Mutational and genotype–phenotype correlation analyses in 28 Polish patients with Cornelia de...

� 2006 Wiley-Liss, Inc. American Journal of Medical Genetics Part A 140A:1531–1541 (2006)

Mutational and Genotype–Phenotype CorrelationAnalyses in 28 Polish Patients With Cornelia

de Lange Syndrome

Jiong Yan,1 Gulam Mustafa Saifi,1 Tomasz H. Wierzba,2 Marjorie Withers,1 Gabriel A. Bien-Willner,1

Janusz Limon,3 Paweł Stankiewicz,1 James R. Lupski,1,4,5 and Jolanta Wierzba6*1Department of Molecular and Human Genetics, Houston, Texas

2Department of Physiology, Medical University of Gdansk, Poland3Department of Biology and Genetics, Medical University of Gdansk, Poland

4Department of Pediatrics, Baylor College of Medicine, Houston, Texas5Texas Children’s Hospital, Houston, Texas

6Department of Pediatrics, Hematology, Oncology and Endocrinology, Medical University of Gdansk, Poland

Received 10 March 2006; Accepted 6 April 2006

Cornelia de Lange syndrome (CdLS) is a multisystemcongenital anomaly disorder characterized by prenatal andpostnatal growth retardation, developmental delay, distinc-tive facial dysmorphism, limb malformations, and multipleorgan defects. Mutations in the NIPBL gene have beendiscovered recently as a major etiology for this syndrome,and were detected in 27–56% of patients. Two groups havefound significant differences in the severity or penetrance ofsome phenotypes between mutation positive and mutationnegative patients. Different clinical features have also beendescribed among patients with missense versus truncatingmutations. In this study, we identified 13 NIPBL mutations in28 unrelated Polish CdLS patients (46.4%), 11 were novel.Mutation positive patients were more severely affected incomparison to mutation negative individuals with respect toweight, height, and mean head circumference at birth, facial

dysmorphism and speech impairment. Analyses of com-bined data from this and the two previous studies revealedthat the degree of growth, developmental delay and limbdefects showed significant differences between patients withand without mutations and between patients with missenseand truncating mutations, whereas only a portion of thesefeatures differed significantly in any individual study.Furthermore, bioinformatic analyses of the NIPBL proteinrevealed several novel domains, which may give furtherclues about potential functions of this protein.� 2006 Wiley-Liss, Inc.

Key words: CdLS; NIPBL; genotype–phenotype correlation;calmodulin-binding domains

How to cite this article: Yan J, Saifi GM, Wierzba TH, Withers M, Bien-Willner GA, Limon J, Stankiewicz P,Lupski JR, Wierzba J. 2006. Mutational and genotype–phenotype correlation analyses in 28 Polish patients with

Cornelia de Lange syndrome. Am J Med Genet Part A 140A:1531–1541.

INTRODUCTION

Cornelia de Lange syndrome (CdLS;OMIM122470)is a multiple congenital anomaly condition withcharacteristic facial features, pre- and post-natalgrowth delay, mental retardation, limb defects,behavioral problems, ocular and hearing impair-ments, and gastrointestinal, genitourinary and car-diac abnormalities [Ireland et al., 1993; Jackson et al.,1993]. The typical facial features include synophrys,long eyelashes, long philtrum, thin upper lip,depressed nasal bridge with anteverted nares and a

downturned mouth. Marked phenotypic variabilityhas been documented with some patients having a

Jiong Yan and Gulam Mustafa Saifi contributed equally.Grant sponsor: National Institute of Child Health and Human

Development; Grant number: PO1 HD39420.*Correspondence to: Jolanta Wierzba, M.D., Ph.D., Department of

Pediatrics, Hematology, Oncology and Endocrinology, Medical Uni-versity of Gdansk, Debinki Str. 7, 80-211, Gdansk, Poland.E-mail: [email protected]

DOI 10.1002/ajmg.a.31305

readily recognizable phenotype while others presentwith mild manifestations, classified as classical andmild type respectively [Ireland et al., 1993; Jacksonet al., 1993; Van Allen et al., 1993; Allanson et al.,1997].

Recently, NIPBL has been identified as a gene res-ponsible for CdLS. The exact function of this genein mammals is unknown. It is homologous toDrosophila Nipped-B gene and Saccharomycescerevisiae sister chromatid cohesion gene, Scc2,and consists of 47 exons encoding for 2,804 or2,697 amino acids [Krantz et al., 2004; Tonkin et al.,2004]. Different types of heterozygous mutationshave been identified including missense, splice site,nonsense and frameshift, with the mutation detec-tion rates ranging from 27 to 56% [Borck et al., 2004;Gillis et al., 2004; Krantz et al., 2004; Tonkin et al.,2004; Bhuiyan et al., 2005; Miyake et al., 2005]. It hasbeen suggested that there are significant phenotypicdifferences between NIPBL mutation positive andmutation negative CdLS patients, and there is a trendthat for some clinical features individuals withmissense mutations have a milder phenotype thanthose with truncating mutations [Gillis et al., 2004;Bhuiyan et al., 2005].

We identified 13 mutations, including 11 novelones, in 28 sporadic Polish CdLS patients. Missense,nonsense, and frameshift mutations were detected,but the majority (8/13) were frameshift changes.Genotype–phenotype correlations have beenassessed for the patients described herein as well asafter combining our data with those reported byGillis et al. [2004] and Bhuiyan et al. [2005]; thecombined data revealed additional genotype/phe-notype correlations. We also identified several newdomains in NIPBL by bio-informatic analyses.

MATERIALS AND METHODS

Subjects and Clinical Evaluation

Patients were recruited through the CdLS Associa-tion Poland. All patients were examined by a clinicaldysmorphologist experienced with CdLS features,and each patient completed an extended diagnosticprogram. Evaluations consisted of clinical history,family history, physical examinations includingcardiac, neurologic, endocrinologic, gastroenterolo-gic, ophthalmologic, and otolaryngologic examina-tions. Electrocardiogram, echocardiogram, completegastro-intestinal X-ray with small bowel followthrough and esophageal pH study were performedin the patients. Abdominal ultrasonography wasemployed to detect gastrointestinal defects espe-cially for young children in whom the regurgitationwaves were visible, as well as to assess urogenitaldisorders. Endoscopy was performed in 15/28 cases.Brainstem auditory evoked potential examinationand tympanometric examination were carried out in

all the patients to detect hearing loss. All individualswith a classic phenotype and three with a mildphenotype had an EEG. Five patients had a skeletalsurvey (babygram) in the neonatal period and 23 hadradiographic documentation (hip and upper limb;vertebral column was included for older children).Psychological testings aimed to reveal autoaggres-sive and generally hyperactive behavior, stereotypicand autistic features were implemented. All caseswere sporadic. Studies were approved by theInstitutional Review Boards of the host institutions.

Phenotype Stratification

Patients were classified as having a classical or mildphenotype according to generally accepted criteria[Preus and Rex, 1983; Jackson et al., 1993; Kline et al.,1993; Van Allen et al., 1993; Allanson et al., 1997].Birth weight, height and head circumference werecompared to standard curves and to the standards forthe gestational age. Postnatal height and headcircumference were compared to the standardgrowth curve for individuals with CdLS [Kline et al.,1993].

Postnatal growth was classified according to thecentiles of CdLS growth curves [Kline et al., 1993]:>c75 (above 75th centile on CdLS growth curves),c25-75, and <c25. Limb malformations were classi-fied according to Gillis et al. [2004]: class I (mild)—noreduction defect; class II (moderate)—partial reduc-tion defect/oligodactyly (>2 digits of each hand);and class III (severe) severe reduction defect/oligodactyly (<¼ 2 digits on either hand). Develop-ment was also classified according to Gillis et al.[2004]. In class I (mild) motor milestones weredelayed at less than 2 years compared to the normalstandards, with development of speech and com-munication skills in older individuals. Class II(moderate) was defined if delay in reaching motormilestones exceeded 2 years behind normal devel-opmental standards and limited speech and com-munication was developed. In class III (severe) aprofound delay in achieving motor milestones with alack of meaningful communication was found.

Mutation Analysis

DNA was isolated from peripheral blood lym-phocytes using QIAamp DNA Blood Mini Kit(Qiagen, Valencia, CA) and Puregene DNA isolationkit (Gentra Systems). Fifty-three pairs of primerswere designed with ExonPrimer (http://ihg.gsf.de/ihg/ExonPrimer.html) and Primer3 (http://frodo.wi.mit.edu/cgi-bin/primer3/primer3_www.cgi) toamplify the NIPBL coding region (cDNA sequence:NCBI NM_133433 and NM_015384; genomicsequence: NT_006576). Sequences of these primersare available at the laboratory website (http://www.imgen.bcm.tmc.edu/molgen/lupski/). PCR was

1532 YAN ET AL.

American Journal of Medical Genetics Part A: DOI 10.1002/ajmg.a

performed following the manufacturer’s instructions(Qiagen, HotStarTaq) with annealing temperaturesranging from 53 to 608C. PCR products were purifiedwith Exo/SAP (Amershan). Either BigDye Primer orBigDye terminator chemistry (Applied Biosys-tems, Foster City, CA) was used to perform thesequencing on a 377 DNA sequencer or a 3730 DNAsequencer. The chromatograms were analyzedusing Sequencher (Gene Codes). Mutations wereconfirmed with the loss or creation of restrictionenzyme sites in mutant chromosomes, identifiedusing NEB Cutter (http://tools.neb.com/NEBcut-ter2/index.php) and PCR Designer (http://cedar.genetics.soton.ac.uk/public_html/primer.html). Theconservation analysis of amino acids was performedwith BLAST (NCBI) and Clustalw (http://www.ebi.ac.uk /clustalw/index.html).

Genotype–Phenotype Correlations

Genotype–phenotype correlations were analyzedusing contingency—table analysis performed forthe defined categories (mild, moderate, severe, orclassic vs. mild) and were focused on the presenceversus absence of mutation in NIPBL and on thenature of mutations: missense versus truncatingmutations. Fisher exact test and Chi-square test wereemployed for the analysis of noncontinuous fixedvariables and were performed for our data aloneor with the combined data set, respectively. Forcomparison of continuous variables (mother’s age,gestation, Apgar score, weight and height at birth,head circumference) between mutation positiveand mutation negative groups, Student t-test wasused after testing for a normal distribution withShapiro–Wilk test. Significant differences wereaccepted at P< 0.05.

NIPBL Protein Analysis

Motifs and domains were identified in the NIPBLprotein using a number of web-based prog-rams: BLOCKS: http://blocks.fhcrc.org; CDD: http://www.ncbi.nlm.nih.gov/Structure/cdd/cdd. shtml;COILS: http://www.ch.embnet.org/software/COILS_form.html; ELM: http://elm.eu.org; ExPASy: http://us.expasy.org;HEAT:http://www.embl-heidelberg.de/�andrade/papers/rep/search.html; InterPro: http://www.ebi.ac.uk/interpro; MotifScan:http://myhits.isb-sib.ch/cgi-bin/motif_scan; PESTFind: https://emb1.bcc.univie.ac.at/content/view/21/45/; Pfam:http://www.sanger.ac.uk/Software/Pfam/; PRINTS:http://www.bioinf.man.ac.uk/dbbrowser/PRINTS/;PROSITE:http://us.expasy.org/prosite/; PSORT:http://www.psort.org/; SMART: http://smart.embl-heidelberg.de/; TargetP: http://www.cbs.dtu.dk/services/TargetP/;TMPRED:http://www.ch.embnet.org/software/TMPRED_form.html

RESULTS

Clinical Phenotype

Twenty-eight CdLS patients were studied, 21 ofwhich had a classical phenotype including all 13patients with NIPBL mutations (Fig. 1). The averageage was 10.9 years (range 3–27). Five parents(17.9%) reported problems with pregnancy includ-ing bleeding in the first trimester and oligohydram-nios, and five experienced previous spontaneousabortions or stillbirth. The major clinical featureswere similar to those that have been reportedpreviously and are summarized as Table I.

Mutation Analysis

Sequencing exons 2–47 and the flanking intronicregions of the NIPBL gene in all 28 patients revealed13 (46%) disease associated mutations (Table IIA).All the mutations were private and consisted of eightframeshift, three nonsense (1 amber UAG, 1 ochreUAA, and 1 opal UGA), 1 in-frame deletion of threenucleotides, and one missense mutation. Mutationswere further confirmed with the loss or creation ofrestriction enzyme sites (Fig. 2 and data not shown).Only one mutation occurred at a CpG dinucleotide,and all nucleotide replacement changes were trans-versions. Five of the eight frameshift mutations andthe in-frame deletion occurred in short nucleotiderepeats of at least three repeatingunits ranging in sizefrom three to nine bases in length (Table IIB). Onlytwo of the mutations we identified (c.5167C>CþTand c.6653-6655delATA) have been previouslyreported [Borck et al., 2004; Gillis et al., 2004]. Themutations were distributed throughout the gene withonly two exons having more than one mutation. Themutations were not detected in 100–112 controlindividuals and were all documented to be de novoexcept for one patient (c.6475G>GþT) in whomparental DNAs were not available. Mutations cau-sing premature truncation of proteins (i.e., frameshiftor nonsense alleles) account for 11 of the 13 chan-ges identified. These most likely represent loss-of-function mutations due to nonsense mediateddecay.

In the absence of a functional assay, the patho-genicity of the missense (p.D246G ) and the in-framesingle amino acid deletion (p.2218delN) mutationsare challenging to assess. We initially evaluated thepathogenicity of the missense mutation (p.D246G)by Granthams’ index for amino acid substitu-tions [Grantham, 1974]. Our analysis suggests thatthis change is potentially pathogenic. Additionally,pathogenic mutations are often associated withprotein regions that are conserved between ortho-logs and paralogs [Miller and Kumar, 2001]. Wesubsequently studied the amino acid conservation ofregions surrounding these nontruncating point

MUTATION ANALYSIS IN POLISH CdLS PATIENTS 1533


mutations. D246 is conserved in all species examinedexcept for Danio rerio, in which it is replaced by aconservative amino acid, glutamic acid (Fig. 3A). Insuch instances it has been suggested to take theGrantham value for the most diverse amino acidchange across species and compare it with that of themutation [Abkevich et al., 2004]. Our results showthat the chemical difference for the cross-specieschange is much smaller and lies in the rangeobserved with polymorphisms as compared to thatshown by the D246G change. N2218 that is deletedby the in-frame deletion mutation, is conserved in allspecies examined (Fig. 3B). It should be noted that inhumans, thenext aminoacid is alsoNandwould thustake the place of N2218. However, at the positioncorresponding to human 2219, there is always anamino acid across species—never a space in thisposition in the alignment—suggesting that thepresence of an amino acid at this position is crucial.Thus, bioinformatic analyses also suggest that thesetwo amino acid mutations are potentially patho-genic. Eight polymorphisms have been identified, inwhich six were formerly reported [Borck et al., 2004;Gillis et al., 2004] (Table III). The remaining two wereboth inherited from an unaffected parent, with oneSNP residing in an exon.

Genotype–Phenotype Correlations

Gillis et al. [2004] evaluated the associationbetween the genotype and the severity of the CdLSphenotype including limb defects, growth anddevelopment. They found that patients with NIPBLmutations had a significantly more severe phenotypefor growth and development. The severity of limbdefects had a similar trend although it was notstatistically significant. When patients with missenseversus nonsense and frameshift mutations werecompared, the latter showed a more severe pheno-type in all categories except possibly in growthretardation. Bhuiyan et al. [2005] failed to findsignificant differences between themutationpositiveand mutation negative patients and between patientswith missense and truncating mutations for thesethree features. However, they found that patientswith classical facial features were more likely to havea mutation and there were other significant findingsbetween the mutation positive patients and mutationnegative individuals including the presence of adepressed nasal bridge, widely spaced teeth, a shortneck, gastro-esophageal reflux, and a squint.

We performed a similar correlation analysis for ourdata, as well as for the combined data set including

FIG. 1. Photographs of CdLS patients. Mutation positive patients were more likely to have a classical phenotype compared to those without a mutation. Shown hereare patients having a classical phenotype with a frameshift mutation (A), a nonsense mutation (B), a missense mutation (C), without a mutation (D) and a mutationnegative patient with a mild phenotype (E).

1534 YAN ET AL.


those from Gillis et al. and Bhuiyan et al. (Table I). Inour study, patients with mutations had significantlymore severe prenatal growth retardation (lowweight, height, and head circumference at birth)compared with those without mutations. They werealso more likely to have a classical facial phenotype(Fig. 1), in fact all of them had the facial type, and hada more severe speech defect. Other significantfindings for the patients carrying mutations includeda specific narrowing of the bitemporal diameter andpresence of anteverted nostrils. Hyperactivity andscoliosis were observed more often in the mutationnegative patients. Limb defects, growth and devel-opment did not show significant difference in ourstudy, but when the combined data were analyzed,patients with NIPBL mutations manifested a signifi-cantly more severe phenotype. Since the number ofmissense mutations was small in our study, only

combined data were analyzed for comparisonbetween patients with missense and truncatingmutations. Again, all three features, limb defects,delayed growth and development, showed signifi-cant differences, with patients carrying truncatingmutations having a more severe phenotype.

NIPBL Protein Analysis

We analyzed NIPBL protein sequence usingseveral web-based programs for the presence ofknown protein domains and motifs. Our analysis issummarized in Table IV and Figure 4. We identified acaldesmon domain extending from residues 489–1025. This domain is found in Caldesmon 1, acalmodulin- and actin-binding protein that plays avital role in the regulation of smooth muscle andnonmuscle contraction and cytoskeleton formation

TABLE I. Major Features of the Patients and Phenotype–Genotype Correlationsa

Feature

This study Combinedb Combinedb

Mþ M� Mþ M� Missense Truncating

Number of patients 13 15 93 94Gender (M:F) 5:8 10:5 19:18 16:14Mean weight at birth (g) 1983* 2533Mean height at birth (cm) 46* 50.6Mean head circumference at birth (cm) 29.1* 30.7Growth>c75 0 4 9* 24 5* 4c25-P75 9 10 36 31 9 27<c25 4 1 29* 8 3* 26

Limb malformationNo reduction (I) 10 12 61 76 19* 42Partial reduction (II) 1 1 6 10 1 5Severe reduction (III) 2 2 20* 10 0* 20

Developmental delayMild (I) 0 4 4* 16 3* 1Moderate (II) 5 4 27 36 8 19Severe (III) 8 7 57* 38 8* 49

FaceClassical type 13* 8 33* 15 5 28Mild type 0 7 4 11 2 2

SpeechAbsent 10 7Words 3 3Sentences 0* 5

Behavioral problemsAutoaggression 9 9Stereotype 11 12Hyperactivity 7* 14Autism 7 6 19 16Ptosis and ocular abnormalities 7 8Hearing Impairment 6 4Heart defects 6 4Urological abnormalities 5 3Genital abnormalities (boys) 5 10Gastrointestinal reflux 11 10

Mþ, mutation positive; M�, mutation negative.aExcept for the mean weight, height and head circumference at birth, all the numbers in the table are the number of patients. The total numbers of patients for eachcategory for the combined data are different because not every patient was included for comparison in the Bhuiyan and Gillis’ study and some features compared inthese two studies were different.bData combined this study with Bhuiyan et al. [2005] and/or Gillis et al. [2004].*The percentage of patients with the clinical feature on the left is significantly different between either Mþ and M� or missense and truncating mutations (P< 0.05).



TABLE IIA. Summary of NIPBL Mutations

Mutation* Location Type Amino acid change Inherited or de novo

c.416insTTAT Exon 5 Frame-shift p.S139FfsX11 de novoc.737A>AþG Exon 7 Missense p.D246G de novoc.2321_2322delAG Exon 10 Frame-shift p.E774EfsX2 de novoc.2920_2921delAA Exon 10 Frame-shift p.R974EfsX17 de novoc.3526insA Exon 13 Frame-shift p.E1176RfsX8 de novoc.5167C>CþTa Exon 26 Nonsense p.R1723X de novoc.6475G>GþT Exon 37 Nonsense p.E2159X NAc.6653-6655delATAb Exon 39 in-frame p.2218delN de novoc.6874insT Exon 40 Frame-shift p.H2292SfsX46 de novoc.7096C>CþT Exon 42 Nonsense p.Q2366X de novoc.7438-7439delAG Exon 44 Frame-shift p.R2480KfsX3 de novoc.7559delA Exon 44 Frame-shift p.N2520IfsX1 de novoc.7949insT Exon 46 Frame-shift p.I2650IfsX9 de novo

NA, parents not available.aReported by Gillis et al. [2004].bReported by Borck et al. [2004].*Mutation nomenclature as per den Dunnen and Antonarakis, [2000].

FIG. 2. NIPBL mutation detection and confirmation with restriction enzyme digestion. The results for two mutations (c.5167C>T and c.7559delA) are shown.Individuals are labeled with four-digit identification number. Because of c.7559delA, one of the two ApoI sites in the PCR products is lost in the patient but not in theparents. The c.5167C>T results in the loss of a BslI site (CCNNNNNNNGG) in the patient. M, marker; UD, undigested. [Color figure can be viewed in the online issue,which is available at www.interscience.wiley.com.]

1536 YAN ET AL.


[reviewed in Gusev, 2001]. The caldesmon domain isknown to bind to tropomyosin, myosin and phos-pholipids. Caldesmon 1 is a potent inhibitor of theactin-tropomyosin activated myosin Mg ATPase andserves as a mediating factor for calcium-dependentinhibition of smooth muscle contraction [Marstonand Redwood, 1993; Chalovich et al., 1998]. Wefurther identified two calmodulin binding helicalmotifs—one at each end of the caldesmon domain.Calmodulin is the archetype of calcium-modulatedproteins whose functions include roles in growth,cell cycle, signal transduction and the synthesis andrelease of neurotransmitters (OMIM, Calmodulin).Calmodulin regulates L-type calcium channels intransduction of local calcium entry into diversesignaling pathways [Mori et al., 2004]. A calponindomain was identified within the caldesmon domainin NIPBL. Calponin protein is implicated in the

regulationof smooth muscle contraction [Szymanski,2004]. Calponin can bind to actin, calmodulin,troponin C, and tropomyosin. The interaction ofcalponin with actin inhibits actomyosin Mg ATPaseactivity, similar to caldesmon 1. A calponin domain isalso found in smooth muscle protein (SM22) andDrosophila synchronous flight muscle protein.

NIPBL contains a nuclear localization signal(amino acid residues 1108–1125), and its intracel-lular localization is predicted to be nuclear [Nair andRost, 2005]. PSORT predicts it to be nuclear orcytonuclear and TargetP predicts it to be a signalingmolecule. If it has to function as a signaling molecule,then it should have a nuclear export signal (NES) toofor it to exit from the nucleus [la Cour et al., 2004]. Inaccordance with this, we find a predicted NES fromresidues 1578–1610 (Table IV, Fig. 4). Further, wefind an RGD tripeptide (residues 834–836) that isfound in fibronectin and is crucial for its interactionwith its cell surface receptor, an integrin [Ruoslahtiand Pierschbacher, 1986; D’Souza et al., 1991]. Takentogether, these observations suggest that NIPBL is apotential signaling molecule that shuttles betweenthe nucleus and cell surface.

We identified an 11 amino acid repeat, [SG]RPETPKQKG[ED], that is present 5–6 times inNIPBL(Fig. 4). This repeat is an extended variant of the oneidentified recently [Strachan, 2005]. The repeat isconserved in NIPBL homologs. It consists of a core of7 amino acids (the ones in bold) that are absolutelyconserved in NIPBL. The eight amino acid core,RPETPKQK, is found in multiple copies in mouse,rat, cat, and chick whereas the seven amino acidcore, PETPKQK is also seen in multiple copies inXenopus. The pentapeptide core, PETPK, isobserved in multiple copies in FALZ, a humanbromodomain PHD finger transcription factor andin collagen-like surface proteins of several Strepto-cocuus species. The presence of this repeat and itssmaller variants in multiple copies separated by gapsof similar lengths in different proteins and acrossdiverse species such as bacteria and mammalssuggests its functional importance. This is substan-tiated by the evidence that FALZ is highly similar tothe largest subunit of the Drosophila NURF (nucleo-some remodeling factor) complex, which catalyzes

TABLE III. Summary of NIPBL Polymorphisms

Polymorphism Location

c.535G>AþG Exon 6c.610þ102A>AþG Intron 6a

c.2021A>AþG Exon 10a

c.3575-17A>AþG Intron 13b

c.4560þ77A>AþG Intron 21a

c.6764-35C>CþG Intron 39a

c.6955-9delT Intron 40a

c.7410þ81delCTT Intron 43

aReported by Gillis et al. [2004].bReported by Borck et al. [2004].

TABLE IIB. Nucleotide Repeats at the Sites of Mutations

Mutation Repeats

c.2321_2322delAG AGAG*c.2920_2921delAA AAAc.3526insA AAAAAAc.6653-6655delATA ATAATAATAc.6874insT TTTTTTc.7438-7439delAG AGAGAGc.7559delA AAAc.7949insT TT*

*Not counted as repeat unit.

FIG. 3. Conservation of amino acids changed by c.737A>G, p.D246G (A)and c.6653_6655delATA, p.2218delN (B) across different species. Hs, Homosapiens; Mf,Macaca fascicularis; Cf, Canis familiaris; Bt, Bos taurus; Mm,Musmusculus; Rn, Rattus norvegicus; Gg, Gallus gallus; Xl, Xenopus laevis; Tn,Tetraodon nigroviridis; Dr, Danio rerio. [Color figure can be viewed in theonline issue, which is available at www.interscience.wiley.com.]



nucleosome sliding on DNA and interacts withsequence-specific transcription factors thereby facil-itating chromatin remodeling that is required fortranscriptional regulation.

Two potential PEST sequences were found inNIPBL (Table IV, Fig. 4). Their high PEST-positivescores (>5) indicate their likely functional signifi-cance. PEST sequences are found in several calmo-dulin binding proteins that include enzymes,transcription factors and structural proteins [Barnesand Gomes, 1995]. These proteins are susceptible toproteolysis by calpains, which are Ca(2þ)-depen-dent proteases and regulate their targets by limitedproteolysis [Kakkar et al., 1999]. Another domainidentified includes a DNA binding domain,NUMOD3, which is found in homing endonucleases.This DNA binding affinity may be bringing abouttarget specificity that is essential for it to act as atranscription regulator.

DISCUSSION

Comparable to what has been previously reported,we identified mutations in 46% (13/28) Polishpatients with sporadic CdLS [Gillis et al., 2004, 47%;Tonkin et al., 2004, 50%; Bhuiyan et al., 2005, 56%].The 13 mutations included eight frameshift, threenonsense, one in-frame deletion of three nucleo-tides, and one missense mutation, of which 11 were

novel. The frequency of truncating mutations (85%)in our cohort was higher than reported (75%)[Strachan, 2005]. The mutations were distributedover the gene without obvious clustering. We did notsystematically examine for rearrangement muta-tions, but none has been reported in previous studiesthat specifically assayed for them by FISH. In half ofthe CdLS patients, we failed to detect coding regionmutations. The NIPBL regulatory mutations were notevaluated, but genetic heterogeneity is also anappealing explanation.

In a genome-wide linkage exclusion analysis, threecandidate regions, 2q37, 10p13, and 14q24, were notexcluded [Krantz et al., 2004]. Chromosome rearran-gements involving other genomic regions have alsobeen detected in CdLS patients, including regions onchromosomes 1, 3, and 18 [Borck et al., 2004;DeScipio et al., 2005].HEAT repeats, protein–proteininteracting motifs that are found in condensins,cohesions and other complexes with chromosome-related functions, have been identified in NIPBL[Krantz et al., 2004; Strachan, 2005]. A meta-analysisof all known missense mutations [Strachan, 2005;Fig. 4 this study] shows a clustering of such mutationsin and around the HEAT repeats. NIPBL is homo-logous to Saccharomyces cerevisiae sister chromatidcohesion gene, Scc2, which forms a complex withSCC4 and is required to load the cohesin complex tochromosomes [Ciosk et al., 2000; Krantz et al., 2004;

TABLE IV. Motifs, Conservation and E-values

Motif Name

Consensus sequence

E- value/score/confidenceMotif sequence (bold implies identity/similarity with consensus)

Calmodulin binding peptide 1 . . .[SACLIVTM]..[ILVMFCT]Q.{3}[RK].{4,5}[RKQ].. 0.001ERFSKEVQDKDKPLKKRKQD

Calponin . . .[SACLIVTM]..[ILVMFCT]Q.{3}[RK].{4,5}[RKQ].. 0.45GGATGGNRPASQETGSTGNGSRPALM

11 aa repeat srPETPKQKge 0.053srPETPKQKge

Asr HHKK[qa]HAK[AK]A[Pa]QKAQAAKK 2HDNRRDSGKPSTEKKPEVSKH

Cell attachment sequence RGDRGD

Calmodulin binding peptide 2 . . .[SACLIVTM]..[ILVMFCT]Q.{3}[RK].{4,5}[RKQ].. 0.018KLSLDDVQKLIKDREDKSR

NLS [KR]{2}X{10,12}[KR]{3} 84% confidenceKKDDDKAWEYEERDRRSS

PEST 1 RSPSDSDMEDYSPPPSLSEVAR 12.97 (PESTfind score)Octapeptide P[Gk]KED[NG]NK 7.8

NAEEDSNKNES XXXXE[EL]LLXL[ESL]XX[LM]X[LMIV]X{6}SX[SLE]X{6}[SE][SL] 0.621 (NES score)

WPAAELLLSLLGRLLVHQFSNKSTEMALRVASLHEAT 1 AQSFDIYLTQILRVLGENAIAVRTKAMKCLSEVVAVDPS 0.0008HEAT 2 PQLAEQYYDMLIERILDTGISVRKRVIKILRDICIEQPT 7.4e-6HEAT 3 YDWFEQLLQNLLKSEEDSSYKPVKKACTQLVDNLVEHILK 1.4e-5HEAT 4 VNLKIQVLKNLQTYLQEEDTRMQQADRDWKKVAKQEDLKEM 7.0e-6HEAT 5 LLPVLLELLKDEDPEVRLAAAEALSELAEVLGEEXVEEH 4.7e-5NUMOD3 X[Gn]kX[HL][ST][Ed]E[Ts]KXk[Mil][Sr] 0.012-0.0001

ANSKITEEVKRSIVPEST 2 HLDPDEEEEEGEVSASTNAR 16.34 (PEST find score)

A PEST score of 5 or greater is considered as a very good indicator of PEST potential.

1538 YAN ET AL.


Tonkin et al., 2004]. The Drosophila homolog,Nipped-B, is also required for sister chromatidcohesion and regulates cut, ultrabithrox, and Notchsignaling through a postulated architectural role infacilitating enhancer-promoter interaction [Rollinset al., 1999; Rollins et al., 2004]. There are potentiallya large number of genes whose products may beinteractingwithNIPBL that can act as eithermodifiersor potential disease genes for CdLS.

Because of the clinical heterogeneity of CdLS, Gilliset al. [2004] evaluated the associations between theNIPBLmutations and the severity of thephenotype interms of growth retardation, developmental delay,and limb defects, and found that mutation positivepatients tend to have a more severe phenotypealthough the difference was not statistically signifi-cant for the limb defects, and patients with truncatingmutations also had a more severe phenotype exceptfor the growth retardation. These features failed toshow significant differences in the study by Bhuiyanet al. [2005]. However, they evaluated other featuresand some of them showed significant differences

including the facial phenotype. Our study also failedto detect differences in terms of growth retardation,developmental delay, and limb defects, but mutationpositive patients had a more severe phenotype inprenatal growth, facial dysmorphism, and speechdelay. When we combined the data from these threestudies, mutation positive individuals as well aspatients with truncating mutations demonstrated astatistically more severe phenotype in growthretardation, developmental delay, and limb defects.It seems when more data are compiled, genotype–phenotype correlations become more obvious,although these correlations may not be the casewhen an individual patient is considered.

Other than the HEAT repeats, our protein domainanalysis of NIPBL has revealed a nuclear localizationsignal, a NES and an RGD tripeptide, indicating thatNIPBL is a potential signaling molecular. The findingof a DNA binding domain, NUMOD3, is consistentwith the potential role of NIPBL in the chromosome-related functions. The analysis also revealed acaldesmon domain, a calponin domain and two

FIG. 4. NIPBL protein sequence analysis. Different domains and their locations are shown in the horizontal bar representing the linear protein. Amino acid sequencesfor theNIPBL specific repeats are shown in thebox at the lower right. Publishedmissensemutations are shown byfilledblack circles above the proteinbar [Bhuiyan et al.,2005;Miyake et al., 2005; Strachan, 2005]. The twomissensemutations reportedherein are eachmarkedwith a vertically downarrow (#).Note that allmissense mutationscluster around the HEAT repeats or immediately after the NLS. The specific conservation of each motif (i.e., homology to consensus sequence) and statistical confidenceare shown as Table IV. Individual motifs identified and purported functions are delineated in the text. [Color figure can be viewed in the online issue, which is available atwww.interscience.wiley.com.]



calmodulin-binding motifs. In smooth muscle, cal-modulin is activated by Ca(2þ). Activated calmodu-lin can bind with caldesmon or calponin preventingthem from inhibiting actomysin ATPase and thusresulting in muscle contraction [Marston and Red-wood, 1993; Walsh, 1994]. Alternatively, activatedcalmodulin results in the phosphorylation of caldes-mon and calponin and alleviates their inhibitoryeffect on muscle contraction [Pritchard and Moody,1986; Winder and Walsh, 1993]. We postulate thatNIPBL, like caldesmon/calponin, is modulated bycalmodulin. Decreased levels of caldesmon andcalponin are associated with smooth muscle con-tractility [Tani and Matsumoto, 2004]. We thereforehypothesize that a decrease in NIPBL may beassociated with increased smooth muscle contrac-tility. This potentially could explain gastrointestinalhyperactivity which is the most frequent signassociated with gastroesophageal reflux (GER),present in 23/28 (85%) of GER-positive CdLS patients[Luzzani et al., 2003].

Our findings add to the body of evidence showingNIPBL mutations cause CdLS and further supportpotential genotype–phenotype correlations. Never-theless, the absence ofNIPBLmutation in over half ofthe patients reported in the literature suggests thatthere may be another gene(s), in which mutation canresult in a CdLS phenotype.

ACKNOWLEDGMENTS

We thank the patients and their families for theirgenerous co-operation and the Polish Cornelia deLang Association for their support. This work wassupported in part by a grant from the NationalInstitute of Child Health and Human Development(PO1 HD39420).

REFERENCES

Abkevich V, Zharkikh A, Deffenbaugh AM, Frank D, Chen Y,Shattuck D, Skolnick MH, Gutin A, Tavtigian SV. 2004.Analysis of missense variation in human BRCA1 in the contextof interspecific sequence variation. J Med Genet 41:492–507.

Allanson JE, Hennekam RCM, Ireland M. 1997. De Langesyndrome: Subjective and objective comparison of theclassical and mild phenotypes. J Med Genet 34:645–650.

Barnes JA, Gomes AV. 1995. PEST sequences in calmodulin-binding proteins. Mol Cell Biochem 149(150):17–27.

Bhuiyan ZA, Klein M, Hammond P, van Haeringen A, MannensMMAM, Van Berckelaer-Onnes I, Hennekam RCM. 2005.Genotype–phenotype correlations of 39 patients with Cor-nelia de Lange syndrome: The Dutch experience. J Med Genet10.1136/jmg.038240 [Epub ahead of print].

Borck G, Redon R, Sanlaville D, Rio M, Prieur M, Lyonnet S,Vekemans M, Carter NP, Munnich A, Colleaux L, Cormier-Daire V. 2004. NIPBL mutations and genetic heterogeneity inCornelia de Lange syndrome. J Med Genet 41:e128.

Chalovich JM, Sen A, Resetar A, Leinweber B, Fredricksen RS, LuF, Chen Y-D. 1998. Caldesmon: Binding to actin and myosinand effects on elementary steps in the ATPase cycle. ActaPhysiol Scand 164:427–435.

Ciosk R, Shirayama M, Shevchenko A, Tanaka T, Toth A,Shevchenko A, Nasmyth K. 2000. Cohesin’s binding tochromosomes depends on a separate complex consisting ofScc2 and Scc4 proteins. Mol Cell 5:243–254.

D’Souza SE, Ginsberg MH, Plow EF. 1991. Arginyl-glycyl-asparticacid (RGD): A cell adhesion motif. Trends Biochem Sci16:246–250.

den Dunnen JT, Antonarakis SE. 2000. Mutation nomenclatureextensions and suggestions to describe complex mutations: Adiscussion. Hum Mutat 15:7–12.

DeScipio C, Kaur M, Yaeger D, Innis JW, Spinner NB, JacksonLG, Krantz ID. 2005. Chromosome rearrangements inCornelia de Lange syndrome (CdLS): Report of a der(3)t(3;12) (p25.3;p13.3) in two half sibs with features of CdLSand review of reported CdLS cases with chromosomerearrangements. Am J Med Genet Part A 137A:276–282.

Gillis LA, McCallum J, Kaur M, DeScipio C, Yaeger D, Mariani A,Kline AD, Li H-H, Devoto M, Jackson LG, Krantz ID. 2004.NIPBL mutational analysis in 120 individuals with Cornelia deLange syndrome and evaluation of genotype–phenotypecorrelations. Am J Hum Genet 75:610–623.

Grantham R. 1974. Amino acid difference formula to help explainprotein evolution. Science 185:862–864.

Gusev NB. 2001. Some properties of caldesmon and calponin andthe participation of these proteins in regulation of smoothmuscle contraction and cytoskeleton formation. Biochemistry(Mosc) 66:1112–1121.

Ireland M, Donnai D, Burn J. 1993. Brachmann-de Langesyndrome. Delineation of the clinical phenotype. Am J MedGenet 47:959–964.

Jackson L, Kline AD, Barr MA, Koch S. 1993. de Lange syndrome:A clinical review of 310 individuals. Am J Med Genet 47:940–946.

Kakkar R, Raju RVS, Sharma RK. 1999. Calmodulin-dependentcyclic nucleotide phosphodiesterase (PDE1). Cell Mol Life Sci55:1164–1186.

Kline AD, Stanley C, Belevich J, Brodsky K, Barr M, Jackson LG.1993. Developmental data on individuals with the Brach-mann-de Lange syndrome. Am J Med Genet 47:1053–1058.

Krantz ID, McCallum J, DeScipio C, Kaur M, Gillis LA, Yaeger D,Jukofsky L, Wasserman N, Bottani A, Morris CA, NowaczykMJM, Toriello H, Bamshad MJ, Carey JC, Rappaport E,Kawauchi S, Lander AD, Calof AL, Li H-H, Devoto M, JacksonLG. 2004. Cornelia de Lange syndrome is caused by mutationsin NIPBL, the human homolog of Drosophila melanogasterNipped-B. Nat Genet 36:631–635.

la Cour T, Kiemer L, Mølgaard A, Gupta R, Skriver K, Brunak S.2004. Analysis and prediction of leucine-rich nuclear exportsignals. Protein Eng Des Sel 17:527–536.

Luzzani S, Macchini F, Valade A, Milani D, Selicorni A. 2003.Gastroesophageal reflux and Cornelia de Lange syndrome:Typical and atypical symptoms. Am J Med Genet Part A119A:283–287.

Marston SB, RedwoodCS. 1993. The essential role of tropomyosinin cooperative regulation of smooth muscle thin filamentactivity by caldesmon. J Biol Chem 268:12317–12320.

Miller MP, Kumar S. 2001. Understanding human diseasemutations through the use of interspecific genetic variation.Hum Mol Genet 10:2319–2328.

MiyakeN,Visser R,Kinoshita A,YoshiuraK-I, NiikawaN,KondohT, Matsumoto N, Harada N, Okamoto N, Sonoda T, NaritomiK, Kaname T, Chinen Y, Tonoki H, Kurosawa K. 2005. Fournovel NIPBL mutations in Japanese patients with Cornelia deLange syndrome. Am J Med Genet Part A 135A:103–105.

Mori MX, Erickson MG, Yue DT. 2004. Functional stoichiometryand local enrichment of calmodulin interacting with Ca2þ

channels. Science 304:432–435.Nair R, Rost B. 2005. Mimicking cellular sorting improves

prediction of subcellular localization. J Mol Biol 348:85–100.

1540 YAN ET AL.


Preus M, Rex AP. 1983. Definition and diagnosis of theBrachmann-De Lange syndrome. Am J Med Genet 16:301–312.

Pritchard K, Moody CJ. 1986. Caldesmon: A calmodulin—bindingactin—regulatory protein. Cell Calcium 7:309–327.

Rollins RA, Morcillo P, Dorsett D. 1999. Nipped-B, a Drosophilahomologue of chromosomal adherins, participates in activa-tion by remote enhancers in the cut and Ultrabithorax genes.Genetics 152:577–593.

Rollins RA, Korom M, Aulner N, Martens A, Dorsett D. 2004.Drosophila Nipped-B protein supports sister chromatidcohesion and opposes the stromalin/Scc3 cohesion factor tofacilitate long-range activation of the cut gene. Mol Cell Biol24:3100–3111.

Ruoslahti E, Pierschbacher MD. 1986. Arg-Gly-Asp: A versatilecell recognition signal. Cell 44:517–518.

Strachan T. 2005. Cornelia de Lange Syndrome and the linkbetween chromosomal function, DNA repair and develop-mental gene regulation. Curr Opin Genet Dev 15:1–7.

Szymanski PT. 2004. Calponin (CaP) as a latch-bridge protein—anew concept in regulation of contractility in smooth muscles.J Muscle Res Cell Motil 25:7–19.

Tani E, Matsumoto T. 2004. Continuous elevation of intracellularCa2þ is essential for the development of cerebral vasospasm.Curr Vasc Pharmacol 2:13–21.

Tonkin ET, Wang T-J, Lisgo S, Bamshad MJ, Strachan T. 2004.NIPBL, encoding a homolog of fungal Scc2-type sisterchromatid cohesion proteins and fly Nipped-B, is mutated inCornelia de Lange syndrome. Nat Genet 36:636–641.

Van Allen MI, Filippi G, Siegel-Bartelt J, Yong S-L, McGillivary B,Zuker RM, Smith CR, Magee JF, Ritchie S, Toi A, Reynolds JF.1993. Clinical variability within Brachmann-de Lange syn-drome: A proposed classification system. Am J Med Genet47:947–958.

WalshMP. 1994. Calmodulin and the regulationof smoothmusclecontraction. Mol Cell Biochem 135:21–41.

Winder SJ. Walsh MP. 1993. Calponin: Thin filament-linkedregulation of smoothmuscle contraction. Cell Signal 5:677–686.



Mutational and genotype–phenotype correlation analyses in 28 Polish patients with Cornelia de...

Documents

Transcript of Mutational and genotype–phenotype correlation analyses in 28 Polish patients with Cornelia de...