Publications

We have identified a polymorphic insertion in the lipoprotein lipase (LPL) gene in a captive baboon colony. Mapping and nucleotide (nt) sequence analysis of the polymorphism showed that it is due to the presence or absence of an Alu repetitive element in intron 7 of the baboon LPL gene. This polymorphic Alu repeat has not been reported in humans, and we did not detect the repeat in a survey of the LPL intron 7 gene region in other non-human primates. Comparison of nt at diagnostic positions in this Alu insertion with different Alu subfamily consensus sequences showed that it most closely resembles the young AluY subfamily. These data suggest that this polymorphic Alu repeat inserted independently in the baboon lineage.

We have isolated two lipoprotein lipase (LPL)-encoding cDNA (LPL) clones from a baboon cardiac cDNA library, one of which spans a region from nucleotide (nt) 705 of the coding sequence to the poly(A) tail (2.8 kb). We used reverse transcription followed by PCR (RT-PCR), and anchor-ligated rapid amplification of cDNA ends (RACE) to amplify the remaining 5' region of the LPL transcript. Sequence comparisons reveal that the baboon nt sequence is 95% identical to the human cDNA sequence (ranging from 97.5 to 92.7% in the coding and noncoding regions, respectively). Less than 2% of nt substitutions cause changes between baboon and human amino acid (aa) sequences. The aa in the catalytic triad residues, the heparin-binding site in exon 6, as well as aa in positions where missense mutations cause LPL deficiency, are identical in baboons and humans. Characterization of the tissue-specific expression of LPL using Northern blots of total RNA showed that spinal cord expressed the most LPL transcripts of all baboon tissues examined. Like humans, baboons have two transcript sizes of approx. 3.6 and 3.4 kb in most tissues that express LPL, and sequencing of the 3' untranslated region (UTR) shows this is due to two polyadenylation sites. In contrast, only the larger 3.6-kb transcript is detected in RNA isolated from central nervous system (CNS) tissues. We used RT-PCR to show that the polyadenylation signal that produces the 3.4-kb message is present in CNS LPL transcripts, but is not utilized.

We present results from an association study between RFLPs in the lipoprotein lipase (LPL) gene and lipid and insulin levels. The study population consisted of 102 Hispanic men and 97 Hispanic women. The subjects were genotyped for two previously reported RFLPs detected with the restriction enzymes HindIII and PvuII. The frequencies of the RFLPs in the Hispanic population are similar to those seen in other Caucasian populations. Strong linkage disequilibrium was detected between the sites in Hispanics. Genotypes were used separately in analyses of variance with fasting serum triglycerides, total cholesterol, high density lipoprotein (HDL)-cholesterol, low density lipoprotein (LDL)-cholesterol, HDL2/HDL3-cholesterol, and insulin levels, as well as two measures of adiposity: waist-hip ratio and body mass index. Men and women were analyzed separately. Mean fasting insulin levels of the LPL PvuII genotypes were significantly different from each other in Hispanic men. The mean fasting insulin level of men who were homozygous for the presence of the PvuII site (+/+) was 9.20 +/- 0.24 mu units/ml, men who were heterozygous had a mean level of 10.54 +/- 0.20 mu units/ml, and men who were homozygous for the absence of the site (-/-) had a mean of 12.91 +/- 0.30 mu units/ml. This effect was not seen in Hispanic women. These results suggest that the regulation of LPL by insulin may be different in Hispanics with different LPL PvuII genotypes.

Lipoprotein lipase (LPL) plays a crucial role in plasma lipoprotein processing by catalyzing the hydrolysis of core triglycerides of chylomicrons and very low density lipoproteins. Several polymorphic restriction sites have been reported in the LPL gene, including those identified by the enzymes HindIII and PvuII. We have determined the HindIII and PvuII polymorphisms in diabetic (D) and non-diabetic (ND) Hispanics (D = 195; ND = 384) and non-Hispanic Whites (D = 76; ND = 539) from the San Luis Valley, Colorado. Both polymorphisms showed comparable gene frequencies between diabetics and non-diabetics, and between the two ethnic groups. The HindIII and PvuII polymorphisms were in strong linkage disequilibrium in both Hispanics and non-Hispanic Whites (P < 0.001). We estimated whether the two DNA polymorphisms have significant impact in determining interindividual differences in plasma levels of total cholesterol, HDL-cholesterol, LDL-cholesterol, triglycerides, fasting glucose, and fasting insulin. Plasma triglyceride levels varied significantly among the HindIII genotypes in the normoglycemic sample. There was a clear gene dosage effect among the three HindIII genotypes, with the (-/-) genotype having the lowest and the (+/+) genotype having the highest triglyceride levels; these levels were intermediate in the (+/-) genotype. The average effect of the (-) allele of the HindIII polymorphism was to lower triglycerides by 12.85 mg/dl in non-Hispanic White males, 8.06 mg/dl in non-Hispanic White females, 10.91 mg/dl in Hispanic males, and 12.47 mg/dl in Hispanic females. The HindIII polymorphism also showed a significant association with HDL-cholesterol levels in the normoglycemic sample.(ABSTRACT TRUNCATED AT 250 WORDS)

The purpose of this work was to examine the influence of apolipoprotein gene variation on plasma lipid levels in a population of Mayan Indians of the Yucatán Peninsula, Mexico. Four restriction enzymes: XmnI, PstI, SstI, and PvuII, were used to detect restriction fragment length polymorphisms (RFLP) within the region of the apolipoprotein AI/CIII/AIV gene cluster. The frequencies of these polymorphisms in this Mayan population were similar to those reported for other Amerindian populations, but differed widely from those reported for Caucasian populations. The XmnI and SstI RFLPs were informative for association studies in this population, and we analyzed their influence on the quantitative variation of plasma cholesterol and triglycerides. Using a nonparametric analysis of variance, it is shown that the presence of the XmnI restriction site had a significant effect in lowering plasma cholesterol, whereas the presence of the restriction site for SstI had a significant effect in raising plasma triglycerides. Consequently, genetic indicators of both low and high risk for lipid-related diseases, such as atherosclerosis and coronary heart disease, seem to be present within the same gene region in this Mayan population.

The Dogrib, an Amerindian tribe residing in the Northwest Territories of Canada, were typed for DNA and protein polymorphism at the apolipoprotein A-I/C-III/A-IV gene cluster. Variation was seen at three previously described RFLPs detected with the enzymes SstI, PstI, and XmnI, though frequencies of these polymorphisms differ significantly from those reported in other populations. They exhibit no variation at two previously reported PvuII sites. No variation was seen in the APO A-I or APO A-IV gene products, with the Dogrib showing the most common isoelectric-focusing/immunoblot patterns of other world populations. Haplotype frequencies computed from inferred haplotypes and by maximum likelihood estimation did not differ significantly. The extent of nonrandom association of these sites is highly significant (P less than .00001), though pairwise analysis shows significance between the SstI and XmnI sites only. Levels of fasting triglyceride and fasting total cholesterol were determined for each individual. Analysis of covariance shows that fasting triglyceride levels in women vary significantly with the XmnI genotype. These results suggest that genetic variation at the APO A-I/C-III/A-IV gene cluster may be a useful tool for the study of quantitative lipoprotein variation in the Dogrib.

Two DNA markers, the met oncogene and the anonymous probe, pJ3.11, previously reported to be tightly linked to cystic fibrosis (CF), were used for linkage analysis in 26 families with two or more individuals affected with CF. A new high frequency polymorphism was identified using BanI and the pmetD probe. The results of linkage analysis were as follows: between met and CF, lod score of 18.2 at theta of .009; between pJ3.11 and CF, lod score of 12.1 at theta of 0; and between met and pJ3.11, lod score of 16.7 at theta of 0. These data indicate that most or all of CF is due to an abnormality at a single locus and that the DNA markers are useful for prenatal diagnosis and heterozygote detection within affected families.