Abstract
Protein tyrosine phosphatase 1β (PTP-1β) is involved in the regulation of several important physiological pathways. It regulates both insulin and leptin signaling, and interacts with the epidermal- and platelet-derived growth factor receptors. The gene is located on human chromosome 20q13, and several rare single nucleotide polymorphisms (SNPs) have been shown to be associated with insulin resistance and diabetes in different populations. As part of our ongoing investigations into the genetic basis of hypertension, we examined common sequence variants in the gene for association with hypertension, obesity and altered lipid profile in two populations of Japanese and Chinese descent. We re-sequenced all exons, selected intronic sequences and the promoter region in 24 individuals from our cohort. Fourteen SNPs were discovered, and six of these spanning 78 kb were genotyped in 1553 individuals from 672 families. All six SNPs were in linkage disequilibrium, and we found strong association of common risk haplotypes with hypertension in Chinese and Japanese (P < 0.0001). In addition, individual SNPs showed association to total plasma cholesterol, LDL-cholesterol and VLDL-cholesterol levels, as well as obesity measures (body mass index). This analysis supports that PTP-1β affects plasma lipid levels, and may lead to obesity and hypertension in Japanese and Chinese. Given similar associations found in other populations to insulin resistance and diabetes, this gene may play a crucial role in the development of the characteristic metabolic changes seen in patients with the metabolic syndrome.
INTRODUCTION
Hypertension is a major health problem in developed countries throughout the world, and represents a major risk factor for stroke, myocardial infarction and renal disease. In addition, hypertension often occurs in combination with other metabolic complications such as hyperlipidemia, obesity and insulin resistance. This combined disorder is often called the metabolic syndrome (1). Several studies have suggested the contribution of heritable factors to the etiology of the metabolic syndrome (2–4). However, to date no specific gene mutations have been identified that account for a significant proportion of patients affected by the syndrome.
Several hormonal and enzymatic pathways have been the primary target in the search for genetic alterations in patients with the metabolic syndrome. The metabolic pathways regulated by insulin have been of particular interest, as disturbances in the intricate regulation of these pathways may lead to the development of insulin resistance and diabetes. Insulin levels affect several major metabolic pathways in the human body. They regulate circulating levels of glucose, free fatty acids and amino acids by altering their uptake and release in liver, muscle and adipose tissue, to maintain near constant levels of these metabolites in the blood. The effect of insulin is initially mediated by binding of the peptide hormone to the insulin receptor. The tyrosine kinase activity of the receptor then initiates numerous signaling cascades to assert the hormone’s metabolic effect.
Several kinases and phosphatases have been shown to influence the activity of the insulin receptor. In particular, the interaction with protein tyrosine phosphatase 1β (PTP-1β) is highly specific and suggests that PTP-1β may be important in down-regulating the activity of the insulin receptor (5).
Apart from its role in the regulation of insulin signaling, PTP-1β has been shown to dephosphorylate JAK2, a kinase essential in leptin signaling (6,7). Its involvement in this additional metabolic pathway affecting lipid metabolism and obesity illustrates the essential role of PTP-1β in glucose and lipid metabolism. Furthermore, PTP-1β inactivates the epidermal- and platelet-derived growth factor receptors after endocytosis of the active receptors from the cell membrane (8).
The importance of PTP-1β for several crucial metabolic pathways has been illustrated in mice deficient for this specific phosphatase (9). These animals had lower blood glucose concentrations and reduced circulating insulin levels when compared with their normal littermates. On a high-fat diet, PTP-1β-deficient mice were resistant to weight gain, and remained insulin-sensitive, whereas their wild-type littermates became insulin resistant and obese.
In humans, PTPN1, the gene coding for PTP-1β, is located on human chromosome 20q13. The 10 exons of the gene span more than 74 kb of sequence. The open reading frame includes 1305 bp and codes for a protein of 435 amino acids (10). The gene is located in a genomic region that has been identified in multiple linkage studies as a QTL for obesity and diabetes (11–13).
Sequence variants in the human gene have been implicated in both insulin resistance and diabetes. Mok et al. (14) showed association of a single nucleotide polymorphism (SNP) in exon 8 with impaired glucose tolerance and type 2 diabetes in a Canadian Indian population. Similarly, Echwald et al. (15) showed association of a non-synonymous SNP (P387L) in PTPN1 to type 2 diabetes in a Danish population. Recently, Di Paola et al. (16) identified a nucleotide insertion in the 3′-UTR associated with insulin resistance in obese individuals. Although these studies suggest a role of PTP-1β in diabetes, insulin resistance and possibly obesity, the effect of common sequence variants and haplotypes on other features of the metabolic syndrome (such as plasma lipid levels and blood pressure) has not been investigated so far.
The Stanford Asia-Pacific Program for Hypertension and Insulin Resistance (SAPPHIRe) is a collaborative study as part of the Family Blood Pressure Program of the National Heart, Lung and Blood Institute of the National Institutes of Health to investigate the genetic determinants of hypertension and insulin resistance in Chinese and Japanese. The study has collected over 1300 sib pairs either concordant for high blood pressure or discordant. Detailed descriptions of the study cohort can be found elsewhere (17–19). In short, subjects were between 35 and 60 years old, and of Chinese or Japanese ancestry (all four grandparents Japanese or Chinese). Hypertension was defined as systolic blood pressure >160 mmHg, or diastolic blood pressure >95 mmHg, or taking two medications for high blood pressure (stage II hypertension). Alternatively, the subject could be taking one medication for high blood pressure with a systolic blood pressure >140 mmHg or a diastolic blood pressure >90 mmHg. Low-normal blood pressure was defined as blood pressure in the bottom 30% of the age- and sex-adjusted blood pressure distribution. The average number of affected (hypertensive) sibs per family was 1.72, with an average of 2.31 total sibs per family.
It was the goal of the present study to investigate the possible role of common genetic variation in PTPN1 on the development of hypertension, hyperlipidemia and obesity in the SAPPHIRe study cohort. Given the alteration in the lipid profile and feeding behavior of mice lacking a functional copy of the gene, we were interested to see whether common genetic variation in the human PTPN1 gene was associated with altered blood pressure, plasma lipid levels or body mass index (BMI) as a measure of obesity. We re-sequenced all exons, the promoter region and several intronic regions in multiple individuals from our Asian study cohort, and identified new SNPs throughout the genomic region of PTPN1. Here, we present our results on the association analysis of these SNPs and the resulting haplotypes with characteristic features of the metabolic syndrome in the SAPPHIRe cohort.
RESULTS
SNP discovery
In order to identify common sequence variants (SNPs) in PTPN1, we re-sequenced all exons, the promoter region (~10 kb upstream of exon 1) and about 7 kb of intronic sequence (500 bp segments at 5 kb intervals across intron 1, and ~2 kb of intronic sequence flanking the remaining exons) in 24 individuals from our study cohort and eight DNA samples from the Coriell Polymorphism Discovery Resource (PDR8, Coriell Cell Repository, Camden, NJ, USA). Only variants that were identified in at least two successfully sequenced samples were considered for further analysis. In total, we identified 14 SNPs spanning 78 kb of genomic sequence. The average distance between SNPs was 5939 bp (24–43 296 bp). The position of all SNPs relative to the exons of PTPN1 is illustrated in Figure 1. Two of the newly discovered SNPs had an estimated minor allele frequency of <10% (on the basis of the sequencing assessment of 32 DNA samples). Three of the SNPs had been submitted to dbSNP, the NCBI database for SNP discovery data, by other groups (Table 1). One of the SNPs (g.69220T>C) is a synonymous change located in exon 8 at codon 303. All other SNPs are located in non-exonic regions of the genomic interval.
Figure 1.
Genomic organization of the PTPN1 gene region on human chromosome 20. Exons are indicated by boxes, white boxes indicate translated regions. The size of the interspersed introns is given in base pairs above the line. Block arrows indicate the approximate location of all SNPs used in our analysis, small arrows indicate additional SNPs discovered by our re-sequencing efforts. Two SNPs with minor allele frequencies of <10% (Table 1) are not shown, they are located close to SNPs g.8235G>T and g.67543A>G, respectively.
Table 1.
SNP allele frequencies
| SNP name | dbSNP | Minor allele frequency (%) | |||
|---|---|---|---|---|---|
| Sequencing | Chinese | Japanese | |||
| g.–7077 G>C | 40.9 | 30.1 | 24.5 | ||
| g.–4022 G>A | rs6012953 | 33.3 | 33.6 | 28.3 | |
| g.6930 C>T | 40.0 | ND | ND | Not tested | |
| g.8211 C>T | 7.1 | ND | ND | Low MAF | |
| g.8235 G>T | 39.1 | 38.3 | 28.5 | ||
| g.10985 A>G | 28.0 | ND | ND | Not tested | |
| g.54281 T>A | 45.0 | 37.6 | 36.7 | ||
| g.56860 C>A | 28.0 | ND | ND | Not tested | |
| g.58585 T>C | rs2038526 | 39.1 | 33.6 | 35.3 | |
| g.64840 T>C | 37.5 | ND | ND | Assay failed | |
| g.67543 A>G | 37.5 | ND | ND | Not tested | |
| g.67694 G>A | 6.3 | ND | ND | Low MAF | |
| g.69220 T>C | rs2282146 | 12.6 | 20.6 | 20.2 | |
| g.70139 G>T | 36.0 | ND | ND | Not tested | |
For all 14 SNPs discovered by re-sequencing, the minor allele frequency is listed on the basis of re-sequencing information, and of genotyping data for both ethnic groups. dbSNP reference numbers are listed for three SNPs which match records in this NCBI database. SNPs that have not been tested in genotyping our cohort, that were excluded owing to low minor allele frequency, or that could not be genotyped successfully owing to assay failure are annotated accordingly. ND, not determined.
Analysis of SAPPHIRe cohort
We selected six SNPs for genotyping in our study cohort, and obtained complete genotyping data for these SNPs from 1553 subjects from 672 families. Priority in the selection of SNPs for genotyping was given to SNPs previously reported in dbSNP and SNPs in exons of PTPN1. Three SNPs had been submitted to dbSNP by other investigators (g.–4022G>A, g.58585T>C and g.69220T>C), and one SNP (g.69220T>C) was located in exon 8. However, this SNP did not alter the amino acid sequence of PTP-1β. Two SNPs with a minor allele frequency of <10% were not considered for genotyping, and five other SNPs were not selected as they were in close proximity (within 2 kb) to other genotyped SNPs. In addition, the design for one assay failed, resulting in six successfully genotyped SNPs for the PTPN1 gene region. Information about the 14 SNPs is summarized in Table 1.
From our study cohort, we obtained complete genotyping data for all six SNPs from 1130 hypertensive and 423 low-normotensive individuals. The mean age of these subjects at examination was 51 years. Of the individuals 54% were female. Of the subjects 75% were of Chinese descent, 25% of Japanese descent. Mean values for BMI, and plasma levels of triglycerides (TG), total cholesterol (CHOL), HDL-cholesterol (HDL), LDL-cholesterol (LDL) and VLDL-cholesterol (VLDL) are listed in Table 2 for both ethnic groups. For comparison, the mean values for hypertensives and low-normotensive individuals in our cohort are also presented.
Table 2.
Population data of the SAPPHIRe cohort
| Chinese | Japanese | |
|---|---|---|
| Number of sibs | 1171 | 382 |
| Hypertensive | 806 | 324 |
| Low-normotensive | 365 | 58 |
| Age (mean ± STD) | 49.6 ± 8.2 (35.0; 77.0) | 55.0 ± 7.9 (35.0; 81.0) |
| BMI (kg/m2) (mean ± STD) | 25.3 ± 3.4 (15.4; 35.0) | 26.4 ± 3.7 (17.4; 35.0) |
| Triglycerides (mg/dl) (mean ± STD) | 130.8 ± 83.6 (16.3; 840.3) | 171.4 ± 101.5 (45.0; 726.0) |
| Total cholesterol (mg/dl) (mean ± STD) | 188.4 ± 37.2 (85.7; 348.7) | 202.2 ± 36.6 (105.3; 369.0) |
| HDL-cholesterol (mg/dl) (mean ± STD) | 43.7 ± 11.8 (15.7; 98.7) | 49.2 ± 14.4 (23.3; 107.0) |
| LDL-cholesterol (mg/dl) (mean ± STD) | 119.4 ± 35.3 (0.0; 265.6) | 120.0 ± 33.3 (29.4; 275.6) |
| VLDL-cholesterol (mg/dl) (mean ± STD) | 25.1 ± 13.6 (3.3; 79.9) | 31.6 ± 15.5 (9.0; 79.7) |
| Hypertensive | Low-normotensive | |
| Number of sibs | 1130 | 423 |
| Age (mean ± STD) | 52.2 ± 8.3 (35.0; 81.0) | 47.5 ± 7.9 (35.0; 76.0) |
| BMI (kg/m2) (mean ± STD) | 26.3 ± 3.4 (15.4; 35.0) | 23.7 ± 3.1 (16.7; 34.0) |
| Triglycerides (mg/dl) (mean ± STD) | 153.1 ± 95.0 (19.7; 840.3) | 107.2 ± 64.0 (16.3; 447.0) |
| Total cholesterol (mg/dl) (mean ± STD) | 194.4 ± 37.2 (89.0; 369.0) | 184.7 ± 37.6 (85.7; 333.0) |
| HDL-cholesterol (mg/dl) (mean ± STD) | 44.1 ± 12.3 (15.7; 107.0) | 47.5 ± 13.5 (19.7; 98.7) |
| LDL-cholesterol (mg/dl) (mean ± STD) | 120.8 ± 34.2 (0.0; 275.6) | 116.1 ± 36.2 (18.6; 257.8) |
| VLDL-cholesterol (mg/dl) (mean ± STD) | 28.8 ± 14.6 (3.9; 79.9) | 21.1 ± 12.0 (3.3; 76.3) |
Data are listed separately for both ethnic groups, and for hypertensive and low-normotensive individuals. Numbers in parentheses indicate the minimum and maximum observed values for the measurement.
All SNPs had minor allele frequencies between 20.2 and 38.3% in both ethnic groups, as estimated by TRANSMIT for affected and unaffected individuals, considering family relationship in the estimate. These values were similar to the allele frequency estimates on the basis of our initial sequencing data. Only two SNPs (g.–4022G>A, g.8235G>T) showed significant differences in allele frequency between Chinese and Japanese (P<0.05). The frequency data are summarized in Table 1.
To determine the extent of linkage disequilibrium (LD) in our sample set, |D′| was calculated for all pairs of SNPs according to Lewontin (20). Only data from 587 unrelated individuals were included in the calculations, and |D′| was calculated separately for the two ethnic groups. A schematic diagram of all pairwise comparisons between the SNPs in the region is shown in Figure 2A. As is evident from the graphic, there is significant LD between the six SNPs. The average |D′| for our Japanese cohort is 0.60, slightly higher than the average LD in the Chinese samples (|D′| = 0.55). In Japanese, SNPs g.54281T>A and g.58585T>C are in complete LD (|D′| = 1). Likewise, SNPs g.–7077G>C and g.8235G>T are in complete LD in Chinese.
Figure 2.
Linkage disequilibrium and haplotype structure of PTPN1 in Japanese and Chinese. (A) Pairwise linkage disequilibrium (|D′|) in the Japanese and the Chinese cohort. Values are calculated for unrelated individuals only. (B) Estimated common haplotypes in the SAPPHIRe cohort. All estimated haplotypes with a frequency of >5% in at least one of the two populations are listed. Black squares represent the rarer allele, white squares the common allele for each SNP.
Given the amount of LD found between the six SNPs, we compared the distribution of the estimated haplotypes across the entire gene region including all six SNPs. Overall, 33 haplotypes were predicted in our Chinese cohort, and 21 haplotypes in the Japanese cohort. Only eight haplotypes with a frequency of at least 5% in one or both of the ethnic groups were identified. Of these, only one haplotype (consisting of the major allele for all six SNPs) was found at a frequency of >10% in both populations. Overall, the eight haplotypes illustrated in Figure 2B account for 83% of all independent chromosomes in our Chinese cohort, and for over 88% in our Japanese cohort.
SNP association analysis
In order to determine the association of individual SNPs with quantitative phenotypes characteristic of the metabolic syndrome, we used variance-component analysis as implemented in the program SOLAR (21,22) to assess the relationship of our six SNPs with BMI, and plasma levels of triglycerides, total cholesterol, HDL-cholesterol, VLDL-cholesterol and LDL-cholesterol. Additional covariates included gender, ethnicity and age. Results of these analyses are summarized in Table 3. No association was found with plasma levels of triglycerides and HDL-cholesterol and any of the six SNPs. Furthermore, two SNPs (g.–4022G>A and g.69220T>C) did not show association with any of the phenotypes analyzed. Two SNPs were marginally associated with BMI, a measure of obesity (g.54281T>A and g.58585T>C) using a dominant model (P<0.05). In addition, SNP g.–7077G>C was associated with plasma total cholesterol (P = 0.0124) and LDL-cholesterol levels (P = 0.0084) in a recessive model. Finally, SNPs g.–7077G>C and g.8235G>T were marginally associated with plasma VLDL levels (P<0.05). As shown in Table 3, there is no evidence for an additive effect of alleles of any of the SNPs. Though individual comparisons reach nominal significance, there is no evidence for a significant difference of homozygotes to both heterozygotes and the opposite homozygotes.
Table 3.
Variance-component analysis
| g.–7077 G>C | g.–4022 G>A | g.8235 G>T | g.54281 T>A | g.58585 G>C | g.69220 T>C | |||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| 1/1 + 1/2 vs 2/2 | 1/1 vs 1/2 + 2/2 | 1/1 + 1/2 vs 2/2 | 1/1 vs 1/2 + 2/2 | 1/1 + 1/2 vs 2/2 | 1/1 vs 1/2 + 2/2 | 1/1 + 1/2 vs 2/2 | 1/1 vs 1/2 + 2/2 | 1/1 + 1/2 vs 2/2 | 1/1 vs 1/2 + 2/2 | 1/1 + 1/2 vs 2/2 | 1/1 vs 1/2 + 2/2 | |
| Dominant models | ||||||||||||
| BMI | 0.6075 | 0.8899 | 0.8193 | 0.2345 | 0.5472 | 0.2542 | 0.7225 | 0.0146 | 0.5096 | 0.0180 | 0.8115 | 0.1508 |
| TG | 0.4139 | 0.0711 | 0.7132 | 0.7314 | 0.2915 | 0.0736 | 0.6608 | 0.3234 | 0.1086 | 0.8479 | 0.3731 | 0.2751 |
| CHOL | 0.0124 | 0.9535 | 0.1421 | 0.9031 | 0.3998 | 0.1505 | 0.5310 | 0.6411 | 0.5053 | 0.2887 | 0.7195 | 0.8880 |
| HDL | 0.6846 | 0.6807 | 0.8945 | 0.9093 | 0.1211 | 0.9010 | 0.9580 | 0.6445 | 0.8998 | 0.1503 | 0.7541 | 0.8408 |
| VLDL | 0.7628 | 0.0240 | 0.8931 | 0.2374 | 0.8845 | 0.0396 | 0.5880 | 0.6530 | 0.1763 | 0.5873 | 0.3157 | 0.3300 |
| LDL | 0.0084 | 0.5887 | 0.1292 | 0.7496 | 0.9755 | 0.5403 | 0.4685 | 0.7521 | 0.2825 | 0.5744 | 0.9518 | 0.6746 |
| 1/1 vs 1/2 | 1/1 vs 2/2 | 1/1 vs 1/2 | 1/1 vs 2/2 | 1/1 vs 1/2 | 1/1 vs 2/2 | 1/1 vs 1/2 | 1/1 vs 2/2 | 1/1 vs 1/2 | 1/1 vs 2/2 | 1/1 vs 1/2 | 1/1 vs 2/2 | |
| Additive models | ||||||||||||
| BMI | 0.8026 | 0.8586 | 0.2076 | 0.7223 | 0.4056 | 0.2996 | 0.0141 | 0.4914 | 0.0360 | 0.1254 | 0.2240 | 0.5398 |
| TG | 0.1338 | 0.2544 | 0.5704 | 0.8733 | 0.1340 | 0.0899 | 0.5045 | 0.9199 | 0.9338 | 0.0900 | 0.2920 | 0.4437 |
| CHOL | 0.5579 | 0.0271 | 0.7696 | 0.1968 | 0.1947 | 0.2385 | 0.8885 | 0.5096 | 0.9818 | 0.1466 | 0.8975 | 0.7257 |
| HDL | 0.6893 | 0.7735 | 0.7001 | 1.0000 | 0.5660 | 0.3102 | 0.9017 | 0.9588 | 0.3361 | 0.9274 | 0.8305 | 0.8719 |
| VLDL | 0.0418 | 0.4075 | 0.1954 | 0.8534 | 0.0393 | 0.2422 | 0.7888 | 0.8065 | 0.7357 | 0.1142 | 0.3168 | 0.3878 |
| LDL | 0.9131 | 0.0119 | 0.8719 | 0.1648 | 0.4939 | 0.8386 | 0.9040 | 0.4846 | 0.2557 | 0.5515 | 0.5483 | 0.9169 |
All P-values listed here are not corrected for multiple testing. The analysis was performed separately for each SNP and the different lipid parameters. However, because neither the SNPs (due to LD) nor the lipid phenotypes are completely independent, an accurate correction of the individual P-values is difficult. Nonetheless, the individual associations are at best only marginally significant given the number of analyses performed. Despite the lack of strong association, the repeated association of SNPs in this gene region with measures of plasma lipids and measures of obesity suggest that genetic variation in PTPN1 modulates lipid metabolism and predisposes to obesity in humans.
Transmission disequilibrium test using hypertension status
As our cohort had initially been ascertained using defined threshold criteria for hypertension [systolic BP (SBP) ≥160 mmHg, diastolic BP (DBP)≥95 mmHg, taking two medications for high blood pressure, or taking one medication with either SBP≥140 mmHg or DBP≥90 mmHg] and low-normal blood pressure (bottom 30% of the age- and sex-adjusted blood pressure distribution), we analyzed any potential association of any of the six SNPs and the resulting haplotypes using hypertension status as a binary variable. The analysis was performed using the generalized transmission disequilibrium test as implemented in the program TRANSMIT (23,24) using both individual SNPs and haplotypes. The results are summarized in Table 4. Only one SNP (g.54281T>A) is significantly associated with hypertension in our cohort (P = 0.0206). However, using haplotypes constructed for adjacent SNPs, the association is strongest for the haplotype including all six SNPs across the gene (P = 0.0002). The results look similar when Chinese and Japanese are analyzed separately (Table 4).
Table 4.
Transmission disequilibrium test
| SNP name | P-Values for order of association | |||||
|---|---|---|---|---|---|---|
| 1 | 2 | 3 | 4 | 5 | 6 | |
| All individuals | ||||||
| g.–7077 G>C | 0.5872 | |||||
| 0.7752 | ||||||
| g.–4022 G>A | 0.5104 | 0.698 | ||||
| 0.6324 | 0.1198 | |||||
| g.8235 G>T | 0.3156 | 0.0926 | 0.044 | |||
| 0.0466 | 0.0478 | 0.0002 | ||||
| g.54281 T>A | 0.0206 | 0.0008 | 0.321 | |||
| 0.0004 | 0.0342 | |||||
| g.58585 T>C | 0.9546 | 0.0068 | ||||
| 0.2378 | ||||||
| g.69220 T>C | 0.1642 | |||||
| Chinese cohort only | ||||||
| g.–7077 G>C | 0.6684 | |||||
| 0.8238 | ||||||
| g.–4022 G>A | 0.9042 | 0.141 | ||||
| 0.5972 | 0.175 | |||||
| g.8235 G>T | 0.7258 | 0.216 | 0.0282 | |||
| 0.173 | 0.0804 | 0 | ||||
| g.54281 T>A | 0.0534 | 0.0084 | 0.3164 | |||
| 0.0092 | 0.2722 | |||||
| g.58585 T>C | 0.7656 | 0.0036 | ||||
| 0.0826 | ||||||
| g.39220 T>C | 0.0676 | |||||
| Japanese cohort only | ||||||
| g.–7077 G>C | 0.0162 | |||||
| 0.1348 | ||||||
| g.–4022 G>A | 0.132 | 0.2308 | ||||
| 0.081 | 0.212 | |||||
| g.8235 G>T | 0.0126 | 0.0248 | 0 | |||
| 0.0064 | 0.2022 | 0.3106 | ||||
| g.54281 T>A | 0.1214 | 0.0048 | 0.0128 | |||
| 0.2448 | 0.0296 | |||||
| g.58585 T>C | 0.2516 | 0.092 | ||||
| 0.0494 | ||||||
| g.69220 T>C | 0.5112 | |||||
Using TRANSMIT (24,25), association was tested between SNPs and hypertension status in the SAPPHIRe cohort. SNPs were tested either individually (column 1), or as haplotypes of two, three, four, five or six neighboring SNPs. Significance levels were calculated using bootstrap methods. The P-values listed are for global tests of association, equivalent to multiple-degrees of freedom Chi-square tests. The results are, at least informally, suggestive of association in the region: of 21 tests shown, nine have a P-value of <0.05, and four have a P-value of <0.01. Italic indicates P < 0.05, and bold highlights P < 0.01.
On the basis of the output of the program TRANSMIT, of all 33 predicted haplotypes, two haplotypes (haplotypes 1 and 7; Fig. 2B) are most significantly over-transmitted in hypertensives [Chi-square (1df) of 16.34 and 15.92, respectively]. In contrast, haplotype 4 is under-transmitted (Chi-square 16.13). Overall, nine association tests of SNPs or haplotypes reached statistical significance (P < 0.05) as assessed by bootstrapping (Table 4). The two risk haplotypes (haplotypes 1 and 7) combined have a frequency of 42.89% in Chinese, and 54.53% in Japanese for our specific study cohort.
This analysis suggests a role of PTP-1β in essential hypertension in Asians, and identifies two risk haplotypes that may contribute to the effect in Japanese and Chinese.
DISCUSSION
SNPs in PTPN1, the gene for PTP-1β, have been reported to be associated with insulin resistance and diabetes in different human populations (14–16). However, these studies only investigated the effect of individual SNPs with low population frequencies (<2%) on phenotypes related to glucose metabolism. The phenotypes observed in mice deficient for PTP-1β, however, would suggest that the gene plays an integral role in regulating lipid metabolism and weight gain as well. Therefore, it is clear that PTP-1β plays an important role in a multitude of metabolic and physiological pathways, and it is conceivable that genetic sequence variants in the gene could contribute to a multitude of phenotypes and influence any of a number of human diseases, including dyslipidemia, obesity, diabetes and resulting cardiovascular complications.
It was the intent of our study to analyze common sequence variation in PTPN1, and to study the association of individual SNPs and resulting haplotypes with alterations in the lipid profile and with blood pressure as a measure of cardiovascular problems in patients of Japanese and Chinese descent. This cohort was recruited for an ongoing study on the genetic basis of hypertension, and individuals were ascertained for blood pressure as well as for a number of plasma lipid and obesity-related measures.
Our analysis identified two risk haplotypes that were over-transmitted in hypertensive individuals. No individual SNP reaches a similar level of significance. The two haplotypes combined have a frequency of 40–50% in our study population, supporting our hypothesis that common genetic variation in PTPN1 affects cardiovascular health, possibly as a consequence of alterations in lipid metabolism.
This is further supported by the results from our association analysis of PTPN1 SNPs with alterations in plasma lipid measures and BMI. Although the association of individual SNPs is at best marginally significant given the number of tests performed, the repeated association with a variety of related phenotypes suggests that PTP-1β plays a role in lipid regulation in Asian populations, and that common genetic variants alter the lipid profile. Interestingly, the strongest association to BMI, a common risk factor for hypertension and the metabolic syndrome with its cardiovascular complications, is seen with SNP g.54281T>A, one of the SNPs that differentiate the over- and under-transmitted haplotypes in our study cohort. In addition, even though we do not find any association with plasma triglyceride or HDL-cholesterol levels, the two alterations of the lipid profile used to define individuals with the metabolic syndrome, the suggestive associations are indicative of abnormal lipid metabolism, similar to the effects seen in knockout mice. This further supports a role for PTP-1β in the complex pathophysiology of the metabolic syndrome.
It remains to be seen whether additional studies will confirm the results of our association for other populations. Given the complex genetic contributions to phenotypes such as BMI or plasma lipid levels, the role of PTPN1 haplotypes will also have to be evaluated in conjunction with risk haplotypes in other genes. Whereas the region on human chromosome 20q13 has been repeatedly identified in linkage studies for obesity and diabetes (11–13), other genes in this region have been suggested as responsible for this linkage (25,26), and the results of both our studies and the analysis of other genes needs to be expanded to examine the relationship of these different genes and their combined role in glucose and lipid metabolism, as well as in obesity.
Questions will remain about the functional SNPs that are responsible for this association result. The association of different SNPs with lipid parameters would suggest that the SNPs examined in this study are not the causal SNPs, and our associations are the result of linkage disequilibrium between the SNPs used in our study and functional variants. However, we have extensively re-sequenced all exons, flanking intronic sequence and the promoter region of PTPN1, and did not find any additional variants. We also did not find variants previously reported in other studies (14–16), possibly owing to their low minor allele frequencies in our Asian cohort. Only one of our 14 SNPs is located in an exon, and this SNP is a synonymous change. In addition, none of the SNPs we identified are located in an evolutionarily conserved segment of the human genome on the basis of sequence alignment with mouse, rat and fugu. Though this does not exclude the possibility of a direct function of any of the intronic SNPs, it makes it less likely that they are located in a regulatory element. However, given the size of the genomic interval spanned by this gene, it is conceivable that the large intronic regions harbor additional regulatory elements that have not been uncovered and may be specific to humans.
Alternatively, the effect seen in our association may be caused directly by an individual SNP in one of the risk haplotypes identified in our analysis by an unknown mechanism. Regardless of the actual mechanism of action, the involvement of PTP-1β in the development of an altered lipid profile and in essential hypertension in Asians offers new treatment options. Over the past several years, numerous synthetic inhibitors of PTP-1β have been reported that can be used to reduce or inhibit the activity of the phosphatase (5,27,28). These inhibitors may now offer a new approach to treating hypertension or dyslipidemia in Japanese or Chinese.
MATERIALS AND METHODS
Sequencing and discovery of single nucleotide polymorphisms
On the basis of the genomic sequence available in GenBank (accession number NT_011362), we designed primers to amplify all exons and flanking intronic sequences of PTPN1. In addition, we designed primers to amplify ~10 kb upstream of the first exon, and amplified segments of ~500 bp spaced at 5 kb intervals from intronic sequence. All regions were amplified from 24 individuals of the SAPPHIRe cohort (16 hypertensives, eight low-normotensive controls) as well as from eight samples of the Coriell Polymorphism Discovery Resource Panel (PDR8, Coriell Cell Repository, Camden, NJ, USA). All PCR products were sequenced using Big Dye Terminator sequencing chemistry (Applied Biosystems, Foster City, CA, USA) in both directions according to standard protocols. Resulting sequencing traces were analyzed using phredPhrap (29) and Polyphred (30) to identify SNPs. A graphical overview of the structure of the gene and the location of the SNPs used in this study is shown in Figure 1.
Genotyping and analysis of pairwise linkage disequilibrium
Six SNPs were genotyped on the SAPPHIRe cohort using Invader technology (31,32). Genotypes were assigned automatically as described previously (32). For each pair of SNPs, the standardized pairwise linkage disequilibrium parameter |D′| (21) was calculated on the basis of the genotyping data for unrelated individuals in our sample (individuals selected from different families, 73.2% hypertensives in Chinese, 86.6% in Japanese cohort) using the computer program GDA (33). As two SNPs had significant differences in allele frequencies between the two populations, |D′| was calculated separately for Japanese and Chinese. The results of this analysis are summarized in Figure 2A.
In addition, eight haplotypes with a frequency of >5% in at least one of the two populations were identified across the genomic region using the six SNPs in our analysis. The haplotypes and their frequencies are illustrated in Figure 2B.
Association analysis
Quantitative measures of plasma lipid parameters (plasma triglyceride, total cholesterol, HDL-cholesterol, LDL-cholesterol and VLDL-cholesterol) and obesity (BMI) were regressed on age, gender, ethnicity, and, where applicable, on BMI. Residuals from the regression were used in variance-component analysis using the program SOLAR (21,22). Here, the variability among the phenotypes from individuals in families is expressed in terms of fixed effects from covariates, residual polygenic effects and residual non-genetic variances. In our analysis, we compared differences in the outcome variables between 1/1 + 1/2 and 2/2 or between 1/2 + 2/2 and 1/1 sibs. The analysis results for additive models are also provided in Table 3.
Association of individual SNPs and PTP haplotypes with hypertension was assessed using the generalized transmission disequilibrium test for haplotypes as implemented in the program TRANSMIT (23,24). This program accounts for phase uncertainty in its calculations. All tests are based on a score vector that is averaged over all possible configurations of parental haplotypes and transmissions consistent with observed genotyping data. Data from siblings may be used to narrow down the range of possible parental genotypes and haplotypes. The TRANSMIT program provides a test for excess transmission for each haplotype and global tests of association. The transmitted and untransmitted haplotypes were compared for their ‘similarity’ in this approach. Significance was determined using 5000 bootstrap samples as described in the program. The P-values listed in Table 4, from the output file of TRANSMIT, are for global tests of association, equivalent to a multiple-degree of freedom Chi-square test. The global test was restricted to the common haplotypes to ensure the validity of Chi-square testing.
Supplementary Material
Acknowledgments
We would like to thank K. Sheppard and D. Zierten for help with sequencing; J. Allen, J. Bushard and N. Tran for excellent technical assistance with SNP genotyping and X. Liu, A. Indap, N. Vo and D. Flowers for computer assistance at the Stanford Human Genome Center. We thank all participants in the SAPPHIRe study for their support. This paper is written on behalf of members of the Stanford Asia-Pacific Program for Hypertension and Insulin Resistance (SAPPHIRe). This work was funded by a grant from the Family Blood Pressure Program of the National Heart, Lung and Blood Institute, National Institutes of Health and the National Health Research Institutes, Taiwan.
References
- 1.Reaven GM. Banting lecture 1988. Role of insulin resistance in human disease. Diabetes. 1988;37:1595–1607. doi: 10.2337/diab.37.12.1595. [DOI] [PubMed] [Google Scholar]
- 2.Kissebah AH, Sonnenberg GE, Myklebust J, Goldstein M, Broman K, James RG, Marks JA, Krakower GR, Jacob HJ, Weber J, et al. Quantitative trait loci on chromosomes 3 and 17 influence phenotypes of the metabolic syndrome. Proc Natl Acad Sci USA. 2000;97:14478–14483. doi: 10.1073/pnas.97.26.14478. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Groop L. Genetics of the metabolic syndrome. Br J Nutr. 2000;83:S39–48. doi: 10.1017/s0007114500000945. [DOI] [PubMed] [Google Scholar]
- 4.Isomaa B. A major health hazard: The metabolic syndrome. Life Sci. 2003;73:2395–2411. doi: 10.1016/s0024-3205(03)00646-5. [DOI] [PubMed] [Google Scholar]
- 5.Kennedy BP, Ramachandran C. Protein tyrosine phosphatase-1B in diabetes. Biochem Pharmacol. 2000;60:877–883. doi: 10.1016/s0006-2952(00)00305-1. [DOI] [PubMed] [Google Scholar]
- 6.Zabolotny JM, Bence-Hanulec KK, Stricker-Krongrad A, Haj F, Wang Y, Minokoshi Y, Kim YB, Elmquist JK, Tartaglia LA, Kahn BB, et al. PTP1B regulates leptin signal transduction in vivo. Dev Cell. 2002;2:489–495. doi: 10.1016/s1534-5807(02)00148-x. [DOI] [PubMed] [Google Scholar]
- 7.Cheng A, Uetani N, Simoncic PD, Chaubey VP, Lee-Loy A, McGlade CJ, Kennedy BP, Tremblay ML. Attenuation of leptin action and regulation of obesity by protein tyrosine phosphatase 1B. Dev Cell. 2002;2:497–503. doi: 10.1016/s1534-5807(02)00149-1. [DOI] [PubMed] [Google Scholar]
- 8.Haj FG, Markova B, Klaman LD, Bohmer FD, Neel BG. Regulation of receptor tyrosine kinase signaling by protein tyrosine phosphatase-1B. J Biol Chem. 2003;278:739–744. doi: 10.1074/jbc.M210194200. [DOI] [PubMed] [Google Scholar]
- 9.Elchebly M, Payette P, Michaliszyn E, Cromlish W, Collins S, Loy AL, Normandin D, Cheng A, Himms-Hagen J, Chan CC, et al. Increased insulin sensitivity and obesity resistance in mice lacking the protein tyrosine phosphatase-1B gene. Science. 1999;283:1544–1548. doi: 10.1126/science.283.5407.1544. [DOI] [PubMed] [Google Scholar]
- 10.Forsell PA, Boie Y, Montalibet J, Collins S, Kennedy BP. Genomic characterization of the human and mouse protein tyrosine phosphatase-1B genes. Gene. 2000;260:145–153. doi: 10.1016/s0378-1119(00)00464-9. [DOI] [PubMed] [Google Scholar]
- 11.Ghosh S, Watanabe RM, Hauser ER, Valle T, Magnuson VL, Erdos MR, Langefeld CD, Balow J, Jr, Ally DS, Kohtamaki K, et al. Type 2 diabetes: evidence for linkage on chromosome 20 in 716 Finnish affected sib pairs. Proc Natl Acad Sci USA. 1999;96:2198–2203. doi: 10.1073/pnas.96.5.2198. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Lee JH, Reed DR, Li WD, Xu W, Joo EJ, Kilker RL, Nanthakumar E, North M, Sakul H, Bell C, et al. Genome scan for human obesity and linkage to markers in 20q13. Am J Hum Genet. 1999;64:196–209. doi: 10.1086/302195. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Soro A, Pajukanta P, Lilja HE, Ylitalo K, Hiekkalinna T, Perola M, Cantor RM, Viikari JS, Taskinen MR, Peltonen L, et al. Genome scans provide evidence for low-HDL-C loci on chromosomes 8q23, 16q24.1–24.2, and 20q13.11 in Finnish families. Am J Hum Genet. 2002;70:1333–1340. doi: 10.1086/339988. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Mok A, Cao H, Zinman B, Hanley AJ, Harris SB, Kennedy BP, Hegele RA. A single nucleotide polymorphism in protein tyrosine phosphatase PTP-1B is associated with protection from diabetes or impaired glucose tolerance in Oji-Cree. J Clin Endocrinol Metab. 2002;87:724–727. doi: 10.1210/jcem.87.2.8253. [DOI] [PubMed] [Google Scholar]
- 15.Echwald SM, Bach H, Vestergaard H, Richelsen B, Kristensen K, Drivsholm T, Borch-Johnsen K, Hansen T, Pedersen O. A P387L variant in protein tyrosine phosphatase-1B (PTP-1B) is associated with type 2 diabetes and impaired serine phosphorylation of PTP-1B in vitro. Diabetes. 2002;51:1–6. doi: 10.2337/diabetes.51.1.1. [DOI] [PubMed] [Google Scholar]
- 16.Di Paola R, Frittitta L, Miscio G, Bozzali M, Baratta R, Centra M, Spampinato D, Santagati MG, Ercolino T, Cisternino C, et al. A variation in 3′ UTR of hPTP1B increases specific gene expression and associates with insulin resistance. Am J Hum Genet. 2002;70:806–812. doi: 10.1086/339270. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Ranade K, Wu KD, Risch N, Olivier M, Pei D, Hsiao CF, Chuang LM, Ho LT, Jorgenson E, Pesich R, et al. Genetic variation in aldosterone synthase predicts plasma glucose levels. Proc Natl Acad Sci USA. 2001;98:13219–13224. doi: 10.1073/pnas.221467098. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Ranade K, Shue WH, Hung YJ, Hsuing CA, Chiang FT, Pesich R, Hebert J, Olivier M, Chen YD, Pratt R, et al. The glycine allele of a glycine/arginine polymorphism in the beta2-adrenergic receptor gene is associated with essential hypertension in a population of Chinese origin. Am J Hypertens. 2001;14:1196–1200. doi: 10.1016/s0895-7061(01)02213-0. [DOI] [PubMed] [Google Scholar]
- 19.Ranade K, Hsuing AC, Wu KD, Chang MS, Chen YT, Hebert J, Chen YI, Olshen R, Curb D, Dzau V, et al. Lack of evidence for an association between alpha-adducin and blood pressure regulation in Asian populations. Am J Hypertens. 2000;13:704–709. doi: 10.1016/s0895-7061(00)00238-7. [DOI] [PubMed] [Google Scholar]
- 20.Lewontin RC. On measures of gametic disequilibrium. Genetics. 1988;120:849–852. doi: 10.1093/genetics/120.3.849. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Almasy L, Williams JT, Dyer TD, Blangero J. Quantitative trait locus detection using combined linkage/disequilibrium analysis. Genet Epidemiol. 1999;17:S31–36. doi: 10.1002/gepi.1370170706. [DOI] [PubMed] [Google Scholar]
- 22.Fulker DW, Cherny SS, Sham PC, Hewitt JK. Combined linkage and association sib-pair analysis for quantitative traits. Am J Hum Genet. 1999;64:259–267. doi: 10.1086/302193. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Clayton D. A generalization of the transmission/disequilibrium test for uncertain-haplotype transmission. Am J Hum Genet. 1999;65:1170–1177. doi: 10.1086/302577. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Clayton D, Jones H. Transmission/disequilibrium tests for extended marker haplotypes. Am J Hum Genet. 1999;65:1161–1169. doi: 10.1086/302566. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Love-Gregory LD, Wasson J, Ma J, Jin CH, Glaser B, Suarez BK, Permutt MA. A common polymorphism in the upstream promoter region of the hepatocyte nuclear factor-4 alpha gene on chromosome 20q is associated with type 2 diabetes and appears to contribute to the evidence for linkage in an ashkenazi jewish population. Diabetes. 2004;53:1134–1140. doi: 10.2337/diabetes.53.4.1134. [DOI] [PubMed] [Google Scholar]
- 26.Silander K, Mohlke KL, Scott LJ, Peck EC, Hollstein P, Skol AD, Jackson AU, Deloukas P, Hunt S, Stavrides G, et al. Genetic variation near the hepatocyte nuclear factor-4 alpha gene predicts susceptibility to type 2 diabetes. Diabetes. 2004;53:1141–1149. doi: 10.2337/diabetes.53.4.1141. [DOI] [PubMed] [Google Scholar]
- 27.Asante-Appiah E, Kennedy BP. Protein tyrosine phosphatases: the quest for negative regulators of insulin action. Am J Physiol Endocrinol Metab. 2003;284:E663–670. doi: 10.1152/ajpendo.00462.2002. [DOI] [PubMed] [Google Scholar]
- 28.Johnson TO, Ermolieff J, Jirousek MR. Protein tyrosine phosphatase 1B inhibitors for diabetes. Nat Rev Drug Discov. 2002;1:696–709. doi: 10.1038/nrd895. [DOI] [PubMed] [Google Scholar]
- 29.Ewing B, Green P. Base-calling of automated sequencer traces using phred. II Error probabilities. Genome Res. 1998;8:186–194. [PubMed] [Google Scholar]
- 30.Nickerson DA, Tobe VO, Taylor SL. PolyPhred: automating the detection and genotyping of single nucleotide substitutions using fluorescence-based resequencing. Nucl Acids Res. 1997;25:2745–2751. doi: 10.1093/nar/25.14.2745. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Mein CA, Barratt BJ, Dunn MG, Siegmund T, Smith AN, Esposito L, Nutland S, Stevens HE, Wilson AJ, Phillips MS, et al. Evaluation of single nucleotide polymorphism typing with invader on PCR amplicons and its automation. Genome Res. 2000;10:330–343. doi: 10.1101/gr.10.3.330. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32.Olivier M, Chuang LM, Chang MS, Chen YT, Pei D, Ranade K, de Witte A, Allen J, Tran N, Curb D, et al. High-throughput genotyping of single nucleotide polymorphisms using new biplex invader technology. Nucl Acids Res. 2002;30:e53. doi: 10.1093/nar/gnf052. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33.Weir BS. Genetic Data Analysis II. Sinauer Associates Inc; Sunderland, MA, USA: 1996. [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.