A homozygous human WNT11 variant is associated with laterality, heart and renal defects Open Access

Wnt signaling plays important roles during vertebrate development, including left-right axis specification as well as heart and kidney organogenesis. We identified a homozygous human WNT11 variant in an infant with situs inversus totalis, complex heart defects and renal hypodysplasia, and used Xenopus embryos to functionally characterize this variant. WNT11^c.814delG encodes a protein with reduced stability that lost signaling activity in vivo. This is remarkable, because the variant encodes a truncated ligand with nearly identical length and predicted structure to dominant-negative Wnts. Furthermore, we demonstrate that alteration of the truncated C-terminal end can restore stability and signaling activity similarly to Xenopus dominant-negative Wnt11b. Our study also suggests similar functions for WNT11 in human development as those described in model organisms. Therefore, biallelic WNT11 dysfunction should be considered a novel genetic cause of syndromal human phenotypes presenting with congenital heart defects and renal hypoplasia, with or without laterality defects. The work presented here enhances our understanding of human development and structure-function relationships in Wnt ligands.

Bilaterians have an overall bilateral symmetric body plan, but several internal organs display asymmetries along the left-right body axis (Fliegauf et al., 2007; Namigai et al., 2014). The most common organ arrangement is called situs solitus, whereby, in humans, the lung consists of two lobes on the left and three on the right side, the liver is positioned on the right side and the heart is tilted towards the left side (Fliegauf et al., 2007). Laterality defects, mis-arrangements of left-right asymmetric organs, occur in ∼1 in 10,000 human live births (Afzelius, 1999; Lin et al., 2014; Torgersen, 1950). These include situs inversus totalis (SIT), heterotaxia and isomerisms (Fliegauf et al., 2007). Whereas SIT is a deviation from the norm without direct health consequence, heterotaxia and isomerisms can result in misplacements, duplications or absence of individual organs, congenital heart defects (CHDs) and structural defects of the great vessels (Fakhro et al., 2011; Lin et al., 2014).

Organ asymmetries are initiated during development by the left-right organizer (LRO), which breaks bilateral symmetry, leading to asymmetric gene expression guiding further organ morphogenesis (Blum et al., 2014; Hamada, 2020). The LRO is a monociliated epithelium, and its transient appearance, structure and function are remarkably conserved across most vertebrate species (Blum et al., 2009b). In the frog Xenopus laevis, the LRO is called the gastrocoel roof plate (GRP) (Schweickert et al., 2007; Shook et al., 2004). Central cells of the GRP project motile cilia, which are asymmetrically positioned at the posterior poles of cells and beat in a clockwise fashion to generate an extracellular leftward fluid flow (Walentek et al., 2012). Cells in the lateral margins of the GRP harbor non-motile sensory cilia required to sense fluid flow (Gopalakrishnan et al., 2023; Schweickert et al., 2007). Lateral GRP cells also express the signaling ligand Nodal1 and its inhibitor Dand5 (also called Coco in Xenopus) (Schweickert et al., 2010). Leftward flow reduces coco expression, leading to release of Nodal1 repression exclusively on the left side. Nodal1 then induces an asymmetric gene expression cascade in the left lateral plate mesoderm by activating its own expression, as well as expression of its feedback inhibitor lefty (also called lefty1 in Xenopus) and the transcription factor pitx2c (Blum et al., 2014; Schweickert et al., 2001, 2010). Pitx2c remains active on the left side of the embryo during subsequent asymmetric organ development.

Xenopus is particularly suited to studying early development and laterality defects, because embryos develop externally and are large in size, gastrulation stages are reached within a few hours, and tissues such as the left or right side of the GRP can be selectively targeted by micro-injection and manipulations (Blum et al., 2009a; Walentek et al., 2012).

In humans, laterality defects can occur as a consequence of defective motile cilia, for which a vast number of responsible genes have been identified (Antony et al., 2022; Fliegauf et al., 2007; Raidt et al., 2023; Reiter and Leroux, 2017). Furthermore, defects in genes related to cell signaling pathways have also been found to result in laterality defects (Minegishi et al., 2023; Mohapatra et al., 2009; Sempou and Khokha, 2019). Nevertheless, causative genes for laterality defects can only be identified in less than half of the affected individuals (Antony et al., 2022).

The Wnt pathway is a signaling pathway with multiple branches affecting gene expression (e.g. canonical Wnt/β-catenin pathway), cell polarity [e.g. Wnt/planar cell polarity (PCP) pathway] and morphogenesis (e.g. Wnt/calcium pathway) (MacDonald et al., 2007; Minegishi et al., 2023; Niehrs, 2012; Niehrs and Acebron, 2012; Semenov et al., 2007; Sokol, 2015; Walentek et al., 2012, 2018, 2013). In Xenopus development, Wnt/β-catenin is required for dorsoventral axis induction (a prerequisite for left-right axis development) and for generation of motile GRP cilia by activating the transcription factor foxj1 (Walentek et al., 2012). Wnt/PCP signaling is required for posterior positioning of GRP cilia required to generate an asymmetric leftward fluid flow (Walentek et al., 2012). Other Wnt pathway branches affect gastrulation movements and normal GRP morphogenesis, which has also been shown to affect the development of motile versus non-motile GRP cilia and lateral GRP gene expression (Chien et al., 2018; Kühl et al., 2000; Tada and Smith, 2000; Walentek et al., 2013).

Particularly, Xenopus Wnt11b has been shown to act through canonical and non-canonical Wnt signaling branches (Cha et al., 2008, 2009; Kofron et al., 2007; Li et al., 2008; Tada and Smith, 2000; Tao et al., 2005), and has been identified as a regulator of gastrulation movements (Tada and Smith, 2000), GRP morphology, Wnt/PCP and left-right axis formation (Walentek et al., 2013). Wnt11b loss of function in the GRP, induced by morpholino-mediated knockdown or overexpression of a dominant-negative (dn)Wnt11b construct, also reduced expression of nodal1 and coco, indicating that Wnt11b affects left-right symmetry breakage primarily by interfering with GRP morphogenesis and the signaling cascade downstream of the cilia-dependent fluid flow (Walentek et al., 2013). The association of Wnt11b with laterality has recently been confirmed in Xenopus wnt11b knockout animals (Houston et al., 2022). Notably, Wnt11 is also expressed in the LROs of other species, such as the Hensen's node in chicken and the node of mice, suggesting a role for Wnt11 in the determination of left-right asymmetry across vertebrates (Chapman et al., 2004; Kispert et al., 1996).

Besides regulating left-right axis development, Wnt11 also functions in other organ systems. Overexpression of dnWnt11b induced kidney developmental defects in Xenopus (Dichmann et al., 2015), and Wnt11^−/− mice develop hypoplastic kidneys and CHDs with complete penetrance (Majumdar et al., 2003; Zhou et al., 2007). Thus, Wnt11b has been related to laterality defects in Xenopus, and knockout of Wnt11 has been identified as a cause of heart and kidney defects in mice.

Here, we identified a homozygous human WNT11 variant (c.814delG, p.Glu272Asn*13) in an infant with SIT, tetralogy of Fallot (TOF) and severe bilateral renal hypodysplasia. A biallelic WNT11 variant has, to our knowledge, not been identified in humans before, and WNT11 dysfunction has not previously been identified as a cause of these human developmental defects. Based on this discovery, we used Xenopus embryos as a model system to functionally characterize this novel WNT11 variant. We found that WNT11^c.814delG encodes a truncated ligand with reduced stability, which lost its ability to activate Wnt signaling as well as to regulate morphogenesis. Thereby, the resulting protein functionally differs from dominant-negative-acting truncations of similar length described in the literature (Tao et al., 2005; Dichmann et al., 2015), despite only minor differences in the amino acid sequence. Together, we identified a new human WNT11 variant associated with a syndromale developmental phenotype, functionally characterized the altered WNT11 protein, and discovered how differences in the WNT11 protein can dramatically alter its functions.

We performed exome sequencing including copy number variant (CNV) analysis in a first-born infant of consanguineous southeast Asian descent with SIT including dextrocardia, in combination with a complex CHD (TOF: ventricular septal defect, overriding aorta, pulmonary stenosis and right-ventricular hypertrophy) and severe bilateral renal hypodysplasia, with end-stage renal disease in infancy. This revealed a homozygous WNT11 frameshift variant, c.814delG, p.Glu272Asn*13 (Fig. 1A; Fig. S1A). The variant was absent from gnomAD, with only a missense variant at the same amino acid position 272, Glu272Lys, reported in one of 1,614,052 alleles. In total, as of October 2024, there are 38 different stop or frameshift variants in WNT11 reported in gnomAD. All of these are heterozygous and the majority represent private alleles, all with a very low allele frequency (<0.0001 minor allele frequency), suggesting that WNT11 has non-redundant functions and that biallelic loss-of-function variants will result in a human phenotype. In the affected individual, no additional putative disease-causing variants or CNVs were detected. This indicated the WNT11^c.814delG variant as a likely cause in our individual and suggested WNT11 dysfunction as a novel cause of congenital defects in humans.

To investigate whether WNT11 variants could be found in additional patients with overlapping phenotypes, we performed Sanger sequencing of WNT11 coding and splice site regions (positions +/−20) of 39 cases with non-cystic renal dysplasia, 13 cases of severe renal hypoplasia, and one case of renal hypoplasia and heart defect. This did not reveal any putative causative WNT11 variants. Furthermore, we did not identify any additional likely disease-causing WNT11 variants using exome sequencing in a cohort of 37 cases with laterality defects (Antony et al., 2022) or in three additional cases with renal hypoplasia. Likewise, we did not identify any additional cases with biallelic WNT11 variants through our national and international collaborators or through a genematcher search.

Interestingly, the mother of the affected child also presented with SIT, whereas the father was not affected. Segregation analysis within the family revealed that both parents were heterozygous for the WNT11^c.814delG frameshift allele (Fig. 1A,B). To exclude an additional underlying monogenic cause for the laterality defect in the mother, we also performed exome sequencing for her. However, this did not reveal any additional plausible genetic causes.

We investigated the predicted protein structures of wild-type (wt) WNT11 and WNT11^c.814delG using Alphafold v2.1.2 (Jumper et al., 2021), which revealed the stereotypical hand-like structure of the full-length Wnt ligand (Willert and Nusse, 2012), while the ligand encoded by WNT11^c.814delG was lacking the C-terminal ‘index finger’ in the protein structure (Fig. 1C). Interestingly, C-terminally truncated Wnt ligands were previously shown to exhibit dominant-negative effects on Wnt signaling in vivo, including an engineered Xenopus laevis dnWnt11b construct that affected gastrulation movements, morphogenesis and laterality in embryos (Hoppler et al., 1996; Tada and Smith, 2000; Walentek et al., 2013). Therefore, we compared wt human WNT11 and Xenopus Wnt11b structures, which confirmed the expected high degree of similarity (Fig. 1C). In contrast, comparison of dnWnt11b and WNT11^c.814delG predicted structures suggested that changes in the amino acid sequence due to the frameshift in WNT11^c.814delG altered the structural prediction for the last 12 amino acids, with unknown consequences for protein function (Fig. 1C,D).

In summary, these results revealed a novel homozygous human WNT11 variant associated with complex developmental defects and predicted structural alterations of the encoded Wnt ligand.

Two Wnt11 paralogs exist in Xenopus: Wnt11b is maternally deposited, expressed in the superficial layer of the gastrula organizer and the blastopore as well as the developing somites; Wnt11 (also called Wnt11r) is expressed in the developing nervous system and the cement gland (Dichmann et al., 2015; Garriock et al., 2005; Matthews et al., 2008; Tada and Smith, 2000; Walentek et al., 2013). Wnt11 functions in development have been extensively studied in Xenopus embryos, and Wnt11b has been shown to regulate gastrulation movements, morphogenesis and left-right axis specification after both gain and loss of function (Walentek et al., 2013). We therefore chose to test how human WNT11 and WNT11^c.814delG affect early Xenopus development in comparison to Xenopus Wnt11b and dnWnt11b in vivo.

We injected mRNAs encoding Wnt11b, dnWnt11b, WNT11 and WNT11^c.814delG (Fig. S1A), targeting the prospective dorsal mesoderm, and analyzed the impact on morphogenesis in gastrula and neurula [stage (st.) 13 and 20] embryos. As previously described, overexpression of Xenopus full-length Wnt11b and dnWnt11b impaired gastrulation movements, leading to high frequencies of blastopore and neural tube closure defects (>70%) (Fig. 2A,B) (Walentek et al., 2013). Overexpression of full-length WNT11 induced morphological defects at similar frequencies to Wnt11b, confirming a high degree of functional homology between Wnt11b and WNT11 (Fig. 2A,B). In contrast, overexpression of WNT11^c.814delG induced morphological defects at much lower frequencies (<40%), indicating functional differences to all other constructs, including dnWnt11b (>70%) (Fig. 2A,B). Mortality rates in embryos were similar across manipulations, excluding the possibility that fewer morphological defects were recovered after WNT11^c.814delG overexpression due to early embryonic lethality (Fig. S1B).

Next, we tested whether WNT11 and WNT11^c.814delG overexpression can cause laterality defects in Xenopus. wnt11b, dnwnt11b, WNT11 and WNT11^c.814delG mRNAs were targeted to the dorsal mesoderm and LRO (Schweickert et al., 2012; Walentek et al., 2012), and left-right axis specification was analyzed using asymmetric pitx2c gene expression at st. 30-32 by whole-mount in situ hybridization (WMISH). In control embryos, pitx2c expression was found to be predominantly (>80%) expressed exclusively on the left side of the embryo (Fig. 2C,D). Gain of wt Wnt11b or WNT11 induced ectopic right-sided pitx2c, leading to bilateral expression in the majority of embryos (>50% bilateral), whereas dnWnt11b caused loss of pitx2c expression as the most frequent phenotype (>30% absent) (Fig. 2C,D), in line with published results (Walentek et al., 2013). Importantly, overexpression of WNT11^c.814delG not only led to embryos developing no visible morphological defects, but also to no left-right axis defects being detected in the majority of manipulated embryos (>70% left-sided expression) (Fig. 2C,D).

Taken together, in vivo functional tests strongly suggest that WNT11^c.814delG encodes a hypomorph or loss-of-function ligand that lost its ability to activate Wnt signaling. Hence, overexpression of WNT11^c.814delG did not affect embryonic morphogenesis or left-right axis specification.

Structural predictions suggested differences between dnWnt11b and WNT11^c.814delG, and in vivo assays indicated that WNT11^c.814delG encodes a ligand that lost signaling ability. This was remarkable, because various similar C-terminal truncations of Wnt ligands were shown to generate dominant-negative-acting signaling molecules (Dichmann et al., 2015; Hoppler et al., 1996; Tada and Smith, 2000). We therefore wondered whether the differences in length or differences in the amino acid sequence were the cause of the observed functional differences between dnWnt11b and WNT11^c.814delG.

To investigate the molecular basis of functional differences between truncated Wnt ligands, we generated WNT11 constructs lacking the frame-shifted portion (WNT11^Δ814-1062) or with a corrected sequence, but the same extent of truncation (WNT11^Δ850-1062) as WNT11^c.814delG (Fig. 3A; Fig. S1C). We also generated a shortened Xenopus dnWnt11b construct (wnt11b^Δ811-1059) analogous to WNT11^Δ814-1062 (Fig. 3A; Fig. S1C) for comparison. mRNAs encoding the newly generated constructs were targeted to the dorsal mesoderm and LRO, and effects on asymmetric pitx2c expression as well as morphology were compared to effects induced by WNT11^c.814delG and dnWnt11b. Like in previous experiments, WNT11^c.814delG overexpression did not affect asymmetric pitx2c expression or embryonic morphogenesis compared to controls (Fig. 3B-E). In striking contrast, removal of frameshift amino acids (WNT11^Δ814-1062) generated results comparable to dnWnt11b (>50% absent pitx2c expression and >70% of embryos with morphological defects) (Fig. 3B-E). Correction of the frameshift amino acids (WNT11^Δ850-1062) or shortening of the dnWnt11b ligand (wnt11b^Δ811-1059) both caused left-right and morphological defects, however at lower rates than dnWnt11b or WNT11^Δ814-1062 (Fig. 3B-E). Thus, these results revealed striking functional differences in Wnt11-type ligands that differed only in a few amino acids at their truncated C-termini.

The loss of pitx2c induction in the lateral plate mesoderm indicated that the modified truncated ligands (especially WNT11^Δ814-1062) recapitulate effects specifically observed after overexpression of dnWnt11b due to reduced nodal1 expression in sensory GRP cells. These were distinct from gain-of-function effects induced by WNT11 or wnt11b overexpression (Fig. 2C,D) (Walentek et al., 2013). To gain further evidence that WNT11^Δ814-1062 encodes a dominant-negative-acting ligand, we overexpressed WNT11, WNT11^Δ814-1062 or dnwnt11b using DNA injections targeted to the dorsal mesoderm and LRO, and analyzed nodal1 expression in the GRP at st. 17 (Fig. S2A,B). We used plasmids for overexpression to manipulate embryos only after zygotic genome activation (st. 8), leading to later onset and lower levels of overexpression. This sensitized our assay and reduced the degree of gastrulation defects, thereby preserving GRP morphology and facilitating the analysis of the nodal1 domains. This revealed that nodal1 expression was not significantly changed after overexpression of WNT11, and that dnwnt11b overexpression induced significantly greater reduction in nodal1 expression than WNT11, as expected (Fig. S2A,B). WNT11^Δ814-1062 increased the number of embryos with reduced nodal1 domains compared to WNT11, but at lower rates than dnwnt11b (Fig. S2A,B). This supported the conclusion that WNT11^c.814delG encodes a functional ligand similar to dnWnt11b, but with weaker activity.

Wnt11b has been shown to regulate non-canonical as well as canonical (β-catenin-dependent) signaling in early Xenopus development as well as in foregut formation (Cha et al., 2008; Li et al., 2008; Tao et al., 2005). To test whether WNT11^Δ814-1062 can influence canonical Wnt activity, we next injected Xenopus embryos from a Wnt/β-catenin signaling reporter line (pbin7LEF::dGFP; Haas et al., 2019) unilaterally targeting the right neural plate and analyzed Wnt-driven GFP expression at st. 20 by epifluorescent microscopy. Additionally, we used overexpression of Xenopus wnt3a (Kiecker and Niehrs, 2001) as a positive control to stimulate reporter activity. In control embryos, reporter activity was bilaterally symmetric, while Wnt3a increased reporter activity as expected (Fig. S2C,D). In contrast, overexpression of WNT11^c.814delG did not profoundly alter reporter activity on the injected side, whereas WNT11^Δ814-1062 strongly and WNT11^Δ850-1062 weakly reduced Wnt reporter activity (Fig. 4A,B). However, overexpression of WNT11 also inhibited canonical Wnt reporter activity (Fig. S2C,E). Hence, this assay cannot unambiguously distinguish between activating or dominant-negative functions in canonical Wnt signaling.

The observed functional differences in C-terminally truncated WNT11 proteins could be due to changed molecular signaling properties of the ligands (e.g. loss of dimerization ability) or could lead to altered expression levels, e.g. by triggering protein degradation (e.g. due to protein misfolding). To investigate which scenario applied to WNT11^c.814delG, we overexpressed WNT11, WNT11^c.814delG, WNT11^Δ814-1062 or WNT11^Δ850-1062 and assessed protein levels as well as dimerization using sodium dodecyl-sulfate (SDS)-polyacrylamide gel electrophoresis (PAGE) and western blotting. To ensure specific reactivity of the anti-WNT11 antibody to the overexpressed human WNT11 constructs, we included uninjected as well as wnt11b-injected controls in these assays. Uninjected and wnt11b-injected samples showed only minor reactivity to the anti-WNT11 antibody, whereas a clear band around the expected 40 kDa was detected in WNT11-injected samples (Fig. 4C; Fig. S3A). In contrast to WNT11, WNT11^c.814delG and WNT11^Δ850-1062 showed reduced abundance (Fig. 4C; Fig. S3A), linking their reduced functional potencies to reduced protein concentrations. The modified ligand WNT11^Δ814-1062 was, in turn, detected at the expected size (∼30 kDa) and at levels similar to the wt WNT11 ligand (Fig. 4C; Fig. S3A). Comparison of western blot results between reducing (standard) and non-reducing (allowing detection of Wnt ligand dimers; Cha et al., 2009) sample preparation conditions further indicated that all detectable ligands, including WNT11 and WNT11^Δ814-1062, were able to form ligand dimers in vivo (Fig. 4D; Fig. S3B,C), a prerequisite for signaling activity.

Thus, we conclude from these results that changes in the last 12 amino acids in WNT11^c.814delG render the ligand unstable, leading to reduced protein levels and loss of normal or dominant-negative signaling activity.

Our functional studies to characterize the novel human WNT11^c.814delG variant revealed that it encodes a loss-of-function allele: overexpression of WNT11^c.814delG did not affect morphogenesis, left-right axis patterning and canonical Wnt signaling activity in Xenopus embryos. This was surprising, because the variant encodes a ligand with nearly identical length and predicted structure to the Xenopus dominant-negative Wnt11b construct (283 versus 282 amino acids, respectively) as well as to a dominant-negative-acting Wnt11b ligand resulting from mis-splicing of 296 amino acid length (Dichmann et al., 2015; Tada and Smith, 2000). This lack of activity is not caused by functional differences of Wnt11 homologs across species, because overexpression of full-length WNT11 yielded similar results to Xenopus Wnt11b. Instead, our results suggest that the functional differences are due to changes in the composition of the last 12 frameshift amino acids. Removal of the 12 frameshift amino acids (WNT11^Δ814-1062) restored activity, resulting in effects similar to dnWnt11b. Moreover, replacing these 12 amino acids with the wt sequence (WNT11^Δ850-1062) resulted in a ligand with weaker activity. Our results further indicate that the lack of signaling activity by the WNT11^c.814delG variant is due to reduced protein levels, while C-terminally truncated ligands can generally still dimerize (Cha et al., 2008).

Interestingly, removal of 12 amino acids from Xenopus dnWnt11b (wnt11b^Δ811-1059) also resulted in weaker dominant-negative activity of the ligand, and it was described that the dominant-negative activity of Wnt ligands depends on their length (Hoppler et al., 1996; MacDonald et al., 2014). The first dominant-negative Wnts were constructed by C-terminally truncating the ligand behind a particular conserved cysteine (C13). In search of a dominant-negative Wnt8, the authors noted that a ligand 30 amino acids shorter than dnWnt8, which lacked the conserved cysteine, had only weak dominant-negative activity (Hoppler et al., 1996). Subsequently, dnWnt11b was designed with the premature stop positioned directly after the conserved cysteine residue (Tada and Smith, 2000). Nevertheless, our results reveal that the dominant-negative behavior of Wnts cannot be dependent on a specific length or the presence of the cysteine residue, as WNT11^Δ814-1062 and Wnt11b^Δ811-1059 do not contain it. Instead, we propose that a combination of ligand length and amino acid composition is required for protein stability and dominant-negative activity, perhaps by allowing certain post-translational modifications, including glycosylation and acylation (Willert and Nusse, 2012), which are indispensable for Wnt structure and secretion (Komekado et al., 2007; Kurayoshi et al., 2007). Amino acid changes might increase susceptibility for recognition by protein control mechanisms, such as unfolded protein response or destabilization (Teng et al., 2010). In contrast, nonsense-mediated decay of the mRNA seems rather unlikely, because we overexpressed synthetic mRNAs at high levels and the stop codon in WNT11^c.814delG is positioned 120 bases upstream of the last exon-exon junction, a region that is typically insensitive to nonsense-mediated decay (Nagy and Maquat, 1998).

In line with our experimental results, the affected individual was homozygous for WNT11^c.814delG, and both heterozygous parents were healthy individuals (except for SIT in the mother), arguing against a dominant-negative function of WNT11^c.814delG in humans. The phenotypic features observed in the affected child, including TOF, renal hypodysplasia and SIT match phenotypic features previously identified in Wnt11^−/− mice and after manipulating Wnt11b in Xenopus, e.g. CHDs with complete penetrance, including transposition of the great arteries, double-outlet right ventricle, persistent truncus arteriosus and ventricular septum defects in mice (Zhou et al., 2007). These CHDs all belong to the group of outflow tract defects that also includes TOF (Neeb et al., 2013). Wnt11^−/− mice also displayed hypoplastic kidneys, whereby the tips of ureteric buds were partially lost and ureteric branching was affected (Majumdar et al., 2003). Nagy et al. (2016) observed renal hypoplasia in 25% of Wnt11 knockout animals with additionally reported secondary glomerular cysts. Lack of WNT11 resulted in reduced cell proliferation and increased apoptosis in the cortex as well as reduced expression of Wnt9b, Six2, Hox10 and Foxd1 at embryonic day (E)16.5. Additionally, they observed reduced tubular convolution, prompting the authors to hypothesize that disturbed conversion extension movements contribute to the observed renal phenotype (Nagy et al., 2016). wnt11b^−/− Xenopus showed laterality defects (∼25%), in line with a previous study on the role of Wnt11b in Xenopus (Houston et al., 2022; Walentek et al., 2013). These results, generated in model organisms in combination with the patient case presented in this study, strongly suggest similar functions of WNT11 in human development.

Furthermore, biallelic loss of function of WNT11 should be considered a novel underlying genetic cause of syndromal human phenotypes presenting with CHDs and renal hypoplasia/dysplasia, with or without laterality defects. Congenital anomalies of the kidneys and urinary tract (CAKUT) and cardiac malformations are known to co-occur in a number of human syndromes, many of which result from chromosomal aberrations (Sanna-Cherchi et al., 2018). However, bilateral renal hypoplasia overall is rarely observed, with an approximate incidence of 1 in 10,000 live births, fetal deaths/stillbirths and terminations of pregnancy for fetal anomalies (https://eu-rd-platform.jrc.ec.europa.eu/eurocat/eurocat-data/prevalence_en) (Harewood et al., 2010). Interestingly, autosomal dominantly inherited WNT5A dysfunction also results in right-ventricular outflow tract obstruction and renal anomalies with incomplete penetrance in addition to skeletal dysplasia features [Robinow syndrome; Online Mendelian Inheritance in Man (OMIM) #180700)] (Person et al., 2010). Wnt5 ligands share functional similarities with Wnt11 ligands, and, in mice, Wnt5a loss of function likewise causes renal developmental defects, including unilateral or bilateral kidney agenesis and hypoplasia, with altered pattern of ureteric tree organization as well as duplex kidneys with reduced penetrance (Pietilä et al., 2016). It has also been suggested that both WNT11 and WNT5A are required for proper secondary heart field development whereby they govern non-apoptosis-related Caspase activities and promote cardiac progenitor development (Bisson et al., 2015).

It remains elusive why the heterozygous mother of the affected boy presented with SIT. Functional haploinsufficiency during laterality development of the mother in combination with additional stressors could have contributed to the development of SIT, although there is no indication for haploinsufficiency in heterozygous Xenopus wnt11b mutants (Houston et al., 2022). Although whole-exome sequencing of the mother did not reveal any putative genetic cause for laterality defects, genetic abnormalities are generally identified in less than half of cases (Antony et al., 2022). Therefore, the mother could coincidentally have SIT independent of the WNT11 variant.

In conclusion, this work revealed WNT11 dysfunction as a novel likely cause of a complex developmental phenotype in humans and elucidated the functional properties of Wnt ligands in vivo. This enhances our understanding of the structure-function relationship in Wnt ligands of clinical relevance.

DNA samples were collected after obtaining informed consent from affected individuals or parents as part of the clinical diagnostic pathway at Radboud University Medical Center Nijmegen (Innovative diagnostic program). All clinical investigations were conducted in accordance to the principles expressed in the Declarations of Helsinki. Human subject research was approved by the Ethics committee of the Institute of Child Health, University College London, London, UK (GOSH R&D number 11MM03, REC number 08/H0713/82), the ethics committee Arnhem-Nijmegen, The Netherlands (ethical approval no. 2006-048) and the ethics committee of Freiburg University, Freiburg, Germany (votum no. 122/20).

Exome sequencing and sequence analysis were performed as described previously (Loges et al., 2018). Exomic sequences from DNA samples were enriched using a SureSelect Human All Exon V.6 Kit (Agilent Technologies, 5190-8892) according to the manufacturer’s protocol, followed by sequencing on a Hiseq PE150 (Illumina). Read alignment and variant calling were performed using Burrows–Wheeler (BWA)/GATKpipeline using default parameters with the human genome assembly hg19 (GRCh37) as reference. Vtools and ANNOVAR software were used to store and annotate variants. Following alignment and variant calling, serial variant filtering was performed for variants with a minor allele frequency equal or less than 1% in the Exome Aggregation Consortium, 1000 Genomes Project, and esp6500 and gnomAD databases, coding variants or variants within 5 bp of exon-intron boundaries. Obligate loss-of-function variants – such as canonical splice variants, frameshift and stop mutations – were prioritized over missense variants; however, missense variants were not excluded from the analysis. CNV calling was performed using Exome Depth (Plagnol et al., 2012), and BAM files were visually inspected for homozygous CNVs in all genes known to cause laterality defects or renal hypo(dysplasia). All the variants passing these filters were subsequently inspected at aligned read level with the aim of avoiding false call due to misalignment or low depth of coverage.

Genomic DNA was isolated by standard methods using a Qiagen kit for blood samples. Genomic DNA amplification was performed in a volume of 50 µl containing 30 ng DNA, 50 pM of each primer, 2 mM deoxynucleotide triphosphates and 1.0 U GoTaq DNA polymerase (Promega, M3001). PCR amplifications were carried out by an initial denaturation step at 94°C for 3 min, and 33 cycles as follows: 94°C for 30 s, 58-60°C for 30 s, and 72°C for 70 s, with a final extension at 72°C for 10 min. PCR products were verified by agarose gel electrophoresis, purified and sequenced bidirectionally. Sequence data were analyzed using CodonCode software. Primer sequences are listed in Table S1.

Protein structure predictions were created on the Galaxy EU platform, using Alphafold 2 (v.2.1.2) (https://toolshed.g2.bx.psu.edu/repository?repository_id=de07f280bfbbbd77&changeset_revision=c0e71cb2bd1b) (Community, 2024; Jumper et al., 2021) and adding the amino acid sequences into the Fasta Input. For each protein, ‘Model 1’ from the top five predicted structures was selected. Models were displayed in Cartoon or Surface representations, and snapshots of the models were taken from approximately the same angles.

Alignments of amino acid sequences were made using MultAlin (Corpet, 1988) and the following symbol comparison table: Blosum62-12-2. High consensus value (red) was set to 90% and low consensus value (blue) to 50%.

Wild-type and transgenic Xenopus laevis were obtained from the European Xenopus Resource Centre (EXRC) at the School of Biological Sciences, University of Portsmouth, Portsmouth, UK, or Xenopus1. Frog maintenance and care was conducted according to standard procedures in the AquaCore facility, University Freiburg, Medical Center (RI_00544) and based on recommendations provided by the international Xenopus community resource centers National Xenopus Resource (RRID:SCR_013731) and EXRC, as well as by Xenbase (RRID:SCR_003280) (Fisher et al., 2023). This work was done in compliance with German animal protection laws and was approved under Registrier-Nr. G-22/43 by the federal state of Baden-Württemberg.

X. laevis eggs were collected and in vitro fertilized, then cultured and microinjected by standard procedures (Sive et al., 2007, 2010). Embryos were injected with mRNAs at four- to eight-cell stage using a PicoSpritzer setup in 1/3× Modified Frog Ringer's solution (MR) with 2.5% Ficoll PM 400 (GE Healthcare, 17-0300-50), and were transferred after injection to 1/3× MR containing gentamycin (Carl Roth, cat. no.: 0233.4). Drop size was calibrated to ∼7-8 nl per injection.

Human WNT11 (HWNT11-pCMV6-Entry) was obtained from OriGene (RC219688) (matching ENST00000322563.8, NM_004626.3, CCDS8242) and subcloned into pCS108 (using Cla1 and EcoR1 enzymes; NEB, R0197 and R3101) to serve as templates for in vitro mRNA synthesis using the primers listed in Table S2. Xenopus Wnt11b and dnWnt11b constructs were derived from Dichmann et al. (2015) and Tada and Smith (2000). The Xenopus Wnt3a-encoding construct was a gift from C. Niehrs (Kiecker and Niehrs, 2001). All variants were generated by site-directed mutagenesis (NEB, E0554S) using the primers listed in Table S2. mRNAs/DNAs encoding WNT11, WNT11^c.814delG, WNT11^Δ814-1062, WNT11^Δ850-1062, Wnt11b, dnWnt11b and Wnt11b^Δ811-1059 (mRNA, 70-100 ng/µl; DNA, 2 ng/µl), and Wnt3a (5 ng/µl) were bilaterally injected (unless specified) together with centrin4-gfp or membrane-gfp (50 ng/µl) or Rhodamine Dextran (Invitrogen, 11590226) as lineage tracers. All mRNAs were prepared using a mMessage Machine kit using Sp6 (Invitrogen, AM1340) supplemented with RNAse Inhibitor (Promega, N251B).

For whole-mount in situ hybridization against nodal1 and pitx2c, the antisense probe was derived from Schweickert et al. (2001). Embryos were fixed in MEMFA [100 mM MOPS pH 7.4, 2 mM EGTA, 1 mM MgSO₄, 3.7% (v/v) formaldehyde] overnight at 4°C and stored in 100% ethanol at −20°C until use. DNAs were purified using a PureYield Midiprep kit (Promega, A2492) and linearized before in vitro synthesis of anti-sense RNA probes using T7 or Sp6 polymerase (Promega, P2077 and P108G), RNAse inhibitor and digoxigenin-labeled ribonucleotide triphosphates (Roche, 3359247910 and 11277057001). Embryos were in situ hybridized according to Harland (1991), stained with BM Purple (Roche, 1442074001) and imaged.

Embryos were staged according to Nieuwkoop and Faber (1994) (Zahn et al., 2022). Embryos that could not be assessed for pitx2c expression patterns (owing to missing axes or coiled morphology) were not included in the statistics for pitx2c expression. Analysis of nodal1 expression was conducted on dorsal explants. Images of embryos for morphological evaluations and after in situ hybridization were generated using Zeiss Stemi508 with Axiocam208-color, and images were adjusted for color balance, brightness and contrast using Adobe Photoshop.

Canonical Wnt-reporter activity was assessed in transgenic embryos [pbin7Lef::dGFP; line Xla.Tg(WntREs:dEGFP)^Vlemx] (Haas et al., 2019) by comparing fluorescence between the manipulated and non-manipulated control side on images generated using a Zeiss AxioZoom setup. Images were adjusted for brightness and contrast using ImageJ (Schindelin et al., 2012) and Adobe Photoshop.

Eight to 15 embryos were placed in an Eppendorf tube without fluid and stored at −20°C. For standard western blotting, 100 µl 1× lysis buffer [20 mM Tris-HCl pH 8, 150 mM NaCl, 2 mM EDTA, 1× Protease Inhibitor (Roche, 04693116001), 1% NP40 (Sigma-Aldrich, I8896)] was added, the embryos were smashed by pipetting up and down several times, and the samples were centrifuged at 4°C at 21300 g for 15 min. For the dimerization assay, as has been previously described (Cha et al., 2009), embryos were instead smashed by pipetting up and down with 90 µl ice-cold phosphate-buffered saline (PBS) with 0.1% Triton X-100 and 10 µl Protease Inhibitor (Roche, 04693116001), and the samples were centrifuged at 4°C at 14,400 g for 10 min.

The supernatant was transferred to a fresh Eppendorf tube. Then, 4× Laemmli buffer (50 ml 4× buffer, 1 M Tris-HCl, pH 6.8, 4 g SDS, 20 ml glycerol, 10 ml 2-mercaptoethanol, 0.1 g Bromophenol Blue) was added to the supernatant to a 1× final concentration. For standard western blotting, 4× Laemmli buffer containing 5% 2-mercaptoethanol (Roth, 4227.2) was used, and the samples were cooked at 95°C for 5-10 min on a shaking plate. For the dimerization assay, 4× Laemmli buffer without 2-mercaptoethanol (non-reducing condition) or 10% 2-mercaptoethanol (reducing condition) was used, and the samples were not cooked.

For two gels, 10 ml of 8% separating gel was made of 2.5 ml 4× Tris SDS (Roth, 2326), pH 8.8, 2 ml 40% acrylamide (Sigma-Aldrich, A7802), 5.4 ml H₂O, 40 µl tetramethylethylenediamine (TEMED; Roth, 2367.1), 100 µl 10% ammonium peroxodisulphate (APS; Roth, 9592.2). For two gels, 5 ml of 4% collecting gel was made of 1.25 ml 4× Tris SDS, pH 6.8, 0.625 ml 40% acrylamide, 3.11 ml H₂O, 50 µl TEMED, 50 µl 10% APS. A 1× running buffer [25 mM Tris-HCl, pH 8, 192 mM glycine (Roth, 3187) in distilled water] was used for electrophoresis. Then, 10-12 µl Precision Plus Protein Western C Standards Ladder (Bio-Rad, 161-0376) and 10-20 µl of each sample were loaded. The electrophoresis was run at 120 V for 1-2 h at room temperature.

Semi-dry transfer onto an activated polyvinylidene fluoride membrane (Thermo Fisher Scientific, 88518) was conducted in 1× Towbin buffer with 0.1% SDS for 30 min (25 mM Tris-base, 192 mM glycine, 1% SDS) using a PerfectBlue Semi-Dry Electroblotter Sedec M (VWR, 700-1220). Transfer was performed at a constant current of 10 mA for 90 min. Membranes and gels were stored in 1× TBStw (100 mM Tris-base, 500 mM NaCl, 1% Tween 20) at 4°C until further use.

Membranes were blocked for at least 45 min using 5% non-fat dry milk (Roth, T145.3). The following primary antibodies were used at 1:1000 and incubated overnight at 4°C: Rabbit Polyclonal Anti-WNT11 (Thermo Fisher Scientific, PA5-21712, RRID:AB_11154198) and, as a loading control, Mouse Monoclonal Alpha-Tubulin (Cell Signaling Technology, 3873, RRID:AB_1904178). The membrane was then washed in 1× TBStw for 4×20 min. The following secondary antibodies were used at 1:3000 and incubated for 2 h at room temperature: HRP-linked Anti-Mouse IgG (Cell Signaling Technology, 7076, RRID:AB_330924) and HRP-linked Anti-Rabbit IgG (Cell Signaling Technology, 7074, RRID:AB_2099233). Afterwards, the membrane was washed in 1× TBStw for 6×10 min. Membranes were incubated with a mixture of 500 µl peroxide solution and 500 µl Luminol/enhancer solution [both from Clarity™ Western ECL Substrate (Bio-Rad, 170-5061)] for 5 min in the dark at room temperature. Membranes were imaged using the Odyssey XF Imaging System by LI-COR. Afterwards, membranes were washed and stored in TBStw. Membranes were stripped for loading control (α-Tubulin) re-probing with acid stripping buffer (for 100 ml: 5 ml 2 M glycine-HCl, pH 2.1, 500 µl Tween 20, 700 µl 2-mercaptoethanol) for 1 h on a tumbler. The membranes were again rinsed in distilled water and washed 3×5 min with TBStw. To validate effective membrane stripping, the membranes were imaged before re-probing. Results shown in Fig. 4C,D were obtained from two batches (biological replicates, different parents) each.

The membrane image shown in Fig. 4D was loaded into Fiji/ImageJ and used to define regions of interest (ROIs) of the bands, which are depicted in Fig. S3C. Signal intensity within ROIs was quantified intensity using the ‘measure’ function. The areas and intensity values are depicted in Table S3. Wnt11 band intensity was divided by α-Tubulin band intensity (normalization), and fold intensity was calculated by dividing normalized values by the control (uninjected control) value.

Stacked bar graphs were generated in Microsoft Excel. Sample sizes for all experiments were chosen based on previous experience and used embryos derived from at least two different females. No randomization or blinding was applied. For statistical evaluation, chi-squared (χ²) test was performed using Microsoft Excel.

For some of the in situ, morphological and Wnt-reporter analyses, the same embryos were used in multiple graphs.

We thank S. Schefold for expert technical help; S. Arnold for support and discussions; C. Softley for help with setting up western blots; Xenbase and EXRC for Xenopus resources; Light Imaging Center Freiburg, Bio Imaging Core Light Microscopy (BiMiC) and AquaCore for microscope/animal resources; and B. Grüning and the Freiburg Galaxy Team for bioinformatics platform and support.