CUL3‐Related Neurodevelopmental Disorder: Expanding the Prenatal Phenotype

ABSTRACT

Objective

Pathogenic variants of the CUL3 gene are known to cause a neurodevelopmental disorder with a partially described prenatal phenotype. This study further characterizes and expands the spectrum of prenatal sonographic findings associated with the disorder to improve prenatal diagnosis and counseling.

Methods

This multi‐center case series adds seven new cases from 5 pedigrees with pathogenic CUL3 variants identified through exome sequencing. We analyzed the new data and integrated the findings with a comprehensive review of 18 prenatal cases in the literature.

Results

Our analysis of a combined cohort of 25 prenatal cases confirms that intrauterine growth restriction and increased nuchal translucency are frequent, nonspecific findings. The most significant novel finding in our series was cerebellar hypoplasia, which was identified in three of the new cases. Other findings included cardiac anomalies, abnormal sulcation, and skeletal abnormalities.

Conclusion

This study expands on the known prenatal phenotype of CUL3‐related neurodevelopmental disorders, proposing cerebellar hypoplasia as a new sonographic marker. The presence of cerebellar hypoplasia should significantly raise suspicion for a CUL3‐related neurodevelopmental disorder. This finding provides a strong rationale for pursuing exome sequencing and serves as critical evidence for the clinical interpretation of CUL3 variants.

Summary

• What is already known about this topic?

  • ◦The main established prenatal CUL3‐related neurodevelopmental disorder is Fetal growth restriction (FGR).

• What does this study add?

  • ◦Cerebellar hypoplasia might be a more specific prenatal finding
  • ◦This study widens the currently known prenatal spectrum of the CUL3‐related neurodevelopmental disorder.

1. Introduction

The gene CUL3 is associated with two distinct syndromes. One is Neurodevelopmental disorder with or without autism or seizures (#MIM 619239) and the other is Pseudohypoaldosteronism, type IIE (#MIM 614496). Both are inherited in an autosomal dominant pattern.

An association between CUL3 and autism was proposed in two studies from 2012 [1, 2]. In 2020, Dong et al. showed the cul3‐deficient mice had social deficits and anxiety‐like behaviors (symptoms that are well known in autism), with CUL3 playing a critical role in neural development and synaptic transmission [3]. A 2023 study reported that the Cul3 gene in mice showed higher expression levels in the dorsal striatum and cerebellum compared with the averaged whole‐brain expression [4]. In 2020, Nakashima et al. described three patients with de novo CUL3 variants [5]. All three patients had developmental delays and two suffered from intractable seizures. Recently, a larger cohort of 37 patients with heterozygous CUL3 variants was published by Blackburn et al. [6] Almost all the patients had a loss‐of‐function variant, six of which were located in a splice site. Most variants were de novo, nine were inherited from mostly affected parents and the inheritance of four remaining variants was unknown. A functional assay showed that 4E‐BP1 (EIF4EBP1), a prominent substrate of CUL3, fails to be targeted for proteasomal degradation in patient‐derived cells.

While most reported variants of CUL3 were either de novo or inherited from an affected parent, a case of an affected individual that inherited a pathogenic CUL3 variant from an unaffected father was reported [7]. In additional cases of CUL3 inherited variants, the reported parental phenotypes were relatively mild, including learning difficulties, mild intellectual disability, or speech delay. Of the 37 families reported by Blackburn et al. [6], almost all patients had global developmental delays and dysmorphic features (including high forehead, long face and others) and 36% were diagnosed with autism spectrum disorder. Abnormalities of the hands and feet were noticed in 51%. In terms of prenatal findings, 35% of patients had fetal growth restriction (FGR). Brain MRI scans, performed in about half of the cases did not show obvious recurrent abnormalities. The present study proposes additional prenatal phenotypes that seem to be associated with CUL3‐related neurodevelopmental disorders. We report two unrelated families carrying CUL3 variants, presenting with cerebellar hypoplasia and FGR, together with additional cases showing ultra‐sonographic prenatal findings. Based on these additional cases and a review of previously reported CUL3 prenatal phenotypes, we aim to describe the phenotypic spectrum of prenatal CUL3‐related phenotypes.

2. Methods

Two cases of prenatally detected CUL3 variants with prenatal phenotypes that included cerebellar hypoplasia were counseled at the Meir Medical Center Genetics Institute, in Kfar Saba, Israel (Cases 1–2). Following this observation, we approached additional genetic laboratories in the country that perform prenatal exome sequencing to ascertain further cases presenting with prenatal phenotypes associated with CUL3 variants. We received data for three additional cases (cases 3–5).

3. Results

A review of the two prenatal cases presented here, together with five additional novel cases (including two siblings of the initial cases) and 18 previously published CUL3‐related neurodevelopmental disorder cases bearing prenatal findings, is depicted in Table 1.

TABLE 1.

Novel and previously published cases of CUL3‐related neurodevelopmental disorder with congenital/prenatal findings.

No.	Test model	Variant (NM_003590.5)	Classification (present study)	Inher‐itance	Gestational age/age at presentation	Sex	Prenatal/congenital phenotype	Family history	Study a
1A	Trio ES	c.1378–2A > T	LP: PVS1, PM2	Mat	2.5 years	M	GDD, SGA, microcephaly	IVF pregnancies; same variant in both cases	This study
1B	Trio ES				25w + 3 days	F	FGR, CH, early sulcation anomaly		This study
2A	Quad ES	c.1549_1552del, p.Ser517Profs a 23	p: PVS1,PS4, PM2	Pat	32w	F	CH, Fetal cystic hygroma, VSD, short long bone	IVF pregnancies, one additional ended with TOP due to malformations	This study
2B					31w + 5 days	M	FGR, CH, Microcephaly, skeletal dysplasia	Same as above	This study
3	Trio ES	c.963G > C, p.Leu321Phe	LP: PS2,PM2, PP2,	DN	15w	F	Increased NT, Type I truncus arteriosus	Not available	This study
4	Trio ES	c.1610 + 2T > C	LP: PVS1,PS2, PM2	DN	12w + 3 days	F	Increased NT, Fetal cystic hygroma, retrognathia	Not available	This study
5	Trio ES	c.2293A > C, p,Thr765Pro	LP: PS2,PM2,PP2	DN	33w + 2 days	F	FGR, short corpus callosum	Not available	This study
6	Trio ES	c.1664dupG, p.Ser556fs	p: PVS1,PS2, PM2	DN	29w	F	Head circumference −1.5SD		38735835
7		c.173A > G, p.Tyr58Cys		DN	2 years	M	Oligohydramnios, absent thumb, microcephaly		30311385
8	Trio ES	c.854T > C, p.Val285Ala		DN		F	ASD, PVS		32341456
9		c.394C > T, p.Gln132 a					LVO		33084842
10	Trio ES	c.433_436del;p.Iso145Phe fs a 23		DN			LVO: Hypoplastic mitral valve, hypoplastic aortic annulus, aortic stenosis, coarctation, CHDP, scoliosis		23665959
11		c.493_494del, p.Leu165Ilefs a 37		DN	2 years	M	FGR		39301775, P13
12		c.602_603dup, p.Glu202Metfs a 25		Mat (aff.)	11 years	M	FGR		39301775, P7
13		c.743_744dup, p.Met249 a		Mat	8 years	M	FGR		39301775, P9
14	Proband ES	c.827_828del, p.Ile276Serfs a 28		DN	14 years	M	FGR, Increased NT, VSD		39301775, P23
15	ES	c.1538G > A, p.Trp513 a		Mat (unaff.)	7 years	M	FGR, Increased NT		39301775, P18
16	Trio ES	c.1549_1552del; p.Ser517Profs a 23		DN	21w + 6		Mild FGR, CHD (two‐vessel umbilical cord, PAA, mildly hypertrophic RV, mod. TR, and reverse flow in the ductus venosus, hepatic vein, and IVC)		37665043
17		c.1549_1552del, p.Ser517Profs a 23		Pat (aff. brother also a carrier)	12 years	F	FGR		39301775, P12
18	Trio ES, proband GS	c.1549_1552del, p.Ser517Profs a 23		Mat (unaff.)	37 years	F	FGR		39301775, P29
19		c.1549_1552del, p.Ser517Profs a 23					CHD		33084842
20		c.1765A > T, p.Lys589 a		DN	17 years	M	FGR		39301775, P10
21		c.1767_1768insA, p.His590Thrfs a 17		DN	13 years	M	FGR, placental calcifications		39301775, P16
22	Trio ES	c.264 + 2T > G		DN	3 years	M	FGR		39301775, P34
23	Trio ES	c.883 + 1G > T		DN	5 years	F	FGR		39301775, P31

Abbreviations: aff.‐ affected, ASD‐Atrial septal defect, CH‐Cerebellar hypoplasia, CHD‐ Congenital heart defect, CHDP‐ Congenital hip dysplasia, DN‐ de novo, ES‐ Exome sequencing, F‐female, GS‐ Genome sequencing, FGR‐ Fetal growth restriction, IVC‐ Inferior vena cava, IVF‐ In vitro fertilization, LP‐ Likely pathogenic, LVO‐Left ventricular obstruction, Mat‐ Maternal, M‐male, NT‐ Nuchal translucency, P‐ Pathogenic, PAA‐Pulmonary artery atresia, Pat‐ Paternal, PVS‐ Pulmonary valve stenosis, RV‐ Right ventricle, SGA‐ Small for gestational age, TOP‐ Termination of pregnancy, TR‐Tricuspid regurgitation, unaff.‐ unaffected, VSD‐ Ventricular septal defect, VUS‐ Variant of uncertain significance.

^a Numbers represent a PMID (PubMed Identifier).

3.1. Case 1

A 40‐year‐old female presented at 19 weeks of gestation for genetic counseling before amniocentesis due to advanced maternal age.

The mother reported learning difficulties as a child; however, she holds a master's degree in Special Education. The father, 45 years old, also completed higher education.

The couple had a 4‐year‐old son from a previous pregnancy (Case 1A). This pregnancy was achieved via in vitro fertilization due to male‐factor infertility. Pregnancy follow‐up was normal. Amniocentesis was performed due to advanced maternal age, with chromosomal microarray analysis (CMA) showing the same 15q11.2 recurrent microdeletion seen in the father's nephew and classified as a likely benign variant. The child was born small for gestational age (2.67 kg at 40 weeks, first percentile [8]) and with microcephaly (33 cm, second percentile [8]). He experienced delayed development, with a General Developmental Quotient test score of 60 at 18 months and a diagnosis of global developmental delay. Autism was ruled out by 2 years of age. He also had mildly dysmorphic features. Trio ES performed postnatally at 2.5 years revealed a maternally inherited CUL3 splice variant, initially classified as a variant of uncertain significance (VUS), NM_003590.5, c.1378–2A > T, chr2‐225367791 T > A (GRCh37/hg19). The variant was not detected in her parents or brother (de novo).

This rare CUL3 variant, absent from gnomAD v4.1, results in an in‐frame exon 10 deletion via acceptor loss as well as alternative acceptor gains at 44 and 137 bases 3′, causing a frame‐shift and premature termination codon (SpliceAI prediction tool) [9]. Hence, this variant receives the criteria PVS1 Very Strong [10, 11], resulting in a “Likely Pathogenic” classification.

To further investigate the CUL3 variant, a focused RNA analysis of blood samples from the child and his mother was performed. Analysis showed partial intron 9 inclusion in ∼10% of the reads, resulting in an out‐of‐frame 46 nucleotide addition to exon 10 (Figure 1A). All the aberrant reads were seen in cis with the suspected variant. This finding was observed in both the proband's and the mother's samples and was not identified in the control samples (Figure 1B). In addition, a larger intronic insertion was observed in 1% of the reads. This finding supports the prediction of both gain and loss of the acceptor site in blood tissue and probably nonsense‐mediated decay in a significant proportion of RNA transcripts, thus supporting the classification of the variant as likely pathogenic.

FIGURE 1.

(A) RNA Sequencing analysis of the splicing variant presented in Case 1 showing aberrant splicing in ∼10% of reads in the patient (first line) and his mother (second line), which includes part of intron 9. Control samples (lines 3 and 4) show normal splicing. (B) IGV (Integrative Genomics Viewer) view of patient and control samples‐variant is colored in green, with an additional segment (part of intron 9) included.

3.2. Current Pregnancy—Case 1B (Fetus)

Pregnancy follow‐up included a normal nuchal translucency (NT), and normal anatomical scans at 16 weeks. The fetus had female genitalia. At 25 + 3 weeks, FGR was detected. The fetus was in the seventh percentile according to Hadlock et al. [12]. The cerebellar size was below first percentile (Figure 2A), suggesting cerebellar hypoplasia. In addition, the cingulate gyrus was not visualized, suggesting an early sulcation anomaly. Chromosomal microarray revealed a paternally inherited 15q11.2 deletion as well as a de novo 1.5 Mb deletion at 12q21.32 (chr12:87341964–88851396, hg19), which was classified as a VUS. It encompasses eight genes, two of which are OMIM morbid genes, associated with autosomal recessive inheritance (TMTC3 MIM# 617218, CEP290 MIM# 610142). Trio ES was also performed, and the fetus was found to carry the maternally inherited CUL3 variant. No additional variants were detected (testing also included specific analysis of the TMTC3 and CEP290 genes). Based on the molecular and anatomical findings, the couple opted for pregnancy termination at 27 weeks. Post‐mortem examination revealed cerebellar hypoplasia with preserved morphology. The axial cerebellar width measured 23 mm (expected: 27 mm), and the head circumference was 22.5 cm (1.0 SD below the mean). Hypoplasia predominantly affected the posterior aspects of the cerebellar lobes. Furthermore, the cerebral hemispheres exhibited mildly deficient operculization of the Sylvian fissures, and histological sections identified focal cortical thickening.

FIGURE 2.

Ultrasonographic scans of presented cases showing prenatal cerebellar hypoplasia. (A) Case 1: Ultrasound at 25 + 3 weeks revealing a cerebellar size consistently with 22 + 1 week. (B) Case 1: Ultrasound at 26 + 2 weeks revealing a cerebellar size consistently with 23 + 4 weeks.

3.3. Case 2

A 39‐year‐old female presented at 16 weeks of gestation in her third pregnancy for genetic counseling before amniocentesis due to FGR.

Due to unexplained infertility, with normal parental chromosome analysis, the couple conceived via IVF. The first pregnancy was terminated due to a maternally inherited deletion that included exons 4 and 5 in SYN1 (chrX:47,464,377–47,464,778, hg19) detected in trio exome sequencing (ES) (Table 1, Case 2a). CMA during the patient's second pregnancy showed the same SYN1 deletion in a male fetus, which led to pregnancy termination at 20 weeks, with normal ultrasound scans up to that point. ES was not performed. The mother reported being in good health with no apparent cognitive difficulties.

During the patient's third pregnancy, the ES from the first pregnancy was revised, and an additional variant was reported. This was a paternally inherited variant in the gene CUL3, c.1549_1552delTCAG, p.Ser517Profs*23, NM_003590.5, classified as likely pathogenic (Table 1, Case 2B). Notably, the father exhibited normal cognitive abilities, with a head circumference at the 90th percentile. Sonographic scans during the patient's third pregnancy were abnormal. At 16 weeks, estimated fetal weight was at the first percentile, and at 21 weeks, FGR was seen, with the fetus at the 0.26 percentile. The cerebellar size corresponded to 19 weeks GA (second percentile). Cerebellar hypoplasia was diagnosed, with the vermis size preserved. The tibia and humerus bones corresponded to 18 + 5 weeks GA. At 26 + 2 weeks, sonographic scans showed severe FGR with cerebellar hypoplasia still apparent (Figure 2B). Fetal echocardiography at that time was normal. CMA showed a normal male analysis without the SYN1 deletion. ES showed the CUL3 variant, previously detected in the first fetus. The pregnancy was terminated at 26 + 3 weeks.

Following outreach to other genetics laboratories, our cohort was expanded by an additional three cases featuring prenatal phenotypes associated with the CUL3 gene. Of the seven new cases in this cohort, cerebellar hypoplasia was reported in three, with abnormal sulcation reported in one of them. Microcephaly was reported in two cases and short corpus callosum was reported in one case. FGR was reported in three of the seven cases, and short long‐bones and suspected skeletal dysplasia were reported in two. Elevated NT and cystic hygroma were detected in three cases and heart malformation in two others.

4. Discussion

The CUL3‐related neurodevelopmental disorder has been described in several papers, including the delineation of prenatal findings. However, while FGR was reported in a third of the cases in previous reports, including the large review by Blackburn et al. [6], the sonographic finding of cerebellar hypoplasia has not yet been described. The present paper presents two unrelated families with fetal findings of cerebellar hypoplasia. We believe this represents a new phenotypic finding of the CUL3‐related neurodevelopmental disorder. Hints for cerebellar hypoplasia can be seen in the animal study conducted by Amar et al. in 2021, where decreased cerebellum volume was seen in a haploinsufficient Cul3 mouse model [13]. Our review of previously published cases, together with those first presented in this paper, show that together with FGR, increased NT is another nonspecific typical prenatal feature of the syndrome. The involvement of the heart and cardiovascular system is also a recurrent feature (as reviewed by Di Francesco et al. [14]) and is sometimes already apparent in the prenatal period. While FGR and increased NT are nonspecific findings, and the cardiovascular involvement is diverse, cerebellar hypoplasia might be a more suggestive finding for the disorder, pointing to the need for sequencing‐based prenatal genetic testing, such as exome or genome sequencing.

Cerebellar hypoplasia is associated with major structural anomalies and adverse outcomes. Among fetuses with FGR, it is associated with fetal demise or adverse neonatal outcomes [15]. Both genetic and prenatal non‐genetic causes such as congenital infections, including CMV [16], have been described as a cause of cerebellar hypoplasia. Genetic causes include chromosomal abnormalities, such as trisomy 13, 18 and 21, partial trisomy 12q, monosomy 21q, and monogenic syndromes [17]. In a 2019 study of 57 probands with cerebellar hypoplasia, the diagnostic yield of monogenic etiologies detected by exome sequencing was 51%, with 29 variants discovered in 22 related genes [18]. Interestingly, CUL3 was not among them. Additional studies are required to assess the prevalence of cerebellar findings among patients with CUL3‐related neurodevelopmental disorder.

Up to now, elevated NT has not been associated with CUL3 pathogenic variants. We report elevated NT/cystic hygroma in three of the seven cases in our cohort. Elevated NT has been linked to multiple genetic chromosomal and monogenic anomalies. A 2023 review of data from eight articles and 324 fetuses with increased NT and normal CMA showed that 8% of fetuses had either pathogenic or likely pathogenic variants on their exome sequencing. However, CUL3 variants were not reported in this review [19].

The present study characterizes the early prenatally detectable features of CUL3‐related neurodevelopmental disorder. These specific sonographic findings are crucial not only for indicating the need for genetic testing but also for aiding the clinical interpretation of CUL3 variants found by exome sequencing [20].

Based on the prenatal findings reported in our cohort, in cases where a CUL3 variant is detected on prenatal exome sequencing, we recommend following the fetus's growth parameters and cerebellar measurements, and to consider performing a prenatal MRI and fetal echocardiogram.

This study has several limitations that warrant consideration. A primary source of bias is that our cohort was composed of fetuses referred for prenatal exome sequencing, a population with inherently increased likelihood of congenital malformations. Consequently, our data cannot be used to establish the true prevalence of sonographic findings in an unselected population of fetuses with pathogenic CUL3 variants, and the rate may be lower than that observed here. Furthermore, prenatal information gathered retrospectively from postnatally diagnosed cases is subject to recall bias and incomplete medical records.

It is noteworthy, however, that pathogenic CUL3 variants are not confined to fetuses with structural anomalies. A recent retrospective study of 1020 low‐risk pregnancies with structurally normal fetuses identified a likely pathogenic CUL3 variant (c.1664dupG p.Ser556fs) [21].

5. Conclusion

This study expands the known prenatal phenotypic spectrum for pathogenic CUL3 variants. We propose that cerebellar hypoplasia may be a more specific indicator of a CUL3‐related neurodevelopmental disorder than common, nonspecific findings such as increased NT or FGR. The presence of cerebellar hypoplasia is therefore clinically significant as it not only provides a strong rationale for pursuing genetic testing but also provides critical evidence for interpreting the pathogenicity of identified CUL3 variants.

Funding

This research received no specific grant from any funding agency in the public, commercial, or not‐for‐profit sectors.

Ethics Statement

The study was approved by the relevant Institutional Review Boards.

Consent

The requirement for informed consent was waived.

Conflicts of Interest

The authors declare no conflicts of interest.

Data Availability Statement

The data supporting the findings of this study are available from the corresponding author upon reasonable request.