On the Origin of Pontocerebellar Hypoplasia: Finding Genes for a Rare Disease

UvA-DARE (Digital Academic Repository)

On the origin of pontocerebellar hypoplasia: Finding genes for a rare disease

Eggens, V.R.C.

Publication date 2016 Document Version Final published version

Link to publication

Citation for published version (APA): Eggens, V. R. C. (2016). On the origin of pontocerebellar hypoplasia: Finding genes for a rare disease.

General rights It is not permitted to download or to forward/distribute the text or part of it without the consent of the author(s) and/or copyright holder(s), other than for strictly personal, individual use, unless the work is under an open content license (like Creative Commons).

Disclaimer/Complaints regulations If you believe that digital publication of certain material infringes any of your rights or (privacy) interests, please let the Library know, stating your reasons. In case of a legitimate complaint, the Library will make the material inaccessible and/or remove it from the website. Please Ask the Library: https://uba.uva.nl/en/contact, or a letter to: Library of the University of Amsterdam, Secretariat, Singel 425, 1012 WP Amsterdam, The Netherlands. You will be contacted as soon as possible.

UvA-DARE is a service provided by the library of the University of Amsterdam (https://dare.uva.nl) Download date:06 Oct 2021 Al ben ik meer dan eens de weg kwijt En rijd ik hele einden om Ik ben op weg om je te vinden Dus wees gerust, ik kom

- Boudewijn de Groot TOE1 mutations cause pontocerebellar hypoplasia and disorders of sex development

Veerle RC Eggens, David Chitayat, Hülya Kayserili, Nicola Foulds, Tessa van Dijk, Bart Appelhof, Kazuhiro Muramatsu, Kimberly A Aldinger, William B Dobyns, David Manchester, Linda de Meirleir, Mary Louise Freckmann, Linda Warwick, Chistina Fagerberg, Maria Kibaek, Marie-Cecile Nassogne, Justin H Davis, Umut Altunoglu, Hirotomo Saitsu, Masaaki Shiina, Kazuhiro Ogata, Kanako Kurata, Peter G Barth, Naomichi Matsumoto, Frank Baas

Manuscript in preparation2 Abstract

Pontocerebellar hypoplasia type 7 (PCH7) – also described as pontocerebellar hypoplasia and disorders of sex development (PCH and DSD) - is a rare disease in which both brain and genital development are affected. Only four cases have been described in literature displaying this rare combination of symptoms. Until know it was uncer- tain whether the combination of symptoms in these patients occurred coincidentally or not. Here, we show that the presence of PCH and DSD is a single gene syndrome. We collected a cohort of ten patients (eight families) presenting PCH and DSD, and identified recessive mutations in the TOE1 gene in all of them. Brain MRI of the patients revealed a small pons and cerebellum and enlarged ventricles. All eight 46, XY patients present some degree of undervirilised genitals.

28 Introduction

Pontocerebellar hypoplasia (PCH) represents a group of autosomal recessive neu- rodegenerative disorders with prenatal onset. Patients show variable hypoplasia of pons and cerebellum and severe motor and mental impairments. Nowadays, ten PCH subtypes have been described (PCH1-10) with each subtype having its characteristic Chapter clinical and/or genetic features in addition to a hypoplastic pons and cerebellum. PCH7 is characterised by PCH plus disorders of sex development (DSD). DSD is a 2 broad group of conditions where defects of gonadal development occur, leading to ambiguous genitals or complete sex reversal. The frequency of 46, XY DSD is 1:20,000 T

[1]. In literature, four cases of PCH plus DSD have been described [2-5]. These patients OE 1 mutations in PCH7 had a small pons and cerebellum, enlarged ventricles and a 46,XY karyotype with feminizing genitalia. So far, it was unclear whether the combination of PCH and genital anomalies was fortuitous or a distinct syndrome. In this study we report eight families (ten patients) with PCH and DSD, and present the target of Egr1 (TOE1) gene as locus for this disease in all described families. TOE1 is a nuclear protein that can bind RNA, has deadenylation activity [6] and can have an inhibitory effect on viral replication [7]. Furthermore, it is essential for Cajal body maintenance, and thus potentially involved in mRNA splicing [8]. The gene has previously been associated with cerebellar abiotrophy in Arabian horses [9]. With the identification of TOE1 mutations in ten patients with PCH and DSD, we confirm the presence of a single gene syndrome causing both brain and genital abnormalities.

Materials and methods

Exome capture and sequencing Family 1 was analysed at the Academic Medical Center, Genomics core-facility in Am- sterdam and sequenced on a SOLiD4 platform according to manufacturer’s protocols. Fragment libraries were prepared followed by a Nimblegens EZexome v2.0 sequence capture. Reads were aligned against hg19 with BioScope1.3. Genomic DNA for patient 7 and both parents were sequenced by the University of Washington Center for Men- delian Genomics. DNA was captured using the Roche Nimblegen SeqCap EZ Human Exome Library v.2.0 library and sequenced with paired-end 50 bp reads on an Illumina HiSeq sequencer. Reads were aligned against hg19 using the Burrows-Wheeler Aligner v.0.6.2. For patient 8, genomic DNA of patient and parental samples was captured using the SureSelect Human All Exon v5 Kit (Agilent Technologies) and sequenced on an Illumina HiSeq2000 (Illumina) with 101 bp paired-end reads. Reads were aligned to GRCh37 with Novoalign (Novocraft Technologies). Single nucleotide variants (SNVs)

29 were called with the Genome Analysis Toolkit’s (GATK) UnifiedGenotyper. Rare (<1% in the NHLBI Exome Sequencing Project Exome Variant Server), deleterious variants were analysed under de novo and recessive mutation models. SNVs in TOE1 were confirmed by Sanger sequencing.

Sanger sequencing Variants in TOE1 in families 2,3,4,5 and 6 were identified by Sanger sequencing. Sanger sequencing of PCR amplified DNA was performed using BigDyeTerminator chemistry (Applied Biosystems) and analysed on an ABI3730xl sequencer. Sequences were analysed using CodonCode Aligner software 3.6.1. Primer sequences are outlined in Supplementary Table 1.

CLK2 knockout mouse CLK2 whole body knockout mice were bred from Clk2flox/flox mice and zp3-Cre trans- genic mice on a C57BL/6 background. Littermates carrying the floxed allele but not Cre were used as controls. Mice were bred at Puigserver Laboratory (Dept. of Can- cer Biology, Harvard Medical School, Boston, USA) who kindly provided fixed brain samples. Standard hematoxylin and eosin staining was performed on sagittal slides.

Morpholino injections in zebrafish Morpholino (MO) antisense oligonucleotides (Gene Tools) were designed to target zebrafish clk2 mRNA (NM_001076751.1) either upstream of the ATG startcodon (5’-CCGGTGCGTTTGTCCCACAGAAAAT-3’) or at the exon 5 splice donor site (5’-CATCAAT- GAACAGCTCTTACTTCTT-3). Standard control MOs were provided by Gene Tools. MOs were injected in 1-2 cell stage embryos using glass needles and a microinjector. Fish injected with splice clk2 MO were checked for alternative splicing: total RNA was extracted with TRIzol (ThermoFisher)/chloroform, cDNA was synthesized using dT-oligos and SuperScript III First Strand Synthesis System (Invitrogen) and PCR was performed with primers 5’-AGAGCCGGTCCATATCATCA-3’ and 5’-CCGAAATCCACAATCCTCAC-3’. mRNA transcription and injections in zebrafish Human CLK2 cDNA cloned in a pOTB7 vector was obtained from the IMAGE Consortium. The CLK2 insert was subcloned into a pCS2+ vector using primers with BamHI and XbaI restriction sites (5’- TCGTACAGGCTACCTGGATCCGCCACCATGCCGCATCCTCGAA-3’ and 5’- GTGTCACTGACTGCACCGTGTCTAGATGGGGGTCAAATGAAG-3’ respectively, restriction sites in bold). The construct was linearized with NotI. For zebrafish clk2 cDNA (NM_001044879.2), a vector with clk2 insert was synthesised at Life Technologies and cloned into a pcDNA3 vector using primers with HindIII (5’- TAAGCAAAGCTTATGC- CACACTCCAGGCGGTA -3’) and XhoI (5’-TGCTTACTCGAGTCACCGGCTGATGTCACGGTT-3’)

30 restriction sites. The construct was linearized with StuI. In vitro transcription was performed using the T7 mMESSAGE mMACHINE Kit (Ambion, Warrington, UK).

In situ hybridisation Plasmids with fgf8 insert were linearised with EcoRV and antisense digoxigenin

(DIG)-RNA labelled probes were synthesized using a SP6 polymerase RNA labelling Chapter kit (Roche). Zebrafish embryos were dechorionated, and fixed overnight in formalin. Whole mount in situ hybridisation procedures were adapted from [10]. Preceding 2

ISH, fish were bleached (10% H2O2; 5% formamide; 2,5% 20x SSC in water) for a few minutes until pigmentation had vanished. Fish were permeabilised using 10 µg/ml T proteinase K in PBS, and incubated over night with 1 ng/µl fgf8 DIG-RNA probe. Stain- OE 1 mutations in PCH7 ing was visualised using a NBT/BCIP solution (Roche).

TUNEL assay Zebrafish embryos were fixated, bleached and permeabilised as described above. TUNEL staining was performed using TDT buffer (Roche), TDT enzyme (Roche) and DIG-UTP (Roche). Fish were incubated over night with anti-DIG-AP antibody (Roche). Staining was visualised using a NBT/BCIP solution.

Recombinant CLK2 protein Murine flag-clk2 cDNA in a pcDNA3 vector was kindly provided by P. Puigserver (Dept. of Cancer Biology, Harvard Medical School, Boston, USA). Mutant flag-clk2 was gener- ated using Site Directed Mutagenesis (Stratagene) and primers 5’-ATGACAACAGAGA- GCATCTATCCATGATGGAAAGGATC-3’ and 5’-GATCCTTTCCATCATGGATAGATGCTCTCTGTT- GTCAT-3, the mutation site indicated in bold. Wild type and mutant flag-clk2 were transfected in HEK293 cells using Xtreme GENE HP DNA transfection reagent (Roche). Recombinant protein was purified by immunoprecipitaion with anti-FLAG M2 affinity beads (Sigma-Aldrich) and analysed by Western blotting (1:3000 anti-FLAG antibody (Stratagene)).

Kinase assay Western blot band intensities of FLAG-CLK2WT and FLAG-CLK2A390S were quantified with Aida Image Analyzer v.4.26 to equalize input levels. Recombinant protein was incubated with 5ug SK6 substrate (Signal Chem) and 33P-ATP (Perkin Elkmer) at 30oC for 15 minutes. The reaction mixture was spotted on P81 ion exchange papers (Mil- lipore) and free 33P ATP was removed by extensive washing with 100 mM phosphoric acid. Incorporated 33P on the ion exchange paper was measured. Recombinant human CLK2 (SignalChem) was used as positive control.

31 Haplotype analysis Exome data from patient 3 was analysed in Ingenuity Variant Analysis (http://www. ingenuity.com) and SNPs with a prevalence <0.5%, 10 mB upstream and downstream of the TOE1 gene were selected. Primers used for Sanger sequencing are listed in Supplementary Table 2.

CADD analysis The CADD score of all identified variants in TOE1 listed in the Exome Aggregation Consortium (ExAC, http://exac.broadinstitute.org) database was determined via http:// cadd.gs.washington.edu/. The chance of having two potentially pathogenic variants was calculated as the square of the total of frequencies of all variants with a CADD score >15.

TOE1 qPCR Total RNA was extracted from fibroblasts. cDNA was synthesized using dT-primers and SuperScript III First Strand Synthesis System (Invitrogen). qPCR analysis was performed on a LightCycler 480 using primers listed in Table S2 and matching probes (Roche).

Structural homology modelling A modelled structure of the DEDD domain of human TOE1 was constructed from the crystal structure of mouse poly(A)-specific ribonuclease (PARN) (PDB code; 3D45)[11]. For homology modelling using Phyre2[12], the DEDD domain of human TOE1 (amino acid residues 43-265) was used as a query amino acid sequence. Images of molecular structures were created using PyMOL.

Results

Candidate genes for PCH7 In search for a genetic basis for PCH7, we performed exome sequencing in a Turkish consanguineous family with two affected siblings (Table 2.1, family 1). The affected female presented a hypoplastic cerebellum on MR images and at the age of 20 years, she was spastic and did not have breast development or menarche. The brother - for whom no MRI was available - showed spasticity, hypertonicity and a micropenis, suggesting the same disorder as his sister. The siblings were classified as PCH7. Exome sequencing was performed on both siblings and their parents and analyzed under an autosomal recessive model. Non-synonymous homozygous variants not present in the dbSNP129 were selected. This approach revealed variants in three candidate genes:

32 c.307G>A (p.A103T) in target of Egr1 (TOE1; NM_025077.3), c.248C>G (p.T83S) in solute carrier family 39 member 1 (SLC39A1; NM_014437.3) and c.1168 G>T (p.A390S) in CDC- like kinase 2 (CLK2; NM_003993.2), all positioned on chromosome 1. The three genes were located in two large shared homozygous regions (2.7Mb around TOE1, 5.9Mb around CLK2 and SLC39A1). TOE1 was initially characterized as a growth suppressor protein acting via Egr1 [13]. The protein has deadenylating activity [6], can inhibit Chapter viral replication [7] and is possibly fulfilling a role in mRNA splicing by maintaining Cajal body integrity [8]. Interestingly, TOE1 mutations are found in Arabian horses with 2 cerebellar abiotrophy and Purkinje cell degeneration [9]. SLC39A1 is involved in zinc homeostasis [14]. CLK2 is a member of the LAMMER kinase family, which have been T shown to phosphorylate proteins of the spliceosomal complex [15]. OE 1 mutations in PCH7 The majority of genes that are currently known to cause PCH are involved in RNA metabolism [16]. Because both TOE1 and CLK2 are potentially involved in mRNA splicing, we decided to follow up these two candidate genes for PCH7.

CLK2 is an unlikely candidate gene for PCH7 To investigate whether CLK2 is essential for brain development, we knocked down clk2 in zebrafish using antisense morpholinos (MOs). Injection of a MO targeting the translation initiation site of clk2 resulted in embryos with altered expression of midhindbrain marker fgf8, and an increased number of TUNEL positive cells in the brain,

A B

uninjected

control MO

clk2a MO

Figure 2.1: CLK2 morpholino injections. Injections of an antisense morpholino targeted against the start codon of clk2 results in altered formation of the midhindbrain boundary as shown by fgf8 expression (A) and increased cell death in the brain as shown by TUNEL staining (B). Grey scalebar=200 μm

33 indicative for cell death (Fig. 2.1). However, we could not reproduce these results by injecting a MO targeted against the exon 5 splice donor site in clk2, while partial alternative splicing of clk2 mRNA was established (Fig. 2.2). In addition, the clk2 ATG MO induced phenotype could not be rescued by with either human CLK2 mRNA or ze- brafishclk2 mRNA (Fig. 2.3). This could be due to biological issues –mRNA is delivered in every cell at the same level, which may not reflect the demand for mRNA. However, the observed brain phenotype can also be an off-target effect of the MO. HIER FIGUUR 2.1 HIER FIGUUR 2.2 HIER FIGUUR 2.3 A B

Figure 2.2: CLK2 splice morpholino injections. Injection of antisense morpholinos targeting the exon 5 splice donor site induces partial alternative splicing (A) but did not elicit a brain phenotype in zebrafish embryos (B).

A B

Figure 2.3: mRNA injections in clk2 MO zebrafish.I njection of neither human clk2 mRNA (A) nor zebrafish clk2B mRNA (B) showed a rescue of the clk2 MO induced phenotype

HIER FIGUUR 2.4 Importantly, CLK2 KO mice do not show signs of brain hypoplasia at the age of 8 weeks, again suggesting that CLK2 is not essential in brain development (Fig. 2.4). At 3 months of age, the mice still behave normally and can produce offspring (personal communication, M. Hatting, Harvard University, USA).

34 A B

Chapter 2 C D T OE 1 mutations in PCH7

Figure 2.4: Brain morphology CLK2 mouse. Gross morphology (A and C) and composition of cell layers (B and D) of the cerebellum in CLK2 knockout mice (cre/cre) is similar to wild types (flox/flox). White scale bar = 500 μm, black scale bar = 100 μm.

In order to investigate whether the p.A390S variant we identified in the patients has an effect on CLK2 protein function, we performed a kinase assay. Recombinant FLAG- CLK2WT and FLAG-CLK2A390S protein was incubated with S6K substrate and 33P-ATP to measure kinase activity. This revealed that the kinase function of CLK2A390S was not abolished (Fig. 2.5). Moreover, sequencing a cohort of over a hundred PCH patients did HIER FIGUUR 2.5 not reveal any other potentially pathogenic variations in CLK2. From these data we conclude that CLK2 is an unlikely candidate gene for PCH7.

5000

4000

3000 CLK2-WT Æ 2000 CLK2-A390S

Dpm Dpm HEK cells 1000

0 1 10 100 1000 10000 -1000 CLK2 dilution Æ

Figure 2.5: CLK2 kinase assay. Recombinant murine flag-CLK2 was incubated with S6K substrate and 33P-ATP. CLK2A390S retains its kinase activity. Dpm=disintegrations per minute.

35 TOE1 variants in PCH patients Since the p.A390S variant in CLK2 did not seem to have an effect in vitro or in vivo, we proceeded to our second candidate gene: TOE1. Unaffected members of our index family were wild type or heterozygous for the p.A103T variant in TOE1 (Fig. 2.6). We sequenced the TOE1 gene in a cohort of over a hundred PCH and PCH-like patients HIER FIGUUR 2.6 with various PCH subtypes and unknown genetic cause. We identified potentially HIER FIGUUR 2.7 damaging variants in TOE1 in ten patients (eight families) (Fig. 2.7).

1o cousins II

1o cousins III

Died at 40 Died at one Died at 39y days of age day of age of age, moderate IV mental retardation but self- supporting

Figure 2.6: Pedigree of index family. Circle indicates a female, square indicates a male. Black: homozygous for p.A103T variant in TOE1; black line, gray filling: heterozygous for p.A103T variant in TOE1; grey line, white filling: wild type at position p.A103 in TOE1; black line, white filling: not tested. Diagonal line: deceased.

p.F239S

p.F148Y p.Y231X p.H319Y p.H319Q p.A103T p.R73S p.V173G p.R253W p.S496F p.Q314VfsX8

1 510 DEDD ZnF NLS

Figure 2.7: TOE1 variants in PCH7 patients. Schematic representation of the TOE1 protein with the DNA- like exonuclease (DEDD) domain, the zinc-finger domain (ZnF) and the nuclear localization signal (NLS) indicated. Dashed arrows indicate compound heterozygous mutations, solid arrows indicate homozygous mutations. Each family is indicated by a different colour.

36 Three apparently unrelated patients (patients 3, 4 and 5) carried the same missense variant (c.518T>G, p.V173G, allele frequency of 0.00001648 in ExAC database). Hap- lotype analysis of the three families revealed a common haplotype of approximately 9Mb, indicative for a common ancestor (Supplementary Table 2). All 46,XY patients in our cohort showed a certain degree of genital undervirilisation in addition to PCH. The two 46,XX cases with cerebellar hypoplasia and variants in TOE1 had normal female Chapter external genitals and are sisters of 46, XY patients with genital abnormalities. The lack of TOE1 variants in 46, XY PCH patients with unaffected male genitals indicates that 2 variants in this gene are exclusively related to PCH7. No compound heterozygous TOE1 variants were present in 9300 cases and con- T trols in exome sequencing data from a motorneuron disease consortium [17]. The OE 1 mutations in PCH7 ExAC database lists various potentially pathogenic variants in the TOE1 gene (CADD score > 15), but none of the identified variants in PCH7 patients is reported as a homozygous variant. Frequencies of compound heterozygous variants cannot be obtained from this database. To estimate the frequency of compound heterozygous cases, we used the reported frequency of variants with a Combined Annotation Dependent Depletion (CADD) score higher or equal to 15. In this way, we estimated the chance of having two pathogenic variants in TOE1 as 8.0x10-6. This very low probability and the fact that we found TOE1 variants in all eight PCH7 families supports the hypothesis that the identified variants in TOE1 are disease causing.

Clinical details of PCH patients with TOE1 mutations

All 46, XY patients with TOE1 mutations presented both brain and genital abnormali- HIER TABEL 2.1 ties and were therefore diagnosed with PCH7 (Table 2.1). All index patients were 46, XY and showed some degree of gonadal undervirilisation. The external genitals ranged in appearance from a nearly normal vagina to a micropenis with hypoplastic scrotum (Fig. 2.8). Internally, some 46, XY did not have gonadal tissue at all, some had ovarian or uterine remnants, and some had atrophic undescended testes. The 46, XX siblings had normal female external genitals. In one 46, XX patient the ovaries could not be observed. Endocrinological investigations in six patients revealed normal to high FSH and LH levels suggesting intact hypothalamic function. In contrast, testosterone levels were low and no response to human chorionic gonadotropin stimulation was observed. This suggests hypergonadotrophic HIER FIGUUR 2.8 hypogonadism, including a well-functioning hypothalamus and dysfunctional gonads. The brain MRI scans of the patients with TOE1 mutations show similarities (Fig. 2.9). HIER FIGUUR 2.9 All present pontocerebellar hypoplasia, enlarged ventricles and in most cases decreased white matter with a thin corpus callosum. Other clinical symptoms included axial hypotonia with increased tone in the limbs and spastic seizures. Development is severely delayed with no or very limited sitting,

37 H, low low H, Y FS Y 253W 253W R FS H= follicle 8 Japanese 14y c.757C>T, c.757C>T, c.955C>T p. p.H319 46,X no penis, penis, no labia normal unilateral unilateral undescended testis absent normal LH, LH, normal high testosterone , , F S Y 496 239 F S 7 Chinese/ Vietnamese/ Cambodian 16y c.716T>C, c.716T>C, c.1487C>T p. p. 46,X no penis, penis, no labia normal ovaries present u 231X Y A , p. , >C, >C, S G Y 73 H 4 R 6 6 Canadian 5y c.219 c.693T> p. 46,X no penis, penis, no labia normal ovaries present normal LH, high high LH, normal FS

, , , A G G Y H anish 5 D 11mo c.957C> c.518T> p.V173 p.H319Q 46,X no penis, labia labia penis, no present, majora minora labia absent absent remnants of of remnants tissue uterine normal LH, high high LH, normal FS A

, , , G G Y ustralian 4 A 2y (death) 2y c.940_941delC c.518T> p. V173 p. p.Q314VfsX8 46,X penis, fused fused penis, labia unclear remnants of of remnants tissue uterine u G A

, , , G G Y H, no no H, R 3 n 3 3 British 24w (death) 24w c.940_941delC c.518T> p. V173 p. p.Q314VfsX8 46,X regressing regressing micropenis, labioscrotal folds absent absent normal normal to response G hC to response

, A A , , Y Y 2 M_025077.3; ** N P_079353.3 ** N M_025077.3; gonadotropin;* G = human chorionic

148 148 II F F 2- Moroccan 7y c.443T> c.443T> p. p. 46,XX normal female normal normal ovaries normal present u

, A A , , Y Y Y 2 148 148

I F F 2- Moroccan 3y (death) 3y c.443T> c.443T> p. p. 46,X no penis, fused fused penis, no labia absent absent u 1 mutations. y=years; OE 1 w=weeks; mo=months; mutations. u=unknown; U S : ultrasound; LH= luteinizing hormone; , , H, no no H, A A , possibly possibly , S FS 103T, 103T, 103T 307 307 II A A G G 1- Turkish 22y c. c. p. p. 46,XX normal female normal ovaries not seen seen not ovaries U on atrophic present high LH LH high and at menarche 20y , , A A Y 103T, 103T, 103T 307 307 H, low low H, I A A G G 1- Turkish 25y c. c. p. p. 46,X micropenis, micropenis, hypoplastic scrotum atrophic testes atrophic absent high LH and and LH high FS testosteron Clinical features of patients with T ucleotide change* ucleotide ormones terus mino acid change** acid mino ge at last last at ge thnicity xternal genitals xternal nternal genitals nternal able 2.1: Patient E A examination N A Karyotype E I U H T stimulating hormone; G n R H= gonadotropin releasing hC

38 walking, speech and social eye contact. Three patients died at the age of 24 weeks, 2 years and 3 years, the other patients vary in age from 1 year old to late twenties.

A B C

Chapter 2

Figure 2.8: Genital abnormalities in PCH7 patients. All 46,XY patients with PCH7 present some degree of genital undervirilisation. Patients 3, 5 and 8 are shown. Patient 3 (A) has a micropenis with labioscrotal T OE 1 mutations in PCH7 folds. Patient 5 (B) is a 46,XY patient with labia majora and no labia minora. Patient 8 (C) has nearly normal female genitals, and undescended testes.

1-II 2-II 3 4

5 6 7 8

Figure 2.9: MR images of patients with TOE1 mutations. Patients show a hypoplastic pons and cerebellum, thin corpus callosum and large ventricles.

Functional studies TOE1 Multiple functions have been attributed to TOE1. It was initially identified as gene under control of early growth response 1 (Egr1), and therefore named target of Egr1. It was assumed to be important in cell cycle and cell growth [13]. Furthermore, the TOE1 protein includes a DEDD domain, suggesting deadenylating activity [6] and it can interact with tumor suppressor protein p53 [18]. More recent papers about TOE1 describe the protein as essential for Cajal body integrity [8] and as inhibitor of viral replication [7]. All these functions do not give an explanation for the cerebellar hypoplasia. However, TOE1 has been previously linked to Purkinje cell degeneration in

39 Arabian horses with cerebellar abiotrophy [9]. How TOE1 is functionally linked to brain and genital development is still unknown. Next to a defective function of TOE1 protein, the sequence variants could have a cis-effect on expression of adjacent genes. The TOE1 gene overlaps on the opposite strand with the gene MUTYH on its 5’ end and the gene TESK2 on its 3’ end [19]. TESK2 is involved in spermatogenesis, which could be is of potential interest in view of the genital phenotype we see in PCH7 patients. Indications for an increase in MUTYH expression is found in Arabian horses with TOE1 mutations [9]. However, we could not HIER FIGUUR 2.10 confirm this in patients’ fibroblasts, nor did we find alternations in TESK2 expression (Fig. 2.10).

3,5

control 1 2,5 control 2

het A103T 2 het p.Q314VfsX8 het p.V173G 1,5 p.V173G + p.Q314VfsX8 Fold change Æ change Fold hom p.A103T 1 hom p.F148Y p.R73S + p.Y231X 0,5

0 TOE1 TESK2 MUTYH

Figure 2.10: Expression levels of genes in the TOE1 locus. Displayed expression levels including error bars are means of three independent experiments.

To evaluate the impact of the TOE1 mutations at the molecular structural level, we constructed a model structure of the DEDD domain of human TOE1 from the crystal structure of mouse PARN protein. Five out of the nine missense mutations were mapped onto the obtained model structure, and two other mutations (both at amino acid H319) were mapped onto the NMR structure of the zinc finger domain of OET 1 (Fig. 11). The p.A103T and p.V173G mutations are located at a hydrophobic core of the DEDD domain and may therefore affect stability and deadenylase activity of TOE1. In case of PARN, the highly conserved p.F148 position is reported to be involved in homodimerization, which is crucial for deadenylase activity [20]. Thus, the p.F148Y mutation of TOE1 may impair protein-protein interactions and enzymatic activity. In the zinc finger domain, the p.H319Q and p.H319Y mutations could disrupt its folding. The two other mutations identified in the PCH7 patients, p.F239S and p.R253W, are

40 unlikely to affect protein folding substantially. However, since deadenylases often function in a multisubunit complex, it cannot be excluded that these mutations cause disturbances of protein-protein interactions. HIER FIGUUR 2.11

A B Chapter 2 T OE 1 mutations in PCH7

Figure 2.11: A modeled structure of the DEDD domain of human TOE1. The model is constructed from the crystal structure of mouse poly(A)-specific ribonuclease (PARN) (PDB code; 3D45) and (B) the NMR structure of the C3H-type zinc finger domain of human OET 1 (PDB code; 2FC6) are shown. The residues at the mutation sites are depicted in purple by van der Waals spheres in A and sticks in B, respectively. A gray arrow indicates the RNA-binding cleft. Two catalytic residues, D64 and E66 (A), and three cysteine residues of ZF (B) are shown as sticks with oxygen and sulfur atoms colored red and yellow, respectively. The zinc ion is shown as a green sphere.

Discussion

In this study, we confirmed the presence of a novel syndrome - pontocerebellar hypoplasia with disorders of sex development – previously proposed as PCH7. We described eight families with this disorder, all harbouring mutations in the TOE1 gene. By excluding a functional effect of the p.A390S variant in CLK2 in our index family and identifying TOE1 mutations in all eight families, we show that PCH7 is not a coincident combination of two syndromes, but an isolated disease. The combination of DSD and neurological anomalies is seen in a few other disorders. For example, patients with Smith-Lemli-Opitz syndrome (OMIM 270400) can present hypoplasia of the corpus callosum and cerebellum, microcephaly and ambiguous genitals in 46, XY. The disease is caused by a deficiency in 7-dehydrocholesterol re- ductase [21]. Prader-Willi syndrome (OMIM 176270) includes decreased fetal activity, hypotonia and mental retardation in combination with hypogonadism in 46, XY and 46, XX. In contrast to PCH7, this hypogonadism is of hypothalamic origin, including low levels of LH and FSH [22]. X-linked alpha-thalassaemia mental retardation syndrome (ATRX, OMIM 301040) is a rare syndrome caused by mutations in the ATRX gene. Symptoms include mental retardation, severe developmental delay, microcephaly and genital abnormalities in the majority of patients ranging from hypospadias to XY sex

41 reversal [23]. Although PCH7 is not the only disorder that includes both brain and genital abnormalities, in none of previously described disorders brain atrophy is as pronounced as in PCH7. Up to now, we can only speculate about the exact role of TOE1 in brain and genital development. TOE1 aberrations have shown to cause Purkinje cell degenerations in Arabian horses. It was proposed that the TOE1 mutation altered expression of the overlapping MUTYH gene, but we could not confirm this in fibroblasts of PCH7 patients. Perhaps different results would be obtained when investigated mRNA levels in brain or gonadal tissue. Taking in account the function of other PCH-genes – in RNA metabolism and protein synthesis - the function of TOE1 in the integrity of Cajal bodies is potentially interesting. Cajal bodies are nuclear organelles that contain high numbers of small nuclear ribonucleoproteins (snRNPs), which are essential for mRNA splicing. It is plausible that TOE1 mutations lead to aberrant mRNA splicing, subse- quently leading to disturbances in protein synthesis. Splicing has also been shown to be important in sex differentiation [24]. This is the first time that TOE1 is linked to genital development. The patients with TOE1 mutations present high LH and FSH, suggesting a functioning pituitary gland. Testosterone is low in 46, XY patients, indicating malfunctioning gonads. Also in a 46, XX patient with TOE1 mutations (patient 1-II) we see aberrations in gonad development, as ovaries were not observed and puberty did not initiate. Early in gonadal development, a bipotential gonad is formed, which can differentiate in either testes or ovaries [25]. Since both 46, XY and 46, XX present gonadal dysfunction, it is plausible that TOE1 mutations cause a defect in formation of the bipotential gonad. Further functional studies could focus on the role of TOE1 in mRNA splicing, deadenylation and other RNA processing events. With the identification of TOE1 as PCH causing gene, yet another link between RNA processing and neurodegeneration is made.

42 References

1. Ostrer H: Disorders of sex development (DSDs): an update. J Clin Endocrinol Metab 2014, 99: 1503-1509. 2. Budde BS, Namavar Y, Barth PG, Poll-The BT, Nurnberg G, Becker C et al.: tRNA splicing endo- nuclease mutations cause pontocerebellar hypoplasia. Nat Genet 2008, 40: 1113-1118. 3. Anderson C, Davies JH, Lamont L, Foulds N: Early pontocerebellar hypoplasia with vanishing testes: A new syndrome? Am J Med Genet A 2011, 155A: 667-672. Chapter 4. Siriwardena K, Al-Maawali A, Guerin A, Blaser S, Chitayat D: XY sex reversal, pontocerebellar hypoplasia and intellectual disability: confirmation of a new syndrome.Am J Med Genet A 2013, 2 161A: 1714-1717. 5. Mahbubul Huq AH, Nigro MA: XY sex reversal and a nonprogressive neurologic disorder: a new

syndrome? Pediatr Neurol 2000, 23: 357-360. T OE 1 mutations in PCH7 6. Wagner E, Clement SL, Lykke-Andersen J: An unconventional human Ccr4-Caf1 deadenylase complex in nuclear cajal bodies. Mol Cell Biol 2007, 27: 1686-1695. 7. Sperandio S, Barat C, Cabrita MA, Gargaun A, Berezovski MV, Tremblay MJ et al.: TOE1 is an inhibitor of HIV-1 replication with cell-penetrating capability. Proc Natl Acad Sci U S A 2015, 112: E3392-E3401. 8. Fong KW, Li Y, Wang W, Ma W, Li K, Qi RZ et al.: Whole-genome screening identifies proteins localized to distinct nuclear bodies. J Cell Biol 2013, 203: 149-164. 9. Brault LS, Cooper CA, Famula TR, Murray JD, Penedo MC: Mapping of equine cerebellar abiotrophy to ECA2 and identification of a potential causative mutation affecting expression of MUTYH. Genomics 2011, 97: 121-129. 10. Thisse C, Thisse B: High-resolution in situ hybridization to whole-mount zebrafish embryos.Nat Protoc 2008, 3: 59-69. 11. Wu M, Nilsson P, Henriksson N, Niedzwiecka A, Lim MK, Cheng Z et al.: Structural basis of m(7) GpppG binding to poly(A)-specific ribonuclease. Structure 2009, 17: 276-286. 12. Kelley LA, Mezulis S, Yates CM, Wass MN, Sternberg MJ: The Phyre2 web portal for protein modeling, prediction and analysis. Nat Protoc 2015, 10: 845-858. 13. de B, I, Wu JX, Sperandio S, Mercola D, Adamson ED: In vivo cloning and characterization of a new growth suppressor protein TOE1 as a direct target gene of Egr1. J Biol Chem 2003, 278: 14306-14312. 14. Franklin RB, Ma J, Zou J, Guan Z, Kukoyi BI, Feng P et al.: Human ZIP1 is a major zinc uptake transporter for the accumulation of zinc in prostate cells. J Inorg Biochem 2003, 96: 435-442. 15. Duncan PI, Stojdl DF, Marius RM, Scheit KH, Bell JC: The Clk2 and Clk3 dual-specificity protein kinases regulate the intranuclear distribution of SR proteins and influence pre-mRNA splicing. Exp Cell Res 1998, 241: 300-308. 16. Rudnik-Schoneborn S, Barth PG, Zerres K: Pontocerebellar hypoplasia. Am J Med Genet C Semin Med Genet 2014, 166C: 173-183. 17. Cirulli ET, Lasseigne BN, Petrovski S, Sapp PC, Dion PA, Leblond CS et al.: Exome sequencing in amyotrophic lateral sclerosis identifies risk genes and pathways. Science 2015. 18. Sperandio S, Tardito S, Surzycki A, Latterich M, de B, I: TOE1 interacts with p53 to modulate its transactivation potential. FEBS Lett 2009, 583: 2165-2170. 19. Makalowska I: Comparative analysis of an unusual gene arrangement in the human chromosome 1. Gene 2008, 423: 172-179.

43 20. Goldstrohm AC, Wickens M: Multifunctional deadenylase complexes diversify mRNA control. Nat Rev Mol Cell Biol 2008, 9: 337-344. 21. Bianconi SE, Cross JL, Wassif CA, Porter FD: Pathogenesis, Epidemiology, Diagnosis and Clinical Aspects of Smith-Lemli-Opitz Syndrome. Expert Opin Orphan Drugs 2015, 3: 267-280. 22. Angulo MA, Butler MG, Cataletto ME: Prader-Willi syndrome: a review of clinical, genetic, and endocrine findings. J Endocrinol Invest 2015. 23. Stevenson RE: Alpha-Thalassemia X-Linked Intellectual Disability Syndrome. In: Pagon RA, Adam MP, Ardinger HH, Bird TD, Dolan CR, Fong CT, Smith RJH, Stephens K, editors. GeneR- eviews® [Internet]. Seattle (WA): University of Washington, Seattle; 1993-2014 (updated 2014) 24. Ohe K, Lalli E, Sassone-Corsi P: A direct role of SRY and SOX proteins in pre-mRNA splicing. Proc Natl Acad Sci U S A 2002, 99: 1146-1151. 25. Eggers S, Ohnesorg T, Sinclair A: Genetic regulation of mammalian gonad development. Nat Rev Endocrinol 2014, 10: 673-683.

44 Supplementary Table 1: Overview of primers. All primers – except those used for qPCR – were flanked by M13 motives for sequencing; forward: 5’-TGTAAAACGACGGCCAGT-3’; reverse: 5’-CAGGAAACAGCTAT- GACC-3’.

name forward reverse

CLK2_ex1 5’-GGAAATGAAGTGACCGTGGA-3’ 5’-CTACGACGCCTCCGCTAGA-3’ CLK2_ex2 5’-GTACCTCCGAGGCTCTGACA-3’ 5’-CCCAACCCAAGCCTAACC-3’ Chapter CLK2_ex3 5’-AGCCAGTCCTTGGAGGAGAT-3’ 5’-TTAACACTGCAGCCCAAATG-3’ CLK2_ex4 5’-TTGTCACAGGGTATGGTAGGG-3’ 5’-CCAATGTTCAGCTGAGAACAAG-3’ 2 CLK2_ex5 5’-GTCCCCCATCATCGTCTGT-3’ 5’-CTGTGACTCAGGTTGGCTTG-3’ CLK2_ex6and7 5’-CTTCCCCTGTGACCATCTGT-3’ 5’-TCTAGGACTCCCCTGTCTGC-3’

CLK2_ex8 5’-CTCCTTGTTGGATGGCTGAT-3’ 5’-TGTTTAGGGGCTTCACCAGT-3’ T OE 1 mutations in PCH7 CLK2_ex9 5’-GGCTTTGCCCTAGGTAACCA-3’ 5’-CAGAAGGCAAAGTGGTGCAG-3’ CLK2_ex10and11 5’-CCAGGCAGGAGATCACAGTA-3’ 5’-AGTAGGCAGGTCCTCAGCAG-3’ CLK2_ex12and13 5’-GCCTGGCATTCTTCTACCTG-3’ 5’-ACCCTCTGCTTGTGAAGAGC-3’ CLK2_ex14p1 5’-CTTGAAGAAGGCAGGGTTTG-3’ 5’-AGGTGAGGGTGGAAACTGTG-3’ CLK2_ex14p2 5’-CAACAAGTTGTGGGACTCCA-3’ 5’-GGAGGACTCCCTTCATACCC-3’ TOE_ex1 5’-AACGGAAGTTCGACCCATC-3’ 5’-GGAAGGAGCAGTCCTCTGAA-3’ TOE_ex2en3 5’-GGCCTAGCTTGGTGTCTGTT-3’ 5’-GGTTGCTGTGGCTAGGAGTT-3’ TOE_ex4 5’-GTCACAGACCTGGGTGGTTT-3’ 5’-AAAAAGATTTGGGAGAGAAATGC-3’ TOE_ex5 5’-TTATGGGGAACAGGTGGTGT-3’ 5’-CCACTTGGAACGCTCTCTCT-3’ TOE_ex6 5’-GGCAGCATTGGGTAAGGTAG-3’ 5’-GCAGGGTTTTGTGGTCTGTT-3’ TOE_ex7and8a 5’-TGGGGCCTGATACGAGTTTA-3’ 5’-GGTTTCTTCAGGCTTGATGC-3’ TOE_ex8b 5’-GAAAAACGGAAGAGGGCTTT-3’ 5’-GATAAAGGAGGCGGTCCAG-3’ SLC39A1 5’-GGAGGGGGTTTGTTTGCTTA-3’ 5’-AACCAACACTCCCAGAGGTG-3’; TOE1_qPCR 5’-cctaccataagggcaatgaca-3’ 5’-ggccattgtgtagcaccag-3’ TESK2_qPCR 5’-ccccagattttctgcaactta-3’ 5’-cctgtaggcggctcagaa-3’ MUTYH_qPCR 5’-atgacaccgctcgtctcc-3’ 5’-gcttctgcctcccttcct-3’ rs968322 5’-CCTGAGTCTTTGCCCAGAAG-3’ 5’-ATGGTGGTGTAGGGCATCTG-3’ rs4660360 5’-TGGGCTTTCAGTGCAATGC-3’ 5’-TCCTCTTTCTCCCTCTCTGG-3’ rs61784349 5’-AGATTGTGAAACCTGGCTTCA-3’ 5’-CTTCTCATTGAGCACCTACCC-3’ rs189962969 5’-TGAGAGAAGGCAGTTTGGGT-3’ 5’-CTGTACAAAGCTGCTGGTCC-3’ rs200725979 5’-CCAACTTGCCACACATCAGG-3’ 5’-GCACTGCTATCAATGACAAGGA-3’ rs370490105 5’-ACAGAAACATTAGAAGTGGAGGA-3’ 5’-TGTGCCAACCTCTACCACAT-3’ rs75705909 5’-GAACTCAGCCTCTCCTCCAG-3’ 5’-CAAACCTTCTATCCCTACCTCCA-3’ rs145119239 5’-GTAGCAGGTCCTTCTCTGGG-3’ 5’-AACATCTTGTCCAGCCCCTT-3’ rs150795467 5’-TGCAGGTTTCCAAATACGGC-3’ 5’-GAGGAGCACTGTCCACCTAG-3’ rs191887693 5’-GTGAGTGCTGGAGGGAGG-3’ 5’-CACATGGGGTTTTGCTGAGC-3’ rs200936475 5’-TTTCAGAAGAGCAGCCTGTG-3’ 5’-CTCCTACAACTGACGGTCCA-3’ rs202091916 5’-CAAGCCGCTGTATAACCGTC-3’ 5’-AGGCCATATATCTGTGTACTGCT-3’

45 Supplementary Table 2: Haplotype analysis of the three patients with a c.518T>G; p.V173G variant in TOE1. The 9Mb homozygous stretch shared amongst the three patients is indicated in red.

SNP distance to TOE1 (Mb) patient 3 patient 4 patient 5

rs145119239 -9.7 1 0 0 rs75705909 -7.5 1 1 1 rs202091916 -5.8 0 0 0 rs200415763 -5.8 0 rs150795467 -5-5 1 rs191887693 -0.3 0 0 0 c.518T>G 0 1 1 1 c.940_941delCA 0 0 0 0 rs200936475 +0.2 0 rs189962969 +0.7 0 0 0 rs4660360 +1.5 1 rs968322 +1.8 1 0 rs200725979 +7.3 0 0 0