| Human Molecular Genetics | Pages |
©1999 Oxford University Press |
Splicing modulation of integrin [beta]4 pre-mRNA carrying a branch point mutation underlies epidermolysis bullosa with pyloric atresia undergoing spontaneous amelioration with ageing
Introduction
Results
Clinical observations
Expression of integrin [alpha]6[beta]4 in the proband's keratinocytes
Identification of a point mutation in the paternal [beta]4 allele
Identification of a point mutation in the maternal [beta]4 allele
The branchpoint site mutation 3986-19T->A causes illegitimate splicing of [beta]4 pre-mRNA
Epigenetic factors modulate splicing of the [beta]4 pre-mRNA with mutation 3986-19T->A
Discussion
Materials And Methods
Cells
Electron microscopy and immunohistochemistry
Western blot analysis
Northern blot analysis
RT-PCR amplification of [beta]4 integrin mRNA
Detection, verification and inheritance of the genetic mutations
Construction of ITGB4 microgenes
DNA transfection and analysis of the IGTB4 mRNA transcripts
Ribonuclease protection mapping of the integrin [beta]4 transcripts
Acknowledgements
References
Splicing modulation of integrin [beta]4 pre-mRNA carrying a branch point mutation underlies epidermolysis bullosa with pyloric atresia undergoing spontaneous amelioration with ageing
Received June 9, 1999; Revised and Accepted July 23, 1999
A general improvement with ageing has been reported in a few cases of epidermolysis bullosa with pyloric atresia (PA-JEB), an autosomal recessive skin disease characterized by extensive disadhesion of epithelia. In a patient who improved from severe to mild PA-JEB, asearch for mutations in the integrin [beta]4 gene (IGTB4) detected heterozygosity for a novel base substitution 3986-19T->A in the putative branchpoint sequence of intron 31, and a point mutation 3802+1G->A in the donor splice site of intron 30 previously associated with severe PA-JEB. Analysis of mRNA showed that the intronic mutation prevents legitimate splicing of the [beta]4 pre-mRNA. Functional splicing can be restored in vitro by seeding the proband's keratinocytes on feeders of irradiated fibroblasts. Study of mRNA in wild-type keratinocytes transfected with IGTB4 minigenes containing intron 31 with or without mutation 3986-19T->A, confirmed the causative role of the intronic mutation in PA-JEB, and highlighted the influence of feeders on the maturation process of the mutated [beta]4 pre-mRNA. Our results show that in a context of overall reduction of the [beta]4 mRNA levels, activation of the legitimate splice site in the aberrant [beta]4 pre-mRNA underlies the transient severity of the condition. The results also point to the relevance which the interaction between epithelial and stromal cells may have in modulating expression of integrin receptors.
INTRODUCTION
Junctional epidermolysis bullosa (JEB) is a heterogeneous group of autosomal recessive bullous disorders of the skin and mucous membranes characterized by blister formation after trauma. In the blisters, the plane of the mesenchymal-epithelial separation lieswithin the lamina lucida of the basement membrane zone (BMZ). Based on morphology and body distribution of the lesions, the different forms of JEB have been classified into defined clinical subgroups (1). The severe forms of JEB manifest at birth with widespread blistering and erosions of the integument and oral mucosa. Involvement of the gastrointestinal, genitourinary and respiratory epithelia is also observed. Fluid loss, growth retardation and anaemia are also common. On the basis of clinical and molecular genetic findings, a variant of severe JEB associated with congenital pyloric atresia was designated PA-JEB (MIM 226730) (1,2). Congenital aplasia of the skin, and urethrovesical occlusion, are clinical signs often observed in severe PA-JEB. PA-JEB may also manifest as a mild skin disorder in which aplasia cutis is absent and blister tendency is less pronounced than in the severe form of the disease (3,4). In mild PA-JEB, surgical treatment of pyloric atresia results in long-term survival.
In addition to the severe and mild forms of PA-JEB, rare cases have also been reported in which neonatal PA-JEB undergoes general improvement with ageing (3,5). In the transient forms of PA-JEB, remission of blistering is observed with increasing age, although skin fragility may persist in different degrees.
PA-JEB has been associated with mutations in the genes ITGA6 or ITGB4 that encode the [alpha]6 and [beta]4 subunits of integrin [alpha]6[beta]4, respectively (2,6). Integrin [alpha]6[beta]4 is an adhesion cell receptor localized in the hemidesmosomes (HDs), specialized adhesion structures adjoining the basal cells to the underlying mesenchyme in stratified and transitional epithelia (7). The HDs associate intracellularly with the cytokeratin intermediate filaments and extracellularly with the anchoring filaments of the BMZ. Integrin [alpha]6[beta]4 plays a critical role in the assembly and stability of HDs, and abnormal expression of this protein results in malformation or absence of HDs (2,6,8).
[beta]4 is unique among the integrin [beta] subunits, because of its unusually long cytoplasmic domain (~1000 amino acids) which is connected to the cytokeratin network and harbours the domains essential to the interaction of integrin [alpha]6[beta]4 with the cellular components(collagen type XVII and plectin) of the HDs (9,10). Such interactions lead to the nucleation of the HD in vivo and the formation of the HD-like adhesion structures (SACs) that are assembled in vitro by keratinocytes. Integrin [alpha]6[beta]4 is also detected in epithelia that do not possess HDs, which may reflect a multifunctional role of the protein (7). In this respect, five distinct and tissue-specific variants of [beta]4 have been identified, but the functional role of these polypeptides, all generated by alternative splicing of the ITGB4 RNA transcript, has not yet been elucidated. The mechanisms regulating alternative splicing of the [beta]4 pre-mRNA are also obscure.
In this study, we demonstrate that compound heterozygosity for integrin [beta]4 gene mutations underlie PA-JEB in a patient presenting dramatic amelioration of the neonatal disorder. Our results suggest that cellular factors can influence the effect of splice site mutations in genes associated with severe EB.
RESULTS
Clinical observations
The proband was a 14-year-old boy, the child of the non-consanguineous union of clinically non-affected parents with an older healthy son. At birth, the patient presented with the hallmarks of severe PA-JEB including extensive skin blistering, pyloric atresia and urethrovesical occlusion (5). Immunohistological analysis of lesional skin located the cleavage plane of the blisters within the lamina lucida of the BMZ. As the child grew, the blistering tendency markedly decreased and the proband's skin and epithelia acquired resistance to trauma.
At the age of 14, induction of blisters required prolonged rubbing of the skin. Ultrastructural examination of the skin showed mature HDs connected to the cytokeratin intermediate filaments and presenting poorly formed sub-basal dense plates. The anchoring filaments of the lamina lucida were clearly visible (Fig. 1A). Immunostaining of non-involved skin, performed with a panel of antibodies directed against the antigens of the dermal-epidermal junction, revealed a slightly reduced immunoreactivity of integrin [alpha]6 and a marked decrease in integrin [beta]4 immuno- staining (Fig. 1B). The reactivity of all the other known components of the BMZ was comparable with that observed with normal control skin (data not shown).
Figure 1. Identification of the defective expression of integrin [alpha]6[beta]4 in the PA-JEB patient. (A) Electron microscopy of traumatized mild PA-JEB skin shows slight widening of the lamina lucida (asterisk) between well-formed hemidesmosomes (hd), connected to the keratin intermediate filaments (if) and bridged to the lamina densa (ld) by anchoring filaments (af). Bar, 20 µm. (B) Immunofluorescence analysis of non-involved areas of the proband's skin (a and c) using the anti-[beta]4 pAb 439-9B (a and b) and the anti-[alpha]6 mAb GoH3 (c and d) shows reduced staining of the DEJ compared with the skin of a healthy control (b and d). Bar, 48 µm. (C) Northern analysis of control (lane 1) and mild PA-JEB (lane 2) keratinocytes using a 32P-labelled [beta]4 cDNA probe shows a reduced expression of integrin [beta]4 mRNA in the proband. G, hybridization signal of a GAPDH 32P-labelled probe. (D) Western analysis of extracts obtained from control (lane 1) and mild PA-JEB (lane 2) keratinocytes using an anti-[beta]4 pAb and the anti-tubulin mAb T-4026 shows a reduced expression of the integrin [beta]4 in proband's cells.
Expression of integrin [alpha]6[beta]4 in the proband's keratinocytes
The expression of integrins [alpha]6 and [beta]4 was assessed by northern blot analysis of total RNA purified from cultures of the proband's and control keratinocytes. Hybridization with a probe for integrin [alpha]6 resulted in a signal of normal intensity (data not shown), whereas hybridization with a cDNA encoding integrin [beta]4 was markedly reduced in the proband (Fig. 1C). Consistent with these observations, immunoblotting analysis of the PA-JEB keratinocytes with the anti-[beta]4 polyclonal antibody detected a band 70% less intense than that observed in the controls (Fig. 1D).
Identification of a point mutation in the paternal [beta]4 allele
RT-PCR amplification of total RNA purified from cultures of the proband's keratinocytesusing pairs of primers producing overlapping amplimers, followed by heteroduplex analysis, showed heteroduplex bands for the cDNA fragment spanning nt 3403-4118 (GenBank accession no. X52186) (11; data not shown). Direct sequencing of the PCR amplimer identified two superimposed sequences suggestive of heterozygosity for the corresponding alleles. Careful inspection of the sequence revealed an in-frame 51 bp deletion (nt 3752-3802) located at the 3[prime] end of exon 30 (Fig. 2A). This deletion predicts the synthesis of a [beta]4 polypeptide with an internal deletion of 17 residues ([Delta]17-[beta]4), previously identified in a case of severe PA-JEB (2). PCR amplification of exons 30 and 31 of the proband's genomic DNA (GenBank accession nos U66529-U66541) (12) and sequence analysis detected a heterozygous 3802+1G->A transition at position +1 of exon 30 of ITGB4 (Fig. 2B). The presence of mutation 3802+1G->A was searched for in the genomic DNA of the members of the family. Hybridization with a wild-type and mutated allele-specific oligonucleotide revealed that the proband and the proband's father were heterozygous for the mutation (Fig. 2C).
Figure 2. Identification of mutation 3802+1G->A in the proband's paternal allele of the ITGB4 gene. (A) Sequencing of the cDNA amplimers spanning nt 3403-4118 (715 nt) of the [beta]4 cDNA sequence. In comparison with the wild-type cDNA sequence (top), a 51 bp segment corresponding to nt 3752-3802 is missing (bottom). (B) Sequencing of the genomic DNA from the PA-JEB patient revealed a heterozygous 1 bp transition (G->A) at the 5[prime] donor splice site of exon 30 and designated mutation 3802+1G->A. (C) (Top) Pedigree of the PA-JEB kindred. (Bottom) Inheritance of the [beta]4 integrin allele bearing the mutation 3802G->A in the proband's family. A DNA fragment corresponding to the mutated region was PCR amplified using genomic DNA from each member of the family as a template. Hybridization with a wild-type and a mutated ASO demonstrated that the proband and the proband's father are heterozygous carriers for the mutated allele. C, unrelated healthy control. (D) Schematic representation of the misplicing event caused by mutation 3802+1G->A in [beta]4 pre-mRNA. Mutation 3802+1G->A disrupts the 5[prime] donor splice site of exon 30. Activation of a putative cryptic donor splice (GTGAG) within exon 30 leads to an aberrant splicing resulting in the in-frame deletion of 51 bp affecting the 3[prime] end of the exon.
Identification of a point mutation in the maternal [beta]4 allele
Because the heteroduplex analysis of the proband's mRNA failed to detect sequence variations in the maternal transcripts, we hypothesized that the maternal [beta]4transcripts were under-represented. We therefore specifically synthesized two overlapping cDNA fragments homologous to the full-length transcripts of thematernal allele using a long-range allele-specific RT-PCR procedure (see Materials and Methods). The resulting 1.7 kb cDNA fragment spanning the 3[prime] sequence (nt 3754-5495) of the [beta]4 cDNA was cloned in the bacterial vectorpTAg.Sequencing of 10 distinct clonesdetected a wild-type cDNA sequence in six of them, and a 38 bp out-of-frame deletion in the 5[prime] region of exon 32 (nt 3986-4023) in the remaining samples (Fig. 3A). The deletion results in a downstream premature stop codon (TGA) at nt 4061-4063 and predicts a truncated [beta]4 polypeptide terminating at residue 1324. PCR amplification of the genomic DNA of the proband using primers selected on the basis of the cDNA sequence flanking the 38 bp deletion followed by sequence analysis identified a heterozygous T->A transversion (3986-19T->A) located at position -19 of exon 32 (11) (Fig. 3B). Allele-specific oligonucleotide (ASO) hybridization analysis confirmed the heterozygous status of the proband. The proband's mother, the maternal uncle and grandmother were heterozygous carriers of the mutation (Fig. 3C). The possibility that the T->A transversion represents a polymorphic variation in the population was tested by screening of PCR-amplified genomic DNA from 100 unrelated healthy individuals. The mutation was not found in any of these samples.
Figure 3. Identification of mutation 3986-19T->A in the maternal [beta]4 allele. (A) Nucleotide sequencing of the cDNA reverse-transcribed from the [beta]4 mRNA produced by the maternal allele identified a 38 bp out-of-frame deletion (nt 3986-4023) (bottom) absent in the wild-type controls (top). (B) Sequencing of the genomic DNA from the PA-JEB patient (bottom) revealed a heterozygous one base-pair transversion (T->A) at position -19 of exon 32 of gene ITGB4 that disrupts the putative branching point sequence of intron 31. In the electropherogram of the wild-type DNA sequence (top), the putative branching adenosine is underlined. (C) Inheritance of mutation 3986-19T->A in the PA-JEB kindred. A genomic DNA fragment (300 bp) containing the mutation was PCR-amplified from each member of the family. Hybridization with a wild-type and mutated ASO demonstrated the Mendelian inheritance of the mutated allele. C, unaffected control.
The branchpoint site mutation 3986-19T->A causes illegitimate splicing of [beta]4 pre-mRNA
To confirm the disruptive effect of mutation 3986-19T->A on maturation of the mRNA coding for integrin [beta]4, we constructed two recombinant microgenes expressing the mRNA transcript encoded by exons 31 and 32 of ITGB4 (Fig. 4A). Genomic DNA fragments (401 bp) encompassing exon 31 (182 bp), intron 31 (88 bp) and exon 32 (131 bp) of ITGB4 were specifically PCR-amplified using the proband's DNA as a template. The amplimers were inserted in the eucaryotic expression vector pcDNA3 to obtain the microgene construct pµ[beta]4-M, carrying mutation 3986-19T->A, and the wild-type counterpart pµ[beta]4-WT. The two constructs were transfected into normal human keratinocytes seeded on plastic dishes. RT-PCR analysis of the microgene transcripts isolated from the cell cultures was performed using primers I31L and Sp6R (Fig. 4A). In the cells transfected with plasmid pµ[beta]4-WT, a 402 bp cDNA band was observed that corresponded to the microgene mRNA transcribed from exons 31 and 32 (Fig. 4A, lane 1). A faint 490 bp band was also detected, which corresponded to the pre-mRNA transcripts containing intron 31. In the case of keratinocytes expressing plasmid pµ[beta]4-M, a unique 490 bp band was found, that resulted from the retention of the mutated intron 31 of IGTB4 (Fig. 4A, lane 2). The microgene cDNAs were further characterized by clonal analysis. cDNAs amplified from the keratinocyte cultures expressing the microgene pµ[beta]4-M were cloned and amplified in E.coli. Plasmid DNAs isolated from 50 distinct colonies were analysed by direct nucleotide sequence. A total of 48 colonies contained the 490 bp cDNAs retaining the mutated intron 31, while two of them contained cDNAs carrying the 38 bp deletion corresponding to the deletion detected in the proband's keratinocytes (data not shown). These results confirm that besides intron retention, mutation 3986-19T->A induces illegitimate splicing of the [beta]4 pre-mRNA by activation of a cryptic splice site present in exon 32.
Figure 4. Expression of integrin [beta]4 in the patient's keratinocytes is modulated by epigenetic factors. (A) Transfection of wild-type human keratinocytes using an artificial [beta]4 microgene carrying mutation 3986-19T->A. (Top) Schematic representation of the [beta]4 microgenes. Genomic DNA fragments spanning exons 31-32 were cloned into a eucaryote expression vector to obtain microgenes pµ[beta]4-M, carrying mutation 3986-19T->A, and the wild-type counterpart pµ[beta]4-WT. Positions of primers I31L and SP6R are indicated. The transcription products of pµ[beta]4-WT (402 bp) and pµ[beta]4-M (490 bp) are represented. (Bottom) RT-PCR analysis of the mRNAs transcribed from pµ[beta]4-WT (lanes 1, 3 and 5) and pµ[beta]4-M (lanes 2, 4 and 6) by transfected human keratinocytes grown on plastic (lanes 1 and 2) or on feeders (lanes 3 and 4) and by irradiated 3T3-J2 fibroblasts transfected with pµ[beta]4-WT (lane 5) and pµ[beta]4-M (lane 6). C, RT-PCR amplification in the absence of RNA template. The migration position of the spliced (402 bp) and unspliced (490 bp) cDNAs are indicated on the right. The bottom frame shows the results of RT-PCR amplification of GAPDH mRNA from the cell culture analysed. (B) Immunofluorescence of normal (a and b), mild PA-JEB (c and d) and severe (e and f) PA-JEB keratinocytes grown on feeders (a, c and e) and on plastic (b, d and f) using pAb 450-11A. The mild PA-JEB keratinocytes grown on a feeder display the dotty staining characteristic of cells assembling SACs, while on a plastic support the cells display a diffuse staining. Independently from the culture conditions, wild-type keratinocytes nucleate SACs, whereas the keratinocytes obtained from an unrelated severe PA-JEB patient expressing the [Delta]17-[beta]4 polypeptide do not assemble SACs. Bar, 24 µm. (C) Ribonuclease protection mapping of the [beta]4 mRNAs expressed by the proband's keratinocytes. Total RNA purified from wild-type (lane 1) and mild (lane 2) PA-JEB keratinocytes grown on a feeder were hybridized with an antisense [beta]4 riboprobe spanning the region of the [beta]4 mRNA affected by the paternal ([Delta]51) and the maternal ([Delta]38) mutations. The presence of a protected 359 bp band both in the control and the patient's keratinocytes denoted the synthesis of a wild-type [beta]4 transcript. No protected band is detected with the [Delta]51-[beta]4 transcript synthesized by an unrelated PA-JEB patient (2) (lane 3). Yeast tRNA was used as negative control (lane 4). Migration of undigested [beta]4 riboprobe (lane 5) and that of molecular weight markers (lane M) are shown. (Bottom) mRNA protected by a 255 bp [beta]-actin riboprobe. (D) Mutation 3986-19T->A affects the legitimate branchpoint sequence of intron 31. Activation of a putative cryptic branchpoint (GGCTCAC) and acceptor splice site (CAG) within exon 32 results in an out-of-frame deletion of 38 bp affecting the 5[prime] end of the same exon.
Epigenetic factors modulate splicing of the [beta]4 pre-mRNA with mutation 3986-19T->A
Mouse 3T3-J2 keratinocytes constitute the optimal support of keratinocyte growth in vitro (13). Irradiated J2 fibroblasts secrete both extracellular matrix proteins that favour keratinocyte attachment and growth factors that stimulate proliferation (14 and references therein). The patient's primary keratinocytes cultured on J2 feeders were immunoreactive to the antibodies directed against integrin [alpha]6[beta]4. The fluorescence presented the basal `Swisscheese-like' pattern characteristic of keratinocytes assembling SACs (15) (Fig. 4B, a and c). Conversely, the patient's keratinocytes grown on plastic showed a diffuse labelling, suggesting that integrin [alpha]6[beta]4 did not concentrate into SACs (Fig. 4B, b and d). Indeed, this staining was similar to the labelling pattern observed in the keratinocytes of an unrelated patient with severe PA-JEB which produced low amounts of abnormal [beta]4 polypeptides ([Delta]17-[beta]4) unable to nucleate into SAC (2,16) (Fig. 4B, e and f). Transfer of the proband's keratinocytes from a feeder layer to plastic, or to extracellular matrix(ECM) produced by wild-type human keratinocytes grown on feeder layers, induced loss of the `Swisscheese-like' pattern. Feeding of these cells with culture medium conditioned by wild-type human keratinocytes grown on feeder layers did not restore the fluorescence (data not shown). The staining pattern was not influenced by the temperature at which the cells were grown (data not shown). In all the cell cultures, immunoreactivity of laminin-5 was as strong as in wild-type human keratinocytes (data not shown).
To assess whether the presence of SACs correlated with synthesis of wild-type [beta]4 transcripts, total RNA isolated from the proband's keratinocytes grown on a feeder layer was analysed by RNase protection assay. The antisense riboprobe (359 nt) spanned exons 30-32 (nt 3721-4080) and encompassed the regions of 51 and 38 nt deleted in the paternal and maternal [beta]4 mRNAs, respectively. In control and PA-JEB mRNAs, the riboprobe protected a unique cDNA fragment of 359 nt, which revealed the presence of wild-type [beta]4 mRNA transcripts in the patient cells (Fig. 4C, lanes1 and 2). No protected [beta]4 mRNA was observed in the case of the PA-JEB keratinocytes grown on plastic (data not shown). These data, therefore, demonstrated that the mutated ITGB4 allele carrying the transversion 3986-19T->A can generate wild-type RNA transcripts in cells grown on a feeder layer.
Since homozygosity for the paternal mutation 3802+1G->A is associated with lethal PA-JEB (12) it is conceivable that the wild-type mRNA detected in the patient's cultured keratinocytes is synthesized from the maternal allele, and that the culture conditions influence the steady-state level of this transcript. To verify this possibility, pµ[beta]4-WT and pµ[beta]4-M were transiently transfected into wild-type human keratinocytes seeded on feeders. The microgene cDNAs produced were analysed by RT-PCR amplification of total RNA isolated from the cell cultures and by electrophoresis on an agarose gel. With the keratinocytes transfected with plasmid pµ[beta]4-M, two bands of 402 and 490 bp of comparable intensity were detected (Fig. 4A, lane 4) instead of the single 490 bp cDNA found in cells grown on plastic (Fig. 4A, lane 2). Conversely, in the cells expressing plasmid pµ[beta]4-WT, growth on feeders did not influence the size or intensity of the microgene cDNAs (Fig. 4A, lanes 1 and 4). These results demonstrated that the culture conditions modulate the rate of illegitimate splicing driven by the 3986-19T->A mutation in the branching-point sequence of intron 31 of IGTB4.
DISCUSSION
Genetic mutations of integrin [alpha]6[beta]4 cause PA-JEB, a recessively inherited genodermatosis characterized by disadhesion of the integument and pyloric atresia. Mutations in [beta]4 underlie most of the PA-JEB cases elucidated so far. Premature termination codons are associated with the lethal variants of the disease, while missense and splice site mutations enabling synthesis of abnormal [beta]4 polypeptides cause the non-lethal forms (17,18).
In this study, we describe a patient presenting severe PA-JEB at birth, followed by an amelioration of the symptoms during childhood. The mild skin fragility correlated with impaired functions of integrin [alpha]6[beta]4 following the downregulation of [beta]4 expression at the mRNA and protein level. Mutation analysis disclosed compound heterozygosity for IGTB4 gene mutations affecting splicing of the [beta]4 pre-mRNA and underlying this rare form of PA-JEB.
The mutation 3802+1G->A, affecting the paternal allele, is known to cause lethal PA-JEB (12). This genetic defect disrupts the 5[prime] donor splice site of intron 30 in the IGTB4 gene and activates a putative cryptic splice site (GTGAG) within exon 30 (Fig. 3D). This leads to the production of an aberrant [beta]4 mRNA with an in-frame 51 nt deletion at the 3[prime] end of exon 30, and results in the synthesis of an abnormal [beta]4 polypeptide ([Delta]17-[beta]4) with a 17 amino acid deletion in the second fibronectin-like repeat of the cytoplasmic tail (2). In a recent study, the aberrant polypeptides [Delta]17-[beta]4 were shown to associate with integrin [alpha]6 into [alpha]6[beta]4 heterodimers that fail to nucleate functional HDs and cause defective keratinocyte adhesion which results in severe PA-JEB (16).
The transversion 3986-19T->A is a novel mutation affecting the putative branchpoint site (GGCTCAC) of intron 31 of the maternal IGTB4 allele. This thymidine substitution results in the maturation of aberrant mRNAs generated by activation of putative cryptic branchpoint (CCTGAC) and acceptor splice sites (cag) located within exon 32. The abnormal mRNA shows an out-of-frame deletion of 38 bp in the 5[prime] end of exon 32 leading to a downstream premature termination codon (PTC) and rapid decay of the mutated RNA transcripts. Because the paternal mutation 3802+1G->A in ITGB4 causes severe PA-JEB, we hypothesized that the mild clinical phenotype of the proband is determined by the nature of the branchpoint mutation 3986-19T->A affecting the maternal allele. Substitution of the conserved thymidine in the branchpoint sequence is predicted to hamper the Watson-Crick pairing between the abnormal [beta]4 pre-mRNA and the 5[prime] end of the U2 small nucleolar RNA (snRNA) (19). Further interaction between U2 and the U6 snRNA that constitutes a critical step for exon definition and lariat formation is hindered (20). As a result, assembly of the spliceosome and scrutiny of the coding sequences of the maternal [beta]4 pre-mRNA are likely to be altered.
Indeed, RNase protection assays of the RNA transcripts isolated from primary cultures of the proband's keratinocytes demonstrated that a fraction of the mutated [beta]4 pre-mRNA transcribed from the maternal allele escapes incorrect splicing and generates a wild-type mRNA. Production of wild-type [beta]4 polypeptide and synthesis of functional [alpha]6[beta]4 heterodimers is consistent with the immunoreactivity of the proband's skin to anti-integrin [alpha]6[beta]4 antibodies and the assembly of mature HDs connected to the intermediate filaments of the cytoskeleton. The low expression level of functional [alpha]6[beta]4 molecules also correlates well with the residual fragility of the integument.
A limited number of human diseases have thus far been associated with genetic mutations affecting branchpoint sites, possibly because the branchpoint consensus sequence (YNYTRAY) is loosely conserved in eucaryotes (21). Indeed, only the thymidine at position4 and the adenosine at position 6 are preserved during evolution (22). Recently, it has been reported that substitution of the conserved thymidine in the branchpoint sequence of intron 4 of the lecithin:cholesterol acyltransferase (LCAT) gene induces intron retention and causesLCAT deficiency (23). In our patient, mutation 3986-19T->A did not induce retention of intron 31 in the [beta]4 mRNA. Transient transfection of IGTB4 minigenes that contained intron 31 with or without mutation 3986-19T->A in normal human keratinocytes confirmed the causative role of this intron mutation in PA-JEB. In fact, intron retention was observed in vitro in the majority of the mutated minigene RNA transcripts, while out-of-frame deletion of 38 bp in exon 32 was detected in a residual fraction of mRNAs. In remains unclear why in the cells transfected with the mutant minigene, intron retention rather than activation of illegitimate splicing constitutes the main effect of the intronic T->A transition. It is likely that distal regulatory sequences that favour activation of the cryptic splicing of the [beta]4 mRNA are deleted in the short minigene construct (24). Nevertheless, the production of low amounts of the internally deleted minigene mRNA mirrors the effect that mutation 3986-19T->A exerts in vivo on maturation of the [beta]4 pre-mRNA, and confirms the role of this intronic mutation in the aberrant expression of thematernal ITGB4 allele. These ex vivo observations also demonstrate that substitution of the thymidine in the branchpoint consensus sequence can activate alternative cryptic splice sites and promote differential splicing when alternative branchpoint sequences are present in the downstream exonic sequences. These results, therefore, are consistent with previous reports describing use of alternative splice sites in genes with mutations affecting putative branchpoint sites of human pre-mRNA (25,26).
The transfection experiments also helped us to interpret the molecular events underlying the benign outcome of PA-JEB in the proband. Analysis of theproband's keratinocytes revealed that epigenetic factors could bias the effect of the [beta]4 intronic mutation by acting on the splicing machinery, because the cell cultures expanded on feeders activated synthesis of wild-type [beta]4 mRNA and nucleated SACs. The finding that legitimate splicing of the mutated minigene transcripts was restored by seeding the transfected keratinocytes on a feeder layer, correlated with the identification of functional [beta]4 mRNA in the proband's keratinocytes and with the presence of mature HDs in the skin.
Information on the expression of integrin [beta]4 in the proband's skin before clinical improvement and ultrastructural normalization of the BMZ observed with ageing was not available. However, it is tempting to speculate that in the context of an overall reduction of the [beta]4 mRNA levels, activation with ageing of the legitimate splice site in the aberrant pre-mRNA transcribed from the maternal ITGB4 allele underlies the transition from severe to mild PA-JEB. This possibility would agree with the observation that in Nagase analbuminogenic ratsbearing a genetic mutation at a 5[prime] splice site of the albumin gene, ageing is associated with changes in the splicing pattern of the corresponding aberrant pre-mRNA (27). Moreover, a splice site mutation abolishing the obligatory consensus 3[prime] acceptor splice sitein COL7A1 has recently been associated with the a case of transient bullous dermolysis of the newborn, a skin blistering disorder that rapidly ameliorates in infancy (28). Taken together, these observations suggest that subsets of splice mutations underlie the clinically transient phenotypes observed in inherited pathologies. In light of our results, we propose that improved splicing of specific mutant genes may depend on the production/activation of cellular factors along with ageing. At present, the nature of the factor(s) produced by the irradiated feeders that modifies splicing of the mutated [beta]4 allele in our PA-JEB patient remains to be elucidated. However, the observation that splicing activation requires a close interaction between the mutated keratinocytes and the J2 fibroblasts excludes a role of the ECM, and indicates that the possible candidates are cell surface proteins or labile cytokines exerting a labile paracrine effect.
With respect to the clinical heterogeneity, diagnosis, pathogenesis and therapy of epidermolysis bullosa, our data contribute to the establishment of a correlation between specific mutations and phenotypic characteristics. They also highlight the relevance that interactions between epithelial and stromal cells may have in modulating the expression of cell surface receptors.
MATERIALS AND METHODS
Cells
Primary human keratinocytes were isolated from skin biopsies. The cells were cultured on feeder layers of irradiated mouse 3T3-J2 fibroblasts, on keratinocyte ECM, or directly on plastic dishes in the presence of DMEM/F12 (2:1) medium (Life Technologies, Cergy-Pontoise, France) supplemented with 10% fetal calf serum, 4 mM glutamine, 5 µg/ml insulin, 0.4 µg/ml hydrocortisone, 2 nM cholera toxin, 10 ng/ml epidermal growth factor, 2 nM triiodothyronine and 0.18 mM adenine (13). The ECM was prepared by treating keratinocyte cultures grown on 3T3-J2 feeders with phosphate-buffered saline (PBS), pH 7.4, 10 mM EDTA for 10 min at 37°C. After detachment of the cells, the ECM was rinsed three times with PBS containing 1 mM CaCl2 and 1 mM MgCl2.
Electron microscopy and immunohistochemistry
Ultrastructural examination of involved and non-involved PA-JEB skin was performed as described elsewhere (2). Immunofluorescence analysis of skin biopsies and keratinocyte cultures has been detailed previously (29). Expression of integrin [beta]4 was monitored using monoclonal antibodies (mAb) 439-9B and 450-11A (30). Immunomapping of the basement membrane components was performed using mAb GB3 raised against native laminin-5 (31), mAb HD121, directed against HD1 (32), mAb FP1 and 1A8C, specific to bullous pemphigoid antigens of 230 kDa (33) and 180 kDa (34), respectively. Anti-integrin [alpha]6 was mAb GoH3 (35).
Western blot analysis
SDS-PAGE was performed using 7.5% polyacrylamide gels (36). Blotting onto nitrocellulose membranes and immunostaining with an anti-[beta]4 polyclonal antibody (37) and mAb T-4026 (Sigma, St Louis, MO) specific to tubulin was achieved following standard procedures using the ECL detection system (Amersham France, Les Ulis, France).
Northern blot analysis
Total RNA was extracted from cultured keratinocytes using the RNable extraction kit according to the manufacturer's recommendations (Eurobio, Les Ulis, France). Twenty micrograms of total RNA was electrophoresed in a 1% agarose-formaldehyde gel, and blotted onto a nylon membrane (Amersham France). The membranes were hybridized with a 1.9 kb integrin [beta]4 and a 983 bp GAPDH 32P-random-labelled cDNA probes obtained by RT-PCR amplification of total RNA purified from wild-type human keratinocytes. For amplification of the [beta]4 cDNA (GenBank accession no. X52186) (11), the primer pair was (L) 5[prime]-GCGACTATGAGATGAAGGTG-3[prime], (R) 5[prime]-AGGATGGAGTAGCTGAGGAG-3[prime], and the PCR cycling conditions: 94°C, 3 min; 94 °C, 20 s; 58°C, 45 s; 72°C, 2 min (30 cycles). For the GAPDH cDNA probe (GenBank accession no. M33197), primers were (L) 5[prime]-AGATCCCCTCCAAAATCAAGT-3[prime], (R) 5[prime]-TAGGCCCCTCCCCTCTTCA-3[prime].
RT-PCR amplification of [beta]4 integrin mRNA
To search for mutations, total RNA was extracted from PA-JEB and control keratinocytes grown on a plastic support. Ten micrograms of total RNA was reverse-transcribed in the presence of oligodeoxythymidine primers as recommended by the manufacturer (Promega, Charbonnière, France). To detect the RNA transcript of the paternal ITGB4 allele, PCR reactions were performed in the presence of Taq DNA polymerase (Life Technologies), using 0.1 µl of reverse transcription product, and 11 pairs of oligonucleotide primers synthesized on the basis of the integrin [beta]4 cDNA sequence (11). Heteroduplex analysis was performed using the conformation-sensitive gel electrophoresis method (38). The specific primers used to detect the 3802+1G->A mutation were: (L) 5[prime]-AACCCCAATGCTAAGGCCG-3[prime], (R) 5[prime]-TCAGTGTCATCGGAGACG-3[prime]. The PCR cycling conditions were: 94°C, 3 min; 94°C, 20 s; 58°C, 45 s; 72°C, 30 s (30 cycles). Direct sequencing of the PCR products was performed usingan ABI PRISM Dye Terminator Cycle Sequencing kit (Perkin Elmer, Foster City, CA). To detect the maternal mutation, 1 µl of the reverse transcription product was submitted to two long-range allele-specific PCR amplifications spanning the entire reading frame of the [beta]4 cDNA, using primers homologous to nucleotide sequences comprised within the 51 bp deletion found in the transcripts of the paternal [beta]4 allele. Amplifications were performed using the Expand Long Template PCR System (Boehringer Mannheim, Meylan, France). To amplify the [beta]4 cDNA fragment bearing the maternal 38 bp deletion, the following oligonucleotides were used: (L) 5[prime]-TGAGATCACAGCCTTACGAG-3[prime], (R) 5[prime]-AGGATGGAGTAGCTGAGGAG-3[prime]. The PCR cycling conditions were: 94°C, 3 min; 94°C, 20 s; 58°C, 45 s; 68°C, 1.5 min (10 cycles), 94°C, 20 s; 58°C, 45 s; 68°C, 1.5 min (20 cycles, with a 20 s increment in extension time at each stage). The amplicon was then subcloned into a pTAg vector according to the manufacturer's recommendations (R&D Systems, Abingdon, UK) and sequenced.
Detection, verification and inheritance of the genetic mutations
PCR reactions of genomic DNA (100 ng) purified from the consenting members of the PA-JEB kindred and of unrelated controls were performed using standard conditions (39). To detect mutation 3802+1G->A, theprimers were: (L) 5[prime]-CACTGACAGCACTCTTCCTG-3[prime], (R) 5[prime]-CAGGAAGAGTGCTGTCAGTG-3[prime] (GenBank accession no. U66539), and to detect the mutation 3986-19T->A, theprimers were: (L) 5[prime]-GCCCAAGAGGCCCATGTCCA-3[prime], (R) 5[prime]-GTGGCCCCTTTGCCTTTTGT-3[prime] (GenBank accession no. U66538). To determine the segregation of these genetic mutations in the kindred, a standard ASO hybridization protocol was used (2). For mutation 3802+1G->A, the ASOs were 5[prime]-TGACAACCGTAAGAACC-3[prime] (wild-type) and 5[prime]-TGACAACCATAAGAACC-3[prime] (mutant); and for the mutation 3986-19T->A, 5[prime]-CCCTGGCTCACTCCC-3[prime] (wild-type) and 5[prime]-CCCTGGCACACTCCC-3[prime] (mutant).
Construction of ITGB4 microgenes
DNA fragments (401 bp) of the ITGB4 gene were obtained by PCR amplification of the region encompassing exon 31, intron 31 and exon 32, using the proband's DNA as a template and a primer pair synthesized on the basis of the cDNA sequence of the gene (GenBank accession no. X52186). Primer (L) (designated as I31L) was: 5[prime]-ATGCGACCTATTGGGCCCATGAG-3[prime] (nt 3803-3821 in the cDNA) and comprised four bases (underlined) generating a methionine codon in-frame with the open reading frame of exon 31. Primer (R) was 5[prime]-AGTGTCATCGGAGACGCTGGGCCT-3[prime] (nt 4093-4117 in the cDNA). PCR amplification was performed using Pfu DNA polymerase following the supplier's recommendations (Stratagene, La Jolla, CA), and the cycling conditions were: 94°C, 3 min; 94 °C, 20 s; 55°C, 45 s; 72°C, 1 min. One hundred micrograms of amplification products were ligated to 50 ng of expression vector pcDNA3 (InVitrogen, Groningen, The Netherlands) digested with the restriction enzyme EcoRV (New England Biolabs, Hitchin, UK) using 2 × 105 U of T4 DNA ligase (New England Biolabs) and amplified in competent E.coli, strain XL1 blue (Stratagene). Screening of transformants was performed by PCR analysis of total bacterial DNA extracted from 30 isolated colonies using primer I31L and a pcDNA3 specific reverse primer Sp6R: 3[prime]-GCATTTAGGTGACACATATAGAATAG-5[prime]. PCR conditions were: 94°C, 10 min; 94°C, 45 s, 60°C, 45 s, 72°C, 45 s (30 cycles); 72°C, 2 min. Direct sequence analysis of the amplimers identified the pµ[beta]4-WT microgene construct carrying the wild-type intron 31 of ITGB4 and microgene pµ[beta]4-M carrying mutation 3986-19T->A.
DNA transfection and analysis of the IGTB4 mRNA transcripts
Plasmid DNA was purified on affinity columns (Qiagen, Courtaboeuf, France) and subconfluent secondary cultures of wild-type human keratinocytes seeded in six-well plates (2 × 105 cells/well), either directly on plastic or on a feeder of lethally irradiated mouse 3T3-J2 fibroblasts, were transfected using 2.5 µg of plasmid DNA and the Fugene-6 transfection kit (Boehringer Mannheim). Twenty-four hours later, the cells were harvested. Total RNA was extracted and submitted to RT-PCR amplification of pµ[beta]4-WT and pµ[beta]4-M microgene transcripts using the primer pair (L) I31L and (R) Sp6R. The RT-PCR products were analysed by electrophoresis on a 2% agarose gel. Two microlitres of the PCR reactions was used to subclone the pµ[beta]4-WT and pµ[beta]4-M microgene cDNAs in the bacterial plasmid pTOPO 2.1 (InVitrogen). After trans- formation of competent E.coli XL1 blue, 50 white colonies were isolated and lysed, and the nucleic acids directly submitted to PCR amplification using primers I31L and Sp6R in a final volume of 25 µl. The size of the amplimers was checked by electrophoresis in a 2% agarose gel. The PCR products were submitted to direct nucleotide sequencing.
Ribonuclease protection mapping of the integrin [beta]4 transcripts
A 359 bp cDNA probe (nt 3721-4080) (11) encompassing the deletions detected in the [beta]4 transcripts of the PA-JEB patient, was synthesized by RT-PCR using total RNA extracted from control keratinocytes. The primers were: (L) 5[prime]-CTGAGCTGGGCTGAGCCGGC-3[prime], (R) 5[prime]-TGGAGAGCGTAGAACGTCA-3[prime]. The amplimer was subcloned into pCR Script vector (Stratagene). The plasmid was linearized and used as a template for in vitro transcription by T3 RNA polymerase (Stratagene), in the presence of 1 µl [[alpha]-32P]CTP, 3000 Ci/mmol, to obtain a 517 nt antisense riboprobe. The reaction product was resuspended in 20 µl of 80% formamide (v/v), 0.1% xylene cyanol (v/v), 0.1% bromophenol blue (v/v), 2 mM EDTA pH 8.0, and loaded on a 4% acrylamide gel, 8 M urea. The gel was exposed for 5 min at 20°C for autoradiography. The riboprobe band was excised from the gel, eluted in 400 µl of 0.5 mM ammonium acetate pH 5.2, 1 mM EDTA, 0.1% SDS for 4 h at 37°C, precipitated with isopropanol, and dissolved in 40 mM PIPES, 400 mM NaCl, 1 mM EDTA, 40% formamide (v/v) pH 6.4. The RNA mapping assay was performed using a RNase Protection kit (Boehringer Mannheim). The riboprobe (5 × 105 c.p.m.) was hybridized overnight at 42°C with an excess (80 µg) of heat-denatured total RNA. After treatment with ribonucleases A and T1, RNA hybrids were precipitated with ethanol and loaded (2 × 104 c.p.m.) on a 4% acrylamide gel.
ACKNOWLEDGEMENTS
We acknowledge S. Caillet, A. Spadafora, H.J. Alder and A. Grima for technical assistance. We thank A. Sonnenberg for providing reagents. This work was supported by grants from EEC BIOMED 2 (BMH4-97-2062), the Programme Hospitalier de Recherche Clinique (France), the DEBRA Foundation (UK and America), the Association Française contre les Myopathies (France), the Ministère de l'Education Nationale (ACCSV, France), the Association pour la Recherche sur le Cancer, and the United States Public Health Service, National Institutes of Health.
REFERENCES
+To whom correspondence should be addressed. Tel: +33 4 93 37 77 79; Fax: +33 4 93 81 14 04; Email: meneguzz{at}unice.fr
This page is run by Oxford University Press, Great Clarendon Street, Oxford OX2 6DP, as part of the OUP Journals
Comments and feedback: jnl.info{at}oup.co.uk
Last modification:
Copyright© Oxford University Press, 1999.
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  F.-O. Desmet, D. Hamroun, M. Lalande, G. Collod-Beroud, M. Claustres, and C. Beroud Human Splicing Finder: an online bioinformatics tool to predict splicing signals Nucleic Acids Res., May 1, 2009; 37(9): e67 - e67. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
K. Gao, A. Masuda, T. Matsuura, and K. Ohno Human branch point consensus sequence is yUnAy Nucleic Acids Res., April 1, 2008; 36(7): 2257 - 2267. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 R Varki, S Sadowski, E Pfendner, and J Uitto Epidermolysis bullosa. I. Molecular genetics of the junctional and hemidesmosomal variants J. Med. Genet., August 1, 2006; 43(8): 641 - 652. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. Kralovicova, M. B. Christensen, and I. Vorechovsky Biased exon/intron distribution of cryptic and de novo 3' splice sites Nucleic Acids Res., September 1, 2005; 33(15): 4882 - 4898. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. Kralovicova, S. Houngninou-Molango, A. Kramer, and I. Vorechovsky Branch site haplotypes that control alternative splicing Hum. Mol. Genet., December 15, 2004; 13(24): 3189 - 3202. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
Y. Gache, M. Allegra, C. Bodemer, A. Pisani-Spadafora, Y. de Prost, J. P. Ortonne, and G. Meneguzzi Genetic bases of severe junctional epidermolysis bullosa presenting spontaneous amelioration with aging Hum. Mol. Genet., October 1, 2001; 10(21): 2453 - 2461. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 F. Spirito, S. Chavanas, C. Prost-Squarcioni, L. Pulkkinen, S. Fraitag, C. Bodemer, J.-P. Ortonne, and G. Meneguzzi Reduced Expression of the Epithelial Adhesion Ligand Laminin 5 in the Skin Causes Intradermal Tissue Separation J. Biol. Chem., May 25, 2001; 276(22): 18828 - 18835. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||