INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Familial primary pulmonary hypertension (PPH) is a rare autosomal dominant disease characterized by distinctive changes in pulmonary arterioles that lead to increased pulmonary artery pressures, right ventricular failure, and death. Without intervention, median survival is less than 3 yr. The pathogenesis of PPH is unknown, but the existence of families with multiple affected members suggests a genetic etiology in PPH. Our previous studies mapped the disease locus, PPH1, to a 27-cM region on chromosome 2q31-q32, with a maximum logarithm of the odds favoring genetic linkage (lod) score of 3.87 with markers D2S350 and D2S364 (1). This locus was also independently localized to a 25-cM region with a maximum two-point lod score of 6.97 at a recombination fraction () of 0 with the marker D2S389 (2). To narrow the minimal genetic region for PPH, we genotyped 44 affected individuals and 133 unaffected individuals (to establish the affected haplotype and define the recombination events) from 17 families with PPH (14 previously unreported), using 33 highly polymorphic microsatellite markers spanning the 27-cM region of chromosome 2q31-33.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

The experimental protocols used in the study were accepted by the Institutional Review Board of the Columbia University College of Physicians and Surgeons. Methods used for clinical examination, as well as the diagnostic criteria used for PPH, have been described elsewhere (1).

DNA for genotyping was extracted from whole-blood samples or formalin-fixed, paraffin-embedded tissue from consenting individuals (1). The genotyping was done using the ABI Prism system and GENOTYPER 1.1.1 software (Perkin-Elmer Applied Biosytems, Norwalk, CT), with 33 highly polymorphic fluorescent microsatellite markers (see Table 1 for the list of markers used). These polymorphic markers were amplified and interpreted as described elsewhere (3, 4). Genetic locations of these markers were obtained from: the sex-averaged chromosome 2 Marshfield map (http://www.marshmed.org/genetics/; accessed June 1999) (5), Genethon map (http://landru.cephb.fr/ceph-genethon-map.html; accessed June 1999) (6), and LDB integrated map (http://cedar.genetics.soton.ac.uk/pub/chrom2; accessed June 1999) (7).

View larger version (22K): [in this window] [in a new window]   Figure 1.   Results of radiation hybrid mapping and the minimal genetic region for PPH1. Solid line indicates regions of sharing between at least two affected individuals in Family 15 and Family 21. The PPH minimal genetic region shared by Families 15 and 21 is depicted as a bar with arrowheads. The order of markers marked with an asterisk (*) was generated by using radiation hybrid mapping. Markers D2S364 and D2S350, which gave a maximum multipoint lod score of 3.87 in a previous study, are marked with a hatchmark (#).

Radiation hybrid mapping of 18 microsatellite markers was done in duplicate on both the Massachusetts Institute of Technology (MIT) GeneBridge 4 (http://carbon.wi.mit.edu:8000/cgi-bin/contig/rhmapper. pl; accessed June 1999) (8) and Stanford TNG radiation hybrid panels (http://www.shgc. stanford.edu/; accessed June 1999) (9). The order and location of these markers were analyzed by sending results to the radiation hybrid mapping servers at MIT and Stanford University (see foregoing hyperlink).

Two-point lod-score calculations were made by use of the MLINK option of the LINKAGE package of programs 5.1 (10). The VITESSE algorithm was used for the multipoint likelihood calculations (11). Autosomal dominant inheritance with disease penetrances assigned according to age- and gender-related incidence data was used as described previously (1). The phenocopy rate was set at 0.000001 and the disease allele frequency was set at 0.00001. The allele frequencies of the markers were obtained from the healthy chromosome (n = 80). Male and female recombination fractions were assumed to be equal. The computer program HOMOG was used to test for genetic heterogeneity (10). The transmission/disequilibrium test (TDT) (12) and the multiallelic extension of the haplotype relative risk (HRR) test (13) were used to look for evidence of linkage disequilibrium.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Table 1 shows the two-point lod scores derived under the assumption of an autosomal dominant model of transmission, a disease allele frequency of 0.00001, and penetrance estimated from age- and gender-related incidence data (1). The order and genetic distances between the markers in Table 1 are derived exclusively from the sex-averaged chromosome 2 LDB integrated map. A maximum two-point lod score of 7.23 at a recombination fraction of zero was obtained for marker D2S307. All of the families investigated in the study appeared to be linked to the 27-cM region of interest on chromosome 2q31-q33, and there was no evidence of genetic heterogeneity with the HOMOG program (Table 1). The results of the TDT and HRR tests were nonsignificant after correction was done for multiple tests.

Haplotypes were determined by assuming the minimal number of recombination events, based on the marker order in Figure 1. Six of the 17 families investigated had at least one recombination event between two affected individuals, somewhere in the 27-cM region. Two of these six families establish the minimum genetic region for PPH1. As illustrated in Figure 2, a recombination event that defines the centromeric boundary was observed in two affected individuals in Family 15, distal to markers D2S2327 and D2S311. In Family 21 a recombination event was found proximal to markers D2S1384 and D2S2237, and defines the telomeric boundary. These observations reduce the candidate interval for the location of PPH1 from 27-cM to the 3.3 cM (Genethon), 3.58 cM (Marshfield), or 4.1 cM (LDB) regions telomeric to D2S311 and centromeric to D2S1384.

View larger version (23K): [in this window] [in a new window]   Figure 2.   Shared haplotypes in the minimal genetic region of PPH1. Partial pedigrees of Families 15 and 21 are shown, and marker allele data are drawn below each individual who was genotyped. The common core haplotypes shared within each family are given in boxes. The marker order is derived from our radiation hybrid mapping data.

To construct physical maps of the 3.3-to-4.1-cM minimal genetic region for PPH1, we performed radiation hybrid mapping on both the MIT GeneBridge 4 and Stanford TNG radiation hybrid panels with the 18 microsatellite markers located within the critical region, in duplicate. The marker order determined from our radiation hybrid mapping data (Figure 1; markers indicated with an asterisk) was not consistent with any of the three genetic recombination maps. However, it was consistent with the marker order that can be derived from the physical maps of the region on MIT yeast artificial chromosome (YAC) contigs WC2.15, WC-497, and WC-630, except for those markers that are not placed in the MIT contigs (http://carbon.wi.mit.edu:8000/cgi-bin/contig/phys_map; accessed June 1999) (14).

Multipoint linkage analysis of the 17 pedigrees with seven markers (D2S2327, D2S2396, D2S1367, D2S309, D2S307, D2S72, and D2S1384) was then performed, using the intermarker distances from the LDB integrated recombination map because it provided the best fit to the physical map (Figure 3). A maximum multipoint lod score of 7.41 was observed close to marker D2S1367. The lod-1 support interval is 2.8 cM in length and is slightly telomeric to marker D2S2396 and centromeric to marker D2S307.

View larger version (8K): [in this window] [in a new window]   Figure 3.   Multipoint linkage analysis of the PPH1 locus, against a fixed set of markers. The markers used are shown above the horizontal axis. The distances and the order between the markers were derived from the sex-averaged chromosome 2 map from the LDB integrated map.

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

We have observed recombination events that substantially reduce the previous 25- to 27-cM candidate interval for PPHI to the region of approximately 3 cM separating D2S311 and D2S1384. Because we had previously observed a recombination event at D2S1384 (1) and confirmed it in Family 21 (Figure 2), the reduction in the minimal genetic region for PPHI has entirely been the result of movement of the centromeric boundary. The crossover in Family 21 was also observed with D2S2237, which is located centromeric to D2S1384 on the Marshfield map, raising the possibility that D2S2237 is the telomeric boundary for the interval containing PPH1. Our radiation hybrid mapping data and the MIT YAC contig map place D2S2237 telomeric to D2S1384, and we therefore think that the latter marker defines the telomeric boundary for the PPH1 interval.

This entire interval lies within chromosome 2q33. The excluded region between D2S1776 and D2S311 encodes the homeobox D cluster and the genes for integrin subunit-4 (CD49D), integrin subunit-v (CD51), tissue-factor pathway inhibitor, titin, and collagen type III -1 and type V -2 polypeptides, in addition to 89 complementary DNA (cDNA) markers (15). The remaining interval, between D2S311 and D2S1384, contains seven known genes and 73 cDNA markers (15). The genes for CD28 and apoptotic cysteine protease are two of these genes.

All 17 pedigrees in our study showed evidence of linkage of PPH to chromosome 2q33, strongly suggesting that the disease is genetically homogeneous, abeit with reduced penetrance. We tested for potential locus heterogeneity using HOMOG, and were unable to find any evidence of heterogeneity. As can be observed in Table 1, none of the markers in the minimal genetic region had substantially greater maximum lod scores under the assumption of heterogeneity (hence the proportion, , of families that appeared linked was equal to or close to 1). Even though there is no statistically significant evidence of genetic heterogeneity for the disease, the possibility of a second, unlinked, locus for familial PPH remains, and will continue to remain, until the PPH1 gene has been examined in all families with familial PPH.

Even though PPH in all of the families in our study appeared to be linked to chromosome 2q33, we did not observe a shared disease haplotype between the familial PPH pedigrees with the current marker density. Likewise, we observed no evidence of linkage disequilibrium using either the HRR test or TDT. Given that our collection of pedigrees was ethnically mixed, this is not surprising. We have used nearly all the microsatellite markers in the minimal genetic region for PPH1, but are hopeful that by constructing a higher density map of microsatellite and single nucleotide polymorphism markers we will observe evidence of shared DNA segments or linkage disequilibrium across families. Success in this endeavor will depend on whether the disease allele in each family is distinct or whether some of the families are related to a common founder. Because the familial form of PPH is rare, and the disease has its onset after the age of reproduction, there is some hope that mutations in the disease gene would have been an infrequent event and that the latter situation would apply.

Tuder and colleagues (16) observed frequent monoclonal endothelial cell proliferation in primary but not in secondary pulmonary hypertension, and suggested that the pathogenesis of PPH may be similar to a neoplastic process. Interestingly, allelic losses of chromosome 2q33 (D2S116) in lung cancers have been reported in several studies (17), and suggest the presence of a tumor-suppressor gene. D2S116 (209.616 cM) maps between D2S1367 (209.394 cM) and D2S309 (209.650 cM) on the LDB integrated map, and all three markers are contained on YAC 849B12 (440 kb) in MIT YAC contig WC2.15. Another 220-kb YAC, 796C3, was reported by the Genethon group to contain both D2S116 and D2S309 (20). This is the region that gave our maximum multipoint lod score (Figures 1 and 3; Table 1).

Although marker D2S1271 (UT7430) gave high two-point scores (Table 1), we did not use this locus in the multipoint analysis because of the extreme variability of its location in all of the maps. It is regarded as a cryptic duplicate of marker GATA161E02 (D2S2979) at 198.65 cM on the Marshfield map. On the LDB integrated map, D2S1271 is placed at 211.014 cM and D2S2979 is at 209.828 cM. Since the marker order is crucial in multipoint linkage analysis (11), we did not include D2S1271. Given the high two-point lod score with D2S1271, we are currently trying to physically place both it and D2S2979 on the MIT YAC contigs.

Although considerable scientific information on biochemical and cellular mechanisms of PPH has been generated in recent years, the central basic cause for the disease remains unclear (21). Therefore, identification of the gene responsible for familial PPH will contribute to understanding of the etiology of PPH, and would provide the means for its earlier diagnosis and perhaps for novel treatments for it. Our findings provide the basis for further attempts to identify PPH1.