A Truncating Mutation of CEP135 Causes Primary Microcephaly and Disturbed Centrosomal Function

A Truncating Mutation of CEP135 Causes Primary Microcephaly and Disturbed Centrosomal Function
of 8
All materials on our website are shared by users. If you have any questions about copyright issues, please report us to resolve them. We are always happy to assist you.
Related Documents
  REPORT  A Truncating Mutation of   CEP135  Causes PrimaryMicrocephaly and Disturbed Centrosomal Function Muhammad Sajid Hussain, 1,2,3,4 Shahid Mahmood Baig, 4 Sascha Neumann, 2 Gudrun Nu¨rnberg, 1,3 Muhammad Farooq, 4 Ilyas Ahmad, 1,4 Thomas Alef, 1 Hans Christian Hennies, 1,5 Martin Technau, 2  Janine Altmu¨ller, 1 Peter Frommolt, 1,5 Holger Thiele, 1 Angelika Anna Noegel, 1,2,3,5, *and Peter Nu¨rnberg 1,3,5, * Autosomal-recessive primary microcephaly (MCPH) is a rare congenital disorder characterized by intellectual disability, reduced brainandheadsize, butusuallywithoutdefects in cerebralcorticalarchitecture, andother syndromic abnormalities.MCPHis heterogeneous.The underlying genes of the seven known loci code for centrosomal proteins. We studied a family from northern Pakistan with twomicrocephalic children using homozygosity mapping and found suggestive linkage for regions on chromosomes 2, 4, and 9. Wesequenced two positional candidate genes and identified a homozygous frameshift mutation in the gene encoding the 135 kDa centro-somal protein ( CEP135 ), located in the linkageinterval on chromosome 4, in both affected children.Post hoc whole-exome sequencingcorroboratedthismutation’sidentificationasthecausalvariant.Fibroblastsobtainedfromoneofthepatientsshowedmultipleandfrag-mented centrosomes, disorganized microtubules, and reduced growth rate. Similar effects were reported after knockdown of   CEP135 through RNA interference; we could provoke them also by ectopic overexpression of the mutant protein. Our findings suggest an addi-tional locus for MCPH at HSA 4q12 (MCPH8), further strengthen the role of centrosomes in the development of MCPH, and placeCEP135 among the essential components of this important organelle in particular for a normal neurogenesis. Autosomal-recessive primary microcephaly (MCPH [MIM251200]) is a neurodevelopmental disorder characterizedbyreducedsizeofthecerebralcortexandmildtomoderateintellectual disability whereas the architecture of the brainis largely normal. Head circumference is already reduced atbirth and usually more than 3 SD below the age- and sex-matched population mean throughout the patient’s life-time. Many patients have a receding forehead. MCPH isheterogeneous; it has seven known loci, MCPH1–MCPH7, 1 each of which is associated with an underlyinggenetic defect: MCPH1 (MIM 251200) is associated withmutations of   MCPH1  (MIM 607117); 2 MCPH2 with orwithout cortical malformations (MIM 604317) is associ-ated with mutations of   WDR62  (MIM 613583); 3,4 MCPH3 (MIM 604804) is associated with mutations of  CDK5RAP2 (MIM608201); 5 MCPH4(MIM604321)isasso-ciated with mutations of   CEP152  (MIM 613529); 6 MCPH5(MIM 608716) is associated with mutations of   ASPM   (MIM605481); 7 MCPH6 (MIM 608393) is associated with muta-tionsof  CENPJ  (MIM609279); 5 andMCPH7(MIM612703)is associated with mutations of   STIL  (MIM 181590). 8 Recently, a new locus at HSA 10q11.23-21.3 was describedin a consanguineous Turkish family, but the authors werenot able to find the disease-causing gene variant. 9 TheMCPH-associated genes described to date have been impli-catedincelldivisionandcellcycleregulation,andmanyof the corresponding gene products are localized to thecentrosome. It has been speculated that they affect neuralprogenitor cell number through disturbed microtubuleorganization at the centrosome, resulting in altered celldivision and cortical development. Despite their similarsubcellular localization, the biological functions of thesegenes may vary. 1,3,4,10 We studied a consanguineous family from northernPakistan; the two affected children had primary micro-cephaly at birth (Figures 1A and 1B). After informedconsent of the parents was obtained, pedigree informationwas documented and clinical data and blood samples werecollected from the affected siblings, their father, and anunaffected sibling. Each affected child had a sloping fore-head. By 5 years of age they showed severe cognitive defi-cits; their speech was not understandable. They wereunable to form sentences and even to say any clear words,although their hearing was not impaired. Other abnormal-itieswerenotapparent.IndividualV-2diedat11years.Thehead circumference of the affected individuals rangedbetween  12 and  14.5 SD compared to the average pop-ulation of the same age and sex. The parents were healthywith normal head circumference. Ethical approval for thisstudy was obtained from the ethics review board at theNational Institute for Biotechnology and Genetic Engi-neering in Faisalabad according to the Declaration of Hel-sinki.Standard procedures were used to extract DNA from theblood samples for homozygosity mapping. Initially, weexcluded all known MCPH loci in the family, using STR markers to demonstrate heterozygosity in the relevantgenomic regions. We then performed a genome-wide 1 Cologne Center for Genomics (CCG), University of Cologne, 50931 Cologne, Germany;  2 Institute of Biochemistry I, Medical Faculty, University of Cologne,50931Cologne,Germany; 3 CenterforMolecularMedicineCologne(CMMC),UniversityofCologne,50931Cologne,Germany; 4 HealthBiotech-nology Division, National Institute for Biotechnology and Genetic Engineering (NIBGE), Faisalabad 38000, Pakistan;  5 Cologne Excellence Cluster onCellular Stress Responses in Aging-Associated Diseases (CECAD), University of Cologne, 50674 Cologne, Germany*Correspondence: (A.A.N.), (P.N.) DOI 10.1016/j.ajhg.2012.03.016.  2012 by The American Society of Human Genetics. All rights reserved. The American Journal of Human Genetics  90 , 871–878, May 4, 2012  871  linkage analysis using the Affymetrix GeneChip HumanMapping 250K Sty Array. Data handling, evaluation, andstatistical analysis were performed as described previ-ously. 11 We observed three peaks on chromosomes 2, 4,and9thatweresuggestiveforlinkagewithmaximummul-tipoint LOD scores of 2.07, 2.53, and 2.53, respectively(Figure S1). The underlying common homozygous regionsdefined a candidate region of 8.6 Mb on chromosome2 (90,183,484–98,795,792; hg19), a candidate region of 18.1 Mb on chromosome 4 (40,631,476–58,702,130;hg19), and a candidate region of 8.8 Mb on chromosome9 (92,274,161–101,122,314; hg19). Altogether theseregions include more than 200 annotated known andpredicted coding genes (UCSC Genome Bioinformatics,hg19). The genes were prioritized with Endeavour andGeneWanderer. 12,13 Highly ranked genes were furtherscrutinized manually with the NCBI, Ensembl, and UCSCgenome databases. Finally, we gave top priority to genesthat were very likely to have important functions relatedto cell division or chromosome segregation and decidedto sequence the following two strong candidate genesfirst: the cell division cycle-14 homolog B gene ( CDC14B [MIM 603505]) at cytoband 9q22.32-q31.1 and the geneencoding the 135 kDa centrosomal protein ( CEP135 [MIM 611423]) at HAS 4p14-q12 (Figure 1C).All exons and the intron-exon boundaries of   CEP135 and  CDC14B  were sequenced in the two affected individ-uals. The PCR products (primers are listed in Table S1) Figure 1. Identification of an MCPH-Causing Mutation in  CEP135  (A) Clinical features of two siblings of a family from northern Pakistan presenting with microcephaly, sloping forehead, and retrogna-thia. Informed consent to publish the photographs was obtained from the subjects’ parents.(B) A simplified pedigree of the Pakistani family. Filled circles indicate individuals with MCPH. Their parents are second cousins. Onlythe core family is shown. DNA was available for individuals IV-1, V-1, V-2, and V-3.(C) G-banded chromosome 4 showing the position of the linkage interval at 4p14-4q12 along with the position of   CEP135 . The homo-zygous region was delimited by the markers SNP_A-2154951 (rs12498424, physical position 40,631,476 bp) and SNP_A-1894332(rs13134527, physical position 58,702,130 bp).(D) Sequence chromatograms of a part of   CEP135  exon 8 as obtained by Sanger sequencing. Traces of the male patient (V-1) and hisfather (IV-1) show the mutation c.970delC in homozygous and heterozygous status, respectively. The mutant sequences are shownalong with a wild-type trace from a control individual.(E)Schematicrepresentationofthegenomicstructureofhuman CEP135. The26exonsof  CEP135 aredrawntoscale,whileintronsshowjust artificial lines. Black boxes represent untranslated regions. The position of mutation c.970delC in exon 8 is indicated.(F) CEP135 structure as predicted by SMART (Simple Modular Architecture Research Tool) database. This software predicted six coiled-coil domains (orange rods) which cover almost the entire region of the protein. The position of mutation p.Q324Sfs*2 in the fourthcoiled-coil domain is indicated. 872  The American Journal of Human Genetics  90 , 871–878, May 4, 2012  were sequenced bidirectionally with a BigDye Terminatorv1.1 cycle sequencing kit on an ABI3730xl automatedDNA sequencer. Sequences were analyzed with DNASTAR (Lasergene) and Mutation Surveyor (SoftGenetics). Wefound no mutation in  CDC14B  but a homozygous singlebase-pair (bp) deletion (c.970delC) in exon 8 of   CEP135 .We found the homozygous 1 bp deletion in both affectedchildren, but not in the unaffected child, whereas thefather was heterozygous for this mutation (Figure 1D),which is compatible with recessive inheritance. Moreover,the mutation is unlikely to be a polymorphism, as it is notlisted in dbSNP, and we could not find it when testing 384healthy Pakistani controls with pyrosequencing. For thispurpose, PCR primers (listed in Table S2) were designedby the PSQ Assay Design program v.1.0.6 (QIAGEN,Hilden, Germany). Pyrosequencing was done accordingto the manufacturer’s instructions on a PSQ HS96A instru-ment (QIAGEN) with the use of PyroMark Gold Q96Reagents (QIAGEN). The data were analyzed by PyroQ-CpG v.1.0.9 analysis software (QIAGEN).In an attempt to identify a second independent muta-tion of   CEP135 , we sequenced  CEP135  in patients fromseven other families affected with MCPH from northernPakistan in which all known MCPH-associated geneswere previously excluded. No  CEP135  mutation was foundin any of these families. Therefore, we decided to performwhole-exome sequencing of the affected boy (individualV-1) of the family presented in Figure 1B in order todemonstrate that there were no other mutations in rele-vant genes that might also explain the phenotype. Wefragmented 1  m g of DNA using sonification technology(Covaris, Woburn, MA, USA). The fragments were end re-paired and adaptor ligated. After size selection, the librarywassubjectedtotheenrichmentprocess.WechosetheSeq-Cap EZ Human Exome Library v2.0 kit from NimbleGen(Roche NimbleGen, Madison, WI, USA) and analyzed thesampleonanIlluminaHiSeq2000sequencinginstrument.About 10Gb of sequence were produced for this sample byloadingitindividuallyononelaneofaflowcellandgener-atingpaired-endreadsof2 3 100bp.Thisresultedinaveryhigh coverage; i.e.,  >  30 3  for nearly 92% of the targetsequences, which comprised about 44 Mb. Primary datawere filtered according to signal purity by the IlluminaRealtime Analysis (RTA) software v1.8. Subsequently, thereads were mapped to the human genome reference buildhg19 via the ELANDv2 alignment algorithm on a multi-node compute cluster. With the use of CASAVA v1.8, PCR duplicates were filtered out, and the output was convertedinto BAM format. Variant calling was performed with theuse of SAMtools (version 0.1.7) for indel detection. 14 Scripts developed in-house at the Cologne Center forGenomics were applied to detect protein changes, affectedsplice sites, and overlaps with known variants. In partic-ular, we filtered the variants for high-quality unknownvariants in the linkage intervals (dbSNP build 132 or the1000 Genomes database; in-house variation database; andpublic Exome Variant Server, NHLBI Exome SequencingProject, Seattle) (Table S3). Only three other homozygousvariants resisted our filter criteria in addition to  CEP135 c.970delC. None of these could be assumed to be relevantfor the phenotype (Table S4). Likewise, we could not finddeleterious mutations in any of the seven known MCPH-associated genes, even when allowing for compoundheterozygosity during filtering.Thedeletionc.970delCof  CEP135 resultsinaframeshift,changing glutamine at position 324 into serine, immedi-ately followed by a premature termination codon(p.Gln324Serfs*2). This truncating mutation is obviouslyincompatible with a normal function of CEP135 if oneconsiders the size of the protein and the position of the mutation (Figures 1E and 1F).  CEP135  consists of 26exons, and its open reading frame codes for a polypeptideof 1,140 amino acids.CEP135 is a conserved  a -helical protein which is presentat the centrosome throughout the cell cycle. Electron-microscopic studies showed its association with thepericentriolar material, an electron-dense material sur-rounding the centrioles. Reducing CEP135 amounts incells via RNA interference caused a disorganization of interphase and mitotic spindles, leading to the hypothesisthat CEP135 has a role in maintaining the structure andorganization of the centrosome and of microtubules. 15 More recently the protein was identified as a centriolarcomponent; it functions in centriole biogenesis andpresumably has a scaffolding role. 16 Centrioles are corecomponentsofanimalcentrosomesandactasbasalbodiesto assemble cilia and flagella. 17 To unravel the effect of the mutation on the cellularlevel, we analyzed control and patient fibroblasts. BiopsiesweretakenfromtheaffectedindividualV-1(Figure1B)andhealthy individuals. Tissues were cleaned with antisepticagent (Betaisodona, Mundipharma) and incubated over-night with Dispase II (1.5 U/ml, Roche) diluted in PBS,pH 6.8, at 4  C to separate the intact epidermis from thedermis. The dermis was incubated in Dulbecco’s modifiedEagle’s medium (DMEM) at 37  C. After one week, fibro-blasts were detected. The pool of fibroblasts was increasedby additional culturing. Primary fibroblasts establishedfrom the patient grew very slowly. To enhance growth,DMEM with 15% fetal bovine serum was used. Patientand wild-type primary fibroblasts were then cultured on12 mm coverslips and fixed with 3% paraformaldehyde.For staining of microtubules, cells were incubated for15 min in tubulin stabilization buffer, which is composedof Hank’s buffer (137 mM NaCl, 5 mM KCl, 1.1 mMNa 2 HPO 4 , 0.4 mM KH 2 PO 4 , 5 mM Glucose, 4 mMNaHCO 3 ) containing 1 mM MES, pH 6.8, 2 mM EGTA,and 2 mM MgCl 2 . Permeabilization was done by 0.5%Triton X-100 in 1 3  tubulin stabilization buffer for 4 minat room temperature. For  g -tubulin detection, the cellswere fixed with prechilled methanol for 10 min at  20  C.Subsequently, the fixed cells were treated three times withtubulin stabilization buffer. Blocking was done for 15 minwith blocking buffer (1 3 PBG: PBS containing 5% BSA and The American Journal of Human Genetics  90 , 871–878, May 4, 2012  873  0.45% fish gelatin). Primary antibodies were diluted inblocking buffer and incubated overnight at 4  C. Thefollowing antibodies were used: mouse monoclonal anti g -tubulin (Sigma-Aldrich, GTU-88; 1:300), rabbit poly-clonal anti-pericentrin (Abcam, ab4448; 1:300), and ratmonoclonal anti  a -tubulin (YL 1/2 1:20). 18 After incuba-tion, samples were treated with 1 3  PBS three times for5 min followed by secondary antibody incubation(1:1,000 diluted in blocking buffer) for one hour at roomtemperature.AlexaFluor568goatanti-mouseIgG(Invitro-gen, A11004), Alexa Fluor 647 donkey anti-mouse IgG(Invitrogen, A31571), and Alexa Fluor 488 goat anti-ratIgG (Invitrogen, A11006) were used as secondary anti-bodies. DNA was detected with DAPI (Sigma-Aldrich,D9564). Finally, the cells were mounted on glass slideswith Gelvatol. Images were taken with a confocal micro-scope (Leica, LSM TCS SP5).Control primary fibroblasts had oval nuclei surroundedby organized microtubules and a single centrosome in Figure 2. Centrosome, Microtubule, andNuclearShapeDefectsin CEP135  Microce-phalic Patient Cells (A) Immunofluorescence staining andconfocal microscopy images of wild-typeprimary fibroblasts with well-organizedmicrotubules and a single centrosome de-tected by  g -tubulin.  g -tubulin (turquoise)and  a -tubulin (green) antibodies wereused. DAPI (blue) was used for DNA stain-ing. Scale bar, 5  m m.(B) Abnormality of centrosome numberin  CEP135  patient primary fibroblasts.Interphase cell showing supernumerarycentrosomes. In this figure three centro-somes were observed as detected by g -tubulin (turquoise). Scale bar, 5  m m.(C) Disorganized microtubule network inpatient fibroblasts. Scale bar, 20  m m.(D) Control primary fibroblast showinga well-shaped ellipsoid nucleus. Scale bar,5  m m.(E) Mutant fibroblast with a dysmorphicnucleus. Scale bar, 10  m m.(F) Graphical representation showingthe number of   CEP135  mutant primaryfibroblasts with dysmorphic nuclei. About20% of mutant fibroblasts harboredmisshapennuclei whereas in control fibro-blasts this number was only ~3%. Threehundred cells of each, wild-type andmutant,were counted. ErrorbarsrepresentSEM, p ¼ 3.44 3 10  03 (Student’s t test). the vicinity of the nucleus duringinterphase (Figure 2A). In thepatient’s primary fibroblasts, thecentrosome number was increased inmore than 18% of the cells (TableS5). In such cells we found 3, 4, or5 centrosomes per cell (Figures 2Band S2). Furthermore, centrosomesappearedfragmented(FigureS2).Themicrotubulenetwork wasfrequentlydisorganized(~55%ofthe cells),whichwasaccompanied by cell shape changes (Figures 2C and S3; Table S5). We also observed misshapen and fragmentednuclei (Figure 2E and S4). Statistical analysis showed that ~20% of the mutant cells harbored misshapen nucleias compared to ~3% in control fibroblasts (Figure 2F; Table S5). Another prominent aspect of the patient’s primaryfibroblasts was the complete loss of centrosomes. Approx-imately 22% of mutant primary fibroblasts were withoutcentrosomes as detected with  g -tubulin, whereas this wasnever observed in control cells (Table S5; Figure S3). Most cells with misshapen nuclei were also devoid of centro-somes (Figure S4).For ectopic expression of wild-type and mutant(c.970delC) CEP135 fused to green fluorescent protein(GFP), we used COS-7 cells. Gateway Technology (Invitro-gen) was employed to clone wild-type (NM_025009.3) CEP135  cDNA. The  CEP135  mutation (c.970delC) was 874  The American Journal of Human Genetics  90 , 871–878, May 4, 2012  introduced into the full-length cDNA with the use of theQuikChange II Site-Directed Mutagenesis Kit (Stratagene)(primers are listed in Table S6) in an attempt to reproducethe patient’s situation. Entry clones in pENTR/TEV/D-TOPO were transformed into Gateway destinationvector pcDNA-DEST53, which has an N-terminal cycle-3GFP tag. Both wild-type and mutant plasmids (10  m g/ m l)were used for transfection of COS-7 cells. The Gene PulserII (Bio-Rad) device was used for electroporation. 72 hrafter transfection, the cells were subjected to immuno-fluorescence.When we analyzed the localization of the proteins andtheir effect on centrosomes and microtubule organization,we observed wild-type GFP-tagged CEP135 at the centro-some where it colocalized with the centrosomal proteinpericentrin. We also detected it in some spots outside of the centrosome that were not positive for pericentrin(Figure 3A). By contrast, we did not detect the mutantprotein on the centrosome; instead, in a few cells weobserved a diffuse cytoplasmic staining (Figure 3B). Over- Figure 3. Distribution of GFP-TaggedWild-Type and Mutant CEP135 (A) N-terminally GFP-tagged humanCEP135 (GFP-CEP135) was transientlyexpressed in COS-7 cells. GFP-CEP135is present in dots of variable size andnumber. A single centrosome was detectedby pericentrin-specific antibodies (tur-quoise). Scale bar, 5  m m.(B) COS-7 cell transfected with GFP-taggedmutantCEP135(GFP-CEP135-mut),wherediffuse staining was detected. Scale bar,5  m m.(C) Abnormal microtubule network inCOS-7 expressing wild-type CEP135. Scalebar, 5  m m.(D) COS-7 cells expressing GFP-CEP135-mut have a disorganized microtubulenetwork. Microtubules appeared brokenand concentrated near the nucleus. Scalebar, 5  m m. expression of both wild-type andmutant GFP-taggedCEP135 inHaCaT cells led to the presence of abnormalmicrotubule networks. The severityof the disorganization was morepronounced in cells expressing themutant protein (Figure 3C and 3D).Cells transfected with GFP-taggedmutant CEP135 also harbored mul-tiple centrosomes (up to 5), whichthen resulted in multipolar spindleformation. This phenotype was notobserved in the cells that overex-pressed the wild-type protein. Inter-estingly, the GFP-tagged CEP135 wasdetected on microtubules (Figure 4).These findings of a disorganizedmicrotubule network and multiple centrosomes in cellstransfected with GFP-tagged CEP135 resembled thoseobserved in mutant primary fibroblast cells.The  CEP135  mutation c.970delC is thought to lead to atruncation of the protein due to the introduction of apremature stop codon (p.Gln324Serfs*2). Alternatively,it might also trigger nonsense-mediated mRNA decay(NMD). 19 We designed RT-PCR experiments to investigatethis (Table S7). The PAXgene Blood RNA system (QIAGEN)was used to extract RNA from patient and control bloodsamples, and cDNA was synthesized with SuperScript IIIreverse transcriptase enzyme (Invitrogen). When primerswere used to amplify a nearly full-length  CEP135  cDNAof~3.4 kb,noPCRproduct wasobtained withthepatient’sRNA, but only with the control sample, suggestingpronounced degradation of mutant  CEP135  mRNA(Figure S5A). In contrast, a smaller PCR product of only385 bp, generated with primers just flanking the site of mutation, was easily obtained from both control andmutant RNA (Figure S5B). A contamination of the mutant The American Journal of Human Genetics  90 , 871–878, May 4, 2012  875
Similar documents
View more...
Related Search
We Need Your Support
Thank you for visiting our website and your interest in our free products and services. We are nonprofit website to share and download documents. To the running of this website, we need your help to support us.

Thanks to everyone for your continued support.

No, Thanks