LBR mutation and nuclear envelope defects in a patient affected with Reynolds syndrome
- Caroline Gaudy-Marqueste1,2,
- Patrice Roll2,3,
- Vera Esteves-Vieira3,
- Pierre-Jean Weiller4,
- Jean Jacques Grob1,
- Pierre Cau2,3,
- Nicolas Lévy2,3,
- Annachiara De Sandre-Giovannoli2,3
- 1Service de Dermatologie, Hôpital Ste Marguerite, Marseille, France
- 2Inserm UMR_S910, Faculté de Médecine la Timone, Marseille, France
- 3Département de Génétique Médicale et Biologie Cellulaire, Hôpital d'Enfants La Timone, Marseille, France
- 4Service de Médecine Interne, Hôpital La Timone, Université Aix-Marseille II, Marseille, France
- Correspondence to Dr Annachiara De Sandre-Giovannoli, Laboratoire de Génétique Moléculaire, Département de Génétique Médicale, Hôpital d'Enfants la Timone, 264 Rue St. Pierre, Marseille 13385, France;
- Received 23 July 2009
- Revised 5 November 2009
- Accepted 9 November 2009
Background Lamins are proteins of the nuclear envelope involved in ‘laminopathies’, an heterogeneous group of diseases sharing clinical similarities with systemic sclerosis (SSc).
Methods In this context, a search was undertaken for mutations in LMNA, encoding Lamins A/C, and ZMPSTE24, LBR, LMNB1, LMNB2, MAN1, SYNE1a and LAP2, encoding Lamins A/C molecular partners, in a Caucasian woman affected with Reynolds syndrome, a particular nosologic entity specifically associating limited cutaneous SSc and primary biliary cirrhosis.
Results Coding regions and intron-exon boundaries of these genes were PCR amplified and sequenced, revealing a single heterozygous missense mutation in LBR exon 9 (c.1114C/T; p.R372C). This variant was absent in 400 control chromosomes. The mutation was predicted to induce a change in Lamin B receptor (LBR) tertiary structure and molecular interactions by bioinformatic tools. Further functional explorations were performed on the patient's fibroblasts and lymphoblastoid cell lines. On the latter, the expression levels of LBR, Lamins A/C, Lamin B1, Lamin B2, and HP1a were conserved. Conversely, in the patient's skin fibroblasts, LBR and the aforementioned molecular partners showed dramatically reduced or abolished expression levels. The immunofluorescence analyses performed on both cell lines corroborated these findings.
Conclusion The fibroblast specific abnormalities observed suggest that this particular LBR mutation might have dominant negative deleterious effects in a tissue specific fashion, possibly through the perturbation of the interactions or stability of the nuclear envelope protein network. LBR mutations might thus be associated with Reynolds syndrome.
- systemic sclerosis
- lamin B receptor
- nuclear envelope
- primary biliary cirrhosis
- Reynolds syndrome
Scleroderma is a rare multisystemic disease of unknown aetiology, characterised by progressive sclerosis of several tissues. Two main forms of the disease can be distinguished: a systemic form, named systemic sclerosis (SSc), which affects both skin and internal organs, and a localised form (LS) confined to skin and subcutaneous tissues. Two further subgroups of disease are individualised among SSc, with limited (lSSc) and diffuse (dSSc) cutaneous forms, depending on the extent of the cutaneous involvement.1 Primary biliary cirrhosis (PBC) is a chronic autoimmune disorder that leads to the progressive destruction of intrahepatic bile ducts. Its diagnosis is made on the basis of three criteria: the presence of antimitochondrial antibodies (AMA), 1.5× elevation of alkaline phosphatase values, and a characteristic hepatic histology. The diagnosis is probable or definite respectively when two or three of these criteria are fulfilled. In 1970, a particular nosologic entity named Reynolds syndrome specifically associating limited cutaneous SSc to PBC was identified.2 3
The existence of familial aggregation4 5 and studies in monozygotic twins6 have supported the evidence of a genetic predisposition to both SSc and PBC, probably involving the interaction of various genes, that leads to the constitution of a genetically receptive host.
Lamins are ubiquitous proteins that polymerise to form the nuclear lamina, a meshwork of intermediate filaments localised under the inner nuclear membrane. Lamins A/C are also major components of the nuclear matrix,7 where they form structures that are less well defined. Two types of Lamins have been described in mammals: A-type (including as major isoforms Lamins A and C), encoded by the LMNA gene through alternative splicing, and B-type (including the major isoforms B1 and B2) respectively encoded by the LMNB1 and LMNB2 genes.8 Beside the fact that they provide shape and mechanical stability to the nucleus, as well as contribute to the mechanical properties of the whole cell through their connection with cytoskeletal proteins,9 10 Lamins are also involved in various fundamental nuclear processes such as mitosis, chromatin organisation, DNA replication and repair, RNA transcription and splicing, and globally, regulation of gene expression patterns through direct or indirect interactions with DNA and a great number of molecular partners (for review see Dechat et al11). The latter include: proteins of the inner nuclear membrane, as the Lamin B receptor (LBR) encoded by the LBR gene, Emerin, some isoforms of Nesprins (encoded by the SYNE1 and 2 genes), some isoforms of the lamina associated polypeptides (LAP) type 1 and 2, MAN1, encoded by MAN1/LEMD3, which interacts with receptor associated-SMADs in the TGF-β/BMP pathway; many transcription factors, as pRb, SREBP, MOK2, OCT-1; proteins of the nuclear pore complexes; chromatin-binding proteins like HP1 and BAF (barrier to autointegration factor) (for review see Zastrow et al12 and Prokocimer et al13).
Mutations of Lamins and some of their partners have been reported to cause different diseases included in the nosologic spectrum of ‘laminopathies’, a heterogeneous group of pathologies involving various tissues in a combined or specific fashion.
Patients affected with certain laminopathies (namely Hutchinson-Gilford Progeria, restrictive dermopathy, atypical Werner syndrome) share clinical similarities with SSc, autoantibodies directed against Lamins A/C have been previously observed in the serum of a patient suffering from linear scleroderma, a form of LS,14 and antibodies against LBR have been described in certain patients affected with PBC.15 On these bases, we screened the coding sequences and intron–exon boundaries of the candidate genes LMNA, ZMPSTE24, LBR, LMNB1, MAN1/LEMD3, SYNE1-alpha and LAP2, respectively encoding Lamins A/C and some of their molecular partners, in a patient associating lSSc and PBC, in the overlap context of Reynolds syndrome. We report in this patient the identification of the first heterozygous LBR missense mutation described in human disease. Functional explorations were subsequently performed in search for Pelger–Huet anomaly (PHA) and other potential alterations in nuclear morphology and composition, allowing us to identify fibroblast specific abnormalities.
Patient and methods
The patient was from a larger cohort of patients previously studied at our laboratory which included three patients with LS and 25 with SSc (14 limited cutaneous forms and 11 diffuse cutaneous forms; for details, see Gaudy-Marqueste et al16). The patient was a Caucasian woman aged 76, suffering from Raynaud's phenomenon since the age of 30 years. When she was 69, she started to complain of asthenia and diffuse arthralgia, while the Raynaud's phenomenon worsened. Subcutaneous nodes diagnosed as erythema nodosum further appeared on her legs. Clinical examination showed telangiectasia of palmar and dorsal faces of both hands without acrosclerosis (figure 1). Biological tests revealed accelerated sedimentation rate (54/99 mm) associated with moderate lymphopenia at 1.29 Giga/l and mild cholestasis (PAL 108 UI/l). Serum protein electrophoresis showed polyclonal hyperglobulinaemia and lymphocyte count showed mild TCD4 lymphopenia. Cholesterol values were normal. Anti-thyroperoxidase autoantibodies were positive, together with antinuclear autoantibodies (1/800 rate), displaying an anti-centromere staining pattern. Other autoimmune markers were present including positive latex reaction (109 UI/ml, normal rates <15), anti-PR3 ANCA (>8 UI/ml positivity rate 3.5), antimitochondrial autoantibodies (AMA, 1/400 rate) and antigastrin IgG autoantibodies. Oesophageal and pulmonary involvement were excluded by respiratory function test, cardiac echography, and oesophageal manometry. Capillaroscopy revealed focal ectasia of capillaries without giant capillaries. Endoscopic examination did not reveal evidence of coeliac disease. The diagnosis of lSSc according to Leroy's classification1 was made on the basis of the clinical and biological assessments. Furthermore, the diagnosis of PBC was made on the basis of the association of AMA and biological cholestasis. The patient was thus subsequently diagnosed as being affected with Reynolds syndrome. Familial history revealed coeliac disease in her son.
Genomic and transcriptional analysis
Patient samples (10 ml of peripheral venous blood) were obtained after informed consent, complying with the ethic guidelines of the institutions involved. Blood samples were used to establish Epstein–Barr virus (EBV) immortalised lymphoblastoid cell lines. Fibroblasts were obtained from a skin biopsy, and cultured in RPMI 1640 medium containing 20% fetal calf serum, 2 mM/ml L-glutamine and 100 UI/ml penicillin–streptomycin (GIBCO BRL, Invitrogen, Life Technologies Corp, Carsbald, California, USA).
DNA extraction and lymphoblastoid cell lines preparation were performed from peripheral blood samples following standard procedures. Polymerase chain reaction (PCR) based amplification and direct sequencing were performed for the LBR, LMNB1, MAN1/LEMD3, SYNE1-alpha and LAP2 genes using conditions and primers available upon request. Denaturing high performance liquid chromatography (DHPLC) (WAVE System 3500 HT® TRANSGENOMIC, San Jose, California, USA) was used as a preliminary screening method for LMNA and ZMPSTE24 sequence variation detection. The patient's samples were loaded together with those of other patients in order to compare the elution profiles. The samples, which were characterised by heteroduplex elution profiles, were further sequenced for variation characterisation, using an ABI-Prism 3130XL (Applied Biosystems, Foster City, California, USA) automatic 16 capillaries sequencer.
Transcriptional analyses (reverse transcriptase (RT)-PCR followed by direct sequencing) were performed on RNA extracted from a lymphoblastoid cell line and a fibroblast cell line issued from a skin biopsy. Total RNA was extracted with Trizol (Roche Diagnostics, Basel, Switzerland) following the company's recommendations and RTs were performed with the M-MLV reverse transcriptase kit (Sigma, Saint Louis, Missouri, USA). Whole LBR cDNA was amplified in overlapping fragments for both tissues cell samples. All chromatograms were visualised and compared to reference sequences with the Sequencher 4.8 software (Gene Codes Corp, Ann Arbor, Michigan, USA).
Semi-quantitative RT-PCR amplification was performed on cDNAs prepared from fibroblast cell lines issued from the patient and a healthy individual (the same used for immunofluorescence and immunoblot) for LMNA, LMNB1, LMNB2, HP1α and two housekeeping genes: HMBS encoding hydroxymethylbilane synthase isoform 1, and TBP encoding TATA box binding protein.
Primers were designed to amplify each of the genes' transcripts between two distant exons with the Primer 3 software (http://frodo.wi.mit.edu/primer3/). The PCR products were then loaded on a 1.5% agarose gel and submitted to electrophoresis.
The GenBank reference sequences used were the following: NM_170707.2 for LMNA, NM_005857.3 for ZMPSTE24, NM_002296.2 for LBR, NM_005573.2 for LMNB1, NM_133650.2 for SYNE1-alpha isoform, NM_014319.3 for LEMD3/MAN1, NM_003276.1 for LAP2α isoform, NM_001032283.1 for LAP2β isoform, NM_032737 for LMNB2, NM_012117 for HP1α, NM_000190 for HMBS and NM_003194 for TBP. The databases used for sequence variation and single nucleotide polymorphism (SNP) identification were: http://www.umd.be, http://www.dmd.nl, http://ncbi.nlm.nih.gov, http://genome.ucsc.edu/. The nomenclature used for the description of sequence variations was the one proposed by the Human Genome Variation Society at http://www.genomic.unimelb.edu.au/mdi/mutnomen/recs.html.
Quantitative PCR amplification
Real-time PCR amplification of DNA extracted from the patient's lymphoblastoid cell line was performed on an ABI 7500 Real-Time PCR System (Applied Biosystems). Primers were designed to amplify three of the five LBR's homozygous sequence variations using Primer Express® software (Applied Biosystems). DNA quality was assessed by spectrophotometry using the A260/A280 ratio (≥1.7–1.9). The reaction was performed using the Applied Biosystems SYBR® Green PCR Master Mix. Threshold cycle values were compared between normal and mutant alleles using the following cycling parameters: 10 min at 95°C (heat activation step); 40 cycles of 15 s at 95°C, 1 min at 60°C. Dissociation curve analyses were performed using the instrument's default settings immediately after each PCR run.
Real time RT-PCR amplification was performed on cDNAs prepared from fibroblast cell lines of both the patient and an healthy control on an ABI 7500 Real-Time PCR System (Applied Biosystems) using a TaqMan® gene expression assay (reference hCG32785) amplifying exons 10 and 11 of LBR. Normalisation of the transcripts' expression levels was obtained by the use of the GAPDH housekeeping gene (reference TaqMan® gene expression assay HCG2005673).
Protein extractions from cultured fibroblasts and nuclear matrix proteins separation from EBV immortalised lymphoblastoid cell lines were performed as previously described.17 18 Total proteins (60 μg/sample) were loaded and separated in 7% Tris-acetate or 10% Bis-Tris precast criterion gels (Biorad, Hercules, California, USA) and transferred on a PVDF membrane (Millipore, Billerica, California, USA). Blots were blocked for 1 h at room temperature in 5% non-fat milk in PBST 0.1% and incubated for 1 h with the following primary antibodies: anti-Lamin A/C, mouse (1:100; JOL2, Chemicon, Millipore, Billerica, California, USA), anti-Lamin B1, rabbit (1:1000; Ab16048, Abcam, Cambridge, UK), anti-Lamin B2, mouse (1:100; sc-56147, Santa Cruz, California, USA), rabbit anti-LBR antibody (1:100; Ab32535, Abcam, Cambridge, UK), HP1α antibody (1:200; 2HP-2G9, Euromedex, Souffelweyersheim, France) and anti-GAPDH, mouse (1:5000; MAB374, Millipore, Billerica, California, USA) as control, both in a blocking solution. After several washes in PBST, the membranes were incubated with HRP conjugated secondary antibodies (1:5000; anti-mouse or anti-rabbit, Bio-Rad). Opti-4CN Detection Substrate Kit (Bio-Rad) was used for detection.
The immunofluorescence protocol employed has been described elsewhere.19 Cultured fibroblasts or EBV immortalised lymphoblastoid cell lines were spread on glass slides by Cytospin (Shandon, Pittsburgh, Pennsylvania, USA). Cells were fixed in 4% paraformaldehyde, 2% sucrose phosphate buffered saline (PBS) for 15 min at room temperature and permeabilised in 20 mM HEPES pH 7.4, 50 mM NaCl, 3 mM MgCl2, 300 mM sucrose and 0.5% Triton X-100 (Sigma-Aldrich, Poole, UK) for 3 min at room temperature. Thereafter, slides were washed in PBS before immunostaining. Primary antibody incubations were performed for 40 min at 37°C at dilutions depending on primary antibodies in PBS supplemented with 2% bovine serum fraction V albumin (BSA) (Sigma-Aldrich, Poole, UK) and followed by washing in PBS. The following primary antibodies were used: anti-Lamin A/C, mouse (1:100; Chemicon, Millipore, Billerica, California, USA), anti-Lamin A, mouse (1:100; Ab8980, Abcam, Cambridge, UK), anti-Prelamin A, goat (1:200; Santa Cruz, California, USA), anti-Lamin B1, rabbit (1:200; Abcam, Cambridge, UK), anti-Lamin B2, mouse (1:100; Abcam, Cambridge, UK), anti-LBR rabbit (1:50; Abcam, Cambridge, UK), anti-HP1α, mouse (1:200; 2HP-2H5, Euromedex, Souffelweyersheim, France) and anti-calreticulin, rabbit (1:100; Stressgen, Assay designs, Michigan, USA). After several washes in PBS, the slides were incubated with secondary antibodies for 20 min at room temperature. Secondary antibodies were obtained from Molecular Probes Company (Life Technologies Corp, Carsbald, California, USA): Alexa 488-conjugated goat anti-mouse (1:400), Alexa 488-conjugated goat anti-rabbit (1:400) and Alexa 594-conjugated donkey anti-goat (1:400). The cells were then washed twice for 10 min in PBS, then incubated with DAPI (Sigma-Aldrich, Saint Louis, Missouri, USA) at 100 ng/ml for 10 min, and finally washed three times for 5 min in PBS. The slides were mounted in FluorSave Reagent (Calbiochem, Merck Chemicals, Nottingham, UK) and observed on a Leica DMR microscope (Leica Microsystems, Wetzlar, Germany) equipped with a CoolSNAP camera (Princeton, Trenton, New Jersey, USA).
Identification of a novel LBR mutation
Sequencing of the coding regions and intron–exon boundaries of the LMNA, ZMPSTE24, LMNB1 and LAP2 genes revealed no potentially pathogenic sequence variant nor any common sequence polymorphism in our patient. Homozygous SNPs were observed in MAN1 (rs10534559) and SYNE1-α (rs2252748, rs2252755, rs2747662, rs2813485, rs2256135 and rs2813566).
Several frequent polymorphisms were also found at the homozygous state throughout the patient's LBR locus: rs4653636 (intron 1), rs1056607 (exon 2), rs1056608 (exon 3), rs2230419 (exon 5) and rs6702433 (intron 7) together with a heterozygous sequence variation in exon 9 (c.1114C>T), which was unreported in public sequence databases (figure 2). The variation was predicted to result in an arginine to cysteine substitution at amino acid 372 (p.R372C, p.Arg372Cys), thus leading to a change from an hydrophilic amino acid (arginine) to an hydrophobic one (cysteine).
The de novo character of the variation could not be tested since the patient's parents were dead, and, given the highly variable expressivity/penetrance observed in SSc and PBC, it was not ethically acceptable to test the asymptomatic son of the patient in a research context. The variant was not found in 400 chromosomes from a control Caucasian population. Additional sequencing of the 5′ promoter sequences of the gene revealed no further sequence variation.
Genomic quantitative PCR amplification of the LBR gene
Genomic quantitative PCR amplification was performed for exons 2 and 3 of the LBR gene, containing two of the five homozygous polymorphic variations, and for exon 9, containing the p.R372C mutation, on DNA extracted from the patient's lymphoblastoid cell lines. The same quantitative profiles were obtained in the patient and controls, excluding the presence of an intragenic LBR constitutional deletion mimicking homozygous sequence polymorphisms (data not shown).
Bioinformatic analysis of the R372C mutation
Submission of the LBR protein sequence (UniprotKB/Swiss-Prot reference number Q14739) to Prosite (at http://www.expasy.ch/prosite/) displayed two sterol reductase family signature sequences located from AA (amino acid) 362 to AA 377 (sterol reductase family signature 1), and from AA 579 to AA 602 (sterol reductase family signature 2), as well as a TUDOR domain (a region of 50 amino acids originally found in the Drosophila Tudor protein, which adopts a particular three dimensional structure allowing methylated histones to be recognised), between AA 4 and 62. An Ergosterol biosynthesis ERG4/ERG24 signature could be identified between AA 205 and 615, making LBR belong to this class of proteins. The phylogenetic conservation of the sterol reductase family signature 1 domain amino acid sequence, including the p.R372 mutated residue, was analysed using BLASTp (http://blast.ncbi.nlm.nih.gov/Blast.cgi), Consseq (http://conseq.tau.ac.il/) and Clustal (http://www.clustal.org/). The p.R372 residue was very conserved, within the ERG24 sterol reductase recognition motif (figure 3).
Prediction of the consequences of the mutation
The effect of the mutation on the protein's structure, function and localisation was predicted through several bioinformatic tools using various algorithms. The p.R372C mutation was predicted to be damaging, when submitted to Pmut (Molecular Modelling and bioinformatics group at http://mmb2.pcb.ub.es:8080/PMut/; prediction score: 0.6480 with mutations predicted as pathological for an index above 0.5), Polyphen (prediction of functional effect of human SNPs at http://genetics.bwh.harvard.edu/pph/; PSIC (Position-Specific Independent Counts) score difference: 2.964 with pathogenicity predicted for a score over 0.5), SIFT (Sorting Intolerant From Tolerant) at http://sift.jcvi.org/: median sequence conservation 2.27, score 0.01 with a prediction of pathogenicity for a score under 0.05); and PANTHER databases at http://www.pantherdb.org/, supplying a subPSEC (substitution Position-Specific Evolutionary Conservation) score of −4.71 744, with deleterious effects on protein function predicted for scores under −3). Furthermore, the R372 residue was predicted to be included in a ‘stabilisation centre element’ by the program Scpred (http://www.enzim.hu/scpred/pred.html), suggesting that the mutation might affect the protein's structural stability or its appropriate folding. The ELM resource (http://elm.eu.org/links.html) further predicted the abolition of a cyclin 1 recognition site (LIG_CYCLIN_1 motif ‘RELNP’) at position 372–376; finally, the creation of a consensus recognition motif for one splicing cofactor, SC35, was predicted by the ESE finder database (http://rulai.cshl.edu/cgi-bin/tools/ESE3/esefinder.cgi?process=home) with a score of 4.77, the threshold value being significant above 2.383.
Functional characterisation of R372C
Granulocytes' nuclear lobulation was analysed on a blood smear of the patient, showing conserved nuclear morphologies and no Pelger-Huët trait (data not shown).
In light of the ESE finder database predictions, the effects of the R372C mutation at the mRNA level were further explored by RT-PCR, performed on mRNAs that had been obtained from both skin fibroblasts and lymphoblastoid cell lines. Transcript amplification between exons 7 and 12 allowed the obtention of an amplicon of about 1180 bp, at the expected size, with no altered size fragments (data not shown). Direct sequencing of the transcripts from both cell lineages confirmed the presence of the heterozygous c.1114C>T variation in LBR exon 9, with wild type and mutated peaks of equal height (figure 4).
The overall expression levels of LBR, Lamins A/C, Lamin A, Lamin B1, Lamin B2 and HP1α were further explored by western blot on both the lymphoblastoid cell line and skin fibroblasts of the patient and an healthy individual. On the first cell line, the same expression levels were observed in the patient and the control for LBR, Lamins A/C, Lamin B1, Lamin B2 and HP1α (figure 5A). Conversely, in the patient's fibroblasts, LBR, Lamins A, C and B2 expression levels were dramatically reduced (figure 5B), while Lamin B1 and HP1α expression appeared completely abolished. These dramatic decreases of expression levels were confirmed on two different extractions from the patient's fibroblasts at passages 3 and 5 (data not shown). In the patient's fibroblasts, Prelamin A (figure 5B) was furthermore undetectable.
Immunostainings on lymphoblastoid cell lines showed conserved expression and localisation for LBR, HP1α, Lamins A, C, B1 and B2, (figure 6). In fibroblasts (figure 7), nuclear shape was slightly affected, with an average of 31% dysmorphic nuclei, versus 8% in control fibroblasts at the same passage, corresponding to previously published values.20–22 Furthermore, abnormal immunostaining patterns were observed for all the aforementioned proteins. Staining of LBR was reduced in about 18% of the fibroblasts, and Lamin B1 and B2 stainings were heterogeneous or highly reduced in 42% and 34% of the patient's nuclei, respectively. An abnormal distribution pattern was observed for HP1α with a marked signal reduction and absent focal distribution in about 40% of cells. The distribution of Lamins A and C (studied with antibodies targeting both isoforms) was abnormal, with 22% of the nuclei showing heterogeneous and/or reduced staining. Immunostaining with a Lamin A-specific antibody showed reduced to absent expression in 18% of the cells. Prelamin A staining was negative in all nuclei of the patient (data not shown). Finally, DAPI counterstaining showed mottled chromatin in several nuclei.
Quantitative RT-PCR amplification of the LBR gene
Quantitative RT-PCR amplification was performed using a TaqMan® gene expression assay designed to amplify LBR exons 10 and 11, on cDNAs prepared from the patient's and a healthy control's fibroblast cell lines. The absolute LBR expression levels were normalised in each case by the GAPDH housekeeping gene expression levels. The same quantitative profiles were obtained in the patient and the control, excluding an impact of the mutation on LBR mRNA expression levels (figure 8).
Semi-quantitative RT-PCR amplification of LMNA, LMNB1, LMNB2, and HP1α
Semi-quantitative RT-PCR amplification (<30 cycles of amplification) of LMNA, LMNB1, LMNB2, and HP1α was performed on cDNAs prepared from the patient's and control's fibroblast cell lines. mRNAs expression was either conserved (ie, for LMNB1), increased (LMNA 4.5×, LMNB2 3×), or slightly decreased (HP1α 0.75×) in the patient versus the control (figure 9). These results globally suggest that the p.R372C mutation does not have an impact on LBR partners' transcription.
LBR, located at chromosome 1q42, encodes a 615 amino acid long protein, weighing 58 kDa, the ‘Lamin B receptor’, which has been one of the first Lamin partners identified.23 LBR is an integral protein of the inner nuclear membrane, fundamental during human development24 and necessary for a complete granulocyte morphological and functional development.25 26 LBR is further known to play an active role in positioning heterochromatin at the nuclear periphery by anchoring it to the lamina and the nuclear membrane and by mediating its interactions with Lamin B.27 The protein is composed of a 208 amino acid long hydrophilic N-terminal domain containing critical regions for interactions with B-type Lamins,27 double stranded DNA,27 HA95,28 HP1 heterochromatin protein,29 importin β,30 and Histones H3/H4 oligomers interacting with both LBR and HP1.31 More recently, interactions of LBR with the heterochromatin methyl binding protein MeCP2 have been described.32
The C-terminal hydrophobic domain of the protein contains eight putative transmembrane segments and exhibits a 3β-hydroxysterol Δ14-reductase enzymatic activity (from amino acids 362 to 377 and 579 to 602). Both the C-terminal and the N-terminal domains of the protein project into the nucleoplasm.
LBR mutations have been reported to cause Pelger–Huet anomaly (PHA), an autosomal dominant haematologic trait characterised by hypolobulated nuclei and abnormal chromatin structure in granulocytes without any clinical symptom,33 and hydrops-ectopic calcification-‘moth-eaten’ (HEM) dysplasia, also known as Greenberg dysplasia, an autosomal recessive chondrodystrophy with a lethal course characterised by fetal hydrops, short limbs, and abnormal chondro-osseous calcifications.24 PHA and HEM were thus subsequently considered as allelic disorders, with non-viable fetuses affected with HEM representing the severe end, and PHA representing the mild end of the spectrum of diseases associated with LBR mutations.24
Our search for mutations in genes of the nuclear envelope was prompted first of all by the clinical features shared by patients affected with scleroderma and patients affected with Lamin-linked progeroid syndromes (ie, fibrous cutaneous modifications, beaked nose, ectopic calcifications, etc), by the frequent observation of antinuclear autoantibodies in the sera of patients affected with SSc and by the previous identification of anti-LBR autoantibodies in the sera of patients suffering from PBC15 and anti-Lamin A/C autoantibodies in the serum of one patient affected with linear scleroderma, a form of localised SSc.14 Furthermore, Lamins A/C have been shown to function as important regulatory factors of nuclear processes downstream of TGF-β1, whose involvement in SSc pathophysiology is well known.34 35 LMNA and ZMPSTE24 sequencing allowed us to exclude the presence of potentially pathogenic variations. The coding sequences of other genes encoding nuclear components were subsequently searched for mutations. Since skin anomalies are observed in HEM24 and because studies performed on mouse models have shown that homozygous icJ mutations, resulting in complete loss of LBR, cause clinical signs including skin defects with marked fibrosis and scaly appearance,36 37 LBR appeared as a particularly good candidate gene.
We report a novel heterozygous mutation in LBR exon 9, leading to a non-synonymous amino acid change, in a patient affected with Reynolds syndrome. The segregation of the mutation in the family could not be studied, but the variation, not described in public databases, was not identified in 400 healthy Caucasian control chromosomes, highly reducing the possibility of a rare sequence polymorphism.
We first asked whether the patient carried a Pelger-Huët trait, which was excluded by polynuclear lobulation analysis. As neutrophil hyposegmentation is known to occur when wild type LBR intranuclear amounts are half the normal ones,33 the normal quantity of LBR expressed in our patients' lymphoblastoid cell lines explains the absence of PHA.
Arginine 372 is highly conserved phylogenetically and among proteins with similar enzymatic function, suggesting a potential role for this amino acid in LBR structure and function. The effect of the p.R372C mutation was analysed using several public databases using different algorithms, all of them predicting it to be pathogenic. The mutation occurs between the fourth and fifth transmembrane domains of the LBR protein, and faces the nucleoplasm. A direct impact of the mutation on LBR binding of Lamin B or chromatin seems excluded, since these functions have been shown to be supported by the N-terminal domain of the protein27; however, the fibroblast specific strong reduction in LBR expression suggests that the mutated LBR proteins have dominant negative deleterious effects on the wild type residual proteins, at least in this cell lineage. Indeed, the R372 residue was also predicted to be involved in the LBR stabilisation centre, potentially affecting the protein's structural stability and possibly leading to an increased proteasome mediated degradation in some tissues.
Since a potential aberrant effect on splicing had been predicted bioinformatically, RT-PCRs were performed on mRNAs extracted from both skin fibroblast and lymphoblastoid cell lines of the patient and a healthy control. This analysis did not provide evidence of any abnormally spliced fragment in either cell line. The c.1114C>T heterozygous mutation was indeed observed in cDNAs issued from both cell lines, with equally represented wild type and mutated peaks. This aspect indicated a balanced expression of the two alleles in both cell lineages, (1) suggesting the absence of a mosaic mutation pattern, and (2) excluding the presence of a fibroblast specific, intragenic LBR compound heterozygous deletion. Furthermore, quantitative RT-PCR amplification of LBR transcripts showed conserved expression levels in our patient's fibroblasts, further arguing against an impact of the observed mutation on the gene's transcription.
The C-Terminal domain of LBR is known to contain a conserved sterol reductase family signature motif (AA 362 to 377) which is highly conserved among different species, and LBR is known to possess a Δ14 sterol reductase activity.38 Since the p.R372C mutation occurs within this domain, an impact on LBR Δ14 sterol reductase activity could be expected. Total cholesterol values and subtype repartition (high density lipoprotein/low density lipoprotein) were normal in the patient's serum, suggesting no impairment in cholesterol biosynthesis. It is currently known that the Δ14 sterol reductase enzymatic activity is redundant in mammals, being also supported by at least one other protein: DHRC14.39 The latter's activity is thus likely to complement the enzymatic deficiency, if any, due to LBR p.R372C mutation. It must be noted that normal cholesterol values are also observed during HEM, which was initially supposed to result from impaired cholesterol biosynthesis,28 and which, conversely, has recently been considered to be due to dysfunction of the Lamin network.37 Since it has been suggested that the sterol reductase domain of LBR acts as a receptor for cell cycle signalling,38 we can also hypothesise an impact on cell cycling regulation of the mutation, which would specifically act in skin. This hypothesis is further supported by the bioinformatic prediction stating that the mutation inactivates a cyclin 1 recognition site.
The results of the immunoblots performed on lymphoblastoid cells of the patient were comparable to those obtained in cell lines issued from the control healthy individual, while those performed on the patient's fibroblast cell lines showed many alterations, notably including a dramatic reduction of LBR, Lamins A and C and Lamin B2 expression levels, with nearly abolished Lamin B1 and HP1α expression. Concordantly, immunostaining with the same antibodies was conserved in lymphoblastoid cells, whereas it was altered in fibroblasts.
Due to ethical concerns and to the difficulty of obtaining a skin biopsy from patients affected with scleroderma, the control cells used in this study were issued from a healthy subject. It could thus be hypothesised that underexpression of Lamins and related proteins might be associated more with SSc than with the LBR mutation. This hypothesis could not be specifically addressed since we did not dispose of other fibroblast cultures, but we plan to test it by collecting further skin biopsies of patients, in which we will search for both nuclear envelope protein unbalances and LBR genomic mutations.
Quantitative and semi-quantitative transcriptional analyses (respectively performed for LBR or its downregulated partners) showed a conserved expression of LBR and LMNB1, an increase of LMNA and LMNB2 expression, and a slight decrease of HP1α expression in the patient versus the control. These results suggest that LBR and its partners' downregulation at the protein level is globally not a direct consequence of an eventual transcriptional downregulation. Conversely, we observed LMNA and LMNB2 increased mRNAs levels that may result from a compensatory mechanism to the decrease of one or more nuclear protein(s), as previously evoked for LAP2α in Lmna null mice.40
Given that inner nuclear membrane proteins, nuclear Lamins and chromosomes are known to be extensively interconnected, the strong reduction of LBR but also of Lamins A/C, Lamin B1, Lamin B2 and HP1α expression in the patient's fibroblasts might also suggest that the mutation has dominant negative deleterious effects not only on its wild type counterparts but also on the stability of several of its partners, in a tissue specific manner. The coding sequences of many LBR partners also did not show any mutation, supporting the hypothesis of a global perturbation of the nuclear envelope protein network due to LBR mutation/reduced expression.
Another hypothesis might be that the LBR mutation silences the expression of other nuclear envelope genes either directly, through altered LBR–chromatin interactions, or indirectly. Indeed, interactions of LBR with the heterochromatin methyl binding protein MeCP2 seem to play an active role in positioning methylated DNA to the nuclear periphery and thereby contributing to gene silencing.32
The mutation could also induce a nuclear fragility leading to altered signal mechanotransduction with changes in the regulation of gene expression.41 LBR itself is indeed known to be regulated by a number of transcription factors such as SP-1, AP-1, AP-2 and NFκB42 and it has been shown that mutations interfering with nuclear architecture and composition can severely affect the function of transcription factors involved in tissue differentiation.43 44 In particular, an impact on the expression or localisation of OCT-1, a transcription factor known to associate with B-type Lamins at the nuclear periphery,45 may be evoked. It has been shown that in mouse fibroblasts lacking normal, full length Lamin B1, Oct-1 is delocalised from the nuclear periphery and that in this context the transcriptional regulation of collagenase and oxidative stress response genes45 46 is altered. Given the extreme reduction of B-type lamin expression observed in the fibroblasts of the patient, a similar dysregulation could be involved.
Finally, the perturbation of this network may have an important impact on the TGF-ß1 signalling pathway, whose activation is known to be linked to SSc (for review, see Verrecchia et al35). Indeed, Lamins A/C have been shown to modulate the effect of TGF-ß1 on collagen production through PP2-A dependent dephosphorylation of pRB and other transcriptional regulators like SMAD2, able to quench TGF-ß1 signalling.34
LBR mutations do not seem to represent a common cause of SSc, given that we could not identify this mutation nor any other clearly pathogenic one in 27 other patients affected with SSc tested at our laboratory, except one unreported heterozygous variation within intron 2 (c.1114C>T) identified in one patient affected with dSSc, but in none of 200 control chromosomes, which could not be functionally tested due to the absence of biological material (data not shown and Gaudy-Marqueste et al16). Our data suggest, however, a deleterious effect of this particular, sporadic LBR mutation in a patient affected by Reynolds syndrome, specifically associating lSSc and primary biliary cirrhosis.2 3 Skin was specifically affected in the patient, while other tissues, like liver, could not be explored.
Since homozygous mutations constitutionally suppressing LBR expression are known to be lethal early in life,23 24 the fibroblast specific LBR reduction may explain the patient's non-lethal phenotype, predominantly involving this tissue. Taken together, our results thus suggest that LBR mutations might represent a rare cause of Reynolds syndrome either by exerting a direct pathogenic effect or by constituting a predisposing factor in conjunction with other genetic and environmental factors. In light of these results, we believe that genomic and functional screening are now warranted in larger cohorts of patients affected with Reynolds syndrome, localised cutaneous scleroderma, and PBC in order to explore an eventual deregulation of LBR or related nuclear envelope partners.
The authors are grateful to the patient for her kind collaboration and to Fabien Angelis and Mike Mitchell, who provided excellent technical support.
Competing interests None.
Patient consent Obtained.
Provenance and peer review Not commissioned; externally peer reviewed.