A novel RLIM/RNF12 variant disrupts protein stability and function to cause severe Tonne–Kalscheuer syndrome

Tonne–Kalscheuer syndrome (TOKAS) is an X-linked intellectual disability syndrome associated with variable clinical features including craniofacial abnormalities, hypogenitalism and diaphragmatic hernia. TOKAS is caused exclusively by variants in the gene encoding the E3 ubiquitin ligase gene RLIM, also known as RNF12. Here we report identification of a novel RLIM missense variant, c.1262A>G p.(Tyr421Cys) adjacent to the regulatory basic region, which causes a severe form of TOKAS resulting in perinatal lethality by diaphragmatic hernia. Inheritance and X-chromosome inactivation patterns implicate RLIM p.(Tyr421Cys) as the likely pathogenic variant in the affected individual and within the kindred. We show that the RLIM p.(Tyr421Cys) variant disrupts both expression and function of the protein in an embryonic stem cell model. RLIM p.(Tyr421Cys) is correctly localised to the nucleus, but is readily degraded by the proteasome. The RLIM p.(Tyr421Cys) variant also displays significantly impaired E3 ubiquitin ligase activity, which interferes with RLIM function in Xist long-non-coding RNA induction that initiates imprinted X-chromosome inactivation. Our data uncover a highly disruptive missense variant in RLIM that causes a severe form of TOKAS, thereby expanding our understanding of the molecular and phenotypic spectrum of disease severity.

www.nature.com/scientificreports/ Thus far, 8 variants that cause TOKAS in distinct unrelated families have been reported [1][2][3] . RLIM variants have been reported in 4 cases out of a cohort of 405 cases of unresolved syndromic X-linked intellectual disability with no known genetic or environmental basis 2 . This data suggests that a significant proportion of unresolved X-linked intellectual disability cases may be caused by RLIM variants, and that other RLIM TOKAS variants await identification.
Here, we report identification of a novel RLIM missense variant, p.(Tyr421Cys), adjacent to the regulatory basic region, which causes a severe form of TOKAS leading to perinatal lethality by diaphragmatic hernia. Inheritance and X-chromosome inactivation patterns clearly implicate RLIM p.(Tyr421Cys) as the causative variant in the affected individual and kindred. We show that the RLIM p.(Tyr421Cys) variant severely disrupts protein expression and function, and is readily degraded by the proteasome. RLIM p.(Tyr421Cys) also displays significantly impaired E3 ubiquitin ligase activity. Together, defects in RLIM p.(Tyr421Cys) protein expression and activity profoundly interfere with RLIM function in Xist long-non-coding RNA induction, a key step in initiating imprinted X-chromosome inactivation. Our data uncover a highly disruptive missense variant in RLIM that causes a severe form TOKAS, thereby expanding our understanding of the molecular and phenotypic spectrum of disease severity.

Results
Clinical information for a patient with an undiagnosed developmental disorder. The male proband was the first child of unrelated parents. The first trimester nuchal translucency/morphology scan and second trimester fetal morphology scan were reported as normal. Polyhydramnios was detected at 36 weeks' gestation and intrauterine growth restriction (IUGR) was reported at 37 weeks. Labour was induced at 39 weeks and delivery was via emergency Caesarean section for subsequent fetal distress. The baby became cyanosed immediately after birth, was unable to be resuscitated and died at 30 min of age.
On examination the baby was symmetrically small for gestational age with a birth weight of 2285 g (< 3rd percentile; Z score − 2.57), length 48.5 cm (~ 5th percentile; Z score − 1.67) and head circumference 32.5 cm (< 3rd percentile; Z score − 2.16). Craniofacial anomalies were identified, including hypertelorism, broad nasal bridge with flat nasal tip, and a high arched palate. He had distal limb hypoplasia, with brachytelephalangy, soft tissue syndactyly between the 2nd and 3rd toes bilaterally and absent or hypoplastic nails on 2nd-3rd fingers and 2nd-4th toes. There were no ocular abnormalities.
Postmortem examination revealed congenital diaphragmatic hernia with aplasia of the posterolateral left hemi-diaphragm, displacement of the mediastinum to the right, herniation of the small and large intestine, stomach, spleen, left lobe of liver, and pancreas into the left hemithorax, absent middle lobe right lung and severe bilateral pulmonary hypoplasia. There was an accessory spleen. The cardiovascular system showed aberrant aortic arch branching with the left subclavian, left common carotid and right common carotid arteries arising from a single brachiocephalic trunk. Urogenital abnormalities included cryptorchidism with pelvic testes, but kidneys were normal. A cavum septum pellucidum was present. The clinical features were considered to be most consistent with Fryns syndrome (MIM #229850). Pallister-Killian syndrome (tetrasomy 12p; MIM #601803) was considered, but fluorescence in-situ hybridisation for chromosome 12p on fetal lung imprints, standard karyotype on cultured fibroblasts and chromosome microarray on DNA extracted from stored fetal tissue (Agilent Sureprint G3 ISCA Targeted Microarray 8 × 60 K) were normal.

Identification of a novel missense RLIM variant and diagnosis of Tonne-Kalscheuer syndrome (TOKAS).
In order to investigate the genetic basis of this case, we performed whole exome sequencing on genomic DNA from the proband and his parents, and identified a maternally inherited hemizygous variant in the X-linked RLIM gene (NM_016120.3) c.1262A>G, p.(Tyr421Cys) (Fig. 1a). RLIM encodes the RLIM/RNF12 E3 ubiquitin ligase that is mutated in the recessive X-linked disorder Tonne-Kalscheuer syndrome (TOKAS) 1-3 , which is characterised by clinical features that significantly overlap with Fryns syndrome. There were no variants in PIGN, in which biallelic loss of function mutations have been reported to cause a Fryns-like syndrome. We hypothesised that the individual was affected by a severe form of TOKAS that is caused by this novel RLIM variant. Consistent with this notion, the p.(Tyr421Cys) variant is located adjacent to the basic region of the RLIM protein, a key regulatory domain that is mutated in several TOKAS kindreds 1-3 . RLIM p.(Tyr421Cys) has not been observed in affected individuals in the literature and was not reported in the ClinVar archive of genomic variation in human health (https:// www. ncbi. nlm. nih. gov/ clinv ar) or the gnomAD population database (https:// gnomad. broad insti tute. org/). As a result, the variant was submitted to ClinVar under the accession number SCV001435291.
Sanger sequencing confirmed the presence of the RLIM c.1262A>G p.(Tyr421Cys) variant in the proband, whilst the mother and maternal grandmother were confirmed as carriers (Fig. 1b). The variant was absent in the mother's unaffected brother (Fig. 1b), consistent with an X-linked inheritance pattern. Furthermore, X-chromosome inactivation analysis showed the mother had a highly skewed X-chromosome inactivation pattern (95%; Supplementary Figure 1 www.nature.com/scientificreports/ Therefore, we exploited RLIM-deficient male mouse Embryonic Stem Cells (mESCs) to investigate expression and function of RLIM p.(Tyr421Cys) in a cellular model. In comparison to wild-type human RLIM, the p.(Tyr421Cys) variant is poorly expressed (Fig. 2a). Quantitative analysis indicates that RLIM p.(Tyr421Cys) expression is reduced to 28.7 ± 11.5% of wild-type (Fig. 2b). However, the mRNA is expressed at similar levels, (Fig. 2c) and the protein is correctly localised to the nucleus (Fig. 2d), suggesting that the RLIM p.(Tyr421Cys) variant specifically interferes with translation and/or stability. We explored whether RLIM p.(Tyr421Cys) is more readily turned over by the proteasome. Indeed, treatment of RLIM p.(Tyr421Cys) expressing mESCs with the proteasome inhibitor MG132 expression rescues RLIM p.(Tyr421Cys) expression to levels approaching that of wild-type RLIM (Fig. 2e), indicating that RLIM p.(Tyr421Cys) more readily undergoes proteasomal degradation. Taken together, these data indicate that the RLIM p.(Tyr421Cys) variant destabilises the protein.

RLIM p.(Tyr421Cys) variant interferes with E3 ubiquitin ligase activity. As other RLIM TOKAS
variants located within the proximal basic region (Fig. 1a) have been shown to disrupt E3 ubiquitin ligase activity 1,6 , we next sought to determine the impact of p.(Tyr421Cys) on RLIM catalytic activity. We expressed and purified RLIM wild-type and p.(Tyr421Cys) from E. coli, and examined the ability of these variants to transfer ubiquitin from a cognate E2 (UBE2D1) to the prototypic substrate REX1, which is a transcription factor that is ubiquitylated by RLIM to initiate X-chromosome inactivation 11 . Wild-type RLIM efficiently promotes REX1 ubiquitylation, as evidenced by appearance of ubiquitylated species of increasing molecular weight upon addition of substrate (Fig. 3a). However, RLIM p.(Tyr421Cys) significantly impairs REX1 ubiquitylation activity (Fig. 3a), suggesting that the p.(Tyr421Cys) variant also has a negative impact on RLIM catalytic activity. Quantitative analysis indicates that RLIM p.(Tyr421Cys) activity is reduced to 32 ±3.7% of wild-type (Fig. 3b). Therefore, similar to other previously reported RLIM TOKAS variants, p.(Tyr421Cys) displays impaired E3 ubiquitin ligase activity.   13 . Thus, we used Xist expression as a readout for RLIM function. As expected, Xist expression in RLIM-expressing Rlim +/y mESCs is very low (Fig. 4a). However, Xist expression is induced by expression of wild-type human RLIM, but not by a catalytically inactive variant (W600Y; Fig. 4a). When compared to wild-type RLIM, Xist induction is severely impaired upon expression of RLIM p.(Tyr421Cys) (Fig. 4a), demonstrating that the p.(Tyr421Cys) variant profoundly interferes with RLIM function. In order to control for expression, we analysed RLIM levels in Rlim -/y mESCs, which do not express endogenous RLIM (Fig. 4b). We also show using Rlim -/y mESCs that loss of RLIM expression and function does not significantly alter cell number accumulation under standard mESC culture conditions (Supplementary Figure 2)

Discussion
Tonne-Kalscheuer syndrome (TOKAS) is a developmental disorder characterised by clinical features including intellectual disability, facial dysmorphism, velopharyngeal abnormalities and diaphragmatic hernia 1-3 . In the most severe cases, diaphragmatic hernia causes death shortly after birth 1 . TOKAS is caused by variants in the X-linked RLIM/RNF12 E3 ubiquitin ligase, which impair catalytic activity to varying extents 1,6 . However, the extent to which phenotypic/disease severity correlates with genotypic severity (i.e. the extent of RLIM functional disruption) remains unclear.
Here, we provide a case report of a male patient who died shortly after birth with diaphragmatic hernia, facial dysmorphism and skeletal abnormalities. Cytogenetic studies showed no evidence of Pallister-Killian syndrome and an initial diagnosis of Fryns syndrome was thought most likely, although TOKAS was not considered as this baby was born several years before the association of diaphragmatic hernia with severe TOKAS was delineated 1 . Research exome sequencing through the Care4Rare Canada Consortium uncovered a missense RLIM variant, p.(Tyr421Cys), which was heterozygous in the mother but not present in unaffected relatives, suggesting that the patient suffered a severe form of TOKAS. These findings raise the important question of whether other individuals diagnosed with Fryns syndrome and/or syndromic diaphragmatic hernia might actually represent incorrectly diagnosed TOKAS patients. www.nature.com/scientificreports/ In this study, we explore the impact of the p.(Tyr421Cys) variant on RLIM protein expression and function. Strikingly, we find that RLIM p.(Tyr421Cys) is poorly expressed in an RLIM-deficient mouse embryonic stem cell model, and is prone to proteasomal degradation, although the protein is correctly localised in the nucleus. This is in contrast to other RLIM TOKAS variants, which show impaired catalytic activity but no major impact on stability 1,6 . Interestingly, analysis of recombinant RLIM p.(Tyr421Cys) indicates that this variant also displays impaired E3 ubiquitin ligase catalytic activity. Although the RLIM p.(Tyr421Cys) variant lies outwith the core catalytic RING domain, this variant may disrupt a poorly understood catalytic function of the basic region 1,6 .
Our data also support the notion that disrupted RLIM p.(Tyr421Cys) protein expression and catalytic activity contribute to impairment of RLIM function. In this regard, the RLIM p.(Tyr421Cys) variant significantly disrupts induction of the Xist lncRNA, which is a key initiating step of X-chromosome inactivation (XCI). However, XCI is unlikely to be a major feature of TOKAS pathogenesis, as this syndrome affects only male patients. Nevertheless, heterozygous female carriers of RLIM TOKAS variants display highly skewed XCI accompanied by relatively mild developmental anomalies 1 , consistent with the presence of a deleterious X-linked variant. Indeed, we confirm that the mother, who is a carrier of the RLIM p.(Tyr421Cys) variant, displays a highly skewed XCI pattern.
A major question arising from this study concerns the developmental functions of RLIM that are disrupted to cause the major clinical features of TOKAS. Work from our group has previously identified a function for RLIM signalling in controlling expression of neuronal genes 14 . We hypothesise that RLIM TOKAS variants disrupt correct neuronal development leading to intellectual disability. However, the relevant gene expression programmes that are disrupted to cause the syndromic features of TOKAS, such as diaphragmatic hernia, are yet to be elucidated. In future, the mechanisms underpinning these syndromic anomalies will be uncovered by transcriptomic analyses of tissue-specific cellular and animal TOKAS models.
In summary, we introduce RLIM p.(Tyr421Cys) as the prototypic member of a new class of RLIM TOKAS variant that profoundly impacts on both protein stability and catalytic activity, which leads to severe TOKAS. This in turn expands our understanding of the molecular and phenotypic spectrum of TOKAS syndrome severity.

Methods
Genomic DNA sequencing and analysis. Trio exome sequencing of genomic DNA from proband and parents was performed through a collaboration with Care4Rare program (http:// care4 rare. ca). Target capture was performed with the Agilent CRE V1.0 and sequencing performed on the Illumina NextSeq 500 using 150 base-pair paired-end reads. Data analysis was done by standard methods and variants annotated using both Annovar and custom scripts to identify whether they affect protein coding sequence, and whether previously seen in dbSNP132, the 100 Genomes dataset (Nov 2011), the NHLBI GO exomes or in the approx. 1500 exomes previously sequenced at the center 15 . Variant reporting relevant to diagnosis was informed by reference to HGMD, dbSNP, online search engines e.g. PubMed and locus-specific databases. Variants seen in > 20 of their controls, or with an allele frequency of > 3% in 100 Genomes or NHBLI were removed. Confirmatory Sanger sequencing and familial testing was performed in the Sydney Genomic Diagnostic laboratory, Children's Hospital at Westmead.
Primary antibodies are anti-mouse RLIM amino acids 1-271 (S691D third bleed; see MRC-PPU Reagents & Services http:// mrcpp ureag ents. dundee. ac. uk for further information and requests), anti-ERK1/2 (Santa Cruz Biotechnology) and anti-REX1 (Abcam). Secondary antibodies are Sheep IgG-horseradish peroxidase (HRP), Mouse IgG-HRP (Cell Signaling Technology) and Rabbit IgG-HRP (Cell Signaling Technology). After secondary antibody incubation, membranes were subjected to chemiluminescence detection with Immobilon Western Chemiluminescent HRP substrate (Millipore) using a Gel-Doc XR + System (Bio-Rad) or to Infrared detection using a LI-COR Odyssey Clx system. Detected protein signals were quantified using Image J (NIH) or Image Studio (LI-COR Biosciences). All unprocessed immunoblots are provided in Supplementary Figure 3. Immunofluorescence. For localisation studies, mESCs were plated in 0.1% gelatin (v/v) coated coverslips and fixed in 4% PFA in PBS for 20 min at room temperature (RT). Cells were permeabilised with a 0.5% Triton X-100 in PBS solution for 5 min at RT  www.nature.com/scientificreports/ RT. RLIM primary antibody (Novus Biologicals) was diluted 1:200 in blocking solution and added to cells for 2 h at RT. Anti-mouse Alexa-488 (Thermo Fisher Scientific) was used as a secondary antibody at 1:500 in blocking solution for 1 h at RT. Actin Red 555 reagent (Thermo Fisher Scientific, one drop per ml of blocking solution) was added together with secondary antibody for actin staining. Hoechst was added at 1:10,000 dilution in PBS for 5 min at RT as nuclear marker. Coverslips were mounted in glass slides using Fluorsave reagent (Millipore). Digital images were acquired in a Zeiss 710 confocal microscope and analysed and processed using ImageJ (NIH), Photoshop CC and Illustrator CC (Adobe).
In vitro ubiquitylation assay. All  RNA extraction and quantitative RT-PCR. RNA was extracted using Omega total RNA extraction kit (column-based system) and obtained RNA converted to cDNA using iScript cDNA synthesis Kit (Bio-Rad). qPCR was performed using SsoFast EvaGreen Supermix (Bio-Rad) in a CFX384 real time PCR system (Bio-Rad). Relative mRNA levels were expressed using the ΔΔCT method and normalized to Gapdh expression. Data was analysed in Excel software and plotted using GraphPad Prism v7. mESC proliferation assay. The CellTiter Glo kit (Promega) was used to measure mESC cell number accumulation (colorimetric MTS assay). Rlim +/y and Rlim -/y mESCs were plated into flat bottom 96 well plate (1000 cells/well) in 100 µl medium. Cells were cultured for the indicated time, whereupon a titration of resuspended mESCs was used to calculate cell number. Each condition was measured in triplicate. CellTiter Glo was prepared as per manufacturer's instructions. 100 µl CellTiter Glo was pipetted into each well and mixed for 2 min on an orbital shaker at room temperature to induce cell lysis. Plates were incubated for 10 min at room temperature to stabilise the luminescent signal. The plate was read on a Perkin Elmer Envision at a Peak emission wavelength of 560 nm, and data analysed in Graphpad Prism.
Approval for human experiments. This patient was enrolled under the research study "Enhanced Care for Rare Genetic Diseases in Canada". Research ethics approval for the study was obtained from the Children's Hospital of Eastern Ontario Research Ethics Board (Ethics study number CTO 1577), which includes approval for sequencing and functional studies. All experiments were conducted in accordance within these ethical guidelines and regulations. Informed consent was obtained from all participants and/or their legal guardians.

Data availability
The datasets generated and/or analysed during the current study are available from the corresponding authors on request.