The recurrent de novo c.2011C>T missense variant in MTSS2 causes syndromic intellectual disability

Summary

MTSS2, also known as MTSS1L, binds to plasma membranes and modulates their bending. MTSS2 is highly expressed in the central nervous system (CNS) and appears to be involved in activity-dependent synaptic plasticity. Variants in MTSS2 have not yet been associated with a human phenotype in OMIM. Here we report five individuals with the same heterozygous de novo variant in MTSS2 (GenBank: NM_138383.2: c.2011C>T [p.Arg671Trp]) identified by exome sequencing. The individuals present with global developmental delay, mild intellectual disability, ophthalmological anomalies, microcephaly or relative microcephaly, and shared mild facial dysmorphisms. Immunoblots of fibroblasts from two affected individuals revealed that the variant does not significantly alter MTSS2 levels. We modeled the variant in Drosophila and showed that the fly ortholog missing-in-metastasis (mim) was widely expressed in most neurons and a subset of glia of the CNS. Loss of mim led to a reduction in lifespan, impaired locomotor behavior, and reduced synaptic transmission in adult flies. Expression of the human MTSS2 reference cDNA rescued the mim loss-of-function (LoF) phenotypes, whereas the c.2011C>T variant had decreased rescue ability compared to the reference, suggesting it is a partial LoF allele. However, elevated expression of the variant, but not the reference MTSS2 cDNA, led to similar defects as observed by mim LoF, suggesting that the variant is toxic and may act as a dominant-negative allele when expressed in flies. In summary, our findings support that mim is important for appropriate neural function, and that the MTSS2 c.2011C>T variant causes a syndromic form of intellectual disability.

A cohort of five unrelated individuals with the same heterozygous de novo variant in MTSS2 (GenBank: NM_138383.2: c.2011C>T [p.Arg671Trp]), present with syndromic mild intellectual disability. Modeling in Drosophila suggested that the variant had decreased normal function and increased toxicity compared to the reference MTSS2 and may act as a dominant-negative allele.

Main text

MTSS2 (MIM: 616951), also known as MTSS1L (MTSS I-BAR domain containing 2), is a member of a family consisting of five proteins sharing a conserved I-BAR (inverse BAR) domain at the N terminus, as well as an actin-binding WH2 (WASP-homology 2) domain at the C terminus.1, 2, 3 I-BAR domains generate inverse membrane curvature and induce formation of plasma membrane protrusions when expressed in cells by binding to the inner leaflet of membranes through their convex lipid-binding interface.^4,5 Based on structural features and phylogenetic relationships, MTSS2 and its paralog MTSS1 belong to the Mim (missing-in-metastasis) subfamily.^1,6 In addition to the I-BAR and WH2 domains, MTSS2 also contains a serine-rich region, three proline-rich motifs, and a leucine zipper motif (Figure 1A).² Human MTSS2 is mainly expressed in the CNS while MTSS1 is expressed at variable levels in most tissues (GTex).⁷ Mouse Mtss2 has been shown to be highly expressed in fetal radial glia, which are multipotent cells involved in neuronal migration, neurogenesis, and gliogenesis in the developing CNS.² In adult mice, Mtss2 is predominantly expressed in the cerebellum but not the hippocampus; however, hippocampal expression can be highly induced by synaptic activity such as exercise and may promote dendritic spine formation of neurons.^2,8 Variants in MTSS2 have not been previously linked to any genetic disorder in OMIM.

Figure 1.

Genetic and clinical information of the c.2011C>T variant in MTSS2

(A) Structure of the MTSS2 protein (GenBank: NM_138383.2; NP_612392.1) and the position of the c.2011C>T (p.Arg671Trp) variant. MTSS2 is composed of an N-terminal I-BAR domain, a serine-rich region, three proline-rich motifs, a leucine zipper motif, and a C-terminal WH2 domain (upper panel). The structure of Drosophila Mim protein (GenBank: NM_001259228; NP_001246157.1) with indication of the corresponding Arg (p.Arg838). The breaks represent stretches of unaligned sequences (lower panel).

(B) Three affected individuals share the following facial dysmorphisms: long upslanting palpebral fissures, bitemporal narrowing, arched eyebrows, thickened ear helices, and epicanthal folds. Individual 1 at 8 years old; individual 2 at 42 years old; individual 4 at 15 months of age.

(C) The human MTSS2 p.Arg671 (GenBank: NP_612392.1) is present in all species listed: mouse (GenBank: NP_941027.1), rat (GenBank: NP_001178487.1), Xenopus (GenBank: XP_031756693.1), zebrafish (GenBank: XP_005170001.1), Drosophila (GenBank: NP_001246157.1), and C. elegans (GenBank: NP_001317862.1).

(D) Real-time PCR shows decreased MTSS2 expression compared to age-matched control subjects in fibroblasts from individuals 1 and 2. Error bars: SEM. p values were calculated by unpaired t test.

(E) Immunoblot of MTSS2 in fibroblasts from individual 1 and individual 2 show no consistent changes of protein levels across affected individuals and control subjects. Total protein served as a loading control. P1, individual 1; P2, individual 2; C1–C6, age- and sex-matched control subjects from 6 individuals.

Through genomic matchmaking, including the use of the MatchMaker Exchange⁹ and one-sided matchmaking strategies,¹⁰ we identified a cohort of five individuals affected by an intellectual disability syndrome who all carry the same de novo heterozygous (GenBank: NM_138383.2: c.2011C>T [p.Arg671Trp]) variant in MTSS2 (Table 1). Informed consent was obtained from the five families. The five individuals range in age from 18 months to 42 years. Individual 2 is the only adult. The five individuals present with mild developmental delay and/or intellectual disability. Individuals 1 and 2 have a developmental coordination disorder. Individuals 1 and 3 have been diagnosed with autism spectrum disorder. Individual 2 developed adult-onset focal absence seizures, which are well controlled with one antiepileptic drug (topiramate). The remaining four individuals have not had any reported seizures, and individual 1 has had normal electroencephalograms. Individuals 2 and 5 have microcephaly, whereas the other three individuals have a relative microcephaly with head circumferences that are low compared to their heights (Table 1). A review of facial photographs from individuals 1, 2, 3, and 4 reveals shared mild dysmorphic features including long upslanting palpebral fissures, bitemporal narrowing, arched eyebrows, and epicanthal folds (Figure 1B).

Table 1.

Main clinical features of five unrelated individuals with a de novo c.2011C>T (p.Arg671Trp) variant in MTSS2 (GenBank: NM_138383.2)

Individual	1	2	3	4	5
Agea	8 yo	42 yo	15 yo	14 months	21 months
Sex	male	female	male	male	male
Ethnicity	European	European	European	European	Chinese
ID or GDDb	mild ID	mild ID	mild ID	GDD	GDD
Autism spectrum disorder	+	–	+	N/Ab	N/Ab
Seizures	–	+	–	–	–
Ophthalmological anomalies	nystagmus, foveal hypoplasia	optic atrophy	nystagmus, ptosis	bilateral iris cysts	nystagmus, ptosis
Sensorineural hearing loss	+	+	Ub	–	–
Distinctive facial featuresc	+	+	+	+	Ub
Microcephaly or relative microcephaly (head circumference centile)d	+ (40%)	+ (<0.1%, −2.9 SD)	+ (2%)	+ (2%)	+ (<0.1%, −2.6 SD)
Height centiled	83%	60%	Ub	91%	70%

^aAge at last clinical evaluation.

^bAbbreviations: ID, intellectual disability; GDD, global developmental delay (individuals 4 and 5 are too young to be evaluated for ID); N/A, not applicable; U, unknown.

^cUpslanting palpebral fissures, epicanthal folds, bitemporal narrowing.

^dCentiles based on WHO growth curves and Nellhaus head circumference curves.^11,12

While all but individual 2 have normal vision, all five individuals present with ophthalmological anomalies (Table 1). Three individuals have nystagmus and two have ptosis, but the other ocular anomalies are not present in more than one individual in this cohort. Individual 1 has a mild form of congenital foveal hypoplasia, and individual 4 has bilateral iris cysts. Individual 2, the only adult, has a history of progressive bilateral optic atrophy, first noted at age 13, and is now legally blind.

Individual 2 developed progressive bilateral sensorineural hearing loss requiring bilateral hearing aids that started in late childhood, and her hearing loss has now been stable for more than 10 years. Individual 1 passed hearing screens at birth and at 19 months of age but was found to have mild bilateral sensorineural hearing loss at 8 years of age. The hearing status of individual 3 is unknown, and individuals 4 and 5 are reported to have normal hearing, but they did not undergo a hearing test.

From the available records, there does not appear to be any clearly consistent brain magnetic resonance imaging (MRI) findings across this cohort. Of the four individuals who have had brain MRIs, two were reported as normal (individuals 3 and 4), two showed dissymmetry of the corpus callosum, dysmorphic hippocampi, and mildly dysmorphic lateral ventricles (individuals 1 and 2; Figure S4), one showed possible mild cerebellar atrophy (individual 1; Figure S4A), and one showed delayed myelination of cerebral white matter and mildly dysmorphic lateral ventricles (individual 5). No anomalies of the white matter or cerebral or cerebellar atrophy were seen on the brain MRI of individual 2 in adulthood. Additional clinical information for all five individuals can be found in Table S1.

Trio exome sequencing for the five individuals identified a de novo heterozygous variant in MTSS2 (GenBank: NM_138383.2: c.2011C>T [p.Arg671Trp]). To gather information about the gene in human and model organisms, we queried the Model organism Aggregated Resources or Rare Variant ExpLoration (MARRVEL).¹³ The loss-of-function (LoF) observed/expected (o/e) score for MTSS2 is 0.15, and the probability of being LoF intolerant (pLI) score is 0.98, suggesting that MTSS2 is intolerant to LoF alleles.¹⁴ The c.2011C>T variant is absent from gnomAD,¹⁴ and is located at an evolutionarily conserved residue (Figures 1C and S1A). Multiple in silico prediction programs (see supplemental methods) predict that this missense change has a deleterious effect on MTSS2, including a CADD score¹⁵ of 25. No other variants in known or novel genes have been retained as plausible candidates by exome analysis for the five individuals. In summary, we suspected that the de novo c.2011C>T variant explains the individuals' phenotypes given the deleterious in silico predictions, absence of this variant in population databases, the involvement of MTSS2 in neuron physiology, and the identification of an overlapping phenotype in five unrelated individuals with the same de novo missense variant. The mechanism by which the c.2011C>T variant affects the function of MTSS2 could be haploinsufficiency, gain-of-function (hypermorph or neomorph) or a dominant-negative action (antimorph), although the recurrent nature of the variant increases the likelihood of the latter two mechanisms.¹⁶

We evaluated the impact of c.2011C>T by assessing mRNA and protein levels from individual 1- and 2-derived fibroblasts compared to age- and sex-matched control subjects. Real-time PCR analysis showed a reduction of MTSS2 transcript level (p = 0.0054 in individual 1 and p = 0.0653 in individual 2; Figure 1D). However, western blot analysis showed variable levels of MTSS2, and the affected individuals’ levels appeared within the normal range (Figure 1E). These data suggest that the variant leads to a decrease in mRNA level, but this may not affect the level of MTSS2 in fibroblasts.

To investigate the function of the MTSS2 variant, we modeled the variant in Drosophila melanogaster. Drosophila mim (missing-in-metastasis) is the ortholog of human MTSS2 and MTSS1 with DIOPT scores¹⁷ of 6/16 and 7/16, respectively. Although the fly Mim contains unaligned sequence stretches and is larger than the human MTSS2 (Figure 1A) as well as MTSS1, Mim and MTSS2 share 46% similarity and 31% identity of the protein sequences. The leucine zipper motif is disrupted by the unaligned stretches, but the overall protein structures are similar, and the major domains of MTSS2 and MTSS1 are present in the fly protein (Figures 1A and S1B).¹⁸ Furthermore, the residue affected by the c.2011C>T (p.Arg671Trp) variant in MTSS2 is conserved in fly Mim (Figure 1C).

First we generated a mim^T2A-GAL4 allele by inserting a CRISPR-Mediated Integration Cassette (CRIMIC) in a shared intron of all mim transcripts.¹⁹ The splice acceptor (SA) allows the T2A-GAL4 to be incorporated into the mRNA and the poly(A) tail leads to transcription termination and truncates the mim mRNA (Figure 2A). In addition, the viral T2A sequence arrests translation but allows the production of GAL4,²⁰ which is under the control of endogenous regulatory elements of mim.^21,22 Therefore, the mim^T2A-GAL4 allele is able to drive expression of any UAS-cDNA in the same pattern as mim.²³ Our real-time PCR data indicate that the transcript levels of two exons which are adjacent to the interrupted intron are not detected in mim^T2A-GAL4/mim^T2A-GAL4 larvae (Figure 2B), indicating that the T2A-GAL4 truncates the mim transcript and is therefore likely a severe LoF allele.

Figure 2.

mim^T2A-GAL4 is expressed in many neurons and some glia

(A) Structure of fly mim and T2A-GAL4 allele. cheb42b and cheb42c are nested genes in the mim locus. The CRIMIC T2A-GAL4 sequence is inserted into a shared intron of all mim transcripts, truncating the transcript and protein while expressing T2A-GAL4.20, 23^, P, attP; F, FRT; SA, splice acceptor.

(B) mim mRNA expression based on real-time PCR of exon L (left) and R (right) that adjacent to the inserted intron is not detected in homozygous mim^T2A-GAL4 mutant larvae when compared to wild-types (w1118). Exon L and R are shared exons of all transcripts that are indicated in (A). mRNA levels were normalized to that of housekeeping gene rpl32. Error bar: SEM. ^∗∗∗p < 0.001 by unpaired t tests.

(C) Whole-mount, projection image of larval central nervous system (CNS) from mim^T2A-GAL4/+; UAS-mCD8-RFP/+ (cell membrane) showing mim is highly expressed in mushroom body (MB) in the central brain (CB), optic lobe (OL), and ventral nerve cord (VNC). Schematic of larval CNS shows the different structures.

(D) Projection image of larval CNS co-stained with neuronal marker anti-Elav from mim^T2A-GAL4/+; UAS-NLS-mCherry/+ (cell nuclei).

(E–H) Single-focal images of mushroom body (MB) (E and G) and ventral nerve cord (VNC) (F and H) co-stained with anti-Elav (E and F) and anti-Repo (G and H).

(I) Projection image of adult central brain (CB) from mim^T2A-GAL4/+; UAS-mCD8-RFP/+ showing mim is highly expressed in mushroom body (MB) and antennal lobe (AL). Below: schematic of adult brain.

(J) Projection image of half of an adult brain from mim^T2A-GAL4/+; UAS-NLS-mCherry/+ co-stained with anti-Elav.

(K–N) Single-slice confocal images of adult mushroom body (MB) and optic lobe (OL) co-stained with anti-Elav (K and L) and anti-Repo (M and N).

Given the neurological deficits in all the identified individuals and given that MTSS2 is highly expressed in the mammalian CNS (GTEx),^2,7 we explored the expression of mim in the fly CNS. We used the mim^T2A-GAL4 to drive expression of UAS-mCD8-RFP to label the membranes of the cells that express mim and found a widespread expression of RFP in the third instar larval and adult brain (Figures 2C and 2I). This included the mushroom body (insect neurons that play a critical role in learning and memory), the optic lobe, and the ventral nerve cord (the spinal cord equivalent in insects). We used the UAS-NLS-mCherry to label the nuclei of the cells that expressed mim (Figures 2D and 2J). By comparing its expression to the pan-neuronal nuclear marker Elav and the nuclear glial marker Repo, the nature of the cells expressing mim could be easily identified. In the larval CNS, mCherry (mim) was expressed primarily in neurons (Figures 2E and 2F) and in some glia in the ventral nerve cord (Figures 2G and 2H). In the adult brain, mCherry (mim) was expressed in most neurons as well as many glia of the central brain and optic lobe (Figures 2K–2N). This was consistent with single-cell RNA-seq data from the Fly Cell Atlas (Figure S2).^24,25 Of note, mim was highly expressed in neurons of the mushroom body (Figures 2C–2E and 2I). In summary, mim was widely expressed in neurons as well as in some glia of developing larval CNS and adult brain, and expression was particularly enriched in the neurons that mediate learning and memory.

The mim^T2A-GAL4 efficiently truncates the mim transcript (Figure 2B), and flies that carry mim^T2A-GAL4 over a chromosomal deficiency allele Df(2R)Exel6051 that lacks mim (∼120 kb; Df for short) all survived to adulthood, suggesting that the gene is not essential for development and viability. However, flies that are homozygous mim^T2A-GAL4/mim^T2A-GAL4 showed a low eclosion rate, as only 10% of the expected number of flies develop into adults at 25°C (Figure 3A). This suggests that the mim^T2A-GAL4 contains an off-target second mutation(s) that reduces the viability of homozygous mim mutants. To assess the function of human MTSS2, we generated transgenic flies that carry UAS-MTSS2 cDNAs. The eclosion rate of mim^T2A-GAL4/mim^T2A-GAL4 mutants was partially but significantly rescued by expression of the human UAS-MTSS2 reference cDNA (from 10% to 50%). This partial rescue suggests that the absence of mim enhances lethality of the off-target mutation(s) and hence reduces the eclosion rate to 10% when homozygous. However, this rescue was not observed by expression of the c.2011C>T variant (Figures 3A and S3A), suggesting that it is a LoF variant.

Figure 3.

Expression of human MTSS2 reference rescues the defects in mim LoF flies, whereas c.2011C>T variant showed decreased rescue ability

(A) Eclosion rates of adult flies of the indicated genotypes. Numbers of analyzed flies are in Figure S3A. NS, p > 0.05; ^∗∗∗p < 0.001 based on chi-squared tests between each genotype.

(B) Lifespan of adult flies with indicated genotypes. n > 60 flies for each genotype; ^∗∗p < 0.01, ^∗∗∗p < 0.001 by chi-squared test for trend between each genotype.

(C) Locomotor activity of flies at 2–7 days post eclosion (dpe) of indicated genotypes measured by DAM assay, n = 32 flies for each genotype (top). Recovery time (s) after bang-sensitivity induced by 15 s vortex of flies at 8 dpe, n > 50 flies for each genotype (bottom). Error bar: SEM. ^∗p < 0.05, ^∗∗p < 0.01, ^∗∗∗p < 0.001 based on one-way ANOVA with Tukey’s multiple comparison test between each indicated genotype.

(D) Electroretinograms (ERGs) of flies at 6 dpe. On (indicated as magenta), Off (indicated as green) transients and amplitudes were quantified. Error bar: SEM. NS, p > 0.05; ^∗p < 0.05, ^∗∗p < 0.01, ^∗∗∗p < 0.001 by one-way ANOVA with Tukey’s multiple comparison test between each indicated genotype.

Although mim^T2A-GAL4/Df mutants showed no obvious defects in eclosion rate, their lifespan was significantly shorter than control flies that are heterozygous mim^T2A-GAL4/+ and carry an empty control UAS (robo1^T2A-GAL4/+; UAS-empty/+) at 29°C. Expression of UAS-MTSS2 reference in the mim^T2A-GAL4/Df mutants fully rescued the lifespan, whereas the c.2011C>T variant only partially rescued the reduced lifespan (Figure 3B).

The expression level of human MTSS2 is highest in the cerebellum and spinal cord (GTEx).⁷ Fly mim is also highly expressed in the ventral nerve cord (Figure 2C), which corresponds to the vertebrate spinal cord. To determine whether mim plays a role in locomotor behavior of flies, we first performed Drosophila Activity Monitoring (DAM) assay²⁶ of adult flies at 25°C. mim^T2A-GAL4/Df mutants showed a significant decrease in locomotor activity when compared to control flies (mim^T2A-GAL4/+; UAS-empty/+). This decrease in activity was fully rescued by expression of UAS-MTSS2 reference allele but not the c.2011C>T variant (Figures 3C and S3B). Next, we did bang-sensitivity testing, which is an assay to assess neuronal dysfunction. Upon vortexing the flies for 10–15 s, wild-type flies recover in less than a few seconds and do not exhibit seizures. However, some flies with genotypes that are susceptible to seizures become paralyzed, uncoordinated, or shake, exhibiting a seizure-like phenotype, and these flies can take a significantly longer time to recover than wild-type flies.27, 28, 29 The mim^T2A-GAL4/Df mutants showed bang-sensitivity and recovered slowly to an upright position after vortexing for 15 s. Expression of UAS-MTSS2 reference partially rescued the bang-sensitivity phenotype and significantly shortened the recovery time; however, c.2011C>T expression had decreased rescue ability (Figure 3C).

Ophthalmological defects appeared to be a common clinical finding in the cohort of affected individuals with the c.2011C>T variant, with one individual having significant progressive optic atrophy. The fly mim is highly expressed in the optic lobe (Figures 2C and 2I) as well as in adult photoreceptors and lamina neurons in the eyes (Figure S2). To examine whether mim is required for visual function, we performed electroretinograms (ERGs) in adult flies. The amplitudes of the ERG traces represent the ability of photoreceptors to sense photons upon light exposure, while the On/Off transients provide a measure of synaptic transmission between photoreceptors and the postsynaptic neurons in the lamina.³⁰ The mim^T2A-GAL4/Df mutants had significantly reduced On/Off transients, but the amplitude at 29°C was not affected (Figure 3D). This suggests that loss of mim does not affect phototransduction but that defects at the synapses between photoreceptors and postsynaptic neurons may be at play. Expression of UAS-MTSS2 reference in mim^T2A-GAL4/Df mutants fully rescued the On-transient decrease and partially rescued the Off-transient decrease, while the c.2011C>T variant showed very limited rescue for both transients (Figure 3D). In summary, mim LoF impairs lifespan, reduces locomotor activity, affects the bang sensitivity response, and impairs synaptic transmission in the visual system of adult flies. These defects were partially or fully rescued by expression of the human reference MTSS2, implicating functional conservation of human MTSS2 with the fly Mim. The c.2011C>T variant has significantly decreased rescue ability in all assays tested when compared to the reference allele, suggesting that it is a partial LoF allele.

The previous assays were focused on rescuing the severe LoF alleles with a reference or mutant cDNA copy. However, given that the variant of interest is a de novo recurrent dominant change, we investigated whether overexpression of the reference and the variant cDNAs induced different or similar phenotypes. This was done by driving the reference and variant MTSS2 cDNAs in mim^T2A-GAL4 heterozygous flies. Expression of UAS-MTSS2 reference in mim^T2A-GAL4/+ flies did not significantly affect lifespan or locomotor activity when compared to the mim^T2A-GAL4/+; UAS-empty/+ controls. However, expression of the c.2011C>T variant caused defects in both assays (Figures 4A, 4B, and S3B), suggesting that expression of the variant is toxic. To further assess the toxicity associated with the c.2011C>T variant, we ectopically expressed it with tissue-specific GAL4s. Pan-neuronal (Elav-GAL4) expression of UAS-MTSS2 c.2011C>T, but not the reference, led to mild but significant climbing defects and bang-sensitivity (Figure 4C). Moreover, expression of UAS-MTSS2 c.2011C>T, but not the reference, in the eye using the GMR-GAL4 caused a decrease in On/Off transients but again did not significantly affect the phototransduction pathway (Figure 4D). These data strongly suggest that the c.2011C>T acts as a dominant-negative or antimorphic allele. Interestingly, the ERG phenotype in flies at 12 days post eclosion (dpe) is slightly more severe than that at 6 dpe (Figure 4D), suggesting that the phenotypes associated with the c.2011C>T variant may become progressively worse with time.

Figure 4.

MTSS2 c.2011C>T variant is toxic when expressed in flies

(A) Lifespan of adult flies with indicated genotypes. n > 100 flies for each genotype; ^∗∗p < 0.01, ^∗∗∗p < 0.001 by chi-squared test for trend between each genotype.

(B) Locomotor activity of flies with indicated genotypes. Error bar: SEM. n = 32 flies for each genotype; ^∗p < 0.05 by one-way ANOVA with Tukey’s multiple comparison test between each indicated genotype.

(C) Climbing and bang-sensitivity assays of flies at 5 dpe, UAS-cDNAs were driven by Elav-GAL4. Error bar: SEM. n > 70 flies for each genotype; ^∗p < 0.05, ^∗∗∗p < 0.001 by one-way ANOVA with Tukey’s multiple comparison test between each indicated genotype.

(D) ERGs of flies expressing UAS-cDNAs under the control of GMR-GAL4. On and Off transients and amplitudes were quantified. Error bar: SEM. NS, p > 0.05; ^∗p < 0.05, ^∗∗p < 0.01 ^∗∗∗p < 0.001 by one-way ANOVA with Tukey’s multiple comparison test between each indicated genotype.

In summary, exome sequencing combined with one-sided and two-sided matchmaking strategies resulted in the identification of five unrelated individuals with the same de novo c.2011C>T variant in MTSS2 and an overlapping phenotype consisting of global developmental delay, mild intellectual disability, ophthalmological anomalies, microcephaly or relative microcephaly, and shared facial features. Two of the five individuals also have hearing loss. An age-related penetrance could explain why some of the clinical features that were of teen or adult onset in the oldest individual in this cohort—severe optic atrophy and seizures—are not present in younger individuals from this cohort (Table 1). The recurrent nature of the specific heterozygous c.2011C>T variant raised the possibility of a gain-of-function or dominant-negative mechanism. The fact that the MTSS2 level in fibroblasts from individuals 1 and 2 was not significantly altered compared to control subjects (Figure 1E) suggests that the variant protein was present in the affected individuals. Our fly studies revealed that mim is expressed in the ventral nerve cord, optic lobe, and eyes (Figures 2C and S2), and that its loss underlies defects in locomotor and visual functions (Figure 3). Importantly, the defects in mim LoF mutants were rescued by human MTSS2, implicating functional conservation between the two orthologs. The MTSS2 c.2011C>T variant behaved as a partial LoF allele in a mim LoF background. Overexpression of the c.2011C>T variant caused similar phenotypes as the LoF, including shortened lifespan, decreased locomotor activity, bang-sensitivity, and abnormal communication between pre and postsynaptic cells (Figures 3 and 4). These data indicate that the MTSS2 c.2011C>T variant may interfere with the normal function of fly Mim by means of a dominant-negative effect.

The later onset of some neurological findings—optic atrophy and seizures—in the adult individual in this cohort suggests that this condition may have a slowly progressive clinical course, in line with our finding in flies that expression of the c.2011C>T variant leads to a progressively worsening ERG phenotype with age (Figure 4D). The expression of MTSS2 mRNA in the mouse hippocampus is highly dynamic and activity dependent,⁸ suggesting that neuronal activity may lead to the production of the reference protein as well as the toxic protein. The expression of MTSS2 and MTSS1 in the adult retina is not high (GTex),⁷ suggesting that the retina does not require as high levels of MTSS proteins as the CNS, and one copy of the reference protein may be sufficient for its function. However, the retina could be sensitive to the presence of the toxic MTSS2 allele; we hypothesize that the toxic protein is induced by synaptic activity in the optic nerves, becoming detrimental with age and causing progressive optic atrophy.

In conclusion, our findings demonstrate that the c.2011C>T variant in MTSS2 causes an autosomal-dominant intellectual disability syndrome through a suspected dominant-negative mechanism. The identification and detailed phenotyping of additional affected individuals across their lifespan will be required to better define the natural history of this condition. If this neurodevelopmental disorder is confirmed to have a progressive nature, this creates an opportunity for potential therapeutic intervention to prevent the neurological deficits with a later age of onset.

Published: September 5, 2022; corrected online October 21, 2022

Web resources

CADD, https://cadd.gs.washington.edu/

DIOPT, http://www.flyrnai.org/diopt

Fly Cell Atlas: Adult brain, https://scope.aertslab.org/#/FlyCellAtlas/FlyCellAtlas%2Fs_fca_biohub_head_10x.loom/gene

GeneMatcher, https://genematcher.org/

gnomAD, https://gnomad.broadinstitute.org/

GTEx, https://gtexportal.org/

L3 brain, http://scope.aertslab.org/#/Larval_Brain/∗/welcome

MARRVEL, http://marrvel.org/

Matchmaker Exchange, https://www.matchmakerexchange.org/

Mutation Taster, http://www.mutationtaster.org/

OMIM, https://omim.org

PhenomeCentral, https://www.phenomecentral.org/

PolyPhen2, http://genetics.bwh.harvard.edu/pph2/

ShinyR-DAM, https://karolcichewicz.shinyapps.io/shinyr-dam

SIFT, https://sift.bii.a-star.edu.sg/

Data and code availability

The variant in MTSS2 was submitted to ClinVar (https://www.ncbi.nlm.nih.gov/clinvar/) (GenBank: NM_138383.2; accession numbers SCV001432151.1). The exome datasets supporting this study have not been deposited in a public repository because of ethical restriction.