Abstract
ABSTRACT. Senior-Løken syndrome is an autosomal recessive disease with the main features of nephronophthisis (NPH) and Leber congenital amaurosis. The gene for adolescent nephronophthisis (NPHP3) was recently localized to chromosome 3q21-q22. The hypothesis was tested that Senior-Løken syndrome (SLS) might localize to the same region by studying a kindred of German ancestry with extended consanguinity and typical findings of SLS. Twenty highly polymorphic markers located in the vicinity of the NPHP3 genetic region were tested. Haplotype analysis revealed homozygosity by descent in affected individuals, and linkage analysis yielded a parametric maximum multipoint logarithm of likelihood of odds (LOD) score of 3.14, thus identifying the first locus for SLS. The SLS1 locus is flanked by D3S1587 and D3S621 and contains a 14-cM interval that contains the whole critical NPHP3 region. Three additional families with SLS were studied, and evidence for genetic heterogeneity in one of them was found. Localization of a SLS locus to the region of NPHP3 opens the possibilities of both diseases arising by mutations within the same pleiotropic gene or two adjacent genes.
Nephronophthisis (NPH) is an autosomal recessive cystic kidney disease that causes progressive renal failure (1). A gene for juvenile NPH (NPH1, MIM 256100) was localized on chromosome 2q12-q13 (2,3). The responsible gene (NPHP1) has recently been identified (4,5). For infantile NPH (NPH2, MIM 602088), a locus was mapped to chromosome 9q22-q31 (6). We recently identified a novel type of NPH that leads to end-stage renal disease at a median age of 19 yr (7), adolescent NPH (NPH3, MIM 604387), in a large Venezuelan kindred. With the use of a homozygosity mapping strategy, the responsible gene (NPHP3) was localized on human chromosome 3q21-q22, which is homologous to the murine renal cystic disease locus pcy on mouse chromosome 9 (8). Clinical signs of NPH3 consist of renal symptoms such as polyuria, polydipsia, secondary enuresis, severe anemia, and progressive renal failure. Renal pathology is characterized by cysts at the corticomedullary junction. Renal histology shows the characteristic triad of irregularly thickened tubular basement membranes, atrophy and dilatation of tubules, and sclerosing tubulointerstitial nephropathy. However, histology is not pathognomonic for NPH3, and diagnosis relies on typical clinical history, sonographic findings, and exclusion of other renal diseases (7). NPH has been associated with several extrarenal associations such as Leber congenital amaurosis (LCA), cerebellar ataxia, skeletal involvement, congenital oculomotor apraxia, and hepatic fibrosis (9–13). The term “Senior-Løken syndrome” (SLS) denotes the association of NPH and LCA (MIM 266900). LCA (MIM 204000/204100) is a clinically and genetically heterogeneous retinal disorder that occurs in infancy and is accompanied by profound visual loss, nystagmus, poor pupillary reflexes, and either a normal retina or varying degrees of atrophy and pigmentary changes (14–16). The electroretinogram is extinguished or severely reduced (17). LCA is inherited as an autosomal recessive trait. To date, four genes for LCA have been identified: retinal guanylate cyclase (RETGC1) on chromosome 17p13 (18,19), retinal pigment epithelium protein (RPE65) on chromosome 1p31 (20,21), cone-rod homeobox (CRX) on chromosome 19q13.3 (22), and aryl hydrocarbon receptor–interacting protein-like 1 (AILP1) on chromosome 17p13 (23). Two additional loci for LCA have been reported on chromosomes 14q24 and 6q11-q16 (24,25). In this study, we pursued a focused approach directed at the NPH3 gene locus; we tested a consanguineous family of German ancestry with SLS for linkage to the NPHP3 region on chromosome 3q21-q22. Extensive pedigree analysis detected multiple consanguineous loops with a remote degree of consanguinity dating back to the 17th century. Because this remote consanguinity controls for many meioses, identification of homozygosity by descent in the two affected individuals allowed localization of a novel gene locus for SLS. Thus the NPHP3 locus harbors a gene or genes responsible for SLS in this family. By studying other families with SLS, we demonstrated genetic heterogeneity.
Materials and Methods
Patients
Five members of a German family originating from northern Germany with SLS were investigated (Figure 1). The two affected siblings of this family had nearly identical clinical and diagnostic findings; therefore, the clinical description is pertinent to both. In infancy, pressing on the globes (the digito-ocular phenomena of Franceschetti-Bamatter) played a prominent part in childhood behavior and brought them to medical attention. On ophthalmologic investigations, they had poor visual acuities in the order of 0.05 to 0.1 (normal value 1.0), nystagmus, poor pupillary reflexes, retinal mottling, and high hypermetropia. The electroretinogram was abolished. Later, when visual field testing became possible, they had severe tubelike restriction of the visual fields. Both affected siblings developed renal symptoms such as polyuria, polydipsia, secondary nocturnal enuresis, and progressive renal failure. For the creatinine course, see (Figure 2). Renal replacement therapy had to be implemented at the age of 15 and 12 yr, respectively. Only minor proteinuria of 0.4 g (0.6 g)/24 h was noted. On renal sonography, the kidneys appeared to be of normal to slightly reduced size, with cysts at the corticomedullary junction and increased echogenicity. No urinary tract abnormality was observed. After demonstration of linkage of SLS in F1, three additional families with SLS were studied. In all families, diagnosis of SLS was based on (1) pedigree structure suggestive for autosomal recessive inheritance; (2) profound visual loss, nystagmus, poor pupillary reflexes, and varying degrees of atrophy and pigmentary changes of the retina; (3) abolished electroretinogram; (4) renal symptoms typical for NPH: polyuria, polydipsia, secondary nocturnal enuresis, and progressive renal failure; (5) sonography of the kidneys characteristic of NPH with cysts at the corticomedullary junction or renal biopsy compatible with the diagnosis of NPH3; and (6) absence of other renal disease.
Figure 1. Pedigree of the consanguineous German family F1 with Senior-Løken syndrome (SLS). Haplotypes for markers on chromosome 3q21-q22 are shown from centromeric to telomeric (top to bottom) direction. Flanking markers D3S1587 and D3S621 of the SLS locus are underlined. Flanking markers of the NPHP3 region D3S1292 and D3S1238 are indicated by arrows. Solid symbols denote affected individuals; open symbols denote unaffected individuals. Double horizontal lines between parents indicate consanguinity. Microsatellite marker loci are given in the first column. The haplotype segregating with the disease locus is indicated by a black bar. The year of birth is given below the symbol for each individual.
Figure 2. Creatinine course for the two affected individuals of the German family with SLS (see (Figure 1)). The dotted line represents the serum creatinine course of the affected girl and the other that of her affected brother.
Genetic Study
Venous blood samples from family members were obtained after informed consent was obtained. DNA was prepared according to standard methods. The NPHP1 locus for juvenile NPH on chromosome 2q12-q13 was excluded by haplotype analysis (data not shown). Infantile NPH was excluded by clinical means, because end-stage renal disease was reached after age 3 yr. Haplotype analysis was performed as described elsewhere, by use of highly polymorphic microsatellites residing at the NPHP3 locus on chromosome 3q21-q22 (7). Marker order was based on physical mapping data of the NPHP3 region that have been reported elsewhere (8) and information available at the Weizman Institute for Bioinformatics (URL given below). Parametric two-point and multipoint analysis in family F1 was performed by use of the program ALLEGRO (26). SLS was analyzed as an autosomal recessive trait with complete penetrance and an assumed gene frequency of 0.003, for a conservative estimate. Likewise, only the smallest consanguinity loop for the logarithm of likelihood of odds (LOD) score calculation was used, which resulted in a probably slight underestimation of the maximum LOD score. A total of 20 microsatellite markers were analyzed. Allele frequencies of markers were arbitrarily assigned a value of 1/n, where n refers to the number of alleles at each marker. Numbers of alleles were taken from Dib et al. 1996 (27). Recombination frequencies for males and females were assumed to be equal. The ALLEGRO program was used to perform parametric multipoint linkage analysis against a fixed map of 20 markers by use of the genetic model detailed above (26). Usually, incorrect parameters of the genetic model in linkage analysis does not affect the type I error rate, i.e., will not result in false-positive results but rather in loss of statistical power. This is quite different in homozygosity mapping when marker allele frequencies are concerned. Too low an estimate of the cosegregating marker allele frequencies eventually will lead to a false-positive result. Therefore, as a test for robustness, multipoint LOD scores were recalculated in a second model that used marker allele frequencies of 1/n, where n refers to the observed number of alleles in the examined family, which resulted in allele frequencies ranging from 0.3333 to 0.500 for informative markers.
Results
Haplotype studies of the German kindred F1 showed a result compatible with homozygosity by descent in all affected individuals, covering the whole NPHP3 region (Figure 1). Recombinants observed for D3S1587 and D3S621 defined the critical disease interval, which spans a 14-cM interval. Parametric LOD score calculation achieved a maximum two-point LOD score of Zmax = 1.06 (θ = 0) for marker D3S3637. Multipoint LOD score calculation resulted in a significant parametric LOD score (Zmax = 3.14) in the interval between markers D3S3548 and D3S1309. For detailed results, see (Table 1). When higher allele frequencies were used as a test of robustness, the maximum multipoint LOD score remained stable (Zmax = 3.13).
Multipoint likelihood of odds score results for family F1 between the SLS1 locus and markers from the NPHP3 region on chromosome 3q21-q22a
Haplotype analysis of three additional families tested for the newly identified SLS region are shown in (Figure 3). One family (F353) showed a haplotype analysis compatible with homozygosity by descent in the affected individual, which is consistent with linkage. Haplotype analysis of family F15 excluded linkage to the examined region, because both affected individuals do not share the same haplotype, thus giving evidence for further genetic heterogeneity in SLS. In family F335, both affected children share the same haplotypes, but no homozygosity by descent was observed, which is highly suggestive of absence of linkage in this consanguineous family.
Figure 3. Pedigrees of three additional families with SLS. Haplotypes for markers on chromosome 3q21-q22 are shown from centromeric to telomeric (top to bottom) direction. Flanking markers D3S1587 and D3S621 of the SLS locus are underlined. Solid symbols denote affected individuals; open symbols denote unaffected individuals. Double horizontal lines between parents indicate consanguinity. Microsatellite marker loci are given in the first column. The haplotype segregating with the disease locus is indicated by a black bar. Note that the haplotype analysis of family F353 is compatible with linkage and indicative for homozygosity of descent. Haplotype analysis of family F15 excludes linkage to the examined region.
Discussion
Testing a candidate locus hypothesis, we were able to demonstrate that a gene locus for SLS (SLS1) maps on chromosome 3q21-q22 within a 14-cM interval flanked by D3S1587 and D3S621, which contains the whole critical region of NPHP3 (Figure 1). Our studies provide several lines of evidence for the localization of the first gene locus for SLS. First, parametric LOD-score calculation achieved a significant multipoint LOD score, and multipoint LOD scores remained robust toward the change of allele frequencies. Second, both affected siblings of the family showed identical haplotypes compatible with homozygosity by descent (Figure 1). This is an expected finding in individuals from an inbred population affected by rare autosomal recessively transmitted diseases who inherit identical alleles from a common ancestor. The chromosomal segment of homozygosity by descent—that is, shared by affected relatives—is likely to harbor the responsible gene for the disease because of the rarity of disease alleles in the population (28). Third, by use of a focused approach, the multiple testing problem as it occurs in a total genome scan was avoided, thus minimizing the risk of finding homozygosity by descent by chance (29). Fourth, we found linkage in a candidate region where a gene responsible for NPH is located, which is a key feature of SLS. Haplotype studies of other families with SLS identified a consanguineous family (F353) with a haplotype analysis compatible with homozygosity by descent and linkage to the examined region. If linkage in this family is not spurious, the region of interest would be restricted telomeric to D3S1273 (Figure 3).
Three potential genetic mechanisms may serve to explain the finding of linkage of SLS to the NPHP3 locus. (1) A pleiotropic effect caused by different mutations within a single gene may be responsible for the association of LCA and NPH3. (2) Two genes might be located in close vicinity to each other in the NPHP3 region, which are independently responsible for LCA and NPH3. In this model, SLS would be a result of a contiguous gene deletion syndrome involving both genes (30). (3) Two adjacent genes independently cause SLS and NPH3, and localization of the SLS locus occurred only by chance in vicinity of the NPHP3 region. The latter possibility is the least likely. By haplotype analysis that used a dense set of polymorphic markers, we have not obtained evidence for a deletion yet. Cloning the NPHP3 region identified seven expressed genes (ACPP [acid phosphatase, prostate-specific], TOPB1 [topoisomerase (DNA) II binding protein], EDF1 [endothelial differentiation-related factor 1], TF [transferrin], DGKZ [diacylglycerol kinase-zeta], SLC21a2 [prostaglandin transporter, PGT], and RYK [receptor-like tyrosine kinase]) and eight expressed-sequence tags, most with unknown function (8). On the basis of available functional information for these genes, there is no clear candidate for NPH or LCA. The identification of the first SLS locus hopefully will aid gene identification in NPH3, because the search for potential candidate genes can be based on hypotheses for LCA and NPH3.
The differentiation between LCA with isolated involvement of the eye and SLS has important clinical implications, because patients with LCA without genetic evidence for isolated LCA (LCA types 1 to 6) should have regular measurement of renal function parameters. Otherwise, in these patients, development of chronic renal failure eventually might be missed, and residual visual function may be impaired by hypertension or uremia. It is important to know that renal function may remain normal for a long time, and some patients with SLS develop renal failure in adulthood (31–33). A study that examined several families with SLS excluded linkage to the NPHP1 locus (2). After genetic testing of NPHP1 became available, three families with homozygous deletions of NPHP1 and mild to moderate late-onset tapetoretinal degeneration were reported (34). However, these patients did not meet the diagnostic criteria of LCA. Patients developed only minor visual impairment and therefore do not represent classical SLS. Recently, we found evidence for further genetic heterogeneity in isolated NPH (35). In this study, we give evidence also for further genetic heterogeneity in SLS, because haplotype analysis of family F15 excluded linkage to the examined markers. Obviously, clinical as well as genetic heterogeneity is a prominent feature of the NPH syndromes.
In summary, a candidate locus study and homozygosity mapping strategy has identified the first locus for SLS, which is identical with the NPHP3 locus on chromosome 3q21-q22. Mapping of this SLS locus and subsequent gene identification will aid the understanding of molecular mechanisms that cause Leber congenital amaurosis and adolescent nephronophthisis.
Acknowledgments
We are indebted to the families who participated in this study. The authors would also like to thank Dr. M. Bulla (Münster) and Dr. R. Burghard (Memmingen) for clinical information on affected individuals. H.O. and F.H. were supported by a grant from the German Research Foundation (DFG Om 6/1 to 2, DFG Om 6/2 to 1 and DFG Hi 381/3 to 3) and by a grant from the Zentrum Klinische Forschung, Freiburg (ZKF-A1).
Footnotes
- © 2002 American Society of Nephrology
References
