Skip Navigation


Human Molecular Genetics Advance Access originally published online on July 15, 2003
This Article
Right arrow Abstract Freely available
Right arrow FREE Full Text (PDF) Freely available
Right arrow A corrigendum has been published
Right arrow All Versions of this Article:
12/18/2395    most recent
ddg234v1
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Add to My Personal Archive
Right arrow Download to citation manager
Right arrow Search for citing articles in:
ISI Web of Science (25)
Right arrowRequest Permissions
Google Scholar
Right arrow Articles by McLean, W. H. I.
Right arrow Articles by McGrath, J. A.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by McLean, W. H. I.
Right arrow Articles by McGrath, J. A.
Social Bookmarking
Add to CiteULike   Add to Connotea   Add to Del.icio.us  
What's this?

Human Molecular Genetics, 2003, Vol. 12, No. 18 2395-2409
DOI: 10.1093/hmg/ddg234
© 2003 Oxford University Press

An unusual N-terminal deletion of the laminin {alpha}3a isoform leads to the chronic granulation tissue disorder laryngo-onycho-cutaneous syndrome

W. H. Irwin McLean1,*, Alan D. Irvine2, Kevin J. Hamill1, Neil V. Whittock1, Carrie M. Coleman-Campbell1, Jemima E. Mellerio3, Gabrielle S. Ashton3, Patricia J. H. Dopping-Hepenstal3, Robin A. J. Eady3, Tanvir Jamil4, Rodney J. Phillips5, S. Ghulam Shabbir6, Tahir S. Haroon6, Khawar Khurshid6, Jonathan E. Moore7, Brian Page7, Jonathan Darling8, David J. Atherton9, Maurice A. M. van Steensel10, Colin S. Munro11, Frances J. D. Smith1 and John A. McGrath3

1Epithelial Genetics Group, Human Genetics Unit, Department of Molecular and Cellular Pathology, University of Dundee, Ninewells Medical School, Dundee DD1 9SY, UK, 2Department of Paediatric Dermatology, Our Lady's Hospital for Sick Children, Dublin, Ireland, 3Genetic Skin Disease Group, St John's Institute of Dermatology, The Guy's, King's College and St Thomas' Hospitals' Medical School, St Thomas' Hospital, London, UK, 4Hasham M. Jamil Trust, Burnham Health Centre, Minniecroft Road, Burnham, Burkinghamshire SL1 7DE, UK, 5Department of Paediatrics, Royal Children's Hospital, Melbourne, Australia, 6Department of Dermatology, King Edward Medical College, Nila Gumbad, Lahore, Pakistan, 7Department of Ophthalmology, Royal Victoria Hospital, Belfast, UK, 8Department of Paediatrics, St James's University Hospital, Leeds, UK, 9Department of Paediatric Dermatology, Great Ormond Street Hospital for Children, London, UK, 10Department of Dermatology, University Hospital Maastricht, The Netherlands and 11Department of Dermatology, Southern General Hospital, Glasgow, UK

Received June 25, 2003; Accepted July 8, 2003


   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 RESULTS
 DISCUSSION
 MATERIALS AND METHODS
 REFERENCES
 
Laryngo-onycho-cutaneous (LOC or Shabbir) syndrome (OMIM 245660) is an autosomal recessive epithelial disorder confined to the Punjabi Muslim population. The condition is characterized by cutaneous erosions, nail dystrophy and exuberant vascular granulation tissue in certain epithelia, especially conjunctiva and larynx. Genome-wide homozygosity mapping localized the gene to a 2 Mb region on chromosome 18q11.2 with an LOD score of 19.8 at {theta}=0. This region includes the laminin {alpha}3 gene (LAMA3), in which loss-of-expression mutations cause the lethal skin blistering disorder Herlitz junctional epidermolysis bullosa. Detailed investigation showed that this gene possesses a further 38 exons (76 exons in total) spanning 318 kb of genomic DNA, and encodes three distinct proteins, designated laminin {alpha}3a, {alpha}3b1 and {alpha}3b2. The causative mutation in 15 families was a frameshift mutation 151insG predicting a stop codon 7 bp downstream in an exon that is specific to laminin {alpha}3a. This protein is secreted only by the basal keratinocytes of stratified epithelia, implying that LOC is caused by dysfunction of keratinocyte–mesenchymal communication. Surprisingly, the 151insG mutation does not result in nonsense-mediated mRNA decay due to rescue of the transcript by an alternative translation start site 6 exons downstream. The resultant N-terminal deletion of laminin {alpha}3a was confirmed by immunoprecipitation of secreted proteins from LOC keratinocytes. These studies show that the laminin {alpha}3a N-terminal domain is a key regulator of the granulation tissue response, with important implications not only in LOC but in a range of other clinical conditions associated with abnormal wound healing.


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
RESULTS
 DISCUSSION
 MATERIALS AND METHODS
 REFERENCES
 
Laryngo-onycho-cutaneous syndrome (LOC, OMIM 245660) is a chronic granulation tissue disorder seen in the Punjabi Muslim population. The disease was first characterized by Shabbir in 1986 (1), in 22 patients, all of whom shared significantly similar clinical features that had not been previously described. Eighteen of the 22 patients were children of consanguineous marriages between clinically normal individuals, strongly indicative of autosomal recessive inheritance. Some of the original families examined by Shabbir were studied here. All LOC families reported to date are of Punjabi descent (1–4), including all subjects in this report, suggesting a founder effect and perhaps an unusual defect that has not arisen in other populations. The term laryngeal and ocular granulation in children originating in the Indian subcontinent (LOGIC syndrome), has been suggested as an alternative name for the disorder (3).

Patients with LOC syndrome share common clinical features (Fig. 1). The onset is characterized by a hoarse cry soon after birth, although the physical appearance of these neonates is normal. Granulation tissue develops in the mucosal regions, nail bed and the larynx. Beginning in infancy, chronic skin ulcers and conjunctival lesions appear (1–4). Although eye involvement in LOC syndrome was not mentioned in the original description (1), ocular granulation tissue resembling a particularly aggressive pterygium has been reported in all subsequent patients described (2–4) and was a prominent feature in the patients studied here. In addition, there is failure of tooth enamel formation (2–4) and marked dental malformations (1) (Fig. 1). Many patients diagnosed with LOC syndrome do not live beyond childhood, the most common cause of premature death being acute or chronic respiratory obstruction with secondary pulmonary sepsis (2–4). However, patients who survive the early neonatal period often survive to adulthood (A.D. Irvine and C.S. Munro, personal observation) and symptoms of hoarseness and ulceration slowly improve with age (1). As granulation tissue accumulates in the larynx, a permanent tracheostomy often becomes necessary to ensure a clear respiratory passageway. In addition, granulation tissue can spread to the epiglottis, trachea (3) and even the main bronchi (4). Furthermore, most of the older patients studied here are blind due to corneal pterygium formation.



View larger version (84K):
[in this window]
[in a new window]
  		Figure 1. Clinical features of LOC syndrome in patients from families 2 and 3. (A) Proband in family 3, showing cutaneous erosions on the face, which are slow to heal. Healed lesions often leave behind atrophic, pigmented scars as seen on the right cheek of this patient. (B) Nail dystrophy is common in LOC; in infants erosions are often seen but older patients develop a thickened hyperkeratotic appearance, seen here in right great toe of the proband from family 2. (C) Ocular granulation tissue with lateral symblepharon formation in the proband from family 2. This patient, in his 20s, is blind due to almost complete closure of the eyelids and deep corneal scarring secondary to the invasive fibrovascular growth. (D) Hypoplastic enamel and subsequent carious teeth are common findings in LOC. (proband of family 2).

 
Here, by genomewide homozygosity mapping in consanguineous families, we have localized the LOC gene to a 2 Mb interval on 18q11.2 and identified the causative mutation as an N-terminal deletion of the laminin {alpha}3a isoform.


   RESULTS
TOP
 ABSTRACT
 INTRODUCTION
 RESULTS
DISCUSSION
 MATERIALS AND METHODS
 REFERENCES
 
Clinical details
All living patients and their families were seen and examined by one or more of the authors (A.D.I., C.S.M., R.P. or T.J.). Families from a Pakistani Punjabi background were seen in Scotland, England, Northern Ireland, France and in Pakistan (Lahore and the surrounding Punjab region). Medical records for living patients and possibly affected deceased relatives were examined where available in the local health care setting in conjunction with local health professionals. Owing to the range of health care institutions that the subjects had attended (basic rural health care in some, tertiary referral hospital for others), there was a wide variation in the degree of clinical investigations and medical interventions performed. To ensure a consistent approach, ascertainment of affected phenotype for the purposes of this study was based on clinical history and physical examination and not on the basis of laboratory investigations. After review of the few clinical papers published to date (1–4) and discussion with the original authors, subjects were deemed to be affected with LOC if they fulfilled the following clinical diagnostic criteria: (1) hoarse cry in first few weeks of life or evidence of laryngeal granulation tissue, scarring or stenosis in the first few years of life; (2) nail dystrophy; (3) ocular involvement including aggressive conjunctival pannus formation; (4) dental anomalies (hypodontia, poor enamel formation, excessive caries); (5) slow-healing cutaneous erosions; (6) absence of a history of skin fragility or blistering at any stage. All subjects deemed affected in the study filled all of these criteria. The cardinal features of LOC syndrome are illustrated in Figure 1. Patient DNA and tissue samples were obtained after informed consent with the approval of the relevant local Ethics Committees.

Genetic linkage analysis
To search for the LOC gene, genomewide homozygosity mapping was carried out using the ABI HD5 linkage mapping panel, consisting of 800 fluorescent microsatellite markers, as described previously (5). Affected individuals were genotyped from six apparently unrelated Punjabi families (Fig. 2A). Five of these kindreds are UK citizens of Punjabi ancestry (families 1–4 and 6) and the other family resides in the Lahore region of Pakistan (family 5). Initial screening identified potential loci on chromosomes 5q in the vicinity of D5S400 and 16q, in the region of D16S3061. However, these loci were suspected as false linkage due to lack of shared haplotypes between the families. Analysis of additional LOC families completely excluded these loci (data not shown). No other suspect loci were identified from the initial screen. In case the locus had been missed, the larger gaps in the HD5 panel were filled by synthesis of an additional 60 microsatellites. One of these additional markers, D18S869, showed homozygosity for the same allele (195 bp) in all but three of our 15 LOC families, consistent with the anticipated founder effect (Fig. 2B). Further genotyping was carried out with an additional nine closely spaced microsatellites in the vicinity of D18S869. This revealed an area of shared haplotype for three markers, LAMA3GT1, LAMA3GT2 (see below) and D18S1101 in 17 individuals from 14 LOC families (Fig. 2B). One family remained unlinked (family 15), however, in this case one parent carried the linked haplotype (Fig. 2B). There was no evidence of consanguinity in family 15 and therefore it was assumed that the affected individual in this case might represent a compound heterozygote. All parents of affected individuals in families 1–14 were heterozygous for the linked haplotype and no unaffected siblings of LOC patients were homozygous for this haplotype. For the kindreds where details of consanguinity were known (families 1–7), markers within the region of shared haplotype gave a highly significant two-point LOD score of 9.6 at {theta}=0. The combined two-point LOD score for all kindreds showing linkage (families 1–14) was 19.8 at {theta}=0 for the three markers within the region of shared haplotype.



View larger version (28K):
[in this window]
[in a new window]
  		Figure 2. (A) Pedigrees of consanguineous LOC families 1–6, used for genome-wide homozygosity mapping. Arrows indicate probands and numbers refer to individual probands whose haplotypes are shown in (B), below. Asterisks show individuals from whom DNA was available for study. (B) Haplotypes of affected LOC individuals for microsatellite markers in the vicinity of the LOC locus. Numbers above haplotypes refer to individual LOC patients and are in the format family–patient number. Boxes indicate areas of homozygosity-by-descent.

 
The limits of the LOC locus were defined by visible recombinations seen in families 2, 7 and 12 (individuals 2-1, 7-1 and 12-2; Fig. 2B) with markers D18S869 (centromeric) and D18S480 (telomeric). This gave a critical interval of approximately 2 Mb of genomic DNA (UCSC Genome Map). A total of 12 known and strongly predicted genes were located in this interval. Of these, LAMA3, located centrally, was considered by far the most likely candidate gene on the basis of its biological functions. Laminin {alpha}3 is one of the three polypeptides comprising laminin 5, the major cutaneous basement membrane laminin with a critical function in adhesion of the epidermis to the dermis. Loss-of-expression mutations in LAMA3 cause the skin blistering disorder Herlitz junctional epidermolysis bullosa (H-JEB) (6–8). In addition to mediating cell-substratum adhesion, laminins in general and laminin 5 in particular have important additional functions in the control of cell migration and behaviour (9,10). Therefore, since the main phenotype in LOC is uncontrolled production of granulation tissue in the skin and other epithelial tissues, LAMA3 was considered a prime candidate. Two microsatellites within introns 12 and 41 of the extended LAMA3 gene (described in detail below), designated LAMA3GT1 and LAMA3GT2, respectively, were found to be highly polymorphic in the Punjabi Muslim population. Importantly, the relevant haplotype for these two markers (alleles 290 and 195, respectively) was shared by all LOC affected individuals studied (Fig. 2B), including one copy carried by the presumptive compound heterozygote family 15. Therefore, the LAMA3 gene falls within the region of shared haplotype within the LOC locus.

LAMA3 gene structure and alternate splicing
The LAMA3 gene was initially reported as having 38 exons encoding the laminin {alpha}3a polypeptide (11). However, partial cDNAs have been reported encoding a longer alternate transcript which encodes the larger laminin {alpha}3b polypeptide, whose functions and protein interactions are less well understood (11,12). Since recessive loss-of-function mutations in LAMA3A had already been demonstrated in the severe/lethal skin blistering disorder H-JEB, we initially focused on LAMA3B. Analysis of the human genome sequence with reference to the reported laminin {alpha}3b cDNAs and ESTs revealed an additional 38 exons upstream of the existing 38 LAMA3A exons, giving a total of 76 exons spanning 318 kb of genomic DNA. The full gene structure and transcripts thereof are shown in Figure 3. Exons 1–38 plus 40–76 encode the longer LAMA3B transcript (i.e. exon 39 is skipped in the LAMA3B isoform). In addition, RT–PCR performed on cDNA extracted from normal control keratinocytes revealed that about 20% of the LAMA3B mRNA also lacks exon 10. We propose that the longer LAMA3B isoform should be designated LAMA3B1 (encoding laminin {alpha}3b1) and that the shorter version lacking exon 10 be designated LAMA3B2 (encoding laminin {alpha}3b2). The predicted laminin {alpha}3b1 polypeptide consists of 3333 amino acids and has a calculated molecular weight of 366.3 kDa. The shorter laminin {alpha}3b2 polypeptide comprises 3289 amino acids and has a calculated molecular weight of 361.4 kDa. Promoter sequences immediately upstream of exon 1 (B-isoforms) were strongly predicted by both the TSSG and TSSW algorithms (13). The transcriptional start site of the B isoforms was mapped by RT–PCR (data not shown) and the entire LAMA3B1 and LAMA3B2 cDNAs were amplified in a series of overlapping RT–PCR fragments and fully sequenced. Primers used for RT–PCR are listed in Table 1.



View larger version (26K):
[in this window]
[in a new window]
  		Figure 3. (A) Schematic diagram (to scale) showing the genomic organization of the 76 exons of the LAMA3 gene, which spans ~318 kb on chromosome 18q. Exons involved in transcription initiation and/or alternate splicing are indicated by arrows. (B) Transcripts identified from the LAMA3 gene (spliced exons to scale). These transcripts encode three laminin {alpha}3 polypeptides: {alpha}3a, {alpha}3b1 and {alpha}3b2. The predicted laminin {alpha}3a polypeptide consists of 1724 amino acids and has a calculated molecular weight of 190.3 kDa. The predicted laminin {alpha}3b1 protein consists of 3333 amino acids and has a calculated molecular weight of 366.3 kDa. The shorter laminin {alpha}3b2 polypeptide comprises 3289 amino acids and has a calculated molecular weight of 361.4 kDa. (C) Excerpts of the 5' regions of LAMA3 transcripts showing the start of the predicted open reading frames. The 5'UTR sequences are shown in lower case; and codons are shown in upper case with amino acid translation underneath.

 

View this table:
[in this window]
[in a new window]
  		Table 1. LAMA3 RT–PCR and sequencing primers
 
The LAMA3A transcript starts with exon 39, which includes the initiation codon and 5'-UTR of the LAMA3A transcript, and includes exons 39–78 of the LAMA3 gene. However, the LAMA3A 5'-UTR derived from the human genome sequence (UCSC Browser) was found to have slight differences from the published mRNA sequence and contained an additional in-frame ATG codon (Fig. 4). The sequences surrounding the new ATG conform better to the Kozak consensus sequence than the previously proposed initiation codon (11). This sequence was verified by sequencing of both genomic and RT–PCR fragments, and the likely transcriptional start site predicted by both the TSSG and TSSW algorithms was determined by RT–PCR of keratinocyte mRNA using a series of primers across the region (Fig. 4). The LAMA3A promoter was strongly predicted by both the TSSG and TSSW algorithms to be within intron 38, immediately upstream of exon 39. The entire LAMA3A cDNA was amplified in a series of overlapping fragments and was fully sequenced. Primers used for RT–PCR are listed in Table 1. The predicted laminin {alpha}3a polypeptide consists of 1724 amino acids and has a calculated molecular weight of 190.3 kDa. By RT–PCR, the A-isoform was found to be both specifically and highly expressed by epidermal keratinocytes but not by dermal fibroblasts, whereas the two B-isoforms were expressed by both these cell types (data not shown). Relative expression was not assessed in other tissues. The GenBank accession numbers for the LAMA3 transcripts determined here are AY327114 (LAMA3A) AY327115 (LAMA3B1) and AY327116 (LAMA3B2).



View larger version (35K):
[in this window]
[in a new window]
  		Figure 4. Map of exon 39 of the LAMA3 gene, showing the additional transcribed sequences compared to the previously reported mRNA sequence. A predicted TATA box is shown in bold type, identified by the TSSG and TSSW programs. The predicted start of transcription is shown in bold. Primers used to map the transcription start site are shown as arrows. Primers P2F and P3F were found to amplify RT–PCR keratinocyte cDNA in combination with reverse primers in downstream exons, whereas P1F did not (data not shown).

 
Isoform-specific LAMA3 mutation in LOC
PCR conditions were developed to amplify the individual exons of the LAMA3 gene using intronic primers to facilitate mutation detection based on genomic DNA (primers detailed in Table 2). All 76 exons were screened for mutations in LOC affected individuals by direct sequencing of genomic PCR products, with comparison to normal unrelated controls. In LOC patients from families 1–14, a homozygous 1 bp insertion mutation was detected in exon 39, the exon found exclusively in the keratinocyte-specific LAMA3A transcript encoding the laminin {alpha}3a polypeptide (Fig. 5). The mutation was designated 151insG, numbering from the initiating ATG of our revised LAMA3A sequence (Fig. 4; GenBank accession number AY327114) and leads to a premature termination codon (PTC) 7 bp after the insertion, within this first LAMA3A exon. Numbering from the previously reported mRNA sequences (GenBank accession numbers NM_000227 and X77598), this mutation would be designated 118insG. All parents in families 1–14 were heterozygous carriers of the mutation (Fig. 5). Thus, individuals with LOC are homozygous for a frameshift mutation leading to a PTC within the only exon specific to the LAMA3A isoform. This mutation should have no effect on expression of the LAMA3B isoforms and therefore our initial hypothesis was that LOC is caused by a specific, differential loss of laminin {alpha}3a expression. In contrast, all the LAMA3 mutations reported in H-JEB patients are PTC mutations in exons that are common to both LAMA3A and LAMA3B transcripts (6,8). Therefore, we hypothesized that loss of both laminin {alpha}3a and {alpha}3b expression might cause H-JEB, whereas loss of {alpha}3a alone would cause LOC.


View this table:
[in this window]
[in a new window]
  		Table 2. LAMA3 genomic PCR
 


View larger version (31K):
[in this window]
[in a new window]
  		Figure 5. The homozygous ancestral LOC mutation in exon 39 of LAMA3 (151insG) shared by all affected persons in families 1–14. In addition, the proband in family 15 was heterozygous for this mutation. (A) Direct sequence of a LAMA3 exon 39 PCR product from a normal control. Codons 49–53 of the LAMA3A transcript are shown. (B) Equivalent region as shown in (a) derived from the proband in LOC Family 1 (LOC patient 1-1, Fig. 2), showing homozygous insertion of G at position 151 (151insG in the LAMA3A transcript). This frameshift mutation leads to a premature termination codon 7 bp downstream of the insertion (not seen here). (C) Equivalent region as shown in (a) derived from the father of the proband in LOC family 1. Overlapping sequence traces can be seen due to heterozygosity for the insertion mutation. (D) Equivalent region as shown in (a) derived from the mother of the proband in LOC family 1. Again, overlapping sequence traces can be seen due to heterozygosity for the 151insG mutation.

 
The father in family 15, who carried the shared LOC haplotype (Fig. 2B), was also found to carry the ancestral 151insG mutation, which he had passed on to his affected son, but this was not detected in the mother's DNA. All other exons of the LAMA3 gene were analysed in the family 15 proband and his parents. Both the proband and his mother were found to be heterozygous for nonsense mutation R943X (DNA change 2827C>T) in exon 60, the 22nd exon of the LAMA3A transcript (Fig. 6). This mutation would be numbered R932X in the previously reported sequence. The mutation ablates a SalI restriction enzyme site which was used to confirm the mutation in the proband and his mother and exclude it from his father (Fig. 6). Thus, the proband in family 15 is a compound heterozygote for mutations 151insG and R943X in LAMA3A. This individual would be predicted to lack expression of laminin {alpha}3a but would express 50% of the normal amount of laminin {alpha}3b, again consistent with the hypothesis that LOC is caused by loss of laminin {alpha}3a alone.



View larger version (21K):
[in this window]
[in a new window]
  		Figure 6. Compound heterozygous mutation in the proband from LOC family 15. This patient was heterozygous for the shared haplotype at the LOC locus (Fig. 2) and was subsequently found to be heterozygous for the 151insG mutation (data not shown). (A) Direct sequence of a LAMA3 exon 60 PCR product from a normal control. Codons 941–945 of the LAMA3A transcript are shown. Box indicates arginine codon 943. (B) Equivalent region as shown in (A) derived from the proband in LOC family 15 (LOC patient 15-1, Fig. 2), showing heterozygous nonsense mutation R943X (boxed). (C) Mutation R943X destroys a single SalI restriction enzyme site present in the exon 60 PCR fragment. Lane 1 shows undigested 335 bp exon 60 PCR product. Lanes 2–3 show SalI digests of exon 60 PCR products derived from the mother, father and proband of family 15, respectively. The father's DNA digests fully, giving bands of 230 and 105 bp (lane 3). In contrast, both the mother and the proband's DNA show an additional undigested 335 bp band due to the mutation. Thus, the R943X mutation was inherited maternally, consistent with the haplotype data (Fig. 2).

 
By direct DNA sequencing, the incidence of the common LOC mutation was investigated in 50 normal unrelated British Punjabi Muslims with no family history of LOC syndrome as well as 50 normal unrelated British Caucasians. The mutation was not detected in the British Caucasian control population. However, one Punjabi control was found to be heterozygous for the 151insG mutation, indicating that this mutant LAMA3A allele may be fairly common in the Punjabi population.

Rescue of the LOC mutant transcript
To see if the 151insG mutation would lead to nonsense-mediated mRNA decay, we obtained a skin biopsy from a heterozygous carrier of the LOC mutation; mRNA was extracted and subjected to RT–PCR using a forward primer within exon 39 and reverse primer in exon 42, so that only the LAMA3A transcript would be amplified. mRNA was pre-treated with RNAse-free DNAse I and phenol–chloroform extracted to safeguard against genomic DNA contamination. Direct sequencing of this RT-PCR product (Fig. 7A) was compared to a genomic exon 39 fragment from the same individual sequenced with the same primer (Fig. 7B). Surprisingly, the sequence traces obtained from genomic DNA and cDNA were essentially identical, with the overlapping sequence traces due to the insertion clearly visible in both samples and present at very similar levels. Therefore, this demonstrates that the 151insG mutation does not lead to appreciable nonsense-mediated decay of the LAMA3A transcript.



View larger version (30K):
[in this window]
[in a new window]
  		Figure 7. Mutant LAMA3A transcript escapes nonsense-mediated mRNA decay. (A) Direct sequencing of a LAMA3A-specific RT–PCR product generated with primers L3AP3F and L3AP4R (Table 1) derived from epidermal mRNA from an LOC heterozygote. The forward L3AP3F sequence beginning at codon 49 (AGT) is shown. Heterozygosity for the 151insG mutation is clearly seen as overlapping sequence traces. (B) Direct sequencing of a LAMA3 exon 39 genomic PCR product derived from the same individual as in (A), also sequenced with primer L3AP3F. Note that the heterozygous bases are very similar in signal intensity between the cDNA and genomic DNA samples, for example, where indicated by arrows in each case. Thus, there is no appreciable decay of the mutant message. (C) Genomic sequence in the region of exon 45, downstream of the LOC mutation 151insG in exon 39. The first two in-frame ATG codons that follow the mutation are shown in bold. The first of these, methionine codon 218 is a poor match for the Kozak consensus sequence, a feature required for eukaryotic translation initiation (shown above with matching bases boxed). In contrast, the next ATG, codon 227, is a perfect match for the Kozak sequence (boxed). Therefore, it was predicted that translation might take place from ATG codon 227, leading to an N-terminal truncation of 226 amino acids. Efficient translation of the mutant transcript would prevent nonsense-mediated mRNA decay.

 
Close examination of the LAMA3A mRNA sequence revealed that the wild-type 5'-UTR does not contain any in-frame stop codons (Fig. 4), whereas many 5'-UTRs do, especially longer ones (14,15). We hypothesized that the translation machinery may regard this premature stop codon near the start of the mRNA as a natural feature of the 5'-UTR and therefore translation might occur from another ATG downstream if one existed within the context of a suitable Kozak consensus motif, which is required for efficient translation. The next ATG codon downstream of the frameshift occurs in exon 45 (methionine 218 in the revised protein sequence); however, this codon is a very poor fit for the Kozak consensus (Fig. 7C). In contrast, the next ATG, also in exon 45 (methionine 227) is a perfect fit for the Kozak sequence (14), in fact conforming to the consensus much better than the natural initiation codon of the LAMA3A transcript (Fig. 7C). Thus, it seemed plausible that protein translation would initiate from this codon leading to a truncated laminin {alpha}3a polypeptide lacking the first 226 amino acids. To look for this truncated protein and to see if it would be assembled into trimeric laminin 5 and secreted from keratinocytes, we performed immunoprecipitation of radiolabelled secreted proteins from control and LOC keratinocytes using an antibody specific for the C-terminus of laminin ß3 (16,17). This analysis revealed that indeed a truncated laminin {alpha}3a is synthesized, assembled and secreted by LOC keratinocytes (Fig. 8). The reduction in size of the truncated protein compared to wild-type laminin {alpha}3a was consistent with loss of 226 amino acids (~24 kDa). Thus, the LOC mutant transcript is rescued by a novel mechanism leading to expression of an N-terminally truncated protein. We propose the term ‘delayed initiation codon’ for this novel class of mutation.



View larger version (47K):
[in this window]
[in a new window]
  		Figure 8. Radioimmunoprecipitation of [35S]-labelled secreted proteins from control and LOC keratinocytes with laminin 5 monoclonal antibody K140. Approximate molecular weight values are shown (in kDa). Lane 1, immunoprecipitation from control keratinocytes showing two high molecular weight bands, assumed to be the laminin {alpha}3a and {alpha}3b isoforms (marked A and B). Laminin {alpha}3a normally migrates at ~165 kDa on SDS gels (16). The upper band could also be due to co-immunoprecipitation of other laminin alpha chains. Lower bands may include the laminin ß3 and {gamma}2 chains and other associated proteins. Lane 2, immunoprecipitation from LOC keratinocytes, showing a size reduction in the presumptive laminin {alpha}3a band (marked {Delta}a), of ~24 kDa, the size predicted by loss of the N-terminal 226 amino acids. The fainter levels of the deleted {alpha}3a and the {alpha}3b bands in the LOC sample are likely to be because the LOC keratinocyte culture was much less confluent than the normal control. These results were replicable but lack of LOC cells prevented further labelling experiments to adjust loading.

 
Subtle morphological changes in LOC skin
Using frozen skin biopsy material from an LOC patient (proband in family 1, homozygous for the 151insG mutation) and a normal unrelated control, we performed immunofluorescence staining with a variety of antibodies against all three {alpha}3, ß3 and {gamma}2 chains of the laminin 5 trimer and other components of the cutaneous basement membrane zone. Representative results are shown in Figure 9. Essentially, all antigens examined, including laminin {alpha}3, were present in normal amounts and were distributed normally. The laminin {alpha}3 antibodies used are not able to distinguish laminin {alpha}3a and {alpha}3b isoforms, nevertheless, there was no evidence of a 50% reduction in staining intensity that one might expect if the {alpha}3a isoform was absent. Importantly, normal staining was obtained in LOC skin with monoclonal antibody GB3, which detects trimeric laminin 5 (18), confirming the assembly and secretion of this protein complex.



View larger version (108K):
[in this window]
[in a new window]
  		Figure 9. Indirect immunofluorescence staining of LOC skin (A and C), compared with normal control skin (B and D). Monoclonal antibodies used were GB3 against the laminin 5 trimer (A and B) and K140 against laminin ß3 (C and D). No significant changes in the intensity or distribution of staining was seen between the samples, showing that the mutated laminin {alpha}3a chain is capable of assembly and secretion, confirming the immunoprecipitation data (Fig. 8).

 
The ultrastructure of the dermal–epidermal junction (DEJ) was generally normal, but there were some subtle qualitative changes (Fig. 10). The number of hemidesmosomes and anchoring fibrils, appeared to be normal. However, some of the hemidesmosome plaques were small and lacked normal sub-basal dense plates. These diminutive cell–stromal junctions were associated with reduced numbers of anchoring filaments and focal widening of the lamina lucida, the site of laminin 5 localization within the DEJ. There were no overt splits in the lamina lucida, as would be expected in the junctional forms of epidermolysis bullosa, such as H-JEB.



View larger version (182K):
[in this window]
[in a new window]
  		Figure 10. Electron microscopy of a skin biopsy from the proband in LOC family 1. The ultrastructure of the dermal–epidermal junction (DEJ) was generally normal, but there were some qualitative changes. The number of hemidesmosomes (HD) and anchoring fibrils (AFB) appear normal. However, some of the hemidesmosome plaques are small and lack normal sub-basal dense plates (HDs). These diminutive cell-stromal junctions are associated with reduced numbers of anchoring filaments (AFL) and focal widening (asterisk) of the lamina lucida (LL), although the lamina densa (LD) appeared normal. There are no frank splits in the lamina lucida, as would be expected in the junctional forms of epidermolysis bullosa to laminin 5 ablation mutations, where anchoring filaments are absent.

 

   DISCUSSION
TOP
 ABSTRACT
 INTRODUCTION
 RESULTS
 DISCUSSION
MATERIALS AND METHODS
 REFERENCES
 
LOC syndrome is a rare recessive disorder thus far only described in the Punjabi Muslim population (1–4,19). The disease is characterized by chronic production of vascular granulation tissue which invades and disrupts specific epithelial tissues, notably the conjunctiva and anterior corneal epithelium (4,19), as well as the respiratory mucosae. Here, by genome-wide homozygosity mapping the gene causing LOC syndrome was localized to a small interval on human chromosome 18q. The causative mutation in all but one family was found to be homozygosity for a novel class of mutation in the laminin {alpha}3 gene (LAMA3). The remaining case was a compound heterozygote for this mutation and a complementary null-allele. The ancestral genetic defect is a delayed initiation codon mutation in exon 39 of LAMA3, designated 151insG. Exon 39 is the first exon of the LAMA3A transcript, which encodes the laminin {alpha}3a polypeptide, a component of the laminin 5 heterotrimer, a basement membrane laminin specifically secreted by basal keratinocytes in the skin and other stratified epithelia including airway and ocular epithelia (18,20). Furthermore, the mutated exon is the only one specific to laminin {alpha}3a and is spliced out of laminin {alpha}3b1 and {alpha}3b2 described here, whose functions are currently unknown. Unexpectedly, we found that this frameshift mutation does not lead to appreciable nonsense-mediated decay of the LAMA3A mRNA. The mutant transcript escapes accelerated degradation by utilising an alternative initiation codon downstream (in exon 45) which occurs in the context of a particularly good Kozak sequence (14). Effectively, this leads to an N-terminal deletion of 226 amino acids from the laminin {alpha}3a isoform. Here, we have shown that the mutant polypeptide is synthesized, assembled and secreted efficiently by keratinocytes from LOC patients. In general, PTC mutations located towards the 5' end of mRNA species would be expected to give near-complete absence of protein expression (21). However, we show that, in certain cases at least, PTCs located at the extreme 5' end may escape nonsense-mediated decay if a suitable alternative initiation codon exists within the message, even at a considerable distance downstream. These observations have important implications for predicting the effects of future 5' frameshift or nonsense mutations.

Laminin 5 is a heterotrimer composed of {alpha}3a, ß3 and {gamma}2 chains (18,22–24). It has been shown that these molecules must undergo trimerisation in order to be secreted by keratinocytes (25). We showed here by immunoprecipitation that a mutant form of laminin {alpha}3a lacking the first 226 amino acids (~24 kDa) is able to assemble with laminin ß3 and {gamma}2 to form laminin 5 trimers that are capable of being secreted by cultured keratinocytes. This in vitro result is borne out by the fact that immunohistochemical analysis of LOC patient skin showed no obvious qualitative or quantitative aberrations with antibodies against all three laminin 5 polypeptides. Importantly, staining for antibody GB3 was normal in LOC skin. The GB3 epitope is conformation-dependent and is completely lost in cases of H-JEB when any of the three laminin 5 chains are absent (26,27) and therefore positive staining for GB3 provides further evidence for laminin 5 trimerization and secretion. Ultrastructural analysis was again unremarkable. In particular, there was no overt separation of the skin, consistent with the relatively mild erosive skin phenotype in LOC, rather than the extremely fragile skin as is the case in H-JEB (8,28). Furthermore, anchoring filaments, which are composed largely of laminin 5 and generally absent in H-JEB, were clearly visible in the skin of an LOC patient. From this, we conclude that N-terminally truncated laminin {alpha}3a is both assembly-competent and capable of making anchoring filaments extracellularly in the cutaneous basement membrane zone. Only very subtle ultrastructural changes were seen, such as focal widening of the lamina lucida, demonstrating that the truncated protein is largely able to perform its primary function in mediating keratinocyte–stromal adhesion. In contrast, a genetic mutation that reduces the levels of laminin 5 expression has been shown to lead to a skin blistering phenotype (29). It should be noted that excessive granulation tissue production is in fact seen in H-JEB and non-lethal forms of JEB (30,31), particularly affecting facial skin as is the case in LOC (Fig. 1). Thus, complete or partial ablation of the entire laminin 5 trimer also leads to exuberant granulation tissue as would be expected from our results. However, the severe or lethal skin fragility in JEB patients tends to mask this aspect of the phenotype.

The main phenotype in LOC syndrome is chronic production of vascular granulation tissue often with activation and invasive migration of fibroblasts resulting in neovascularization. However, the defective molecule, laminin {alpha}3a, is synthesized only by basal keratinocytes of stratified epithelia, implying that LOC is a disorder of keratinocyte–mesenchymal communication. Since the structural consequences of this laminin {alpha}3a defect are very subtle, it is likely that the aberrant mesenchymal cell behaviour is due to a lost keratinocyte signalling function that resides in the truncated N-terminal domain of laminin {alpha}3a. Thus, identification of the LOC defect may shed some light on how the granulation tissue response is normally controlled. It has been shown previously that the C-terminus of laminin {alpha}3a interacts with the extracellular domains of the integrin {alpha}6ß4 through a series of G-domain repeats (32,33). These C-terminal repeats can be proteolytically cleaved (34,35) in a manner that affects cell behaviour and migration during wound healing (32). Thus, in the basement membrane of skin and other stratified epithelia, the laminin 5 trimer exists in an inverted cruciform configuration with the N-terminus of laminin {alpha}3a orientated towards the mesenchyme (32). It is therefore possible that the N-terminal domain normally acts as a signal to the underlying mesenchyme that the basement membrane is intact. When a stratified epithelium is damaged such that the basement membrane is breached, cytokines and other soluble factors derived from the epithelium will activate a mesenchymal wound healing response in addition to initiating epithelial repair activity. When the wound is closed and re-epithelialized, the basement membrane is re-established and at this stage, the laminin {alpha}3a N-terminal domain may act as a ‘brake’ to deactivate fibroblasts and/or endothelial cells to terminate the mesenchymal wound healing response. In this model, the negative signal is absent in LOC syndrome, and so the granulation tissue response, once activated by physical trauma in certain delicate epithelia, such as the mucosae or ocular epithelia, cannot be deactivated and continues in an uncontrolled manner. Proteolytic cleavage of the C-terminal G-domains of laminin {alpha}3a has been shown to be important in the control of keratinocyte growth and migration (9,33). Furthermore, synthetic peptides analogous to sequences in the N-terminal domains of a number of classical full-length laminin alpha chains, including equivalent sequences specific to the laminin {alpha}3b chains described here, have been shown to modulate endothelial cell attachment, migration and growth (36). However, similar investigations of the specific N-terminus of laminin {alpha}3a have not yet been performed. Here, this domain is shown to be important in control of mesenchymal cell growth and migration. Thus, the cutaneous basement membrane, which has functions in mediating epithelial–stromal adhesion, can also be regarded as having a double-sided signalling function mediated by the N- and C-termini of laminin {alpha}3a.

The critical interactions through which the N-terminus of laminin {alpha}3a might mediate mesenchymal cell signalling are as yet unknown. Possibilities include a direct interaction with a cell surface receptor expressed by mesenchymal cells or alternatively the domain might interact with other extracellular matrix molecules and the resultant macromolecular complex may influence cell behaviour. Granulation tissue is a major clinical problem, not only in LOC syndrome but in a range of much more common conditions such as rheumatoid arthritis, chronic venous leg ulcers, transplant surgery and others. A fuller understanding of the role of laminin {alpha}3a in this sometimes undesirable biological process might lead to the identification of novel targets for therapeutic intervention.


   MATERIALS AND METHODS
TOP
 ABSTRACT
 INTRODUCTION
 RESULTS
 DISCUSSION
 MATERIALS AND METHODS
REFERENCES
 
Homozygosity mapping
Patient and control lymphocyte DNA was extracted by standard techniques. Genome-wide homozygosity mapping was carried out essentially as described previously (5). Briefly, linkage analysis was initially carried out using four affected persons from consanguineous kindreds (families 1–4) using microsatellite markers from the ABI PRISM Linkage Mapping Set, Version 2 (LMS2) and the ABI HD5 5 cM Mapping Set (PE Biosystems, Foster City, CA, USA). Microsatellite PCR was performed according to the manufacturer's recommended protocol, using AmpliTaq Gold polymerase (PE Biosystems) in a 7.5 µl reaction volume, without multiplexing. Markers were sized on an ABI 377 automated DNA sequencer and data extracted and analysed using Genescan and Genotyper version 2.0 software, as previously described (37). When linkage was not detected using families 1–4, the probands from families 5 and 6 were analysed and an additional 60 fluorescent microsatellites were synthesised to fill gaps in the ABI panel (MWG Biotech AG, Ebersberg, Germany). Genetic linkage was initially scored ‘by eye’ and where significant homozygosity was observed, additional families were analysed and two-point LOD scores were calculated with MLINK algorithm of LINKAGE version 5.1. The mutant allele frequency was assumed to be 0.00001 with 100% penetrance and marker allele frequencies were determined experimentally from a population of 50 unrelated unaffected Punjabi Muslims resident in the UK. Homozygosity was confirmed at the LOC locus in an additional family with documented consanguinity (family 7). In families 8–15, where the nature of the consanguinity was very distant or uncertain, calculations were done assuming that the parents were second cousins to simulate ‘fuzzy’ inheritance (5). Recalculation assuming these parents to be more distantly related (third cousins) did not affect the resultant LOD scores. The two intragenic microsatellite markers within the LAMA3 gene, designated LAMA3GT1 and LAMA3GT2 were amplified using the PCR conditions above and the following primer pairs: LAMA3GT1.L (5' TTC TGG TCC TCT AAG AAT AC 3'), LAMA3GT1.R (5'CTT GAC TGC ATC TCC AAC C 3'); and LAMA3GT2.L (5'CAG GCA TTC TCT TAC CAG C 3'), LAMA3GT2.R (5'GGC TGA GAA GCA ATA AGT G 3'). The product sizes for the LAMA3GT1 and LAMA3GT2 markers were ~284 and ~192 bp, respectively.

Cell culture
Keratinocytes and dermal fibroblasts were cultured from skin biopsy material derived from an LOC patient (proband in family 1) and a normal unrelated control, as previously described (38).

cDNA synthesis
mRNA was extracted from cultured cells or skin biopsy material using the Quick prep micro mRNA purification kit (Amersham-Pharmacia Biotech, St Albans, UK) and reverse-transcribed using oligo dT and AMV reverse transcriptase (Promega, Southampton, UK) for 1 h at 45°C. For reverse transcription of long templates, samples were subjected to multiple rounds of reverse transcription, each time heating to 70°C for 10 min to break up RNA secondary structure and then cooling to 45°C with addition of more reverse transcriptase and further incubation at 45°C for 1 h.

PCR
Unless otherwise stated, PCR was performed as follows. Approximately 50 ng of cDNA or 200 ng of genomic DNA were added to a premix containing PCR buffer [67 mM Tris–HCl pH 8.8, 16.6 mM (NH4)2SO4, 1.5 mM MgCl2, 0.17 mg/ml bovine serum albumin (Sigma, Poole, England), and 10 mM 2-mercaptoethanol], 10 nmol of each dNTP, 20 pmol of each primer in a total volume of 50 µl. After an initial denaturation at 95°C for 2 min, 1 unit of Taq polymerase (Promega, Crawley, UK) was added followed by 35 cycles of 95°C for 10 s, annealing temperature for 10 s, 72°C for 30 s, with a final incubation of 72°C for 5 min. For RT–PCR of LAMA3 transcripts, a series of overlapping PCR fragments were generated using the primer pairs listed in Table 1. For direct amplification of exons, primer pairs were positioned within the introns flanking the exonic sequences (Table 2). The annealing temperatures for each primer pair are displayed in Tables 1 and 2. The PCR products were examined by 1% agarose gel electrophoresis, purified and directly sequenced as above. In some cases, particularly fragments involving the GC-rich 5' end of the gene, PCR was performed using Boehringer Long-Range buffer 1 (Roche Diagnostics, Mannheim, Germany) containing 1.75 mM MgCl2. Reactions were subjected to a hot start with 0.2 µl of Boehringer Long-Range DNA polymerase mix (Roche Diagnostics).

DNA sequencing and mutation detection
PCR products were purified for sequencing using QIA quick PCR purification kit (Qiagen, Crawley, UK). Purified fragments were directly sequenced with either the amplification primers or internal primers as appropriate using the Big-Dye Terminator chemistry on ABI 377 or ABI 3100 automated sequencers (Applied Biosystems, Foster City, CA, USA). Mutation detection was carried out by manual comparison of patient sequences with normal control samples. Direct sequencing also was used to look for mutation 151insG in 50 normal unrelated British Caucasians and 50 unaffected unrelated British Punjabi Muslims resident in the UK with no family history of LOC syndrome. Mutation R943X (see Results) was confirmed in members of Family 15 by SalI restriction enzyme digestion of exon 60 PCR products.

Electronic sequence analysis
DNA sequences were aligned, assembled and analysed using Geneworks version 2.5.1 (Oxford Molecular Systems, Oxford, UK). Promoter and transcriptional start site predictions were performed using the TSSG and TSSW algorithms, accessed via the Medical Research Council Human Gene Mapping Project website (http://www.hgmp.mrc.ac.uk). BLAST analysis of DNA sequences was carried out via the National Center for Biotechnology Information website (http://www.ncbi.nlm.nih.gov). Recent assemblies of the human genome data were accessed through the University of California Santa Cruz website (http://genome.ucsc.edu). Additional microsatellite marker primer sequences and genetic mapping data were obtained from Généthon, Evry, France (http://www.genethon.fr) and The Center for Medical Genetics, Marshfield, WN, USA (http://research.marshfieldclinic.org).

Radioimmunoprecipitation assay
Keratinocytes from control and LOC patients were cultured as above in 24-well plates. Keratinocyte culture media was removed and cells were washed four times with PBS before the medium was replaced with 200 µl methionine/cysteine-deficient DMEM (Life Technologies, Paisley, UK) supplemented with 50 µCi [35S]methionine (Amersham Biosciences UK Ltd, Buckinghamshire, UK) and incubated for 16 h at 37°C. Radiolabelled conditioned medium was removed and micro-centrifuged for 5 min at 14 000 rpm to remove cellular debris before storage in aliquots at -20°C. Monoclonal antibody K140 against laminin ß3 (1 µg) was added to 20 µl radiolabelled keratinocyte conditioned medium and incubated overnight at 4°C with rotation. Twelve micrograms per sample of Protein G sepharose beads (Sigma) were swollen in radio-immunoprecipitation assay (RIPA) buffer (50 mM Tris–HCl, 150 mM NaCl, 0.5% NP-40) and incubated with samples for 3 h at 4°C. Beads with bound antibody–protein complexes were sedimented at 5000 rpm for 5 min and washed three times in RIPA buffer. Beads were resuspended in NuPAGE gel loading buffer plus reducing agent (InVitrogen, Groningen, The Netherlands) and heated to 70°C for 10 min before loading on 4–12% NuPAGE Bis-Tris gels (InVitrogen). Gels were fixed in 10% acetic acid : 50% methanol for 10 min and incubated in ‘Amplify’ (Amersham Biosciences) for 30 min before drying and fluorography with Kodak Hyperfilm for 9 days at -80°C.

Immunohistochemistry
Part of a 4 mm punch biopsy from non-lesional thigh skin was embedded in OCT compound (Agar Scientific, Stansted, UK), snap frozen in liquid nitrogen-cooled isopentane and sectioned in a cryostat for subsequent indirect immunofluorescence microscopy, as described previously (39). Sections of normal skin were incubated with the same primary and secondary antibodies as controls. The primary antibodies included the following: GB3 (anti-laminin 5 trimer, SeraLabs, Loughborough, UK); SE144 (anti-laminin 5 {gamma}2 chain, a gift from Gim Meneguzzi, Nice, France); BM2 (anti-laminin {alpha}3, a gift from Bob Burgeson, Boston, USA); K140 (anti-laminin ß3, a gift from Peter Markinovich, Stanford, CA, USA); PKA1 (anti-laminin {alpha}3, gift from Peter Markinovich, Stanford, CA, USA); GOH3 (anti-{alpha}6 integrin, gift from Stephen Kennel, Oak Ridge, TN, USA); 450-9D (anti-ß4 integrin, gift from Arnoud Sonnenberg, Amsterdam, The Netherlands); HD4-233 (anti-type XVII collagen, gift from Katsushi Owaribe, Nagoya, Japan). Appropriate anti-species immunoglobulin secondary FITC antibody conjugates were obtained from Dako, Glostrup, Denmark.

Electron microscopy
Another part of the skin biopsy (above) was fixed in half-strength Karnovsky solution and osmium tetroxide, as detailed elsewhere (40). Ultrathin sections were stained with uranyl acetate and lead citrate and examined in a JEOL 100CX electron microscope.


   ACKNOWLEDGEMENTS
 
This manuscript is dedicated to the memory of our co-author Professor Syed Ghulam Shabbir, who sadly passed away in 2002. The authors would like to express their thanks to the patients, their families and healthy volunteers from the British Punjabi community without whom this research would not have been possible. Thanks also to Andrew J. Cassidy, DNA Analysis Facility, Human Genetics Unit, Ninewells Medical School, for DNA sequencing and genotyping support services. This work was supported by grants from The Dystrophic Epidermolysis Bullosa Research Association (to W.H.I.M. and J.A.M.); Action Research (J.A.M.); The Wellcome Trust (Senior Research Fellowship in Basic Biomedical Sciences to W.H.I.M. supporting W.H.I.M. and F.J.D.S. and a Prize PhD Studentship supporting K.J.H.); and the Hasham M. Jamil Trust (W.H.I.M. and A.D.I.).


   FOOTNOTES
 
* To whom correspondence should be addressed. Tel: +44 1382425618; Fax: +44 1382425619; Email: wmclean{at}hgmp.mrc.ac.uk Back


   REFERENCES
TOP
 ABSTRACT
 INTRODUCTION
 RESULTS
 DISCUSSION
 MATERIALS AND METHODS
 REFERENCES
 

  1. Shabbir, G., Hassan, M. and Kazmi, A. (1986) Laryngo-onycho-cutaneous syndrome: a study of 22 cases. Biomedica, 2, 15–25.[Medline]

  2. Phillips, R.J., Atherton, D.J., Gibbs, M.L., Strobel, S. and Lake, B.D. (1994) Laryngo-onycho-cutaneous syndrome: an inherited epithelial defect. Arch. Dis. Child., 70, 319–326.[Abstract]

  3. Ainsworth, J.R., Spencer, A.F., Dudgeon, J., Geddes, N.K. and Lee, W.R. (1991) Laryngeal and ocular granulation tissue formation in two Punjabi children: LOGIC syndrome. Eye, 5, 717–722.[Medline]

  4. Ainsworth, J.R., Shabbir, G., Spencer, A.F. and Cockburn, F. (1992) Multisystem disorder of Punjabi children exhibiting spontaneous dermal and submucosal granulation tissue formation: LOGIC syndrome. Clin. Dysmorph., 1, 3–14.[Medline]

  5. Hamada, T., McLean, W.H.I., Ramsay, M., Ashton, G.H., Nanda, A., Jenkins, T., Edelstein, I., South, A.P., Bleck, O., Wessagowit, V. et al. (2002) Lipoid proteinosis maps to 1q21 and is caused by mutations in the extracellular matrix protein 1 gene (ECM1). Hum. Mol. Genet., 11, 833–840.[Abstract/Free Full Text]

  6. Uitto, J., Pulkkinen, L. and McLean, W.H.I. (1997) Epidermolysis bullosa: a spectrum of clinical phenotypes explained by molecular heterogeneity. Mol. Med. Today, 3, 457–465.[CrossRef][ISI][Medline]

  7. Fine, J.D., Eady, R.A.J., Bauer, E.A., Briggaman, R.A., Bruckner-Tuderman, L., Christiano, A., Heagerty, A., Hintner, H., Jonkman, M.F., McGrath, J., McGuire, J., Moshell, A., Shimizu, H., Tadini, G. and Uitto, J. (2000) Revised classification system for inherited epidermolysis bullosa: report of the Second International Consensus Meeting on diagnosis and classification of epidermolysis bullosa. J. Am. Acad. Dermatol., 42, 1051–1066.[ISI][Medline]

  8. Uitto, J., Pulkkinen, L. and Ringpfeil, F. (2002) Progress in molecular genetics of heritable skin diseases: the paradigms of epidermolysis bullosa and pseudoxanthoma elasticum. J. Investig. Dermatol. Symp. Proc., 7, 6–16.[CrossRef][ISI][Medline]

  9. O'Toole, E.A. (2001) Extracellular matrix and keratinocyte migration. Clin. Exp. Dermatol., 26, 525–530.[CrossRef][ISI][Medline]

  10. Lohi, J. (2001) Laminin-5 in the progression of carcinomas. Int. J. Cancer., 94, 763–767.[CrossRef][ISI][Medline]

  11. Vidal, F., Baudoin, C., Miquel, C., Galliano, M.F., Christiano, A.M., Uitto, J., Ortonne, J.P. and Meneguzzi, G. (1995) Cloning of the laminin alpha-3 chain gene (lama3) and identification of a homozygous deletion in a patient with herlitz junctional epidermolysis-bullosa. Genomics, 30, 273–280.[CrossRef][ISI][Medline]

  12. Ferrigno, O., Virolle, T., Galliano, M.F., Chauvin, N., Ortonne, J.P., Meneguzzi, G. and Aberdam, D. (1997) Murine laminin alpha3A and alpha3B isoform chains are generated by usage of two promoters and alternative splicing. J. Biol. Chem., 272, 20502–20507.[Abstract/Free Full Text]

  13. Solovyev, V. and Salamov, A. (1997) The Gene-Finder computer tools for analysis of human and model organisms genome sequences. Proc. Int. Conf. Intell. Syst. Mol. Biol., 5, 294–302.[Medline]

  14. Lewin, B. (2000) Genes VII. Oxford University Press, Oxford.

  15. Cooper, D.N. and Krawczak, M. (1993) Human Gene Mutation. BIOS, Oxford.

  16. Marinkovich, M.P., Verrando, P., Keene, D.R., Meneguzzi, G., Lunstrum, G.P., Ortonne, J.P. and Burgeson, R.E. (1993) Basement membrane proteins kalinin and nicein are structurally and immunologically identical. Lab. Invest., 69, 295–299.[ISI][Medline]

  17. Meneguzzi, G., Marinkovich, M.P., Aberdam, D., Pisani, A., Burgeson, R. and Ortonne, J.P. (1992) Kalinin is abnormally expressed in epithelial basement membranes of Herlitz's junctional epidermolysis bullosa patients. Exp. Dermatol., 1, 221–229.[CrossRef][Medline]

  18. Verrando, P., Hsi, B.-L., Yeh, C.-J., Pisani, A., Serieys, N. and Ortonne, J.-P. (1987) Monoclonal antibody GB3, a new probe for the study of human basement membranes and hemidesmosomes. Exp. Cell Res., 170, 116–128.[CrossRef][ISI][Medline]

  19. Moore, J.E., Dua, H.S., Page, A.B., Irvine, A.D. and Archer, D.B. (2001) Ocular surface reconstruction in LOGIC syndrome by amniotic membrane transplantation. Cornea, 20, 753–756.[CrossRef][ISI][Medline]

  20. Marinkovich, M.P., Keene, D.R., Rimberg, C.S. and Burgeson, R.E. (1993) Cellular origin of the dermal-epidermal basement membrane. Dev. Dyn., 197, 255–267.[ISI][Medline]

  21. Cooper, D.N. (1993) Human gene mutation affecting RNA processing and translation. Ann. Med., 25, 11–17.[ISI][Medline]

  22. Carter, W.G., Ryan, M.C. and Gahr, P.J. (1991) Epiligrin, a new cell adhesion ligand for integrin alpha 3 beta 1 in epithelial basement membranes. Cell, 65, 599–610.[CrossRef][ISI][Medline]

  23. Rousselle, P., Lunstrum, G.P., Keene, D.R. and Burgeson, R.E. (1991) Kalinin: an epithelium-specific basement membrane adhesion molecule that is a component of anchoring filaments. J. Cell Biol., 114, 567–576.[Abstract/Free Full Text]

  24. Burgeson, R.E., Chiquet, M., Deutzmann, R., Ekblom, P., Engel, J., Kleinman, H., Martin, G.R., Meneguzzi, G., Paulsson, M., Sanes, J. et al. (1994) A new nomenclature for the laminins. Matrix Biol., 14, 209–211.[CrossRef][ISI][Medline]

  25. Cheng, Y.S., Champliaud, M.F., Burgeson, R.E., Marinkovich, M.P. and Yurchenco, P.D. (1997) Self-assembly of laminin isoforms. J. Biol. Chem., 272, 31525–31532.[Abstract/Free Full Text]

  26. Shimizu, H., Schofield, O.M. and Eady, R.A. (1991) [Prenatal diagnosis of lethal junctional epidermolysis bullosa by fetal skin biopsy]. Nippon Hifuka Gakkai Zasshi, 101, 539–545.[Medline]

  27. Verrando, P., Schofield, O., Ishida-Yamamoto, A., Aberdam, D., Partouche, O., Eady, R.A.J. and Ortonne, J.P. (1993) Nicein (BM-600) in junctional epidermolysis bullosa: polyclonal antibodies provide new clues for pathogenic role. J. Invest. Dermatol., 101, 738–743.[CrossRef][ISI][Medline]

  28. Christiano, A.M. and Uitto, J. (1996) Molecular complexity of the cutaneous basement membrane zone. Revelations from the paradigms of epidermolysis bullosa. Exp. Dermatol., 5, 1–11.[Medline]

  29. Spirito, F., Chavanas, S., Prost-Squarcioni, C., Pulkkinen, L., Fraitag, S., Bodemer, C., Ortonne, J.P. and Meneguzzi, G. (2001) Reduced expression of the epithelial adhesion ligand laminin 5 in the skin causes intradermal tissue separation. J. Biol. Chem., 276, 18828–18835.[Abstract/Free Full Text]

  30. Lim, K.K., Su, W.P., McEvoy, M.T. and Pittelkow, M.R. (1996) Generalized gravis junctional epidermolysis bullosa: case report, laboratory evaluation, and review of recent advances. Mayo Clin. Proc., 71, 863–868.[ISI][Medline]

  31. Carter, D.M., Lin, A.N., Varghese, M.C., Caldwell, D., Pratt, L.A. and Eisinger, M. (1987) Treatment of junctional epidermolysis bullosa with epidermal autografts. J. Am. Acad. Dermatol., 17, 246–250.[ISI][Medline]

  32. Goldfinger, L.E., Hopkinson, S.B., deHart, G.W., Collawn, S., Couchman, J.R. and Jones, J.C. (1999) The alpha3 laminin subunit, alpha6beta4 and alpha3beta1