GENOMICS

11,857~869

(19%)

Linkage Mapping of Autosomal Dominant Retinitis Pigmentosa (RPI) to the Pericentric Region of Human Chromosome 8 SUSAN HALLORAN BLANTON, JOHN R. HECKENLIVELY,* ANNE W. COTTINGHAM, JACKIE FRIEDMAN, LORI A. SADLER, MICHAEL WAGNER,t LORRAINE H. FRIEDMAN,*,’ AND STEPHEN P. DAIGER’ Graduate School of Biomedical Sciences, The University of Texas Health Science Center, Houston, Texas 77030; *Jules Stein Eye Institute, University of California, Los Angeles, California 90024; and tlnstitute for Molecular Biology, University of Houston, Houston, Texas 77204 Received

June 13, 1991;

INTRODUCTION

Retinitis pigmentosa (RP) is a diagnosis given to a number of clinically similar but genetically distinct

correspondence

should

July 17, 1991

forms of heritable retinal degeneration (Heckenlively, 1988; Marmor et al., 1983). Clinical findings in RP include progressive loss of night and peripheral vision, usually culminating in severe visual impairment or blindness. The diagnosis is confirmed by an abnormal or extinguished electroretinogram (ERG) and characteristic retinal atrophy with deposition of pigment and attenuation of retinal vessels. While only one pathogenetic mechanism that causesRP has been determined (defects in rhodopsin), it is likely that most defects affect the rod photoreceptor cells in the retina, either directly by mutations in photoreceptor proteins or indirectly by mutations in other retinal genes or genes expressed in the retinal pigment epithelium. Survey results suggest an incidence of RP among European and American Caucasians of l/3500 to l/5000 (Boughman et aZ., 1980; Halloran, 1985). Of these cases, roughly 14% are autosomal recessive, 17% are autosomal dominant, 10% are Xlinked, and 42% are isolated; that is, the mode of inheritance is unknown (Heckenlively, 1988). The remaining 17% of cases are syndromic, involving additional clinical findings such as deafness (e.g., in Usher syndrome or Bardet-Biedl syndrome.) ADRP is very heterogeneous clinically. Several clinical classification systems have been proposed (Fishman et al., 1985; Heckenlively, 1988, Lyness et al., 1985; Massof and Finkelstein, 1981). Two types emerge from psychophysical studies: type 1 ADRP (also known as diffuse or D-type) and type 2 ADRP (also known as regional or R-type.) Electroretinographic classification closely parallels the psychophysical, with rod-cone degeneration showing type 1 and cone-rod degeneration usually demonstrating type 2 psychophysical changes. Type 1 families are characterized by an early onset of night blindness, with the onset of symptoms by age 11, by diffuse loss of rod sensitivity, and, if recordable, by loss of rod ERG while cone ERG is more preserved (Heckenlively et al., 1988). Type 2 families have later onset of

Linkage mapping in a large, seven-generation family with type 2 autosomal dominant retinitis pigmentosa (ADRP) demonstrates linkage between the diseaselocus (RPl) and DNA markers on the short arm of human chromosome8. Five markers were most informative for mapping ADRP in this family using two-point linkage analysis. The markers, their maximum lod scores, and recombination distances were ANKl (ankyrin)-2.0 at 16%; DSSS (TLll)-5.3 at 17%; D8S87 [a(CA), repeat]-7.2 at 14%; LPL (lipoprotein lipasef-1.5 at 26%; and PLAT (plasminigen activator, tissue)-10.6 at 7%. Multipoint linkage analysis, using a simplified pedigree structure for the family (which contains 192 individuals and two inbreeding loops), gave a maximum lod score of 12.2 for RPl at a distance 8.1 CM proximal to PLAT in the pericentric region of the chromosome.Based on linkage data from the CEPH (Paris) reference families and physical mapping information from a somaticcell hybrid panel of chromosome8 fragments, the most likely order for four of these five loci and the diseaseslocus is 8pter-LPL-DBSB-DSS87-PLATRPl. (The precise location of ANKl relative to PLAT in this map is not established.) The most likely location for RPl is in the pericentric region of the chromosome.Recently, several families with ADRP with tight linkage to the rhodopsin locusat 3q21-q24 were reported and a number of specific rhodopsin mutations in families with ADRP have sincebeen reported. In other ADRP families, including the one in this study, linkage to rhodopsin hasbeenexcluded. Thus mutations at two different loci, at least, have been shown to causeADRP. There is no remarkable clinical disparity in the expression of diseasecausedby these different loci. 0 lssl Academic Press, Inc.

1 Deceased. 2 To whom

revised

be addressed. 857

o&x3-7543/91

Copyright 0 1991 All rights of reproduction

$3.00

by Academic Press, Inc. in any form reserved.

858

BLANTON

symptoms, typically starting from the mid-twenties to the third decade of life, with regional and combined loss of both rod and cone sensitivities on psychophysical testing (Heckenlively, 1987). While clinical types of ADRP appear to be inherited consistently within families, there is considerable within-family variation in expression, especially in type 2 families (Boughman, 1978; Lyness et al., 1985). Instances of incomplete penetrance (skipped generations) have been reported in type 2 families although this is difficult to assessin this late onset disease (Berson et aZ., 1969; Berson and Simonoff, 1979). ADRP is equally heterogeneous genetically. In 1989 Humphries and colleagues reported tight linkagewith a maximum lod score of 14.7 and no recombinants-between the ADRP locus in an Irish family and the anonymous DNA marker D3S47 (CRI-C17) (McWilliams et al., 1989). Subsequent studies demonstrated tight linkage to rhodopsin in this family, with the D3S47 and rhodopsin loci approximately l-3% recombination from each other (Farrar et al., 1990b). In independent studies, Dryja and colleagues reported a proline-to-histidine mutation in rhodopsin at codon 23 (Dryja et uZ., 1990). Approximately 11% of American ADRP patients have the codon 23 mutation. Recently, additional rhodopsin mutations have been reported in ADRP patients, although not in the original Irish pedigree (Bhattacharya et cd., 1991; Dryja et al., 1991; Inglehearn et al., 1991; Stone et al., 1991; Farrar et al, 1990a). Although roughly one-third of ADRP patients have a deleterious rhodopsin mutation or show tight linkage to rhodopsin, the remainder can be excluded from tight linkage or, at least, they possessno detectable rhodopsin mutation (Farrar et uZ., 1990b; Inglehearn et al., 1990; Jimenez et uZ.,1991; Lester et al., 1990). Of the families excluded from tight linkage to rhodopsin, most can be excluded from 3q altogether although at least one family shows more distant linkage to D3S47 (Olsson et uZ., 1990). Specifically, the ADRP locus in the family presented here has been excluded from much of 3q (Blanton et al., 1990a). The family discussed in this report, UCLA-RPOl, has been the subject of a number of earlier publications. UCLA-RPOl is a large, extended family with over 150 living, affected members, all of whom can trace their diseaseto an affected ancestor living in the early 19th century. The common ancestor and many of her descendants live or lived in the Appalachians in Kentucky and surrounding states. Diagnostic details of the retinitis pigmentosa in UCLA-RPOl have been provided in earlier publications (Heckenlively et al., 1982; Daiger et cd., 1989b). The family has classical type 2 ADRP with relatively late onset of night blindness, usually by the third decade of life, and with slow progression. Characteristic clinical findings include diffuse retinal pigmentation,

ET

AL.

progressive decrease in recordable ERGS, and concentric visual field loss. Funduscopic findings include retinal atrophy, bone-spicule-like pigment deposits, and vascular attenuation. Like other families with type 2 ADRP, though, there is extensive variation in the age of expression of clinical milestones. Some individuals recall development of night blindness in childhood and are severely visually impaired by age 40. Others show unequivocal signs of retinitis pigmentosa in their 60’s but still retain modest visual function. Two deceased obligate carriers, one who died at age 35 and the other at 95, reported no apparent symptoms. One of the long-term goals of this research is to better understand the basis of this wide range of clinical expression. In earlier studies, linkage between the ADRP locus in UCLA-RPOl and amylase was proposed but subsequently withdrawn (Spence et uZ., 1977). Tentative linkage to Rh was reported but further testing with DNA markers excluded both Rh and much of the surrounding region of chromosome 1 (Field et uZ., 1982; Heckenlively et al., 1982; Daiger et al, 1989a). Based on the tentative linkage to Rh, though, the symbol “RPl” was assigned to the disease locus in this family. Finally, exclusion of chromosome 4 and at least one-third of the remaining genome has been reported (Blanton et al., 199Ob; Daiger et aZ., 198913; 1991). MATERIALS

Pedigree Construction

AND

METHODS

undo Clinical Evaluation

The complete, known pedigree of UCLA-RPOl includes over 600 living and deceased individuals. A more compact, subordinate pedigree (Fig. 1) was selected from the larger pedigree for complete clinical evaluation and for linkage testing, with an emphasis on affected individuals. The structure of the complete pedigree, and of the subordinate pedigree, was established by two decades of research by a number of investigators. The pedigree is based on interviews of informative family members and on questionnaires completed by all adult individuals participating in the study. A genealogy of the family produced by one of the family members was a valuable contribution to accurate construction of the pedigree (Breeding, 1982). The pedigree shown in this publication is different in some aspects from earlier published pedigrees (Daiger et uZ., 198913;Heckenlively et uZ., 1982) because new individuals have been added and new diagnostic information is incorporated. Also, there are minor changes in the lineages of earlier generations. Since the genotypes of living members of the pedigree in Fig. 1 have been determined at over 69 loci, the possibilities of pedigree errors or nonpaternity can be

LINKAGE

MAPPING

OF

ADRP

TO

CHROMOSOME

859

8

excluded. The root of the pedigree, for purposes of tracing the ADRP gene, is an affected ancestor who lived in Kentucky from 1803 through 1896. Tentative historical information suggests that her father and paternal grandfather, who lived in the United States, were also affected. Thus the disease in this family can be traced with certainty to the early 19th century and probably into the mid-18th century. Clinical diagnosis was established in all individuals from whom blood was drawn using the following strategy. All individuals included in the study were examined by Dr. Heckenlively. The diagnostic status of individuals who were not at risk (e.g., unaffected spouses) or who had overt retinal disease was confirmed by direct ophthalmoscopy, and in most cases by indirect ophthalmoscopy. Individuals in whom the diagnosis was questionable had careful indirect ophthalmoscopy and fundus photography. Selected individuals had ERGS and fluorescein angiography at the Jules Stein Eye Institute, University of California, Los Angeles. Finally, the diagnostic status of deceased individuals was established by review of clinical records and/or family Bibles, or by information provided by knowledgeable family members who had personal knowledge of the patient. Cell Lines and Preparation

of Genomic DNAs

Blood samples from family members were provided to the Human Genetic Mutant Cell Repository (Camden, NJ) for preparation of transformed cell lines (Daiger et at., 1987; Daiger, 1990). These cell lines are part of the RP Cell Line Collection in the Repository. The cell lines were subsequently grown in Houston and genomic DNAs prepared using a nucleic acids extractor from Applied Biosystems, Inc. (South San Francisco, CA). Details are in earlier publications (Daiger et al., 1989b). DNAs from CEPH reference families were provided by the Centre d’Etude du Polymorphisme Humain, Paris, France, to Dr. Daiger as a participant in the CEPH linkage consortium (Dausset et al., 1990). Genetic Marker

Testing

Tables 1 and 2 list all genetic markers tested for linkage to the ADRP locus in UCLA-RPOl. In addition, several markers listed in Table 2, specifically ANKl, D8S5-S&I, D8S87, and LPL, were tested in the CEPH reference families to refine the linkage map of 8p. The DNA probes listed in Tables 1 and 2 were from a number of sources, cited either in previous reports (Donis-Keller et al., 1987; McWilliams et al., 1989; Daiger et a!., 1989b) or in the Acknowledg-

FIG.

1.

Pedigree

of UCLA-RPOl.

860 ments Table Culture protein ratory

BLANTON

of this article. All of the DNA probes listed in 2 are also available from the American Type Collection (ATCC) (Rockville, MD). The and serologic markers were tested in the laboof Dr. Robert S. Sparkes.

Southern

Gel Analysis

Southern gel analysis was performed as described previously (Daiger et al., 1989b). Those probes containing repetitive elements, such as D8S5 and D8S26, were blocked prior to hybridization using sheared, boiled human placental DNA following a protocol provided by Collaborative Research, Inc. (Bedford, MA). Polymerase

Chain Reaction

(PCR)

Three types of DNA markers listed in Tables 1 and 2 were detected using PCR amplification: VNTR (variable number of tandem repeat) polymorphisms (Nakamura et al., 1987), simple restriction site polymorphisms, and STR (short tandem repeat) polymorphisms (Weber and May, 1989). Following PCR amplification, VNTR-type polymorphisms were detected using ethidium bromide fluorescence after electrophoresis in 0.8% agarose gels (Boerwinkle et al., 1989). Simple restriction site polymorphisms were also detected on ethidium bromide-stained agarose gels, subsequent to restriction digestion (e.g., Rodriguez et al., 1990). STR polymorphisms, both dinucleotide repeats [e.g., (CA),] (Weber and May, 1989) and tetranucleotide repeats [e.g., (TTTA),] (Zuliani and Hobbes, 1990) were detected using thin, vertical polyacrylamide gels (“sequencing gels”) and autoradiography with primers end-labeled with 32P. Additional technical details are given in the references listed in Table 2 and in Innis et al., (1990). Amplification conditions for the STR polymorphisms at the ANKl, D8S87, and LPL loci are given in the references listed in Table 2. Primers for PLAT were 5’-GGA GTT CGA GAC TAG CCT GGA-3’ and 5’-ACT TCA GGC ATG TGC CAC TG-3’. Amplification conditions for PLAT were 20 cycles of denaturation at 94°C for 1 min, annealing at 55°C for 2 min, and elongation at 72°C for 2 min, producing a fragment 111 bp in length disregarding polymorphic variation. PCR primers for detection of genetic variation either were those listed in the original publications or were selected with assistance of the computer programs PRIMERS (Lowe et al., 1990) and NAR (Rychlik and Rhoads, 1989), using GenBank sequence data accessed through the Molecular Biology Information Resource, Baylor College of Medicine (Houston, TX). All primer pairs were purchased either from Genosys Biotechnologies (Woodlands, TX) or from the ATCC. PCR amplification was conducted

ET

AL.

in a DNA thermal (Richmond, CA). Linkage

cycler from Perkin-Elmer

Cetus

Analysis

Two-point linkage analysis was performed using LIPED (Ott, 1976) and multipoint analysis was conducted using the LINKAGE package (Lathrop et al., 1984). Preliminary analyses were run on an 80386/ 80387-based microcomputer and final analyses were run on the VAX 8550 mainframe computer at the School of Public Health, University of Texas Health Science Center (Houston, TX). We faced two problems in linkage analysis in UCLA-RPOl. First, the late onset and incomplete penetrance of the disease gene necessitated correction for penetrance and expressivity. Second, the complete pedigree, with two inbreeding loops and seven generations (Fig. l), was not amenable to complete multipoint analysis, even on a Cray supercomputer (ReedFourquet and Daiger, 1989). Clinical experience with UCLA-RPOl suggests that all obligate carriers, that is, parents of affected children, have a greater than 95% chance of expressing the disease gene by age 40; that some individuals are clearly affected before age 40; and that some individuals who are at risk of inheriting the disease gene may show only equivocal signs of disease. Thus, to deal with the twin problems of penetrance and expressivity, we divided family members into six categories: (1) not at risk; (2) at risk, under age 40, with no evidence of disease; (3) at risk, over age 40, with no evidence of disease; (4) at risk, possible affected; (5) at risk, probably affected, and (6) unequivocally affected. “Possibly” affected and “probably” affected individuals were classified on the basis of showing at least two or at least three, respectively, of the clinical features described by Heckenlively (1988, Chap. 5). From these considerations, the following computer coding system emerged. Individuals not at risk were coded as not possessing the disease gene. Affected individuals and unaffected at-risk individuals over the age of 40 were assigned lifetime penetrances of 95%. Unaffected at-risk individuals under the age of 40 were coded as disease status unknown. Finally, the chance that a gene carrier would be “possibly” affected or “probably” affected was assigned an 80 or 90% probability, respectively. These probabilities translate directly into the LIPED genotype:phenotype matrix. Implementation of this scheme in LINKAGE involved a procedure suggested by Ott (1988). The second problem, computational limitations, was addressed using the following stratagem. All twopoint analyses using LIPED included the entire pedigree structure, with the caveat that no marker could have more than 6 alleles after recoding. For multipoint analysis the family was split into three

LINKAGE

MAPPING

OF

branches, indicated in Fig. 1 by dashed lines. These partitions eliminate the inbreeding loops and require the duplication of two individuals indicated by an asterisk in the figure. Lod scores were calculated for each subunit and then summed. Because we were not able to test all four informative markers simultaneously, lod scores for the disease locus were calculated separately for each pair of contiguous markers. Additionally, there were 10 alleles for PLAT segregating in the family, so that linkage between PLAT and RPl was tested in the divided family only. The PLAT alleles were recoded so that only 6 alleles segregated in each subfamily. Exclusion

Mapping

Exclusion mapping was performed using the EXCLUDE computer program (Edwards, 1987). The EXCLUSION map shown in Fig. 2 includes all data from Table 2 and three markers from Table 3. Allele and Haplotype

Frequencies

Linkage calculations in dominant pedigrees with many untested individuals are very sensitive to the choice of allele frequencies (Daiger and Blanton, 1990). UCLA-RPOl contains 25 unrelated spouses who can be used for determining frequencies. When published allele frequencies were from populations of comparable or smaller size, values from UCLA-RPOl were chosen. When larger populations were tested or when CEPH data were available, these values were used instead of frequencies from UCLA-RPOl. Haplotypes frequencies used for testing linkage to D8S26 were calculated on the basis of published frequencies, assuming linkage equilibrium. Haplotype frequencies for D8S5 were based on observed frequencies (Cottingham et al., 1991). Methods for implementing haplotypes in LIPED and LINKAGE are given in Daiger and Blanton (1990). Physical

Mapping

Physical localization of DNA clones and PCR-detectable sequences on chromosome 8 was accomplished by use of a hybrid cell mapping panel developed by Wagner et al. (1991). The panel consists of human-hamster hybrid cell lines containing fragments of human DNA dividing chromosome 8 into 10 contiguous or overlapping regions spanning the entire chromosome. Physical mapping of DNA probes involved Southern gel blotting of cell line DNAs. Mapping of PCR-detectable markers was achieved by amplification of cell line DNAs with appropriate primers. All Southern gel blots and PCR experiments included genomic DNAs from control hamster and human cell lines.

ADRP

TO

CHROMOSOME

8

861

RESULTS

Exclusion

Map

Table 1 lists genetic markers tested for linkage to the ADRP locus in UCLA-RPOl in earlier studies @pence et al., 1977; Field et al., 1982; Daiger et al., 1989a, 1989b, 1991; Blanton et al., 1990a, 1990b). Three chromosome 8 markers, D8S26, DSS39, and GPT, were also reported in earlier studies but are listed in Table 2. Figure 2 is an exclusion map based on these markers. (An earlier version is shown in Daiger et al., 1991.) An exclusion map shows both the relative probability that the test locus, ADRP in this case, resides on a particular chromosome and the most likely location within that chromosome. The width of the chromosome “box,” and the value to the right of the chromosome number, show the relative probability. The width of the graph contained within the box indicates the likelihood of a particular location. From these preliminary data we concluded that the most probable chromosomal site for ADRP in this family was chromosome 8, with a relative probability of 22%. The most likely location on chromosome 8 is on either side of the centromere. Two-Point

Linkage

Testing

On the basis of evidence from the exclusion map suggesting chromosome 8 as the site of the ADRP locus in UCLA-RPOl we tested a number of additional chromosome 8 markers, listed in Table 2. To improve the information content of the DSS5 locus, originally tested with the HindIII polymorphism only, we screened a panel of DNAs from normal controls for restriction site polymorphisms. We found a polymorphic Sty1 site that is in disequilibrium with the Hind111 site (Cottingham et al., 1991). This restriction site polymorphism permitted construction of D8S5 haplotypes in UCLA-RPOl, with an observed heterozygosity of 55%. Two sites were tested at the D8S26 locus also. Table 3 summarizes the resulting two-point lod scores, including the haplotypic markers DSS5 and D8S26. We also developed a new PLAT-(CA), polymorphic marker (details under Materials and Methods and in a manuscript in preparation) . Five markers listed in Tables 2 and 3 gave suggestive evidence for linkage to the ADRP locus in UCLA-RPOl: ANKl, CA2, D8S26, GPT, and LPL. Three markers gave highly significant evidence of linkage: D8S5, DSS87, and PLAT. The highest twopoint value was for PLAT with a maximum lod score of 10.6 at 7% recombination. The statistically significant lod scores in Table 3 are not sensitive to the choice of genetic parameters for LIPED or LINKAGE. Whether allele frequencies

862

BLANTON

from UCLA-RPOl, CEPH reference families, or published sources are used, maximum lod scores differ by less than 10%. Whether possibly affected and probably affected individuals are included or excluded from the analysis, results differ by less than 5%. Thus, although we believe we have chosen optimal parameters, the conclusions are robust and largely independent of underlying assumptions. Finally, we chose sex-averaged distances between markers, rather than an arbitrary male:female ratio, because, in our opinion, the linkage map of 8p is not sufficiently established to justify this degree of refinement.

Multipoint

Linkage

Testing

Linkage evidence from the chromosome 8 committee report in Human Gene Mapping 10.5 (DonisKeller and Buckle, 1990) and CEPH data (CEPH, 1990; and unpublished) suggest that ANKl, DSS5, DSS87, LPL, and PLAT are in the same linkage group. Physical mapping evidence from Wagner et al. (1991) and that from Human Gene Mapping 10.5 (1990) show all five markers to be on the short arm of chromosome 8, from 8~22 to the centromere. We therefore decided to focus on these markers for multipoint mapping in UCLA-RPOl. Our first need was to establish an accurate linkage map of the five markers. From physical data we knew that LPL must be distal to the other four markers. Although some CEPH data were already available for D8S5, LPL, and PLAT, the specific sites tested were not highly informative and data were not available for the remaining two markers. For these reasons we tested the ANKl-(AC),, DSS5-StyI, D8S87-(CA),, and LPL-(TTTA), polymorphisms in the 40 CEPH reference families. This adds ANKl and D8S87 to the CEPH data set and increases the informativeness of D8S5 and LPL. These data were analyzed for the most likely order and distances using the LINKAGE program package assuming no interference. From our CEPH results and physical mapping data, we were able to establish the following order and distances with high confidence for four of the five markers [distances in CM using the Haldane conversion (Ott, 1985)]: 8pter-LPL-(

18.1)-D8S5(14.8)-D8S87-(

lO.O)-ANKl-8cen.

The proposed linkage order is consistent with the physical map in that LPL is telomeric and is more likely than the second most probable order, reversing ANKl and D8S87, by a factor of over 100,000. A manuscript with further details is in preparation. PLAT is centromeric to DBS87, but its position relative to ANKl could not be determined. The order

ET

AL.

placing PLAT centromeric to D8S87, at a distance of 10.9 CM, was favored by odds of 100,OOO:l over the next favored order, which placed PLAT telomeric to LPL. (We are currently testing the PLAT-(CA), polymorphism in the CEPH families to establish the precise order.) Based on this map we used LINKMAP to determine the lod scores for contiguous pairs of markers versus ADRP. Graphical results are shown in Fig. 3. Because of uncertainty in the order of ANKl versus PLAT, we took as contiguous pairs first D8S87ANKl and then D8S87-PLAT. Note that these results are based on the simplified family structure described under Materials and Methods. Thus the total information provided by the family is reduced. As a result the maximum lod scores in the multipoint analyses are not substantially higher than the two-point lod scores. In all four calculations, shown in Fig. 3, the most likely position for the ADRP locus is centromeric to the five-marker linkage group. The likelihood ratios of the maximum lod scores for a proximal versus a distal location for ADRP in the four sets of calculations are 125:l for LPL-D8S5-ADRP, 5O:l for D8S5D8S87-ADRP, 13:l for D8S87-ANKl-ADRP, and 20,OOO:l for D8S87-PLAT-ADRP. For all four calculations the least likely location for the ADRP locus is between the marker loci. Thus, although current computational limitations preclude a complete multipoint analyss of the five markers and the ADRP locus in UCLA-RPOl, the evidence is strongly in favor of a location for this form of retinitis pigmentosa proximal to both ankyrin and PLAT. The D8S87-PLAT-ADRP multipoint map gives a maximum lod score of 12.2 for ADRP at a distance 8.1 CM proximal to PLAT. This order is favored by over 20,OOO:l over the order ADRP-D8S87-PLAT. Since this is the highest lod score, since these are two of the closest markers to the disease locus, and since PLAT shows the highest two-point lod score, we feel that this is the most accurate estimate of the distance between the four-marker linkage group and the ADRP locus in UCLA-RPOl.

Physical Localization Wagner et al. (1991) assigned a number of DNA markers to physical intervals on human chromosome 8 using a human-hamster hybrid cell panel. Intervals B, C, and D are contiguous and nonoverlapping and cover a region extending from the middle of the short arm (B) to the proximal long arm (D) of chromosome 8. Wagner et al. showed that DBS5 maps to interval C in the proximal portion of the short arm. The same mapping panel was used to map the ANKl-(AC),, D8S87-(CA),, and LPL-(TTTA), sequences (manuscript in preparation). We find that LPL maps to in-

LINKAGE

MAPPING

OF

ADRP

TO

TABLE Polymorphic Locus PND ALPL RH DlS57 PGMl AMY2 FY APOB ACPl D2Sl D2S3 D3S32 D3S47 D3S14 TF D4SlO D4S21 D4S23 D4S18 D4S16 D4S44 D4S40 GC MNS D4S118 D5SlS RDS GLOl BF D6S29 CF ABO ASSP3 RBP3 TYR VWF RBl ESD PI D15Sl PGP HP D18S6 JK C3 LE INSR APOCl D21S58 D21S13 PI

Location 1~36 lp36.1-p34 lp36.2-p34 lp36-p33 lp22.1 lp21 lq21-q25 2p24-p23 2~25 2~25 2q35-q37 3cen 3q21-q24 3q21-q24 3q21-q26.1 4p16.3-~16.2 4p16.l/p15.1 4p16.1-~15.1 4p16.1-~15.1 4p15.1-qll Iqll-qter Iqll-qter 4q12-q13 4q28-q31 4q 5p14 6~21 6~21 6~21.3 6~21 7q31-q32 sq34.1-q34.2 Sqll-q22 lOq11.2 llq14-q21 12pter-p12 13q14 13q14 14q32.1 15q14-q21 16~13 16q22.1 18~11 18qll-q12 19p13.3-~13.2 19 19p13.3-~13.2 19~13.2 21q22.2-q22.2 21q11.2 22q11.2-qter

Markers

Included

CHROMOSOME

1

in Exclusion

Map of ADRP

in UCLA-RPOl”

Exclusion’ (-2.0 LOD)

Z-

Probe-enzyme or typeb JAllO-XhoI;BglI LBKbB-SacI;BclI Serologic YNZ2-RsaI Protein Protein Serologic PCR: VNTR Protein L2.30-BglII;MspI P5-l-32-Hi&II EFD145-RsaI CRI-C17-M@(2) CRI-R208-MspI Protein K082-HindIII(2) GlE5-BamHI GDS5-BgZII 4F2-BamHI 3E5-MspI B3D-PstI B5A-PstI Protein Serologic CRI-C82-M@(2) J044E.B-MspI HRDS8-BglII Protein Serologic HHH157-BamHI PCR: TaqI;PstI Serologic AS419-HindIII PCR: BglII PCR: Mb01 PCR: RsaI PRC: XbaI Protein Protein MSI-14-M& Protein Protein L2.7-PstI Serologic Protein Serologic PCR: NsiI PCR: HpoI PW524.5P-PstI PCR: EcoRI Protein

0.17 0.06 0.21 0.17 0.10 0.04 0.17 0.12 0.12 0.07 0.17 0.01 0.13 0.01 0.07 0.12 0.02 0.01 0.07 0.16 0.14 0.09 0.08 0.07 0.02 0.15 0.14 0.12 0.06 0.01 0.03 0.22 0.16 0.07 0.16 0.10 0.01 0.16 0.07 to.01 0.00 0.19