New Type of Mutations in Three Spanish Families with Choroideremia

Abstract

purpose. Choroideremia (CHM) is an X-linked ophthalmic disease. The gene associated with CHM (REP-1) encodes a ubiquitously expressed protein that is indispensable for the posttranslational activation of retina-specific Rab protein. Different mutations, including large genomic rearrangements involving the REP-1 gene, are responsible for CHM, but they all cause the protein to be truncated or absent. The authors screened 20 Spanish families with clinical diagnoses of CHM to determine the molecular cause of the disease.

methods. First, the authors performed haplotype analyses to determine whether the disease is linked to the REP-1 gene. In families in whom the disease segregated with the CHM locus (n = 14), mutational screening of the REP-1 gene was performed.

results. In 13 of the 14 families in which the phenotype segregated with the CHM locus, the authors identified the mutation associated with the disease. Eight different molecular defects that led to truncation and one that led to complete absence of the REP-1 protein were found in nine families and one family, respectively. Furthermore, the authors identified a novel type of mutation in the REP-1 gene in three families. This novel type of mutation did not result in a truncated or absent protein. Rather, these patients lost different parts of the REP-1 mRNA in-frame that in all the cases encode a conserved protein domain implicated in the interaction with Rab proteins.

conclusions. Based on the different mutations found, the authors propose a four-step protocol for the molecular diagnosis of CHM.

Choroideremia (CHM [MIM-30100]) is an X-linked recessive disease characterized by slowly progressive degeneration of the retina, retinal pigment epithelium, and choroid. Its incidence is estimated as 1 in 100,000. Affected males develop night blindness in their teenage years, followed by loss of peripheral vision because of the progressive visual field constriction and blindness by middle age. ¹ Female carriers generally are asymptomatic, but funduscopic examination shows patchy areas of chorioretinal atrophy that represent areas of clonal origin with different patterns of X inactivation. ² Although females with clinical features similar to those of affected males have been described, in all these cases the disease was associated with an X-autosomal translocation disrupting the gene associated with it. ^{3 4 5 6}  

CHM disease has been linked to a gene (REP-1) located on the Xq 21.2 region, which contains 15 exons that span a genomic sequence of approximately 150 kb. ^{7 8 9} The open reading frame of this gene encodes Rab escort protein (REP)-1, a 653-amino acid protein originally isolated as component A of Rab geranylgeranyl transferase (Rab GGTase). ¹⁰  

REP-1 shares 71% identity with another Rab escort protein, REP-2, which is encoded by the CHM-like gene (CHML) located on chromosome 1q. ¹¹ Both REPs (REP-1 and REP-2) appear to be expressed ubiquitously in systemic organs and bind to newly synthesized Rab proteins to escort them to Rab GGTases, which covalently modify the C termini of the Rab proteins with two geranylgeranyl isoprenoids. ^{12 13 14} Rab proteins are ras-related GTPases that control protein trafficking in secretory and endocytic pathways, and the absence of this posttranslational activation results in a loss of the ability to associate with donor membranes, thus leading to cell death. In persons with CHM, only ocular defects develop, so REP-2 probably compensates to a certain extent for the lack of REP-1 function in other organs. This hypothesis is supported by the observation that one Rab protein, Rab27, which is expressed at high levels in the retina, is recognized more efficiently by REP-1 than by REP-2. Consequently, Rab 27 is dysfunctional in CHM retinas and may trigger progressive chorioretinal degeneration. ¹⁵  

REP proteins contain three sequence-conserved regions (SCR1, SCR2, and SCR3) that bear high sequence similarity to GDP-dissociation inhibitor proteins (Rab GDI). ¹⁶ Both proteins play a role in the trafficking of Rab proteins. Crystal structure analysis of REP proteins reveal two conserved domains, a multisheet domain I, which forms a Rab-binding platform, and a smaller globular α-helical domain II, which may be involved in interaction with membrane proteins. In addition, an insert of 150 amino acids between the two conserved domains has been identified in REP proteins. ¹⁷  

Different mutations, including large genomic rearrangements involving the REP-1 gene, have been described that are responsible for CHM disease, but all these lead to a truncation or a complete absence of REP-1 protein (Preising M, et al. IOVS 1993;34:ARVO Abstract 3760). ^{18 19 20 21}  

Here, we report the results of genetic and molecular analysis of the REP-1 gene in 20 Spanish families with clinical diagnoses of CHM. Significantly, we detected and characterized in three of these families a new type of mutation not previously identified in the REP-1 gene. In these families, the identified mutation does not disrupt the REP-1 protein but presumably affects its function. 

Patients and Methods

Ascertainment of Patients

We analyzed 21 unrelated Spanish families for whom a clinical diagnosis of CHM was suggested. 

Informed consent was obtained from patients participating in this study, and the research protocols were approved by the bioethical committee and were in accordance with the Declaration of Helsinki. 

Genomic DNA was extracted from 15 mL peripheral blood of patients and their relatives by a standard salting-out procedure. RNA from one affected male and one obligate carrier of each family was extracted from 2.5 mL peripheral blood with the use of a blood RNA kit (PAXgene; Qiagen, Hilden, Germany) according to the manufacturer’s instructions. Protein extraction was performed in some families, as previously described. ²²  

Haplotype Analysis

Haplotype analysis was performed using three intragenic polymorphic markers (one SNP in exon 5, one VNTR in intron 9, and one STR in intron 14) and two extragenic markers strongly linked to the REP-1 locus (DXS8076 and DXS1002; Fig. 1 ). An RFLP assay was carried out for the analysis of the SNP in exon 5, using the restriction enzyme HinGI. ²³ For the rest of the makers used, each forward PCR primer was fluorescence labeled, and all markers were separately amplified by PCR in a total volume of 15 μL containing 100 ng genomic DNA, 125 μM dNTP, 10 pmol each primer (forward and reverse), 1× Taq DNA polymerase buffer (500 mM Tris/HCl, 100 mM KCl, 50 mM (NH₄)₂SO₄, 20 mM MgCl₂), and 0.6 U Taq DNA polymerase (FastStart; Roche, Basel, Switzerland). After denaturation at 95°C for 5 minutes, PCR was carried out (GeneAmp PCR System 2700; Applied Biosystems, Foster City, CA) for 10 cycles at 94°C for 30 seconds, 55°C for 30 seconds, and 72°C for 90 seconds and for 15 cycles at 89°C for 30 seconds, 55°C for 30 seconds, and 72°C for 90 seconds, with a final extension time of 30 minutes at 72°C. For the genotyping process, PCR products were electrophoresed in a genetic analyzer (ABI Prism 3100; Applied Biosystems) and analyzed with a software package (GeneMapper v3.5; Applied Biosystems). 

DNA Analysis

Each exon of the REP-1 gene, along with adjacent intronic sequences, was individually amplified from genomic DNA by PCR using previously published intragenic primers. ⁹ PCR amplifications were carried out in 50-μL reactions containing 200 μg genomic DNA, 1× polymerase buffer (500 mM Tris/HCl, 100 mM KCl, 50 mM (NH₄)₂SO₄, 20 mM MgCl₂), 200 μM dNTPs, 20 pmol each primer, and 2.5 U Taq DNA polymerase (FastStart; Roche). After denaturation at 95°C for 5 minutes, PCR was carried out (GeneAmp PCR System 2700; Applied Biosystems) for 35 cycles at 94°C for 90 seconds, annealing temperature for 90 seconds, and 72°C for 90 seconds, with a final extension time of 30 minutes at 72°C. Amplification products were sequenced and analyzed (ABI Prism 3100 Genetic Analyzer; Sequencing Analysis v5.2 software package; Applied Biosystems). 

RNA Analysis

Mutational analysis of the REP-1 gene is best approached at the cDNA level with different specific pairs of primers (available on request). To analyze the REP-1 mRNA, total RNA was extracted from 2.5 mL peripheral blood (PAXgene Kit; Qiagen) and was reverse-transcribed to cDNA (ImPromt Kit; Promega, Madison, WI) using an oligo-p(dT)₁₅ primer according to the manufacturer’s instructions. For the different PCR amplifications, 2 μL patient cDNA was used as template in a total volume of 50 μL that contained 20 pmol each primer (forward and reverse), 1× Taq DNA polymerase buffer (500 mM Tris/HCl, 100 mM KCl, 50 mM [NH₄]₂SO₄, 20 mM MgCl₂), 200 μM each deoxynucleotide, and 2 U Taq DNA Polymerase (FastStart; Roche). Amplification was performed in a thermal cycler (GeneAmp PCR System 2700; Applied Biosystems), with an initial denaturation of 10 minutes at 95°C and 40 cycles of 1 minute at 94°C, 1 minute at annealing temperature, and 2 minutes at 72°C, followed by a final extension of 10 minutes at 72°C. PCR products were subjected to 1% agarose gel electrophoresis. In each reaction, a normal cDNA was also amplified as a control. Fragments displaying an abnormal size were sequenced and analyzed (ABI Prism 3100 Genetic Analyzer; Sequencing Analysis v5.2 software package; Applied Biosystems). 

Southern Blot Analysis

The digested DNA (100 ng) was separated in a 0.8% agarose gel and transferred to a positively charged nylon membrane (NytranN; Schleicher & Schuell BioScience, Keene, NH) by capillary transfer. Probes were radiolabeled using a prime labeling system (Rediprime II Random; GE Healthcare, Chalfont, St. Giles, UK) and α-32-ATP according to manufacturer’s instructions. After overnight hybridization with buffer (PerfectHyb Plus Hybridization; Sigma-Aldrich, St. Louis, MO) the membrane was washed according to manufacturer’s instructions and exposed to x-ray film. ²⁴  

Immunoblot Analysis

For immunoblotting analysis, 200 μg total protein extracted from peripheral blood was separated by 10% dodecyl sulfate polyacrylamide gel (Bio-Rad, Hercules, CA), electrophoresed, and subsequently electrotransferred to nitrocellulose membranes (Hybon ECL; Amersham Biosciences, Piscataway, NJ). After blocking, membranes were incubated with anti–REP-1 monoclonal antibodies (2F1, which recognizes the C terminus of human REP-1 protein), a kind gift from Miguel Seabra. ¹⁵ The membranes were incubated with peroxidase-conjugated secondary antibody and visualized with enhanced chemiluminescence kit (ECL Western Blotting; Amersham Biosciences). 

Results

Haplotype Analysis

After haplotype analysis, we could rule out the implication of the REP-1 gene in the disease in 7 of the 21 families (33.33%), who had been included in this study because of their suspected clinical diagnoses of CHM. 

In the remaining families, according to the haplotype results, 28 of 40 women at risk of being carriers were found to be indeed carriers of the disease, 11 were noncarriers, and 1 was noninformative. 

Mutation Screening

Among the remaining 14 families in whom we could not rule out the implication of the REP-1 gene in the disease, we were able to identify mutations associated with the disease, at least at the mRNA level, in 13 (Table 1) . The spectrum of molecular defects identified in these families was heterogeneous, including four nonsense mutations (c.256 C>T, c.339 T>G, c.745 C>T, and c.907 C>T), one splice-site mutation (IVS 3+1 G>A), two small deletions of 2 bp (c.555_556 Del AG identified in two unrelated families and c.671_672 Del GA), one gross deletion of 1.52 Mb at Xq21.2 (refined by array CGH) that constituted the complete REP-1 gene, one translocation that disrupted the REP-1 gene in one female patient [46,X,t(X;4)(q21;p16)], and a novel type of mutation that did not truncate the REP-1 protein in families CHM-317, CHM-779, and CHM-918. 

In family CHM-317 (Fig. 2A) , we observed three different alternative splicing forms at the mRNA level in affected males. Automated DNA sequencing of the three smallest fragments obtained after amplification of the cDNA with the primers CHM-1 and CHM-B (which comprised the amplification of exons 1–5) showed that the difference in size resulted from the skipping of exon 2, skipping of exons 2 and 3, or skipping of exons 2, 3, and 4 (Fig. 2B) . In the affected males of this family, amplification by PCR of exons 3 and 4 of the REP-1 gene at the genomic DNA level was possible using intronic pairs of primers flanking these exons. PCR consistently failed to amplify when using an intronic pair of primers flanking exon 2 (Fig. 2C) . However, we could confirm by immunoblot analysis that the mutated protein was present in the lymphocytes of peripheral blood of a male patient and a carrier female of this family (Fig. 2D) . All these findings suggested a sequence change in the 3′ splicing acceptor site of intron 1 that we were unable to characterize further. 

In family CHM-779 (Fig. 3A) , we found at the mRNA level that the absence of exon 9 was associated with the disease (Fig. 3B) . We could amplify the 5′ region of this exon by PCR at the genomic level using a forward intronic primer and a reverse primer located in the middle of the exon, but it was not possible to amplify the 3′ region of this exon. In addition, in the affected males of this family, it was not possible to amplify the intragenic polymorphism located in intron 9. Southern blot analysis using exon 10 as a specific probe revealed a deletion of approximately 6 kb between the 3′ region of exons 9 and 10 (Fig. 3C) . 

Finally, in the three affected male patients of family CHM-918, the absence of exons 3 and 4 was detected at the mRNA level (Fig. 4) . With the use of intronic pairs of primers flanking both exons, we confirmed this deletion at the genomic level in the affected males of this family. 

Discussion

Choroideremia is a genetic heterogeneous disease associated with different types of mutations in the REP-1 gene. However, all mutations in this gene reported so far result in the truncation or the complete absence of the REP-1 protein (Preising M, et al. IOVS 1993;34:ARVO Abstract 3760). ^{18 19 20 21} Contrary to that, we report here a novel type of mutation that presumably affects the correct function of the REP-1 protein rather than leading to its truncation or its absence. 

In three of our Spanish families (CHM-317, CHM-779, and CHM-918), the reading frame is maintained but the protein product is missing several amino acids by different mechanisms. In family CHM-317, the disease is associated with a mutation that leads to three aberrantly spliced mRNAs, neither of which contains a premature stop codon. In family CHM-779, the disease is associated with a genomic deletion, which leads to the skipping of exon 9. In family CHM-918, the disease is associated with a genomic deletion of exons 3 and 4. Interestingly, the missing amino acids are part of structurally conserved regions in all these cases. The affected patients in these families lack part of domain I of the REP protein. This domain is implicated in the interaction with Rab proteins and presumably is crucial for the correct function of the REP-1 protein. The amino acids encoded by exon 2 of the REP-1 gene, which are absent in the affected male of family CHM-317, are part of the SCR1 and domain I of the REP protein. The amino acids encoded by exon 9 of the REP-1 gene, which are absent in the affected male of family CHM-779, are part of the SCR3 and domain I of the REP protein. The amino acids encoded by exons 3 and 4 of the REP-1 gene, which are absent in the affected male of family CHM-918, are part of domains I and II of the REP protein. Our studies revealed that in these families, the disease is probably caused by the loss of function of the REP-1 protein rather than by its absence. Future studies on the activity of the REP-1 protein in the lymphoblastoid cultures of the affected patients of these families could determine the role of the different parts of domain I in REP-1 function. ¹⁵  

Apart from the new type of mutations that do not disrupt the REP-1 protein, the spectrum of molecular defects identified in the Spanish families and in other populations is heterogeneous. In this study we found seven different subtle recurrent sequence alterations and two large rearrangements in the REP-1 gene, and another nonsense mutation (c.1049 C>A) was detected in one Spanish CHM family. ²⁵ The deletion c.554 to 555 Del AG was detected in two unrelated families, and the nonsense mutations c.745 C>T and c.907 C>T, along with a splicing site mutation IVS 13+1 G>A, are recurrent because they were reported as being associated with CHM disease in other families. ^{19 20 26} In the other hand, the nonsense mutations c.256 C>T and c.339 T>G and the small deletion c.671_672 Del GA are novel and were not previously reported to be associated with the CHM disease in other families. The large rearrangements identified in the CHM Spanish families included a gross deletion and a balanced translocation between chromosomes X and 4. In family CHM-747, the disease is associated with a deletion of 1.52 Mb at Xq21.2 in a region containing the REP-1 gene, which was detected by haplotype analysis and refined by array CGH. Finally, a translocation [46,X,t(X;4)(q21;p16)], previously described by us, was associated with the disease in a female patient of family CHM-288. ⁶  

Our results expand the repertoire of mutations that cause CHM and lead to the proposal of a protocol for the molecular diagnosis of CHM families in four consecutive steps (Fig. 5) . We propose that after haplotype analysis, the DNA be analyzed, especially, of course, in those sporadic cases in which haplotyping is not possible because of family size limitations or when the haplotype analysis is inconclusive. All mutations detected at the cDNA level must then be characterized at the genomic DNA level and finally at a protein level by immunoblot analysis. We consider this approach to be the most effective method for mutation screening in CHM cases. Using this approach we have detected and characterized the mutation associated with CHM disease in 13 of the 14 families in whom we could not rule out the implication of the REP-1 gene in the disease after haplotype analysis. ²⁷ In only one family (CHM-695), we could not find the mutation associated with the disease. However, alternative diagnoses might be considered in this family, such as XLRP or even ADRP. On the other hand, some authors have proposed that the immunoblot analysis had been sufficient to confirm the diagnosis of CHM. ²² However, this did not hold true for three of our families because the respective mutations did not disrupt the reading frame. 

The CHM phenotype must be distinguished from other retinal dystrophies such as gyrate atrophy or retinitis pigmentosa, and the confirmation of the clinical diagnosis is important in the case of patients with this type of retinal dystrophy. In addition, genetic studies of families with clinical diagnoses of CHM allow the determination of the status of female carriers, which is crucial for the prevention of this type of progressive and still untreatable disease. Moreover, a particular gene therapy approach for CHM disease has been performed and has demonstrated the ability of adeno-associated viral (AAV) vectors delivering the full human cDNA encoding REP-1 to rescue fibroblasts and lymphoblasts of patients with CHM. ²⁸ Knowledge of the molecular mechanism of these diseases will allow the development of these new rational, potentially therapeutic tools. 

 Supported by FIS (04/0193 and PI 060027), EviGenoRet LSHG-CT-2005-512036, Mutua Madrileña, Fundacion Conchita Rábago, and CIBER-ER.

 Submitted for publication September 9, 2007; revised October 30 and November 21, 2007; accepted February 22, 2008.

 Disclosure: M. Garcia-Hoyos, None; I. Lorda-Sanchez, None; P. Gómez-Garre, None; C. Villaverde, None; D. Cantalapiedra, None; A. Bustamante, None; D. Diego-Alvarez, None; E. Vallespin, None; J. Gallego-Merlo, None; M.J. Trujillo, None; C. Ramos, None; C. Ayuso, None

 The publication costs of this article were defrayed in part by page charge payment. This article must therefore be marked “advertisement” in accordance with 18 U.S.C. §1734 solely to indicate this fact.

 

Figure 1.

 

Localization on the X chromosome of the markers used to perform haplotype analysis in our Spanish CHM families.

Table 1.

 

Mutations Detected in Our Cohort of CHM Families

Family	Genomic Level	Mutation mRNA Level	Protein Level
CHM-22	c.555–556 Del AG	c.555–556 Del AG	K 177 fs X 182
CHM-288	46,X,t (X;4)(q21;p16)	—	Truncated protein
CHM-317	Absence of exon 2 amplification	3 different aberrant splicing	Loss of functional domain
CHM-590	c.745 C>T	c.745 C>T	R 239 X
CHM-729	IVS 3+1 G>A	Skipping exon 3	S 39 fs X 45
CHM-779	Del of 6 kb	Skipping exon 9	Loss of functional domain
CHM-747	Del of 1.5 Mb	—	Absent protein
CHM-797	c.339 T>G	c.339 T>G	Y 103 X
CHM-889	c.671–672 DelGA	c.671–672 Del GA	R 214 fs X 221
CHM-844	c.907 C>T	c.907 C>T	R 293 X
CHM-918	Del exon 3 and 4	Del exon 3 and 4	Loss of functional domain
CHM-962	c.555–556 Del AG	c.555–556 Del AG	K 177 fs X 182
CHM-1098	c.256 C>T	c.256 C>T	Q 76 X

Figure 2.

 

Family CHM-317 results. (A) Haplotype analysis. The haplotype that cosegregates with the CHM disease in this family is defined by the alleles located along the black bar. (B) A 1% agarose gel showing the amplified cDNA fragments using the CHM-1 (in exon 1) and CHM-B (in exon 7) primers. The normal control (unaffected male) displays an expected band of 852 bp, and the affected male displays three different bands because of the alternative skipping of exon 2 (band 1 of 786 bp), exons 2 and 3 (band 2 of 713 bp), or exons 2, 3, and 4 (band 3 of 588 bp). (C) Results of exon 2 amplification at the genomic DNA level of all CHM-317 family members participating in this study; in the affected male patients, PCR consistently failed to amplify when an intronic pair of primers flanking this exon was used. (D) Immunoblot analysis of protein isolated from negative control, affected male, and carrier female of family CHM-317. Results show the expected band corresponding to the size of REP-1 protein in all protein samples.

Figure 3.

 

Family CHM-779 results. (A) Haplotype analysis. The haplotype that cosegregates with CHM disease in this family is defined by the alleles located along the black bar. (B) A 1% agarose gel showing the amplified cDNA fragments using the CHM-5 (in the exon 8) and CHM-6 (in the exon 13) primers. The negative control (unaffected male) displays an expected band of 627 bp, the affected male displays a short fragment of 554 bp caused by the skipping of exon 9, and the carrier female exhibits both bands. (C) Southern blot analysis of NcoI-digested genomic DNA, isolated from a negative control, affected male, and carrier female. The expected band of approximately 13 kb was detected in the control and in the female carrier. The additional band observed in the female carrier and in the affected male of approximately 7 kb confirms that in this family, the disease is associated with a deletion of 6 kb between exons 9 and 10.

Figure 4.

 

Family CHM-918 results. (A) Haplotype analysis. The haplotype that cosegregates with CHM disease in this family is defined by the alleles located along the black bar. (B) A 1% agarose gel showing the amplified cDNA fragments using the CHM-1 (in exon 1) and CHM-B (in exon 7) primers. The normal control (unaffected male) displays an expected band of 852 bp, and the affected male displays a short fragment of 654 bp caused by the absence of exons 3 and 4.

Figure 5.

 

Proposed protocol for the molecular diagnosis of CHM families.

The authors thank FIS, EviGenoRet, Mutua Madrileña, Fundacion “Conchita Rábago de Jiménez Díaz,” and CIBER-ER for their support; and Robert Wilkie for the review of the manuscript.