Exclusion of the Locus for Autosomal Recessive Pseudohypoaldosteronism Type 1 from the Mineralocorticoid Receptor Gene Region on Human Chromosome 4 q by Linkage Analysis

Pseudohypoaldosteronism type 1 (PHA1) is an uncommon inherited disorder characterized by salt-wasting in infancy arising from target organ unresponsiveness to mineralocorticoids. Clinical expression of the disease varies from severely affected infants who may die to apparently asymptomatic individuals. Inheritance is Mendelian and may be either autosomal dominant or autosomal recessive. A defect in the mineralocorticoid receptor has been implicated as a likely cause of PHA1. The gene for human mineralocorticoid receptor (MLR) has been cloned and physically mapped to human chromosome 4q31.1-31.2. The etiological role of MLR in autosomal recessive PHA1 was investigated by performing linkage analysis between PHA1 and three simple sequence length polymorphisms (D4S192, D4S1548, and D4S413) on chromosome 4q in 10 consanguineous families. Linkage analysis was carried out assuming autosomal recessive inheritance with full penetrance and zero phenocopy rate using the MLINK program for two-point analysis and the HOMOZ program for multipoint analysis. Lod scores of less than -2 were obtained over the whole region from D4S192 to D4S413 encompassing MLR. This provdes evidence against MLR as the site of mutations causing PHA1 in the majority of autosomal recessive families.

In 1985, Armanini and co-workers demonstrated reduced or absent binding sites for tritiated aldosterone in circulating mononuclear lymphocytes in three patients with PHAl (9, 10).These observations have been confirmed in subsequent studies (11-13) and provide evidence in favor of a defect in hMR in PHAl.However, the possibility exists that they may represent a secondary effect of down-regulation due to high aldosterone levels.Molecular cloning of a complementary DNA (cDNA) encoding hMR (14) has allowed a direct genetic approach to evaluating the hypothesis that PHAl arises from mutations in the gene for this receptor.Analysis has been carried out by 3 independent groups on 7 patients, including the original sporadic case of Cheek and Perry.In this patient and his family, Southern blot analysis using 3 probes from the hMR cDNA revealed bands of normal size and intensity (15).In another patient and his family, hMR cDNA was amplified using 11 pairs of oligonucleotide primers, and total ribonucleic acid from peripheral blood leukocytes was amplified by reverse transcriptase-polymerase chain reaction (PCR).Sequencing of the products showed no deviation from the wild-type sequence (16).Similar results were reported in a study of 5 other patients with PHAl (17,18).Sequencing of the patient's hMR cDNA identified a nonconservative homozygous base change in 4 of the 5 patients, and 1 of them had an additional conservative heterozygous base change.Both of these base changes were in the immunogenic domain and were also found to be present in a considerable number of normal individuals.It was concluded that these sequence variations did not have any pathological significance.The sequences of the first untranslated exon, 0.9 kilobase of the 5'-regulatory region, and the DNA-and ligand-binding receptor domain in these 5 patients were all normal.Full evaluation of the gene for hMR (MIX) by direct mutational analysis awaits characterization of its genomic organization.These results do not, of course, exclude MLR as the site of mutations causing PHAI, particularly in the recessive variety, as genetic (locus or allelic) heterogeneity may exist.
Homozygosity mapping provides a powerful approach to testing the role of a candidate gene in a rare autosomal recessive disorder with a high level of parental consanguinity (19).MLR has been mapped to human chromosome 4 by somatic cell hybridization and regionally localized to 4q31 .l-31.2 by in situ hybridization (20, 21).Linkage analysis was, therefore, carried out in 10 inbred families (see Fig. 1) using 3 simple sequence length polymorphisms spanning the MLR region.

Subjects and Methods
Patients and families

Results
The alleles of the affected individuals at the three loci studied are shown in Table 1.Loci were homozygous on only 10 of 41 occasions among the 14 affected individuals typed at the 3 markers (2 for D4SZ92, 4 for 0451548, and 4 for 045413 were homozygous).The 2-point lod scores are shown in Table 2.The lod score was less than -2 for at least 10 centimorgans (CM) around D4S192 and D4S413 and for 5 CM around D4S1548.The results of the multipoint analysis are shown in Fig. 3.A lod score of less than -2 was obtained over the whole region from D4SZ92 to 045413.

Discussion
These results provide strong genetic evidence against MLR as the site of mutations causing pseudohypoaldosteronism in these families.Linkage analysis in inbred families provides a powerful strategy for examining the role of a candidate gene and for identifying the true location of the gene(s) causing this disease.Aspects of the interpretation of the linkage data and consideration of the possible nonreceptor molecular defects underlying AR PHAl and strategies for future investigations are discussed below.The increased incidence of rare recessive diseases in the offspring of consanguineous matings arises in general from affected individual inheriting disease alleles at a particular locus from a common ancestor.The affected individual is, therefore, said to be homozygous by descent (rather than by state) for the disease allele.This increases the power of linkage data from consanguineous families, a factor exploited in the strategy of homozygosity mapping of disease loci in rare recessive disorders (see below) (19).However, the possibility of a disease chromosome being introduced into a consanguineous family by a married-in unrelated member cannot be totally excluded.This could produce an affected individual who may be heterozygous at the disease locus.This is extremely unlikely, especially if the disease is rare, as is PHAl.Exclusion of a candidate gene is definitively achieved by following the segregation of an intragenic polymorphism, which allows identification of obligate recombinants within families with at least two affected children.As an informative intragenic polymorphism for MLR is not currently available, microsatellite loci spanning the region harboring MLR were used.
The genetic localization of MLR has not been well established.Little genetic data for MLR are available, and therefore, some uncertainty exists concerning the genetic distance between MLR and the loci analyzed.MLR has been physically mapped to 4q31.1-31.2 by fluorescent in situ hybridization.Composite data from existing genetic linkage maps provide good evidence that D4S192 and D4S413 flank the region to which MLR maps, and D4Sl548 is close to the MLR locus.The genetic distance between D4S192 and D4S413 is about 18 CM.Analysis of allele data at the 3 loci by 3 parallel approaches provides significant evidence against a locus for PHAl in this region.Direct inspection reveals homozygosity of affected individuals on 10 of 41 occasions only.The average region of homozygosity expected around the disease locus in affected offspring of a first cousin marriage is at least 15 CM (19).Pairwise linkage analysis generates lod scores of -2 (the threshold accepted for exclusion assuming locus homogeneity) for regions around each locus totalling 25 CM.This includes the putative location of MLR.Finally, linkage analysis incorporating data from the 3 loci simultaneously using the program HOMOZ generates lod scores in the exclusion range for a total of at least 35 CM encompassing MLR.
These results demonstrate that MLR is not a common site of mutations causing PHAl in this group of families.They do not exclude the possibility that MLR is the disease locus in a minority of these families or that a set of linked families might be identified if a much larger resource of AR PHAl families was studied.Moreover, they do not exclude MLR as the locus accounting for PHAl in those families that appear to display AD inheritance, although it is well recognized that allelic heterogeneity can account for variations in the mode of inheritance of diseases arising from a single locus.The evidence arising from cDNA analyses in sporadic patients does not, however, support MLR as the site of mutation in patients who may fall into the AD category, although the possibility that mutations exist in noncoding regulatory regions has not been excluded.
These observations indicate that MLR is not the site of mutations causing PHAl in at least some patients.This is consistent with the observation that nonreceptor defects account for a proportion of cases in several hormone insensitivity syndromes, including androgen insensitivity (30), familial glucocorticoid deficiency (311, and pseudohypoparathyroidism (32).The likely molecular complexity underlying aldosterone function does, of course, provide numerous possible sites at which hormone insensitivity might arise.There is now good evidence that aldosterone exerts an immediate effect on its target cells via a membrane-bound receptor (now known as nongenomic action) (33).This is followed by the classical genomic action of aldosterone, which is thought to act by binding to the cytoplasmic mineralocorticoid receptor, forming an active complex.This complex initiates protein synthesis (aldosterone-induced proteins) by binding to a number of hormone response element of the nuclear DNA.
Many different hypotheses for nonreceptor defects in PHAl have been proposed.The nongenomic actions of mineralocorticoids allow for dysfunction in pathways that do not include the cytoplasmic receptor.At the prereceptor level, there could be interference with aldosterone binding, by competition for the binding site or by cleavage of the receptor.At the postreceptor level, the most obvious candidates include the aldosterone-induced proteins, which include subunits of Na/K-adenosine triphosphatase and the amiloride-sensitive epithelial Na channels.Liddle's syndrome, or pseudoaldosteronism, has recently been shown to arise from mutations in the P-subunit of the epithelial Na channel gene causing excessive sodium reabsorption and hypertension at normal or low levels of circulating aldosterone (34).In so far as these features represent a mirror image of PHAI, it is conceivable that mutations involving these sodium channels may render them nonfunctional or unresponsive to aldosterone.
Homozygosity mapping represents a powerful strategy for examining the role of additional candidate genes in PHAl.This is being pursued using the present family resource in parallel with a genome search using anonymous microsatellite loci in case the causative locus is an as yet unidentified gene.Identification of the gene for PHAl will allow improved diagnosis and treatment in affected individuals.It should also provide new insights into the biology of mineralocorticoid function at a molecular level, information of potential value in understanding more common disease states such as hypertension.
FIG. 2. Physical and genetic maps of human chromosome 4q indicating relative positions of MLR and marker loci D4S192,0491548, and 049413.The map information shown is a composite of published data: a, Refs.20,21, and 27; b, Refs. 25 and 27; and c, Ref. 26.Genetic distance is sex averaged and calculated using the Kosambi function.