppress cyst formation.^58–61 Importantly, our screen also suggested novel candidate protective modifiers, such as ulk1a (Figure 5B). Repeated injections yielded largely reproducible results (Figure 5B).

Figure 5.

MMEJ-based F0 assay revealed protective modifying effects of known and novel genes in kidney cyst formation. (A) Workflow diagram illustrating the screening strategy. (B) Summary of modifying effects for 16 GOIs. An MMEJ-inducing sgRNA targeting a GOI was coinjected with an sgRNA against ift140 into one-cell stage embryos. For each gene, 50–100 injected embryos were scored for pronephric cysts at 5 dpf. Injections were conducted in groups, with each group consisting of approximately 4–5 GOIs (ift140^MJ+GOI^MJ) and one control (ift140^MJ+scramble^MJ). After repeating experiments, gDNA was extracted from four embryos per gene to assess KO efficiency (% KO). ECM, extracellular matrix; GOI, genes of interest; MAPK, mitogen-activated protein kinase; mTOR, mechanistic target of rapamycin.

To assess adult fish kidney phenotypes, we allowed ift140^MJ;mtor^MJ, and ift140^MJ;ulk1a^MJ fish to grow. Although most did not survive, the remaining fish exhibited a significantly lower cyst burden than ift140^MJ fish at 3 months (Supplemental Figure 10), suggesting that mtor and ulk1a inhibition might have protective effects in the adult kidney.

To validate these findings, we used previously generated stable mtor and ulk1a mutants.^33,35 First, ift140-sgRNA was injected into embryos obtained from mtor^+/− and wild-type crosses. ift140^MJ;mtor^+/− embryos developed significantly fewer pronephric cysts than their ift140^MJ siblings (Figure 6A). Similarly, the percentage of ift140^MJ;ulk1a^+/− embryos with pronephric cysts was significantly lower than that of their ift140^MJ siblings (Figure 6B). Further analysis of stable double mutants revealed that while ift140^−/−;mtor^+/− and ift140^−/−;ulk1a^+/− embryos developed cysts comparable with ift140^−/− fish, ift140^−/−;mtor^−/− and ift140^−/−;ulk1a^−/− embryos exhibited a significant reduction in cyst formation (Figure 6, C and D). The differing protective effects of mtor^+/− and ulk1a^+/− in ift140^MJ versus ift140^−/− fish may stem from variations in ift140 KO efficiency between these models.

Figure 6.

The protective role of mtor and ulk1a inhibition was validated through stable mutants and extended to pkd1^−/− embryos. (A and B) Haploinsufficiency of mtor or ulk1a protected against ift140 knockdown-associated pronephric cyst development. Data are from four independent experiments. In each, approximately 100 embryos from mtor^+/−×wt or ulk1a^+/−×wt crosses were injected with ift140 sgRNA, scored for pronephric cysts at 5 dpf, and genotyped for mtor (A) or ulk1a (B). Approximately 50 heterozygous and 50 wt siblings were analyzed per experiment (approximately 200 embryos per group total). (C and D) Inactivation of mtor or ulk1a protected against ift140^e2/e2-associated pronephric cyst formation. Data are from three independent experiments. In each, approximately 150–200 embryos from ift140^+/−;mtor^+/− or ift140^+/−;ulk1a^+/− intercrosses were scored for pronephric cysts at 5 dpf and genotyped for ift140 and mtor (C) or ift140 and ulk1a (D), yielding 7–20 embryos per genotype per experiment. (E) Pharmacologic inhibition of mTor and Ulk1 protected against ift140 knockdown-associated pronephric cyst formation. Data are from three independent experiments. In each, approximately 50 ift140^MJ or wt embryos were incubated with rapamycin (300 nM), SBI-0206965 (6 µM), or vehicle (DMSO) from approximately 40 hpf to 5 dpf, with daily changes of freshly made solutions. Pronephric cysts were scored at 5 dpf. (F) Knockdown of ulk1a protected against pkd1^−/−-associated pronephric cyst formation. Representative images show the glomerulus neck region of the pronephros in wt embryos (arrow) and pkd1^−/− embryos (asterisks), along with quantification of pronephric cysts from three independent experiments. In each, ulk1a sgRNA was injected into embryos from pkd1^+/− intercrosses. At 3 dpf, 48 injected and 48 uninjected embryos were fixed individually, and tails were collected for pkd1 genotyping. Embryos with desired genotypes (7–11 per group per experiment) underwent JB4 embedding, sectioning, and H&E staining to assess pronephric cysts. ift140 KO efficiency: 93.3±6.8%; ulk1a KO efficiency: 74.8±11.5%. g, glomerulus; nc, notochord; rapa, rapamycin.

To increase the translational value of the screen, we validated these findings pharmacologically. Treatment with rapamycin, a well-characterized mTOR inhibitor,²² significantly reduced cyst formation in ift140^MJ embryos from approximately 80% to approximately 20% (Figure 6E). Similarly, treatment with SBI-0206965, a potent ULK1 inhibitor with well-documented safety,⁶² reduced cyst formation to approximately 40% (Figure 6E). Neither drug caused detectable defects in wild-type embryos up to 5 dpf (Figure 6E). These results support the protective roles of mtor and ulk1 inhibition in modulating cystogenesis.

Finally, we hypothesized that some genetic modifiers could be shared across cystic diseases if they target common pathogenic mechanisms. Consistently, inactivation of ulk1a in pkd1^−/− embryos was also protective (Figure 6F). Combined with our previous findings that mtor inhibition prevents cyst formation in pkd1^−/− embryos,²² our results suggest that at least some modifier genes identified in the ift140-based screen may have broader applicability to pkd1-based models.

Inhibition of mtor and ulk1a Both Restored Defects in Single Ciliogenesis, Cilia Bundle Orientation, and Cytoplasmic Microtubule Acetylation

To investigate the mechanisms underlying the protective effects of mtor and ulk1a inhibition, we examined both ciliary and nonciliary defects resulting from ift140 inhibition. Consistent with observation in ift140^e2/e2 mutants, ift140^MJ embryos exhibited reduced length and number of distal single cilia (Figure 7, A and B), misoriented multicilia bundles in the distal MCC region (Figure 7C), and intracellular accumulation of Ac-tub in the proximal MCC region (Figure 7D). However, injection of sgRNAs targeting either mtor or ulk1a rescued all these abnormalities in ift140^MJ embryos, whereas knockdown of mtor or ulk1a alone had no apparent effects (Figure 7). Furthermore, reduced kidney fluid flow in ift140^MJ embryos was partially restored by mtor or ulk1a deficiency (Supplemental Figure 11). These findings suggest that mtor and ulk1a perform similar functions in the regulation of cilia size, cilia orientation, intracellular microtubule stability, and cilia beating-driven kidney fluid flow.

Figure 7.

mtor and ulk1a inhibition rescued ciliary and nonciliary defects in ift140^MJ embryos. (A and B) Knockdown of mtor or ulk1a both restored the length and number of distal single cilia in ift140^MJ embryos. sgRNAs targeting ift140, mtor, or ulk1a were injected individually or in combination, and distal tubular single cilia were analyzed. Representative Ac-tub and α6F immunostaining images are shown in (A), with quantifications of cilium length and number presented in (B). (C) Knockdown of mtor or ulk1a corrected multicilia bundle orientation in the posterior MCC region of ift140^MJ embryos. All ift140^MJ embryos (n=11) exhibited misoriented multicilia bundles, whereas 11 of 12 ift140^MJ;mtor^MJ and nine of 12 ift140^MJ;ulk1a^MJ embryos displayed well-aligned cilia bundles. (D) Knockdown of mtor or ulk1a mitigated aberrant intracellular Ac-tub accumulation (arrow) in the anterior MCC region of ift140^MJ embryos. Intracellular Ac-tub accumulation was observed in ten of 11 ift140^MJ embryos but was markedly reduced in nine of 12 ift140^MJ;mtor^MJ and eight of 12 ift140^MJ;ulk1a^MJ embryos. wt (n=9), mtor^MJ (n=10), and ulk1a^MJ (n=11) fish showed no such abnormalities (C and D). A total of 9–12 embryos per group at 6 dpf were analyzed across two independent experiments. The KO efficiencies for ift140, mtor, and ulk1a are 95.1%±5.3%, 90.4%±7.0%, and 70.1%±9.6%, respectively. Scale bar: 20 µm. *P < 0.001. MCC, multiciliated cell.

Interestingly, mtor and ulk1a inhibition exhibited differential effects on body length regulation in the context of ift140 deficiency. In the same clutches of injected embryos analyzed for cilia defects, ift140^MJ larvae were significantly smaller than uninjected controls at 2 weeks (Supplemental Figure 12A). Knockdown of ulk1a promoted body growth in ift140^MJ fish, whereas knockdown of mtor did not, as mtor knockdown alone suppressed growth (Supplemental Figure 12A). Consistently, ossification of bone structures was restored by ulk1a inactivation but not by mtor inactivation (Supplemental Figure 12B).

Discussion

IFT140 has recently been identified as a causative gene for ADPKD, particularly enriched in patients with milder disease presentations and no known family history of ADPKD.^5,63–66 Consistently, zebrafish ift140^e2/e2 mutants developed pronephric and mesonephric cysts. Also in line with IFT140’s role in skeletal ciliopathy,^1,2,4 zebrafish ift140^e2/e2 mutants recapitulated many skeletal phenotypes observed in mammals with IFT140 mutations. Collectively, the ift140 zebrafish model provides a valuable platform for studying kidney cystogenesis and skeletal development.

By using an MMEJ-mediated mosaic KO approach, ift140 crispants successfully circumvented the early lethality of ift140^e2/e2 fish. Adult ift140 crispants exhibited glomerular and proximal tubular cysts but not distal tubular cysts. Determining whether cysts also originate from collecting ducts will require future identification of a zebrafish collecting duct–specific marker. Although glomerular cysts are commonly observed in early-onset ADPKD cases and have been reported in adult ADPKD, including mouse models and patients,^67–70 their presence in IFT140 patients and viable hypomorphic mouse models remains to be investigated. Our findings highlighted the feasibility of dose-dependent genetic analysis of ADPKD genes in adult zebrafish using MMEJ-mediated mosaic KO. This approach enabled the study of adult phenotypes often missed in heterozygous mutants due to slow disease progression or in homozygous mutants due to early lethality. It offered a rapid alternative to tissue-specific KO while more accurately modeling whole-body genetic lesions seen in patients. For nonlethal mutations, this method would significantly accelerate research by eliminating the need for multigeneration crosses. A limitation is the potential for unclean genetic lesions, including off-target effects from sgRNA injections. Using multiple sgRNAs targeting the same gene can help distinguish shared phenotypes from unique ones.

The mechanisms underlying IFT140-associated cystogenesis remain unclear. As a component of the IFT-A complex, IFT140 facilitates retrograde trafficking of cargos from the cilium tip to the base. Consistent with this role, mutations in IFT140 result in shortened cilia and abnormal accumulation of ciliary proteins.^2–4 In addition, IFT140 has been implicated in anterograde transport of materials into the cilium, as demonstrated in Drosophila, C. elegans, and mammalian models, potentially altering cilia-mediated signaling.^7,71,72 In zebrafish ift140 mutants, kidney cilia were progressively lost, which could contribute to cyst development. However, cyst formed as early as 2 dpf when cilia were only slightly (though significantly) shortened, raising the possibility that defects beyond cilia length may contribute to cyst formation.

In addition to ciliary defects, we observed intracellular microtubule hyperacetylation in proximal tubular cells. This nonciliary phenotype closely resembles findings associated with certain IFT-B proteins, including those seen in cells with IFT88 depletion or mutations in IFT54 and IFT52, as well as in zebrafish ift52 mutants.^51,73–75 Hyperacetylation of microtubules is a posttranslational modification linked to stabilized, long-lived microtubules, which can affect cytoskeletal architecture and intracellular protein trafficking.^76,77 This abnormality may partially account for the misorientation of multicilia bundles and the randomization of basal body docking in the proximal tubular cells of ift140 mutants and crispants. Supporting this, genes involved in cell polarity and cell–cell junctions were among the top differentially expressed in kidney organoids derived from patients carrying IFT140 varients.⁷⁸ In addition, epithelial morphology alterations and mislocalization of E-cadherin and F-actin have been reported in the limb buds of Ift140 missense mutant mice.⁴ Although kidney collecting duct–specific KO of Ift140 leads to kidney cyst formation without affecting the apical localization of basal body,³ further studies are needed to investigate nonciliary defects in human patient samples and proximal tubule–specific Ift140 KO mouse models. In summary, our study is the first to suggest a nonciliary role of an IFT-A protein in kidney cyst development, emphasizing the need for future research to explore its contribution to cystogenesis.

Using an MMEJ-based F0 screening approach, we tested 16 genes implicated in dysregulated signaling pathways in ADPKD and identified both known (e.g., mtor) and novel (e.g., ulk1a) protective modifiers of ift140-associated kidney cysts.^53,54 Further analysis revealed that both mtor and ulk1a inhibition mitigated ciliary and nonciliary defects resulting from ift140 inactivation. Notably, both genes are involved in autophagy regulation. While mTOR suppresses autophagy and ULK1 promotes it, recent studies suggest that Ulk1 depletion can also trigger autophagy activation under certain conditions.^33,79–81 Future investigations to explore the interplay between autophagy and cilia are expected to provide insights into their complex and reciprocal relationship.^82,83 In addition, understanding the regulation and role of intracellular microtubule hyperacetylation will also be a key focus of future research.

Our screen was designed to identify protective modifiers of cystogenesis using the following criteria: (1) high cyst penetrance in ift140 crispants (approximately 90%), achieved by optimizing sgRNA dosage; (2) assay consistency, ensured by testing 4–5 genes plus an ift140-only control per injection session, with approximately 50 embryos per group to provide adequate statistical power; (3) adequate KO efficiency, targeting ≥50% KO for each gene. If efficiency was <50%, sgRNA dose was adjusted or an alternative sgRNA was used. In most cases, an efficient sgRNA was identified after testing two per gene; (4) modifier classification, defining strong candidates as those reducing cyst penetrance to <50%, weak candidates to 50% to 70%, and noncandidates to 70% to 90%. For future large-scale screens, we plan to omit sgRNA efficiency testing and prioritize strong candidates to increase throughput, although this may miss genes targeted by low-efficiency sgRNAs. Although protective modifiers have translational potential, deleterious modifiers are important for understanding cystogenesis mechanisms. Ideally, approximately 50% cyst penetrance in ift140 crispants would allow identification of both, but achieving this consistently is technically challenging, making the current screen unsuitable for detecting deleterious modifiers.

Since mutations in PKD1 account for approximately 85% of ADPKD cases, identifying genetic modifiers of PKD1-associated cystogenesis would have greater clinical relevance. We hypothesized that cystogenesis with different genetic etiologies may share common modifiers if they target similar pathogenic mechanisms. The protective effects of mtor and ulk1a inhibition in pkd1 zebrafish further highlighted the potential of using ift140-based screens to identify modifiers of pkd1-based kidney cysts.²² Overall, we anticipate that the MMEJ-based F0 assay will enable the rapid evaluation of a broader range of candidate modifiers, accelerating research on the genetic factors contributing to ADPKD.

Finally, although the zebrafish embryonic cyst model offers exceptional high-throughput capabilities, its conservation with mammalian PKD pathogenesis requires further validation. Although mtor inhibition has shown benefits across various kidney cyst models, the role of ulk1a inhibition has not been extensively explored. Future studies assessing mtor and ulk1a inhibition in a mouse ift140 model, ideally a viable hypomorphic mutant with global ift140 disruption, as well as ulk1a inhibition in a mouse Pkd1 model, will help determine the extent of conservation between the zebrafish embryonic model and mammalian PKD.

Supplementary Material

jasn-37-474-s001.pdf ^{(1.4MB, pdf)}

jasn-37-474-s002.pdf ^{(4MB, pdf)}

Disclosures

Disclosure forms, as provided by each author, are available with the online version of the article at http://links.lww.com/JSN/F425.

Author Contributions

Conceptualization: Xueying Lin, Xiaolei Xu.

Data curation: Andrew Lavin, Xueying Lin, Ping Zhu.

Formal analysis: Xueying Lin, Ping Zhu.

Funding acquisition: Xueying Lin.

Investigation: Xueying Lin, Ping Zhu.

Methodology: Ping Zhu.

Project administration: Xueying Lin.

Resources: Xueying Lin, Xiaolei Xu.

Supervision: Xueying Lin.

Validation: Xueying Lin, Ping Zhu.

Writing – original draft: Xueying Lin.

Writing – review & editing: Xueying Lin, Xiaolei Xu, Ping Zhu.

Funding

X. Xu and X. Lin: National Institutes of Health (R56 DK137778). X. Lin: Mayo PKD Center Zell PKD Research Innovation Fund.

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. This research was posted on a preprint server. https://doi.org/10.1101/2025.01.02.631132.

Data Availability Statement

Original data generated for the study will be made available upon reasonable request to the corresponding author. Data Type: Research Protocols; Raw Data/Source Data; Published Material; Observational Data. Reason for Restricted Access: No proteomics or transcriptomics data were generated in this study. All zebrafish experiments were conducted at the Mayo Clinic Rochester, and the data generated are presented in the figures within the main text or in the Supplemental Material. Any additional information required for re-examining or analyzing the data shown in this paper will be made available upon request. All data requests will be processed in accordance with institutional policies for noncommercial research purposes.

Supplemental Material

This article contains the following supplemental material online at http://links.lww.com/JSN/F426.

Supplemental Table 1. Sequences of MMEJ-inducing sgRNAs, genetic lesions, and PCR primers for KO efficiency assessment.

Supplemental Figure 1. Expression of ift140 in zebrafish embryos.

Supplemental Figure 2. Reduced body size in ift140^e2/e2 mutants.

Supplemental Figure 3. Inactivation of ift140 did not affect pronephric kidney development.

Supplemental Figure 4. Collagen deposition in the kidneys of adult ift140^MJ fish.

Supplemental Figure 5. The effect of ift140 inactivation on the length of distal single cilia and the orientation of multicilia bundles in the 2 dpf pronephros.

Supplemental Figure 6. Transcript levels of histone acetyltransferase and histone deacetylases.

Supplemental Figure 7. ift140^e2/e2 embryos exhibited normal apico-basolateral polarity in the distal tubular epithelial cells.

Supplemental Figure 8. Impaired kidney fluid excretion in ift140^e2/e2 embryos.

Supplemental Figure 9. The effect of ift140 inactivation on cilia in the Kupffer's vesicle, spinal canal, and nasal pit.

Supplemental Figure 10. Protective effects of mtor and ulk1a inhibition on cystogenesis in the kidney in F0 adult ift140^MJ fish.

Supplemental Figure 11. mtor and ulk1a inhibition partially restored kidney excretion function in ift140^MJ embryos.

Supplemental Figure 12. Effects of ulk1a and mtor inhibition on body size and calcified bone structures in ift140^MJ larvae.

Supplemental Figure 13. Full-length blots corresponding to those shown in Figure 1B.