ABSTRACT
A functional vertebrate kidney relies on structural units called nephrons, which are epithelial tubules with a sequence of segments each expressing a distinct repertoire of solute transporters. The transcriptiona`l codes driving regional specification, solute transporter program activation and terminal differentiation of segment populations remain poorly understood. Here, we demonstrate that the KCTD15 paralogs kctd15a and kctd15b function in concert to restrict distal early (DE)/thick ascending limb (TAL) segment lineage assignment in the developing zebrafish pronephros by repressing Tfap2a activity. During renal ontogeny, expression of these factors colocalized with tfap2a in distal tubule precursors. kctd15a/b loss primed nephron cells to adopt distal fates by driving slc12a1, kcnj1a.1 and stc1 expression. These phenotypes were the result of Tfap2a hyperactivity, where kctd15a/b-deficient embryos exhibited increased abundance of this transcription factor. Interestingly, tfap2a reciprocally promoted kctd15a and kctd15b transcription, unveiling a circuit of autoregulation operating in nephron progenitors. Concomitant kctd15b knockdown with tfap2a overexpression further expanded the DE population. Our study reveals that a transcription factor-repressor feedback module employs tight regulation of Tfap2a and Kctd15 kinetics to control nephron segment fate choice and differentiation during kidney development.
INTRODUCTION
Mammalian kidney organogenesis is unique in that it entails three stages of assembly – the pronephros, the mesonephros and the metanephros – in which each successive version becomes more morphologically complex than the previous (Costantini and Kopan, 2010). Other vertebrates, such as fish and amphibians, undergo two phases of kidney development, the final form being the mesonephros (Romagnani et al., 2013). Importantly, each kidney version across these animals shares a conserved structural unit: the nephron. As such, models such as fish and frogs have surfaced as powerful organisms for studying the genetic regulation of nephron development (Desgrange and Cereghini, 2015; Naylor, Qubisi, and Davidson, 2017).
The nephron is the functional unit of the kidney and comprises a glomerulus and an epithelial tubule, which connect to an arborized collecting duct system for waste excretion (Costantini and Kopan, 2010). The tubule is compartmentalized into a series of proximal, intermediate and distal segments. A key gap in knowledge remains in understanding the developmental signals necessary for the differentiation of segment-specific nephron cell types. More specifically, the genetic regulation of loop of Henle (LOH) formation remains highly understudied. The LOH is partitioned into three limbs and initiates a concentration gradient by transporting water, sodium chloride and potassium ions (Bankir et al., 2020). Zebrafish, being an aquatic species, lack the first two limbs of the LOH, which is consistent with the fact that water conservation is not physiologically requisite (Wingert et al., 2007; Wingert and Davidson, 2008). However, zebrafish do possess a distal early (DE) segment that is analogous to the mammalian thick ascending limb (TAL) of the LOH, which expresses a conserved suite of solute transporters, including the orthologues of human SLC12A1 (also known as NKCC2), KCNJ1, and CLCNK (Wingert et al., 2007; Wingert and Davidson, 2008). Defective TAL transporters are associated with neonatal and classic Bartter Syndrome, ion imbalance, polyuria, and renal failure (Walsh and Unwin, 2012). Accordingly, high-throughput chemical screens in zebrafish have facilitated the discovery of genes and pathways likely linked to renal tubular disorders (Poureetezadi et al., 2016; Chambers et al., 2018; Marra et al., 2019a). More recently, a forward genetic screen and subsequent mutant analysis identified transcription factor AP2α (tfap2a) as a key regulator of the DE/TAL terminal differentiation program (Chambers et al., 2019). TFAP2A belongs to the AP2 family of transcription factors, which generally function in an embryonic context to control proliferation and differentiation (Eckert et al., 2005).
Two Tfap2a gene regulatory network (GRN) candidates are the zebrafish paralogs potassium channel tetramerization domain containing 15a and 15b (kctd15a and kctd15b), which belong to the potassium channel tetramerization domain family. Proteins in this family harbor a conserved BTB/POZ (BR-C, ttk and bab/Pox virus and zinc finger) protein-protein interaction motif situated at the N terminus, however, demonstrate significant structural variability outside this region. The BTB/POZ domain is essential for targeting proteins for ubiquitylation and degradation (Gharbi et al., 2012). A diverse set of biological functions are assigned to these proteins, such as transcriptional repression, gating of ion channels, regulating cytoskeletal elements and acting as adaptor molecules. Mutations in KCTD genes can initiate human diseases such as breast cancer, medulloblastoma, epilepsy, pulmonary inflammation and obesity, highlighting the importance of further functional investigation (Liu et al., 2013). Humans, mice and Xenopus possess one KCTD15 gene; however, zebrafish acquired kctd15a and kctd15b paralogous versions due to an ancient teleost genome duplication event after their divergence from tetrapods (Amores et al., 1998; Postlethwait et al., 1998). Zebrafish Kctd15a and Kctd15b amino acid residues exhibit a high degree of conservation with both splice variants of human KCTD15 protein and nearly identically alignment in the BTB/POZ functional motif (Fig. S1).
In zebrafish, kctd15a and kctd15b inhibit neural crest development by two distinct modules: attenuation of canonical Wnt signaling and direct repression of Tfap2a. KCTD15 strongly inhibits TFAP2A by binding to its proline-rich activation domain (Zarelli and Dawid, 2013a; Wong et al., 2016). Kctd15 is a substrate for SUMOylation; this post-translational modification is associated with transcriptional repression. However, previous studies suggest non-sumoylated Kctd15 functions during neural crest formation (Zarelli and Dawid, 2013b). Genome-wide association studies have revealed connections between Kctd15 and AP2 to metabolic conditions such as obesity, diabetes and eating disorders (Gamero-Villarroel et al., 2017; Smaldone et al., 2018). Contrary to neural crest development, in Drosophila, Kctd15 facilitates SUMOylation of Tfap2b to regulate consummatory behavior and repress the transduction of adipogenesis and insulin signaling pathways (Williams et al., 2012, 2014). These findings authenticate the need for tissue-specific interrogation of Kctd15 function. Here, we demonstrate zebrafish kctd15 paralogs are novel regulators of nephron segment commitment and inhibit DE/TAL differentiation by participating in repressor-mediated genetic feedback with tfap2a.
RESULTS
kctd15a and kctd15b are expressed in the developing pronephros
Zebrafish and frog studies have previously reported kctd15 transcript expression in the pronephros (Dutta and Dawid, 2010; Takahashi et al., 2012). Data from the Genitourinary Development Molecular Anatomy Project (GUDMAP) has also documented KCTD15 expression in nascent mammalian nephrons. Microarray results indicate upregulated Kctd15 expression in developing renal vesicles in E12.5 mice (RID:Q-6PTG). Single cell RNA-sequencing of a week 17 human kidney cortex have revealed elevated KCTD15 expression clustered with the developing nephron populace (RID:16-5HBT) (McMahon et al., 2008; Harding et al., 2011). Taken together, these findings suggested that kctd15a and kctd15b are possible candidates for coordinating nephrogenesis. To expand on these previous studies, we assembled a comprehensive time course of kctd15a and kctd15b spatial expression over the span of kidney development using the genetically tractable zebrafish pronephros model. Using whole-mount in situ hybridization analysis of kctd15a and kctd15b in wild-type animals, transcripts were first detectable in the intermediate mesoderm (IM) at the 10-somite stage (ss) (Fig. 1A, Fig. S2). Throughout the duration of pronephric development, kctd15a and kctd15b showed similar expression patterns within the distal nephron progenitor domain (Fig. 1B). Because the expression patterns of these factors in the developing kidney show minimal deviation from one another, it is highly possible that they function redundantly in this context.
kctd15a and kctd15b are expressed in developing distal nephron precursors. (A) Whole-mount in situ hybridization of kctd15a and kctd15b during wild-type embryogenesis. Black arrowheads indicate expression within the developing renal field. Scale bar: 200 μm. (B) Whole-mount in situ hybridization of kctd15a and kctd15b during wild-type embryogenesis. Black arrowheads indicate expression within distal nephron precursors. Scale bar: 200 μm. (C) Fluorescent in situ hybridization of kctd15a (magenta) and tfap2a (green) in the 22 ss wild-type distal nephron. White rectangle outlines the region in the insets. Scale bars: 5 μm (insets); 35 μm (main image). (D,E) Fluorescent in situ hybridization of tfap2a (green), kctd15a (magenta in D) and kctd15b (magenta in E) at 24 hpf in the pronephros. White rectangle outlines the region shown at higher magnification (bottom panels). Cyan outline indicates an example of a co-expressing cell. Scale bars: 35 μm (top); 5 μm (bottom). Yellow arrowheads indicate intense areas of transcript co-expression in C-E.
kctd15a and kctd15b are expressed in developing distal nephron precursors. (A) Whole-mount in situ hybridization of kctd15a and kctd15b during wild-type embryogenesis. Black arrowheads indicate expression within the developing renal field. Scale bar: 200 μm. (B) Whole-mount in situ hybridization of kctd15a and kctd15b during wild-type embryogenesis. Black arrowheads indicate expression within distal nephron precursors. Scale bar: 200 μm. (C) Fluorescent in situ hybridization of kctd15a (magenta) and tfap2a (green) in the 22 ss wild-type distal nephron. White rectangle outlines the region in the insets. Scale bars: 5 μm (insets); 35 μm (main image). (D,E) Fluorescent in situ hybridization of tfap2a (green), kctd15a (magenta in D) and kctd15b (magenta in E) at 24 hpf in the pronephros. White rectangle outlines the region shown at higher magnification (bottom panels). Cyan outline indicates an example of a co-expressing cell. Scale bars: 35 μm (top); 5 μm (bottom). Yellow arrowheads indicate intense areas of transcript co-expression in C-E.
Next, we wanted to determine how the kctd15a and kctd15b expression domains spatially correlate to tfap2a transcripts in the developing pronephros, as these factors repress Tfap2a in neural crest (Dutta and Dawid, 2010; Zarelli and Dawid, 2013a). Using fluorescent in situ hybridization, kctd15a transcripts were found to be co-expressed throughout the tfap2a+ distal populace in wild-type embryos at 22 ss. (Fig. 1C). At 24 hours post fertilization (hpf), kctd15a and kctd15b were expressed throughout the tfap2a+ pronephric domain (Fig. 1D,E). We confirmed that our kctd15a and kctd15b antisense probes were specific, and found that transcripts were significantly enriched in the pronephros when compared with baseline fluorescent signal detected in the somites (Fig. S2). These experiments demonstrate kctd15a and kctd15b are expressed in developing distal precursors, and reveal that their localization largely overlaps with tfap2a at multiple timepoints during development.
kctd15a and kctd15b regulate the expression of distal nephron markers
In zebrafish, kctd15a/b deficiency leads to an expansion of neural crest fate most likely due to the inability to repress Tfap2a (Dutta and Dawid, 2010; Heffer et al., 2017). Because kctd15a/b function had never been studied in the context of kidney organogenesis, we developed several loss- and gain-of-function strategies to assess their involvement in nephron formation. Knockdown of kctd15a and kctd15b was achieved by injecting morpholinos to disrupt splicing between the exon 1 and exon 2 junctions and validated by RT-PCR analyses (Fig. S3). For a parallel independent loss of function model, we genetically engineered F0 kctd15a, kctd15b and kctd15a/b CRISPR mutants (crispants) by multiplexing sgRNAs targeting different regions of the BTB/POZ functional domain to induce biallelic disruptions (Fig. S4). Although kctd15a/b TALEN mutants have been previously generated for neural crest studies, these animals exhibit mild defects and can survive to adulthood suggesting genetic compensation may be masking true loss-of-function phenotypes (Heffer et al., 2017). Further, the Mouse Genome Informatics database reports that Kctd15 null mice exhibit complete penetrance of preweaning lethality (MGI:5,548,790) (Bult et al., 2019). Therefore, we prioritized experiments on kctd15a/b crispants. T7 endonuclease assays and Sanger sequencing confirmed successful genome editing (Fig. S4). kctd15a/b morphants, kctd15a crispants and kctd15b crispants developed pericardial edema and increased dorsal head pigmentation by 48 hpf, consistent with other live phenotype reports; however, these crispants were embryonic lethal, consistent with possession of more severe defects than previous kctd15a/b TALEN mutants (Figs S3 and S4) (Dutta and Dawid, 2010; Heffer et al., 2017).
Because kctd15 transcripts localized to the developing pandistal pronephric region (Fig. 1), we first decided to survey for alterations in DE and DL marker expression, which are analogous to the mammalian TAL and distal convoluted tubule (DCT), respectively. kctd15a, kctd15b and kctd15a/b morphants and crispants exhibited significant expansion in the expression of the DE markers kcnj1a.1 and slc12a1 when compared with wild-type controls (Fig. 2A-D). In each kctd15-deficient state, slc12a1+ cells encroached into the DL territory marked by slc12a3 (Fig. 2C). Notably, a previous study indicated that when tfap2a is overexpressed, an expansion of DE solute transporter expression occurs within the pronephros (Chambers et al., 2019). These parallel phenotypes highlight the possibility that this alteration of distal nephron fate could be due to inadequate repression of Tfap2a protein by Kctd15a and Kctd15b. Generally, kctd15 crispants manifested slightly milder DE phenotypes when compared with kctd15 morphants, most likely because F0 crispant cellular composition is mosaic in nature and generates a mix of homozygous and heterozygous mutants. Nevertheless, kctd15 crispants recapitulated kctd15 morphant distal nephron phenotypes corroborating the use of MO-mediated knockdown in subsequent experiments. In addition, when we performed incrosses of F0 kctd15b crispants, this produced F1 progeny, which recapitulated the expanded slc12a1 marker expression (Fig. S4). As expected, compound MO knockdown of kctd15a and kctd15b elicited the greatest expansion in expression of DE markers, suggesting that these paralogs maintained redundant functions in teleosts (Fig. 2B,D). Overexpression of kctd15a cRNA produced a significantly decreased DE in conjunction with an expanded DL segment (Fig. 2A-D). kctd15a overexpression also caused a substantial decline in kcnj1a.1+ ionocyte number (Fig. S5). Matching phenotypes occurred upon kctd15b overexpression (data not shown). Next, we assessed Slc12a1 protein expression in kctd15a/b morphants. We found that kctd15a/b morphants exhibit a significant expansion of the Slc12a1 protein domain compared with wild-type embryos (Fig. 2E). Generation of fluorescent intensity profiles confirmed a significant elevation in Slc12a1 signal in kctd15a/b-deficient embryos compared with wild-type controls (Fig. 2F,G). Upon quantifying Slc12a1+ cells, we found that the kctd15a/b morphants formed nearly double the total number of DE cells found in wild-type embryos (Fig. 2H, Fig. S6).
kctd15a/b loss-of-function initiates expansion of DE lineage markers. (A) Whole-mount in situ hybridization of kcnj1a.1 (purple) at 24 hpf in the pronephros of wild type, wild type+kctd15a cRNA, three kctd15 MOs and three kctd15 crispants. Scale bar: 35 μm. (B) Absolute length quantification of the kcnj1a.1 expression domain. Black brackets cluster groups together for collective comparison. (C) Whole-mount in situ hybridization of slc12a1 (purple) and slc12a3 (red) at 24 hpf in the pronephros of wild type, wild type+kctd15a cRNA, three kctd15 MOs and three kctd15 crispants. Scale bar: 35 μm. (D) Absolute length quantification of slc12a1 expression domain. Black brackets cluster groups together for collective comparison. (E) Immunofluorescence of Slc12a1/2 (magenta) and laminin (green) in wild type and kctd15a/b MO at 24 hpf in pronephric cells. White arrowheads mark the limits of Slc12a1/2 pronephric expression. Scale bar: 35 μm. (F) Fluorescent intensity plot of three wild-type (grayscale) and three kctd15a/b MO (blue) individuals. Green dashed line equates to wild-type Slc12a1/2 intensity threshold (au). Red dashed line represents wild-type maximum Slc12a1/2 intensity value (au). (G) Scatterplot of Slc12a1/2 fluorescence intensity values. (H) Quantification of number of Slc12a1+ pronephric cells per nephron. n≥3 embryos quantified for each control and experimental group. *P<0.05; **P<0.01; ***P<0.001. Data are mean±s.d. Absolute lengths were compared using ANOVA. Fluorescent intensities and cell counts were analyzed using unpaired t-tests.
kctd15a/b loss-of-function initiates expansion of DE lineage markers. (A) Whole-mount in situ hybridization of kcnj1a.1 (purple) at 24 hpf in the pronephros of wild type, wild type+kctd15a cRNA, three kctd15 MOs and three kctd15 crispants. Scale bar: 35 μm. (B) Absolute length quantification of the kcnj1a.1 expression domain. Black brackets cluster groups together for collective comparison. (C) Whole-mount in situ hybridization of slc12a1 (purple) and slc12a3 (red) at 24 hpf in the pronephros of wild type, wild type+kctd15a cRNA, three kctd15 MOs and three kctd15 crispants. Scale bar: 35 μm. (D) Absolute length quantification of slc12a1 expression domain. Black brackets cluster groups together for collective comparison. (E) Immunofluorescence of Slc12a1/2 (magenta) and laminin (green) in wild type and kctd15a/b MO at 24 hpf in pronephric cells. White arrowheads mark the limits of Slc12a1/2 pronephric expression. Scale bar: 35 μm. (F) Fluorescent intensity plot of three wild-type (grayscale) and three kctd15a/b MO (blue) individuals. Green dashed line equates to wild-type Slc12a1/2 intensity threshold (au). Red dashed line represents wild-type maximum Slc12a1/2 intensity value (au). (G) Scatterplot of Slc12a1/2 fluorescence intensity values. (H) Quantification of number of Slc12a1+ pronephric cells per nephron. n≥3 embryos quantified for each control and experimental group. *P<0.05; **P<0.01; ***P<0.001. Data are mean±s.d. Absolute lengths were compared using ANOVA. Fluorescent intensities and cell counts were analyzed using unpaired t-tests.
Furthermore, we examined a distal nephron associated cell type: the corpuscle of Stannius (CS). Briefly, the CS is an aggregate of cells that initially develop between the DE and DL segments that eventually leave the pronephric tubule to form a dorsally situated endocrine gland that maintains calcium homeostasis (Cheng and Wingert, 2015). A recent study illustrated that CS cells transdifferentiate from the DE segment by a gland extrusion mechanism (Naylor et al., 2018). kctd15a, kctd15b and kctd15a/b morphants and crispants all possess elevated numbers of CS cells, indicated by increased expression of stc1 as compared with wild type (Fig. 3A,B). Surprisingly, we noticed ectopic stc1+ cells formed at the proximal end of the DE domain that were clearly separated from the principal CS cell cluster in kctd15a/b morphant embryos (Fig. 3A). Upon closer examination, we discovered that kctd15a/b morphants underwent premature CS gland extrusion at 24 hpf, as this group of cells formed a basement membrane and separated from the pronephric tubule (Fig. 3C). As tfap2a deficiency affects CS differentiation, it is not surprising that disrupting the balance of kctd15a/b also alters this process (Chambers et al., 2019). Collectively, our genetic experiments revealed alterations in DE, DL and CS cell differentiation, identifying kctd15 as a novel regulator of distal nephron identity.
kctd15a/b deficiency elevates CS differentiation. (A) Whole-mount in situ hybridization of slc12a1 (red) and stc1 (purple) at 24 hpf in pronephric cells of wild type, three kctd15 MO and three kctd15 crispants. Black arrowheads indicate proximal stray stc1+ cells clearly separate from the main CS cluster in kctd15a/b MO. Scale bar: 35 μm. (B) Quantification of stc1+ cell number per nephron. Black brackets cluster groups together for collective comparison. (C) Immunofluorescence of laminin (green) in the pronephros of 24 hpf wild type and kctd15a/b MO. Premature basement membrane formation occurs in the kctd15a/b morphant, which separates the budding CS cell cluster (white dotted circle) from the pronephric tubule. Scale bar: 5 μm. n≥3. *P<0.05; ***P<0.001. Data are mean±s.d. Cell counts were compared using ANOVA. Data are displayed as individual points distributed about the mean (black line).
kctd15a/b deficiency elevates CS differentiation. (A) Whole-mount in situ hybridization of slc12a1 (red) and stc1 (purple) at 24 hpf in pronephric cells of wild type, three kctd15 MO and three kctd15 crispants. Black arrowheads indicate proximal stray stc1+ cells clearly separate from the main CS cluster in kctd15a/b MO. Scale bar: 35 μm. (B) Quantification of stc1+ cell number per nephron. Black brackets cluster groups together for collective comparison. (C) Immunofluorescence of laminin (green) in the pronephros of 24 hpf wild type and kctd15a/b MO. Premature basement membrane formation occurs in the kctd15a/b morphant, which separates the budding CS cell cluster (white dotted circle) from the pronephric tubule. Scale bar: 5 μm. n≥3. *P<0.05; ***P<0.001. Data are mean±s.d. Cell counts were compared using ANOVA. Data are displayed as individual points distributed about the mean (black line).
kctd15a and kctd15b depletion initiates ectopic DE formation in flanking pronephros segments
Next, we decided to examine the effect of dual kctd15a/b knockdown on the neighboring nephron segment populations, as zebrafish kctd15a and kctd15b share over 90% amino acid sequence identity and likely have overlapping functions (Fig. S1). We performed whole-mount in situ hybridization to mark the proximal straight tubule (PST) and the DL, and discovered that each of these segments were significantly decreased in kctd15a/b morphants (Fig. 4A-D). Upon quantifying the gap between the PST and DL, which correlates with the inferred DE footprint, we found the length of this space was significantly expanded in the kctd15a/b morphants (Fig. 4C). However, the average length of this inferred gap tended to be smaller than the actual dimensions of the DE segment in kctd15a/b morphants (Fig. 2B,D), so we hypothesized that DE cells were also located in the adjacent segment domains. To visualize segment boundaries at higher resolution, we employed fluorescent in situ hybridization to first assess the DE and DL nephron segments marked by kcnj1a.1 and slc12a3, respectively. In 24 hpf wild-type animals, there is a sharp separation between the DE and DL segment compartments (Fig. 4E,F). Conversely, in kctd15a/b morphants, kcnj1a.1+ cells resided within the DL domain (Fig. 4E). This significant overlap of DE and DL segment identities was made evident upon generation of a fluorescent intensity plot depicting a representative kctd15a/b morphant (Fig. 4G). Some of these cells dually expressed kcnj1a.1 and slc12a3, suggesting segment fate infidelity (Fig. 4G). This dual marker expression is highly reminiscent of the resultant phenotype triggered by overexpression of tfap2a (Chambers et al., 2019). These data indicate that kctd15a/b deficiency results in an expansion of the DE lineage at the expense of the DL.
Loss of kctd15a/b sways neighboring pronephric fates to express DE signature. (A) Whole-mount in situ hybridization for trpm7 (purple) and slc12a3 (red) in wild-type and kctd15a/b MO pronephric cells at 24 hpf. Blue arrowheads flank the inferred DE footprint; black arrowheads indicate anterior and posterior limits of the PST and DL. Scale bar: 35 μm. (B-D) Absolute length quantifications of the trpm7 expression domain, inferred DE footprint and slc12a3 expression domain. (E) Fluorescent in situ hybridization of kcnj1a.1 (green) and slc12a3 (red) at 24 hpf in pronephric cells in wild type and kctd15a/b MO. Magenta and cyan arrowheads indicate DE and DL boundaries, respectively. Gray box indicates the area shown at higher magnification below. White outlines indicate individual cells dually expressing kcnj1a.1 and slc12a3. Scale bars: 70 μm (top); 5 μm (bottom). (F) Wild-type fluorescent intensity plot of kcnj1a.1 (green) and slc12a3 (red). (G) kctd15a/b MO fluorescent intensity plot of kcnj1a.1 (green) and slc12a3 (red). White arrows indicate ectopic kcnj1a.1 signal invading the slc12a3 domain. (H) Fluorescent in situ hybridization of trpm7 (red) and kcnj1a.1 (green) at 24 hpf in wild-type and kctd15a/b MO pronephric tubules. Cyan and magenta arrowheads indicate PST and DE boundaries, respectively. Gray box indicates area shown at higher magnification below. White outlines indicate individual cells dually expressing kcnj1a.1 and trpm7. Scale bars: 70 μm (top); 5 μm (bottom). (I) Wild-type fluorescent intensity plot of trpm7 (red) and kcnj1a.1 (green). (J) kctd15a/b MO fluorescent intensity plot of trpm7 (red) and kcnj1a.1 (green). White arrows indicate ectopic kcnj1a.1 signal invading the trpm7 domain. n≥3 embryos quantified for each control and experimental group. ***P<0.001. Data are mean±s.d. Absolute lengths were compared using unpaired t-tests.
Loss of kctd15a/b sways neighboring pronephric fates to express DE signature. (A) Whole-mount in situ hybridization for trpm7 (purple) and slc12a3 (red) in wild-type and kctd15a/b MO pronephric cells at 24 hpf. Blue arrowheads flank the inferred DE footprint; black arrowheads indicate anterior and posterior limits of the PST and DL. Scale bar: 35 μm. (B-D) Absolute length quantifications of the trpm7 expression domain, inferred DE footprint and slc12a3 expression domain. (E) Fluorescent in situ hybridization of kcnj1a.1 (green) and slc12a3 (red) at 24 hpf in pronephric cells in wild type and kctd15a/b MO. Magenta and cyan arrowheads indicate DE and DL boundaries, respectively. Gray box indicates the area shown at higher magnification below. White outlines indicate individual cells dually expressing kcnj1a.1 and slc12a3. Scale bars: 70 μm (top); 5 μm (bottom). (F) Wild-type fluorescent intensity plot of kcnj1a.1 (green) and slc12a3 (red). (G) kctd15a/b MO fluorescent intensity plot of kcnj1a.1 (green) and slc12a3 (red). White arrows indicate ectopic kcnj1a.1 signal invading the slc12a3 domain. (H) Fluorescent in situ hybridization of trpm7 (red) and kcnj1a.1 (green) at 24 hpf in wild-type and kctd15a/b MO pronephric tubules. Cyan and magenta arrowheads indicate PST and DE boundaries, respectively. Gray box indicates area shown at higher magnification below. White outlines indicate individual cells dually expressing kcnj1a.1 and trpm7. Scale bars: 70 μm (top); 5 μm (bottom). (I) Wild-type fluorescent intensity plot of trpm7 (red) and kcnj1a.1 (green). (J) kctd15a/b MO fluorescent intensity plot of trpm7 (red) and kcnj1a.1 (green). White arrows indicate ectopic kcnj1a.1 signal invading the trpm7 domain. n≥3 embryos quantified for each control and experimental group. ***P<0.001. Data are mean±s.d. Absolute lengths were compared using unpaired t-tests.
To probe the rostral DE boundary and its adjacent PST segment, trpm7 expression was assessed. Wild-type animals have a well-defined DE/PST boundary (Fig. 4H,I). Contrary to this, kctd15a/b morphant DE cells occupied the trpm7+ PST region (Fig. 4H,J). Additionally, some cells co-expressed kcnj1a.1 and trpm7, indicating that kctd15a/b-deficiency was potent enough to induce proximal cells to express a distal nephron signature gene (Fig. 4H,J). These findings led us to investigate whether the PST-DE boundary could be regulated in a cell-autonomous fashion by kctd15. We found that at the 24 ss, kctd15a is expressed in the distal-most section of the developing PST segment, and that this zone of overlap approximately spans the length of a somite (Fig. S7). Therefore, kctd15 genes may determine the PST boundary in a cell-autonomous manner during nephron segmentation. Taken together, these data suggest that kctd15a/b factors function to suppress DE lineage in neighboring segment populations during nephrogenesis.
kctd15a/b knockdown expands Tfap2a expression in the developing kidney
To evaluate whether kctd15a/b are regulating nephron differentiation by affecting Tfap2a localization or stability, we performed immunofluorscence experiments on 24 hpf kctd15a/b morphants. kctd15a/b morphants showed expanded pronephric expression of Tfap2a protein compared with wild type (Fig. 5A). Specifically, this expansion of Tfap2a expression in kctd15a/b morphants was characterized by signal detected in the proximal nephron (Fig. 5A). This greatly contrasts with wild type, which exhibits an abrupt cutoff of Tfap2a protein expression at the proximal limit of the Slc12a1+ DE (Fig. 5A). Generation of fluorescent intensity profiles revealed differentially elevated Tfap2a expression in kctd15a/b morphants that was particularly evident within a 100-220 µm region that mapped to the DE territory (Fig. 5B). kctd15a/b morphants also exhibited a significant elevation of Tfap2a fluorescent signal within the 100-220 µm zone (Fig. 5C). Additionally, loss of kctd15a and kctd15b increased nuclear abundance of Tfap2a protein, as we detected elevated mean fluorescent intensity within individual pronephric nuclei (Fig. 5D). Furthermore, kctd15a/b morphants exhibited a statistically significant increase in Tfap2a+ pronephric nuclei compared with wild-type controls (Fig. 5E). We found no differences in nuclear size or number in kctd15a/b MO nephrons when compared with wild type at 24 hpf, and therefore concluded that the changes observed in Tfap2a protein expression are not due to significant alterations in tubule cell morphology (Fig. 5F-H). In summary, these results indicate that, when kctd15a and kctd15b are abrogated, the number of Tfap2a-expressing cells increases in the pronephros and simultaneously parallels an expansion of the Slc12a1+ DE. Our data illustrating alterations in pronephric Tfap2a protein expression reveals a new mechanistic layer that is distinct from previous in vitro studies, which found nuclear extracts exhibited no changes in Tfap2a protein abundance in the absence of Kctd15 and the main mode of inhibition was executed by direct binding to the Tfap2a transactivation domain (Zarelli and Dawid, 2013a). Whole-mount immunofluorescence allowed for cellular changes in Tfap2a protein expression to be tracked in specific nephron regions. Based on these immunofluorescence results, we hypothesized that expanded Tfap2a pronephric expression is the root cause of the DE lineage alterations that occur in kctd15a/b morphants.
kctd15a/b knockdown expands Tfap2a protein expression in the pronephros. (A) Whole-mount immunofluorescence of Tfap2a and Slc12a1/2 in wild type and kctd15a/b MO. Scale bars: 35 μm (top); 5 μm (bottom). White arrowheads indicate the limits of the pronephric Tfap2a expression domain. Red rectangle highlights the region shown at higher magnification below. Dotted lines outline the pronephric tubule. (B) Fluorescent intensity plot of Tfap2a featuring one representative wild-type (grayscale) and kctd15a/b MO (blue) sample. Purple arrows label differential Tfap2a signal peaks. Green bar spans the region of differential expression corresponding to DE locale (100-220 μm). (C) Scatterplot of Tfap2a fluorescent intensities (au) across the 100-220 μm region in three individual wild-type and MO samples. (D) Fluorescent intensity quantification Tfap2a signal within individual nuclei corresponding to the DE locale. Five individual nuclei were measured per sample. (E) Graph depicting the number of Tfap2a+ nuclei per nephron. (F) DAPI stain in the DE region of wild type and kctd15a/b MO. Magenta dots outline the pronephric tubule and yellow outlines a single nuclei. (G) Analysis of nuclear size. (H) Quantification of total number of nuclei per nephron within pandistal region. n≥3 embryos quantified for each control and experimental group. ***P<0.001; N.S., not significant. Data are mean±s.d. Fluorescent intensities and cell counts were analyzed using unpaired t-tests.
kctd15a/b knockdown expands Tfap2a protein expression in the pronephros. (A) Whole-mount immunofluorescence of Tfap2a and Slc12a1/2 in wild type and kctd15a/b MO. Scale bars: 35 μm (top); 5 μm (bottom). White arrowheads indicate the limits of the pronephric Tfap2a expression domain. Red rectangle highlights the region shown at higher magnification below. Dotted lines outline the pronephric tubule. (B) Fluorescent intensity plot of Tfap2a featuring one representative wild-type (grayscale) and kctd15a/b MO (blue) sample. Purple arrows label differential Tfap2a signal peaks. Green bar spans the region of differential expression corresponding to DE locale (100-220 μm). (C) Scatterplot of Tfap2a fluorescent intensities (au) across the 100-220 μm region in three individual wild-type and MO samples. (D) Fluorescent intensity quantification Tfap2a signal within individual nuclei corresponding to the DE locale. Five individual nuclei were measured per sample. (E) Graph depicting the number of Tfap2a+ nuclei per nephron. (F) DAPI stain in the DE region of wild type and kctd15a/b MO. Magenta dots outline the pronephric tubule and yellow outlines a single nuclei. (G) Analysis of nuclear size. (H) Quantification of total number of nuclei per nephron within pandistal region. n≥3 embryos quantified for each control and experimental group. ***P<0.001; N.S., not significant. Data are mean±s.d. Fluorescent intensities and cell counts were analyzed using unpaired t-tests.
kctd15a, kctd15b and tfap2a engage in genetic feedback circuitry
Because we discovered heightened Tfap2a protein signal in the pronephros in response to kctd15a/b deficiency (Fig. 5), we postulated a genetic feedback mechanism might be responsible for this phenotype. In support of this concept, a previous study reported kctd15a/b zebrafish mutants manifest upregulated tfap2a neural expression at the 8-9 ss (Heffer et al., 2017). To probe potential genetic feedback processes, we performed fluorescent in situ hybridization at the 10 ss to detect fluctuations in tfap2a transcript expression. kctd15a/b knockdown led to measurable expansions of tfap2a mRNA expression within the IM populace, which comprises renal progenitors that give rise to the pronephros (Fig. 6A). In wild-type flatmounts, the rostral end of tfap2a IM expression aligns with the somite 8 landmark. However, in kctd15a/b morphants, this expression domain extends in the rostral direction well beyond somite 8 (Fig. 6A). We also noted elevated tfap2a expression in the kctd15a/b-deficient hindbrain (Fig. 6A). We then surveyed tfap2a IM fluorescent intensity and standardized the region of data collection by tracking values from somite 7 to the end of the pre-somitic mesoderm. Plot overlay of tfap2a fluorescent intensity profiles from representative samples revealed morphant IM signal was consistently elevated above wild-type levels (Fig. 6B). Furthermore, tfap2a fluorescent intensity was significantly higher in kctd15a/b morphants compared with wild-type controls (Fig. 6C).
kctd15a, kctd15b and tfap2a participate in genetic crosstalk. (A) Whole-mount fluorescent in situ hybridization of tfap2a (green) and smyhc1 (magenta) in 10 ss wild-type and kctd15a/b MO flatmounts. Yellow box indicates differential tfap2a hindbrain expression. Red arrowheads indicate tfap2a IM stripe expression limits. Numbers label somite 7 and 8 landmarks. White rectangle indicates the region depicted in the panel below at higher magnification. Scale bars: 100 μm (top); 5 μm (bottom). (B) Fluorescent intensity plot of tfap2a IM expression featuring one representative wild-type (grayscale) and kctd15a/b MO (green) sample. Blue dotted line signifies wild-type mean fluorescent intensity threshold. (C) Quantification of tfap2a IM fluorescent intensity (au). Fluorescent intensity values were collected from somite 7 to the end of the pre-somitic mesoderm. (D) Whole-mount in situ hybridization of kctd15a or kctd15b (purple) in wild-type and trm−/− mutants at 24 hpf. Inset features neural expression. Right panel features pronephric expression. Scale bars: 70 μm (left); 35 μm (right). (E) Whole-mount in situ hybridization of kctd15a or kctd15b (purple) in wild-type and hs:tfap2a at 24 hpf. HS+ (red) signifies heat-shock treatment at the 8 ss. Scale bar: 70 μm. n≥3. ***P<0.001. Data are mean±s.d. Fluorescent intensities were analyzed using unpaired t-tests.
kctd15a, kctd15b and tfap2a participate in genetic crosstalk. (A) Whole-mount fluorescent in situ hybridization of tfap2a (green) and smyhc1 (magenta) in 10 ss wild-type and kctd15a/b MO flatmounts. Yellow box indicates differential tfap2a hindbrain expression. Red arrowheads indicate tfap2a IM stripe expression limits. Numbers label somite 7 and 8 landmarks. White rectangle indicates the region depicted in the panel below at higher magnification. Scale bars: 100 μm (top); 5 μm (bottom). (B) Fluorescent intensity plot of tfap2a IM expression featuring one representative wild-type (grayscale) and kctd15a/b MO (green) sample. Blue dotted line signifies wild-type mean fluorescent intensity threshold. (C) Quantification of tfap2a IM fluorescent intensity (au). Fluorescent intensity values were collected from somite 7 to the end of the pre-somitic mesoderm. (D) Whole-mount in situ hybridization of kctd15a or kctd15b (purple) in wild-type and trm−/− mutants at 24 hpf. Inset features neural expression. Right panel features pronephric expression. Scale bars: 70 μm (left); 35 μm (right). (E) Whole-mount in situ hybridization of kctd15a or kctd15b (purple) in wild-type and hs:tfap2a at 24 hpf. HS+ (red) signifies heat-shock treatment at the 8 ss. Scale bar: 70 μm. n≥3. ***P<0.001. Data are mean±s.d. Fluorescent intensities were analyzed using unpaired t-tests.
Next, we sought to determine whether kctd15a and kctd15b expression are affected by tfap2a deficiency; therefore, we employed terminus (trm) mutants, which harbor a G>A mutation that alters the donor splice site of exon 1 in the tfap2a gene (Chambers et al., 2019). trm mutants exhibited considerable reductions in kctd15a and kctd15b mRNA expression at 24 hpf when compared with wild-type controls (Fig. 6D). These visual reductions in kctd15a/b expression were observed in both neural and pronephric tissues. trm−/− mutants do have an intact pronephric tubule, so the diminished kctd15a/b expression observed is not due to loss of this organ (Chambers et al., 2019). We then tested whether overexpression of tfap2a via a heat-shock inducible transgene is sufficient to activate kctd15a/b transcription. hs:tfap2a animals that underwent heat-shock treatment (HS+) at the 8 ss displayed a global escalation of kctd15a and kctd15b expression (Fig. 6E). Surprisingly, hs:tfap2a control animals that did not undergo heat-shock treatment also exhibited increased overall kctd15a and kctd15b expression, but to a lesser extent than heat-shock-treated siblings (Fig. 6E). We believe this is an artifact of a ‘leaky’ tfap2a transgene. Nonetheless, these data illustrate that kctd15a/b transcript expression is extremely sensitive to tfap2a dose. Altogether, our findings expose a previously undescribed level of genetic interaction where tfap2a regulates the transcription of its own repressors kctd15a and kctd15b. Excitingly, these results reveal an autoregulatory-feedback loop operating in the context of nephron differentiation where tfap2a controls its own abundance through the kctd15a/b genetic circuit.
kctd15a and kctd15b counterbalance tfap2a to restrict DE differentiation during pronephros segmentation
To further substantiate that kctd15a and kctd15b functions in nephron development by repressing Tfap2a activity, we disrupted the stoichiometry of these factors by concomitant kctd15b knockdown and induced tfap2a overexpression at the 12 ss via the heat-shock transgene. We decided to pursue kctd15b morphant analysis in this system because kctd15a/b knockdown combined with global tfap2a expression produced increased embryonic lethality and dysmorphism. Interestingly, global tfap2a overexpression disrupted posterior development and produced significantly shorter cdh17+ tubules (Fig. 7A,B). To account for these truncated tubules in subsequent analyses, we calculated a ratio, DE length:tubule length, to statistically compare all groups. We found no significant difference in DE:tubule percentage between wild type and animals that have one copy of the tfap2a heat-shock transgene (hs:tfap2a+/− controls) (Fig. 7A,C). The DE fraction was similarly increased in wild type+kctd15b MO, hs:tfap2a+/−+kctd15b MO and hs:tfap2a HS+ groups (Fig. 7C). Remarkably, pairing kctd15b knockdown with tfap2a overexpression resulted in a drastic expansion of DE marker expression that was more severe than all of the other experimental groups (Fig. 7A,C). In these dual-treated animals, ectopic slc12a1+ cells formed in both the proximal and distal nephron segments (Fig. 7A). These synergistic phenotypes upon concomitant knockdown of kctd15b and overexpression of tfap2a corroborate the interdependence of these factors during nephron segmentation.
kctd15a/b-tfap2a autoregulatory feedback loop balances DE pronephric differentiation. (A) Whole-mount in situ hybridization of slc12a1 (DE, purple) and cdh17 (tubule, red) in hs:tfap2a transgenic background in combination with kctd15b MO treatment at 24 hpf. HS+ indicates application of heat-shock treatment at the 12 ss. Blue arrowheads annotate proximal and distal edges of the slc12a1 expression domain. Scale bar: 35 μm. (B) Quantification of cdh17 tubule expression domain length in control and treatment groups. (C) Bar graph depicting percentage DE occupancy of the tubule. HS−, no heat-shock treatment; HS+, heat-shock treatment. (D) Whole-mount in situ hybridization of slc12a1 in wild-type siblings and trm−/− uninjected controls or trm−/− injected with kctd15a/b MO. Blue arrowheads indicate single DE cells. Scale bar: 35 μm. (E) Quantification of slc12a1 expression domain absolute length per nephron. (F) Penetrance of reduced slc12a1 expression in trm+/− incrosses, uninjected controls versus kctd15a/b MO treatment. (G) Graph depicting the number of DE cells per nephron in trm−/− uninjected controls versus trm−/−+kctd15a/b MO. n≥6 embryos quantified for each control and experimental group. *P<0.05; **P<0.01; ***P<0.001; N.S., not significant. Data are mean±s.d. Absolute lengths and DE:tubule ratios were compared using ANOVA. Penetrances were compared using Fischer's exact test. Cell counts were compared using unpaired t-tests.
kctd15a/b-tfap2a autoregulatory feedback loop balances DE pronephric differentiation. (A) Whole-mount in situ hybridization of slc12a1 (DE, purple) and cdh17 (tubule, red) in hs:tfap2a transgenic background in combination with kctd15b MO treatment at 24 hpf. HS+ indicates application of heat-shock treatment at the 12 ss. Blue arrowheads annotate proximal and distal edges of the slc12a1 expression domain. Scale bar: 35 μm. (B) Quantification of cdh17 tubule expression domain length in control and treatment groups. (C) Bar graph depicting percentage DE occupancy of the tubule. HS−, no heat-shock treatment; HS+, heat-shock treatment. (D) Whole-mount in situ hybridization of slc12a1 in wild-type siblings and trm−/− uninjected controls or trm−/− injected with kctd15a/b MO. Blue arrowheads indicate single DE cells. Scale bar: 35 μm. (E) Quantification of slc12a1 expression domain absolute length per nephron. (F) Penetrance of reduced slc12a1 expression in trm+/− incrosses, uninjected controls versus kctd15a/b MO treatment. (G) Graph depicting the number of DE cells per nephron in trm−/− uninjected controls versus trm−/−+kctd15a/b MO. n≥6 embryos quantified for each control and experimental group. *P<0.05; **P<0.01; ***P<0.001; N.S., not significant. Data are mean±s.d. Absolute lengths and DE:tubule ratios were compared using ANOVA. Penetrances were compared using Fischer's exact test. Cell counts were compared using unpaired t-tests.
To test whether kctd15a and kctd15b might regulate other intermediate factors to control DE development, we performed triple loss-of-function studies. Upon knockdown kctd15a/b in trm−/− mutants, no changes in DE development occurred (Fig. 7D,E). Specifically, trm−/− unininjected controls and trm−/−+kctd15a/b MO groups exhibited a comparable length of slc12a1 expression and number of slc12a1+ cells per nephron. Because kctd15a/b deficiency did not restore the DE differentiation in trm−/− mutants, it is likely that kctd15a and kctd15b solely function through tfap2a. These data underscore Kctd15-Tfap2a as the main genetic axis that controls DE segment fate. In summary, our studies divulge kctd15 paralogs as essential repressors that function during nephrogenesis to stabilize the DE lineage by antagonizing the tfap2a differentiation pathway.
DISCUSSION
During embryonic development, Tfap2 genes are dose-dependent transcription factors and therefore require tight regulation (Hammer et al., 2008; Van Otterloo et al., 2018; Erickson et al., 2018; Chambers et al., 2019; Dooley et al., 2019). Here, our data support a new model where kctd15a and kctd15b fine-tune Tfap2a activity to control nephron segmentation. We found kctd15a/b-deficient nephrons exhibited expanded DE and CS lineages (Fig. 8). This surge in DE and CS fate occurred at the expense of neighboring segments, as PST and DL marker expression were concomitantly decreased. kctd15a/b loss initiated ectopic expression of DE signature genes in adjacent nephron compartments. Frequently, we detected a fraction of cells exhibiting dual segment marker expression insinuating partial fate conversion. This phenomenon of mixed segment profiles was previously documented in Abd-B Hox-deficient and tfap2a-overexpressing nephrons (Magella et al., 2018; Chambers et al., 2019). Previous in vitro experiments established a direct mode of interaction where KCTD15 inhibits AP2 in whole zebrafish embryo and HEK293T lysates (Zarelli and Dawid, 2013a). In light of these mammalian cell line and zebrafish studies, we suspected that the kctd15a/b-deficient nephron phenotypes were caused by increased Tfap2a activity. Interestingly, elevated Tfap2a protein abundance was detected in kctd15a/b-deficient pronephroi, which diverges from previous neural crest studies. We believe that nuclear accumulation of Tfap2a protein (Fig. 5D) is a consequence of genetic feedback circuitry because: (1) kctd15 loss expanded tfap2a mRNA expression in developing renal progenitors (Fig. 6A-C); and (2) tfap2a promotes kctd15a and kctd15b transcription (Fig. 6D,E). In accordance with these observations, Tfap2a can promote its own transcription by binding to an autoregulatory element (Imhof et al., 1999). If Tfap2a is unable to be repressed, positive autoregulation could explain heightened tfap2a transcript production in response to kctd15a/b loss of function.
Proposed model illustrating the role of the Kctd15-Tfap2a interaction in maturing nephron cells. Wild-type (left) and kctd15a/b-deficient context (right) at 24 hpf. Distal early (DE, cyan), corpuscle of Stannius (CS, orange) and neighboring cell populations (gray). Black dashed box indicates the featured distal nephron zone. Cyan arrows indicate expansion of DE differentiation in kctd15a/b-deficient nephron. Magenta circle outlines a single cell featured in the bottom panel. In wild type, Tfap2a promotes kctd15 expression at the transcriptional level and Kctd15 protein (purple) represses Tfap2a (green) activity, allowing for proper nephron segment differentiation. In kctd15a/b deficiency, predicted truncated Kctd15 protein is not able to inhibit Tfap2a activity in neighboring cell populations, allowing for ectopic activation of DE lineage factors and solute transporters (slc12a1 and kcnj1a.1).
Proposed model illustrating the role of the Kctd15-Tfap2a interaction in maturing nephron cells. Wild-type (left) and kctd15a/b-deficient context (right) at 24 hpf. Distal early (DE, cyan), corpuscle of Stannius (CS, orange) and neighboring cell populations (gray). Black dashed box indicates the featured distal nephron zone. Cyan arrows indicate expansion of DE differentiation in kctd15a/b-deficient nephron. Magenta circle outlines a single cell featured in the bottom panel. In wild type, Tfap2a promotes kctd15 expression at the transcriptional level and Kctd15 protein (purple) represses Tfap2a (green) activity, allowing for proper nephron segment differentiation. In kctd15a/b deficiency, predicted truncated Kctd15 protein is not able to inhibit Tfap2a activity in neighboring cell populations, allowing for ectopic activation of DE lineage factors and solute transporters (slc12a1 and kcnj1a.1).
Alternatively, it is feasible that excess pronephric Tfap2a protein is present because it is not properly targeted for degradation. Previous studies have demonstrated that KCTD proteins can function as adaptor molecules for E3 ubiquitin ligases to facilitate post-translational modifications of bound-protein interactors (Pinkas et al., 2017). Furthermore, KCTD family members are substrates for SUMOylation and ubiquitylation, which can alter protein stability and initiate degradation, respectively. There is evidence for SUMO conjugation of TFAP2A, which suppresses transcriptional activation and prevents mesenchymal-to-epithelial transition in breast cancer cells (Eloranta and Hurst, 2002; Berlato et al., 2011; Bogachek et al., 2014, 2015). Future investigation is needed to determine whether Kctd15 facilitates SUMOylation of Tfap2a in nephron precursors. Nonetheless, our kctd15a/b-deficient genetic manipulations truncate the BTB/POZ domain and would prevent direct interactions between Kctd15 and Tfap2a from occurring. Because feedback circuitry is disrupted, Tfap2a activity is left unchecked in neighboring nephron segments, allowing downstream targets like DE-specific lineage factors and solute transporters to be ectopically expressed (Fig. 8). Because the DE is expanded at the expense of the DL and PST in kctd15a/b-deficient animals, we believe Kctd15a and Kctd15b inhibits Tfap2a activity in neighboring segment precursors. In support of our model, kctd15b knockdown, coupled with tfap2a overexpression, yielded a synergistic phenotype where nephron cell composition was shifted toward DE assignment. We postulate that within DE progenitors, Kctd15-Tfap2a crosstalk may result in tight spatiotemporal oscillations of Tfap2a levels as nephron development proceeds and provides a narrow window for induction of DE-specific programs. Future studies are needed to determine whether other co-regulators are present within DE progenitors that attenuate Kctd15 repressor activity, allowing for the expression of downstream Tfap2a transcriptional targets. Taken together, our data reveal a novel feedback circuit operating within the context of kidney development, where a tfap2a-kctd15 transcription factor-repressor module programs DE/TAL differentiation.
Our studies also help to fill a key void in the literature by addressing developmental signaling required for DE/TAL segment formation. Owing to the anatomical limitations of studying mammalian nephron development in vivo, there is a limited knowledge of cell diversity and dynamics within the maturing LOH anlagen. Zebrafish studies provide a complementary launch point to explore how functionality of candidate factors, such as tfap2a, kctd15a and kctd15b, in the present article, and other distally expressed genes, such as mecom, tbx2a, tbx2b, emx1, ppargc1a and irx2a, create nephron cell types in vertebrates (Li et al., 2014a,b; McKee et al., 2014; Marra and Wingert, 2016; Drummond et al., 2016; Morales et al., 2018; Chambers et al., 2018; Marra et al., 2019b). Recently, scRNA-seq and fate mapping in mice revealed early forming LOH (juxtamedullary) and late forming LOH (cortical) have distinct gene signatures (Ransick et al., 2019). LOH outgrowth from the S-shaped tubule is adaptive and oriented by long-range cues from medullary collecting ducts; however, signaling pathways navigating this process have yet to be identified (Chang and Davies, 2019). Current efforts to advance personalized medicine by building patient-derived kidney organoids face challenges in achieving expression of LOH-specific genes (Schutgens et al., 2019). Hence, there is utility in using animal models to assemble a working genetic framework that could enrich for LOH development in culture conditions.
A compelling TFAP2A GRN candidate that warrants further investigation is KCTD1, which shares greater than 78% sequence homology with KCTD15 (Pirone et al., 2019). KCTD1 can directly inhibit TFAP2A as well as suppress canonical Wnt signaling, exhibiting congruence with known KCTD15 molecular mechanisms (Ding et al., 2009; Li et al., 2014a,b). Human KCTD1 mutations cause scalp-ear-nipple (SEN) syndrome, which involves the formation of toxic amyloid-like aggregates thought to influence disease state (Smaldone et al., 2019). With the advent of prenatal screening and genotyping strategies, KCTD1 lesions have been linked to renal hypoplasia (Gray et al., 2019). Along these lines, mice harboring a Kctd1 mutation die perinatally due to kidney failure associated with defective ion homeostasis. Transcriptional profiling of these mutant kidneys identified 102 significantly upregulated genes, including known TFAP2A targets (Kumar et al., 2017). TFAP2A- and TFAP2B-null mice also manifest lethal kidney defects (Zhang et al., 1996; Moser et al., 1997; Moser et al., 2003). Interestingly, recent data indicate that Tfap2b functions upstream of Kctd1 to promote the maintenance of terminally differentiated DCT epithelium in mouse kidneys (Marneros, 2020). During zebrafish pronephros development, tfap2a controls the expression of family member tfap2b (Chambers et al., 2019). Although further investigation is needed in a mammalian system, Tfap2a is likely required for the subsequent activation of Tfap2b and Kctd1 in maturing nephron structures. Ultimately, our genetic studies elucidating TFAP2-KCTD network mechanisms in developing zebrafish nephrons can catalyze candidate identification and refine prenatal screening for renal defects.
MATERIALS AND METHODS
Ethics statement and zebrafish husbandry
Adult zebrafish were maintained at the University of Notre Dame Freimann Life Science Center. All studies were approved and supervised by the University of Notre Dame Institutional Animal Care and Use Committee (IACUC), under protocol numbers 16-025 and 19-06-5412. All wild-type experiments were conducted with the Tübingen strain. Embryonic zebrafish were incubated in E3 medium, staged and fixed as previously described (Westerfield, 1993; Kimmel et al., 1995).
Whole-mount and fluorescent in situ hybridization
Whole-mount and fluorescent in situ hybridization were performed as previously described (Cheng, et al., 2014; Marra et al., 2019c). Digoxigenin and fluorescein anti-sense RNA probes were synthesized by T7, T3 or SP6 in vitro transcription (Roche Diagnostics) from linearized IMAGE clone plasmids. Digoxigenin-labeled probes included kctd15a, kctd15b, kcnj1a.1, slc12a1, stc1, slc12a3 and trpm7. Fluorescein-labeled probes included: tfap2a, slc12a3, slc12a1, kcnj1a.1 and cdh17. Gene expression studies were performed in triplicate with sample size of n>20 for each replicate. Representative samples from each experimental group were imaged and analyzed.
Whole-mount immunofluorescence
Whole-mount immunofluorescence was completed as previously described (Marra et al., 2017; Marra et al., 2019c). For all immunofluorescence experiments, embryos were fixed in freshly diluted 4% PFA for 2 h at room temperature. EM Grade PFA (32%; Electron Microscopy Sciences, 15714) was diluted to 4% concentration in 1×PBS. After fixation, embryos were stored in 100% methanol at −20°C for future use. Primary antibody dilutions consisted of anti-laminin (1:100; Sigma-Aldrich, L9393) (Gerlach and Wingert, 2014), anti-T4 supernatant (1:200; Developmental Studies Hybridoma Bank, Na-K-Cl cotransporter, AB_528406) and anti-Tfap2a (1:50; LifeSpan Biosciences, LS-C87212-100) (Chambers et al., 2019). Secondary antibodies (1:500) included goat anti-rabbit, rabbit anti-mouse and donkey anti-goat (Invitrogen; A11034, A11061, A11055). 4,6-diamidino-2-phenylindole dihydrochloride (DAPI; Invitrogen, D1306) was used for nuclear staining.
Morpholino knockdown and RT-PCR
Morpholino oligonucleotides (MO) were synthesized by GeneTools. MOs were solubilized in DNase/RNase-free water to make 4 mM stock concentrations and stored at −20°C. Splice-blocking MOs for kctd15a and kctd15b genes were designed to target the exon1-intron1 splice sites, and were microinjected at doses of 1 ng and 3 ng, respectively. MO efficacy was validated by RT-PCR. RNA was extracted from pools of 20 embryos, cDNA was synthesized using random hexamers (Superscript IV, Invitrogen) and PCR was performed to amplify target site. Products were run on a 1.5% agarose gel, extracted and sequenced. See Table S1 for specific MO and primer sequences.
gRNA design and crispant generation
gRNAs targeting the BTB-POZ encoding regions in kctd15a and kctd15b were designed using the CHOPCHOP web-based tool (https://chopchop.cbu.uib.no/). kctd15a sgRNA1 and sgRNA2 targeted different regions in exon 3. kctd15b sgRNA1 and sgRNA2 targeted different regions in exon 1. gRNA templates were annealed to a constant oligonucleotide, as previously described (Gagnon et al., 2014), and RNA was synthesized using the T7 Megascript kit (Ambion). Multiplexed gene editing was achieved by injecting a cocktail containing all four gRNAs. Microinjection mix was prepared by combining gRNAs (60 ng/µl) and Cas9 protein (0.8 µM) followed by incubation for 5 min at 37°C. Embryos were injected at the one-cell stage with ∼5 nl of injection mix. T7 endonuclease assay was used to confirm genome editing. For kctd15a and kctd15b crispant verification, primers were designed to flank both sgRNA target sites in exons 3 and 1, respectively. In short, DNA was prepared from individual animals and Accuprime Pfx SuperMix (Invitrogen) was used to amplify target sites. PCR products were column purified with QIAquick PCR purification kit (Qiagen). Purified product (300 ng) and 2 µl 10× NEB Buffer 2 (New England Biolabs) (total volume 20 µl) was rehybridized in a thermocycler using the following program: 5 min at 95°C, ramp down to 85°C at 2°C/s, ramp down to 25°C at 0.1°C/s, and 25°C for 10 min. Rehybridized product was digested with 1 µl T7 endonuclease I enzyme (New England Biolabs) at 37°C for 1 h and separated on a 1.5% agarose gel. See Table S1 for specific gRNA and primer sequences.
Overexpression experiments
A kctd15a pCS2 construct was designed with Not1 and XhoI flanking the open reading frame allowing for in vitro synthesis of full-length sense cRNA using an SP6 mMessage Machine kit (Ambion). kctd15a cRNA (60 pg) was injected into one-cell stage embryos for overexpression. For tfap2a overexpression experiments, the hs:tfap2a transgenic [Tg(hsp70:tfap2a)×24], which was a generous gift from Bruce Riley (Texas A&M University, USA), was employed. To activate the heat-shock-inducible tfap2a transgene, embryos were incubated at 38°C for 30 min at the 8 ss or 12 ss (Bhat et al., 2013; Kantarci et al., 2015). Genotyping for the hs:tfap2a allele was performed as previously described (Chambers et al., 2019). See Table S1 for genotyping primers.
Image acquisition and statistical analysis
Whole-mount in situ hybridization samples were imaged with a Nikon Eclipse Ni with DS-Fi2 camera. Fluorescent in situ hybridization and immunoflurescence samples were imaged with a Nikon C2 confocal microscope. The Nikon elements imaging software polyline tool was employed to quantify absolute lengths of gene expression domains. Confocal z-stacks were processed with Fiji (Image J) software and the plot profile feature was used to collect fluorescent intensity values. All graphical and statistical analyses were completed with GraphPad Prism. A minimum of three samples from control and experimental groups were imaged and scored. Unpaired t-test tests and one-way ANOVA analyses were conducted as appropriate and mean±s.d. was reported.
Acknowledgements
We thank Bruce Riley for sharing the inducible hs:tfap2a zebrafish transgenic. We thank the staff of the Department of Biological Sciences, Integrated Imaging Core, Genomics Core and the Center for Zebrafish Research at the University of Notre Dame for their tremendous dedication and care of our zebrafish aquarium. Finally, we thank the members of our lab for their support, discussions and insights about this work.
Footnotes
Author contributions
Conceptualization: B.E.C., R.A.W.; Methodology: B.E.C., R.A.W.; Validation: B.E.C., R.A.W.; Formal analysis: B.E.C., E.G.C., A.E.G., R.A.W.; Investigation: B.E.C., E.G.C., A.E.G., R.A.W.; Resources: R.A.W.; Data curation: B.E.C., E.G.C., A.E.G., R.A.W.; Writing - original draft: B.E.C., R.A.W.; Writing - review & editing: B.E.C., R.A.W.; Visualization: B.E.C., R.A.W.; Supervision: B.E.C., R.A.W.; Project administration: R.A.W.; Funding acquisition: R.A.W.
Funding
This work was supported by the National Institutes of Health (R01DK100237 to R.A.W.). We are grateful to Elizabeth and Michael Gallagher for a generous gift to the University of Notre Dame on behalf of their family for the support of stem cell research. The funders had no role in the study design, data collection and analysis, decision to publish, or manuscript preparation. Deposited in PMC for release after 12 months.
References
Competing interests
The authors declare no competing or financial interests.