| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Original Article |
Minneapolis Veterans Affairs Medical Center (A.G.), Departments of Medicine (A.G.) and Laboratory Medicine and Pathology (O.A., M.Y.T.), University of Minnesota, Minneapolis, Minnesota 55417; and the Second Department of Medicine (M.N.), Athens University Medical School, 115 27 Athens, Greece
Address all correspondence and requests for reprints to: Angeliki Georgopoulos, M.D., Medicine Service 111M, VAMC, One Veterans Drive, Minneapolis, Minnesota 55417. E-mail: . georg003{at}tc.umn.edu
Abstract
In type 2 diabetes, the threonine (Thr) for alanine (Ala) codon 54 polymorphism of the fatty acid binding protein 2 gene is associated with elevated fasting and postprandial triglycerides and dyslipidemia when compared with the wild type (Ala-54/Ala-54). To assess whether this is the case in patients with type 1 diabetes, who usually do not manifest the metabolic syndrome, we screened 181 patients with similar glycemic control as the type 2 patients. Thirty percent were heterozygous, and 9% were homozygous for the polymorphism. Mean (±SEM) fasting plasma triglyceride levels in patients with the wild type (n = 84), those heterozygous for Ala-54/Thr-54 (n = 44), and those homozygous for the Thr-54 (n = 13) were 1.0 ± 0.07, 1.1 ± 0.17, and 1.2 ± 0.23 mmol/liter, respectively. In addition, there were no differences in total, low-density lipoprotein, high-density lipoprotein, and non-high density lipoprotein cholesterol among the three groups. After a fat load, the postprandial area under the curve of triglyceride in plasma, chylomicrons, and very low-density lipoprotein were similar between the wild type (n = 18) and the Thr-54 homozygotes (n = 12). In conclusion, in contrast to type 2, type 1 diabetes does not interact with the codon 54 polymorphism of the fatty acid binding protein 2 gene to cause hypertriglyceridemia/dyslipidemia. Insulin resistance could account possibly for this difference.
IN CONTRAST TO PATIENTS with type 2 diabetes, those with type 1 diabetes are usually normolipidemic as judged by a fasting lipid profile (1). We have previously reported that the G to A polymorphism of codon 54 of the fatty acid binding protein 2 (FABP2) gene is associated with dyslipidemia and elevated fasting and postprandial triglycerides in type 2 diabetes (2). The mechanism involved is presumed to be an interaction of the polymorphism with type 2 diabetes, resulting in the observed lipid abnormalities. FABP2 is an intracellular protein expressed only in the intestine (3). The polymorphism, which consists of the substitution of Thr for Ala is functional and, as shown in in vitro experiments, it increases the affinity of the FABP2 for long chain fatty acids and is associated with increased triglyceride secretion in a human intestinal cell line (4).
The Thr-54 allele has been associated with increased fat oxidation and insulin resistance in some but not all studies (5, 6, 7, 8, 9, 10, 11, 12). It was shown to be associated with elevated postprandial triglyceride-rich lipoprotein levels in nondiabetic, obese middle-aged subjects homozygous for the Thr-54 allele compared with the wild-type (Ala-54/Ala-54; Ref. 13). However, in a study of young, nonobese, nondiabetic men, there were no differences in fasting or postprandial triglyceride levels between those carrying and those lacking the polymorphism (14). Because type 1 diabetes is usually seen in nonobese, younger, and less insulin-resistant individuals, compared with type 2 diabetes, we decided to investigate whether the codon 54 polymorphism of the FABP2 gene is associated with fasting and postprandial triglyceride elevation or dyslipidemia in type 1 diabetes.
Experimental subjects and study protocol
All study participants were patients with type 1 diabetes as defined using criteria of the American Diabetes Association. To assess the prevalence of the polymorphism, 181 diabetic patients who responded to an announcement placed at the blood drawing room of the Minneapolis Veterans Affairs (VA) Medical Center or who were patients of the diabetic or medical clinics at the VA and at the Second Department of Medicine, Athens University Medical School, were screened. The announcement solicited volunteer patients with diabetes to be screened for our study. All patients signed consent forms approved by the institutional review board. Study participants were 98% Caucasians. Patients on lipid-lowering drugs or steroids and patients with secondary causes of hyperlipidemia except diabetes were excluded from the study. Subjects included in the study (n = 141) were in good health as determined by review of their history and physical examination and routine laboratory studies (normal hematologic, liver, renal, and thyroid studies). Given that the population of the VA is mostly male, to carry out postprandial studies in equal numbers of men and women homozygous for the polymorphism, 42 female subjects from Athens were screened for the polymorphism from blood samples shipped to Minneapolis. After a 12-h fast, blood was drawn for a fasting lipid profile in all 141 study participants. In a smaller number of study participants (n = 30) who were not homozygous for the apolipoprotein (apo) E2 or apo E4 allele, postprandial studies were carried out as described below. Except for four females homozygous for the polymorphism, who underwent the postprandial studies in Athens, all other studies were carried out in Minneapolis. The same protocol was used at both sites. Dr. M. Noutsou, who has done similar studies in Minneapolis during her research fellowship, carried out the studies in Athens. The postprandial lipid protocol is described below. Subjects ingested a shake containing egg white, Sweet and Low, fruit flavor, 55 g of corn oil, and 60,000 IU of retinol per square meter of body surface. We decided not to include sucrose or other carbohydrates in the meal, because they can increase the production of hepatic triglycerides. Blood was drawn through an angiocatheter before and every 2 h for a period of 8 h for measurements of plasma triglyceride and in two triglyceride-rich lipoprotein subfractions: Sf > 400 (chylomicrons), and Sf 20–400 [very low-density lipoprotein (VLDL) and chylomicron remnants]. Subjects were sitting or lying during the duration of the study. No other meal was consumed until the study was completed. Subjects were encouraged to drink water after each blood drawing. To avoid hypoglycemia, 40% of the morning long-acting insulin dose was administered, and blood glucose was monitored at each blood sampling time. If an occasional blood glucose drop of less than 3.3 mmol/liter occurred, it was treated with a glass of fruit juice.
Materials and Methods
Detection of the polymorphism
DNA isolation was carried out from 10 ml of blood drawn in EDTA tubes, using a commercially available kit from Puregene (Gentra Systems, Minneapolis, MN).
DNA amplification and digestion by restriction enzyme were done according to the method described by Baier et al. (5) using digestion by HhaI to detect the Ala 54 Thr polymorphism, which disrupts a HhaI site. DNA was amplified by PCR, using primers 5'-ACAGGTGTTAATATAGTGAAAAG-3'and 5'-TACCCTGAGTTCAGTTCCGTC-3' from exon 2. The PCR buffer contained 2.5 mmol/liter MgCl2, 0.2 mmol/liter of each deoxynucleotide triphosphate (dNTP), and 1.25 U of Amplitaq DNA polymerase. Temperature was cycled 30 times (60 sec at 95 C, 60 sec at 55 C, and 2 min at 72 C, followed by a 3-min extension at 72 C). The PCR product was digested by HhaI at 37 C for 3.5 h or overnight, applied to a 3% agarose gel containing 1 mg/ml of ethidium bromide, electrophoresed for 40 min at 120 V, and photographed under UV light. The PCR products that lack the HhaI site migrate as one 180-bp fragment (those carrying the Thr-54), but PCR products containing the HhaI site are cleaved to two fragments (a 99 bp and an 81 bp).
Apo E genotyping
DNA samples were genotyped using PCR amplification followed by restriction enzyme digestion (15). Each amplification reaction contained PCR buffer with 15 mmol/liter MgCl2, nanogram amounts of genomic DNA, 20 pmol apo E forward (5N TAA GCT TGG CAC GGC TGT CCA AGG A 3N) and reverse (5N ATA AAT ATA AAA TAT AAA TAA CAG AAT TCG CCC CGG CCT GGT ACA C 3N) primers, 1.25 mmol/liter of each dNTP, 10% dimethyl sulfoxide, and 0.25 µl Amplitaq DNA polymerase. Reaction conditions in a thermocycler included an initial denaturing period of 3 min at 95 C, 1 min at 60 C, and 2 min at 72 C, followed by 35 cycles of 1 min at 95 C, 1 min at 60 C, 2 min at 72 C, and a final extension of 1 min at 95 C, 1 min at 60 C, and 3 min at 72 C. PCR products were digested with HhaI and separated on a 10% polyacrylamide nondenaturing gel and stained with silver nitrate. Known apo E isoform standards were included in the analysis.
Handling of blood samples
To avoid lipoprotein degradation, blood samples for lipoprotein analysis were collected in tubes containing 1 mg/ml EDTA, 0.02 mg/ml chloramphenicol, 1.3 mg/ml e-aminocaproic acid, 0.05 mg/ml gentamycin, 0.01 mg/ml benzamidine, and 10 U/ml Trasylol (16). Plasma was separated by centrifugation at 4 C; the plasma samples from Athens were shipped on ice after addition of the cocktail; triglyceride-rich lipoprotein subfractions were separated and analyzed within 5–7 d.
Determination of fasting lipoprotein parameters
Plasma triglyceride and total cholesterol were measured enzymatically by commercially available kits (Waco Chemicals, Richmond, VA, and Boerhinger Mannheim Diagnostics, Indianapolis, IN). The coefficient of variation of these assays is 3.5% for the former and 4% for the latter. High-density lipoprotein (HDL) cholesterol was measured by heparin-manganese precipitation according to the Lipid Research Clinics protocol. These measurements are currently performed in the VA laboratory and are under quality control testing with the Northwest Lipid Research Laboratory (Seattle, WA).
Isolation of lipoprotein subfractions
All isolations were performed under aseptic conditions within 48 h from harvesting of plasma. We used salt density gradient ultracentrifugation according to the method of Lindgren (17) as modified by Redgrave (18) to isolate the triglyceride-rich lipoprotein subfractions as published previously (2). Two subfractions were isolated: Sf > 400 at 4.5 x 106 g x min and Sf 20–400 at 183.2 x 106 g x min.
Determination of retinyl ester and apo B in the isolated subfractions
Retinyl esters were measured in the isolated triglyceride-rich lipoprotein subfractions. Samples were covered with foil to be protected from light and were kept at -20 C. Retinyl acetate was added to the samples as internal standard. The samples were extracted with 4 ml ethanol, 5 ml hexane, and 4 ml water. The hexane (upper) phase was collected and evaporated under nitrogen. The residue was dissolved in 20 µl of ethanol and injected to a Beckman HPLC using a Bondapak C18 column. The mobile phase was 100% methanol, and the flow rate was 2 ml/min (19). The amount of retinyl ester was calculated on the basis of retinyl palmitate standard read at 326 nm with the Gold System software. Apo B measurement in the lipoprotein subfractions was performed at the laboratory of Dr. P. Alaupovic in the Oklahoma Research Foundation (Oklahoma City, OK) by using an electroimmunoassay method (20).
Statistical analysis
Standard statistical methods (21) were used to display and analyze the data. The area under the triglyceride curve was calculated by the trapezoidal rule. The BMDP/Dynamic statistical package (BMDP Statistical Software, Inc., Los Angeles, CA) was used for the analyses. Comparisons between groups were assessed using the nonparametric Mann-Whitney rank-sum test. The effect of FABP2 Thr-54 allele was assessed using an ANOVA (program 2V of BMDP) in which this qualitative factor was entered at three levels (0, absent; 1, heterozygous; 2, homozygous).
The postprandial studies were analyzed by repeated measure ANOVA. Finally, the triglyceride/retinyl ester ratios were log-transformed to normalize their distribution and stabilize the variance (21).
Results
One hundred eighty-one patients with type 1 diabetes were screened for the Thr-54 polymorphism of the FABP2 gene. The population consisted mostly of white men and women, which is typical of the population served by the two study sites. One hundred ten individuals (61%) had the wild type (Ala-54/Ala-54), 54 (30%) were heterozygotes (Ala-54/Thr-54), and 17 (9%) were homozygous for the Thr-54 allele. As mentioned above, we excluded patients on lipid-lowering medications or steroids and those with abnormal liver, kidney, or thyroid function tests. We then assessed in 141 subjects whether dyslipidemia or elevated fasting plasma triglyceride levels were associated with the Thr-54 allele by comparing the lipid profile of the patients who lacked the Thr-54 allele (wild type) with that of the patients who were heterozygous or homozygous for it. As shown in Table 1
, fasting plasma triglyceride levels were similar in all groups. There were no significant differences in mean age, body mass index (BMI), glycohemoglobin, and total LDL, HDL, and non-HDL cholesterol levels among the three groups.
|
, these parameters were similar between the two groups of representative subjects included in the postprandial studies (data not shown). There was no effect of gender or gender by polymorphism interaction on plasma triglyceride. Therefore, the plasma triglyceride levels of both genders were pooled. In addition, the levels of postprandial triglyceride in plasma over time were similar in the Thr-54 homozygotes compared with the wild type (Fig. 1, top panel). The area under the curve (AUC) of the plasma triglyceride levels before or after subtracting the baseline was not different between the two groups. To assess whether the distribution of triglyceride in the two isolated postprandial triglyceride-rich lipoprotein subfractions differed between the two groups, we compared the AUC of the triglyceride levels in the chylomicron subfraction (Sf >400) and the VLDL/chylomicron remnant subfraction (Sf 20–400). As shown in Fig. 1, bottom panel, there were again no differences between the two groups. However, the triglyceride/retinyl ester ratio in the chylomicron subfraction (Sf >400), measured at 4, 6, and 8 h after fat ingestion, was significantly higher in those who were homozygous for the polymorphism compared with the wild type (P = 0.007; F test, ANOVA; Fig. 2). This is consistent with a higher triglyceride content of the chylomicron particles in those carrying the polymorphism. A smaller nonsignificant difference (ANOVA) in the triglyceride to retinyl ester ratio was seen in the VLDL/chylomicron remnant subfraction (Sf 20–400) between the two groups. The mean values for the wild type vs. the Thr-54 homozygotes were as follows: at 4 h, 8.7 vs. 9.5; at 6 h, 8.4 vs. 8.9; and at 8 h, 8.2 vs. 8.6. Lastly, to assess whether the number of particles in the VLDL/chylomicron remnant subfraction (Sf 20–400) differed between the two groups, we measured apo B levels in this subfraction. The AUC of apo B was not significantly different between the two groups (t test); the mean ± SEM was 121.7 ± 41.7 in the wild type vs. 105.4 ± 66.7 mg*h/liter in the Thr-54 homozygotes.
|
|
Our study shows that in patients with type 1 diabetes the frequency of the Thr-54 homozygotes for the Thr-54 polymorphism was similar to that of type 2 diabetes i.e. 9% vs. 11%, respectively (2). These values are similar to those reported by others in both control and diabetic subjects (6, 7, 8, 9, 10, 11, 12). However, our data show that in type 1 diabetes the FABP2 polymorphism is not associated with the hypertriglyceridemia/dyslipidemia phenotype, because the triglyceride, LDL, HDL, and non-HDL cholesterol levels were similar in all study groups (Table 1). This is clearly different from the interaction of the polymorphism with type 2 diabetes, which we reported previously to result in the phenotype of hypertriglyceridemia/dyslipidemia (2). Our results show that the lack of an effect of the FABP2 polymorphism in type 1 diabetes is not due to differences in any confounding factors like BMI, age, or glycemic control, because these parameters were similar among the three genotypic groups with type 1 diabetes (Table 1).
In addition, in type 1 diabetes there were no differences in postprandial plasma and chylomicron triglyceride levels between the Thr-54 homozygotes and the wild type (Fig. 1). This again is in contrast to the results observed in type 2 diabetes where both fasting and postprandial plasma triglyceride and postprandial chylomicron triglyceride were elevated in those homozygous for the polymorphism vs. the wild type (2). The only difference between the two groups in patients with type 1 diabetes was in the particle composition of the chylomicrons (Sf >400). We used the measurement of retinyl esters as a marker of the intestinal origin of the particles, because retinyl esters are formed in the intestine after ingestion of retinol and are not resecreted in the circulation as VLDL after their uptake by the liver (22). We calculated the triglyceride to retinyl ester ratio of the chylomicrons to assess intestinal particle core composition and found it to be enriched in triglyceride in the Thr-54 homozygotes vs. the wild type (Fig. 2). This goes along with the functional alteration of the FABP2, in the presence of the codon 54 polymorphism of the gene described in in vitro studies. Those studies reported that the Thr-54 FABP2 polymorphism was associated with increased affinity for long chain fatty acids, which lead to increased intestinal fatty acid transport and increased intestinal triglyceride secretion when compared with the wild type (4).
What are the differences between the study patients with type 1 and type 2 diabetes that could account for the lack of an interaction of the Thr-54 polymorphism in type 1 diabetes? Glycemic control was similar between the two types, but age and BMI were higher in the study patients with type 2 vs. those with type 1 diabetes. Specifically, the ages of type 1 vs. type 2 diabetes in the three genotypes were as follows: for the wild type, 39 vs. 57 yr; for the Thr-54 heterozygotes, 40 vs. 58 yr; and for the Thr-54 homozygotes, 44 vs. 56 yr. To assess the effects of age, we performed data analysis comparing patients with type 1 diabetes and age below 40 yr (young) vs. those above 50 yr (older). Both Thr-54 homozygotes and heterozygotes were included in this analysis. There were no differences in plasma triglycerides between those carrying the Thr-54 polymorphism and the wild type in either age group. The mean ± SEM plasma triglyceride values were as follows: young wild type, 0.898 ± 0.08, vs. young Thr-54 carriers, 0.964 ± 0.11; older wild type, 1.3 ± 0.132, vs. older Thr-54 carriers, 1.30 ± 0.154 mmol/liter. The mean ages of the young and older type 1 groups were 31 and 56 yr, respectively. It is of note, that the average age of the older type 1 group (56 yr) was very similar to the average ages of all genotype groups with type 2 diabetes (56–58 yr). Therefore, it is unlikely that differences in age can explain the lack of interaction of the Thr-54 polymorphism of the FABP2 gene with type 1 diabetes.
A similar data analysis was performed using data from both type 1 and type 2 diabetes (reported by us previously; Ref. 2) to assess the effect of BMI on the differences between the two types of diabetes in the manifestation of hypertriglyceridemia in the carriers of the Thr-54 polymorphism. Plasma triglycerides were compared in lean (BMI 
26 kg/m2) vs. obese (BMI 
28 kg/m2) study subjects. Both homozygotes and heterozygotes for the Thr-54 polymorphism were included in this analysis. As shown in Table 2, the hypertiglyceridemia phenotype was present in both lean and obese subjects with type 2 diabetes carrying the polymorphism. There was no such effect in the patients with type 1 diabetes. It seems therefore unlikely that BMI could account for the observed differences between type 1 and type 2 diabetes, regarding the manifestation of hypertriglyceridemia in the patients carrying the Thr-54 FABP2 polymorphism.
|
In conclusion, our results show that in contrast to type 2, type 1 diabetes does not interact with the codon 54 polymorphism of the FABP2 gene to cause hypertriglyceridemia/dyslipidemia. Because glycemic control between the two types of diabetes was similar, other factors, like insulin resistance or unknown genetic interactions, may be involved in the development of the diabetic dyslipidemia phenotype.
Acknowledgments
We acknowledge the expert technical help of Laura Salvati of the staff of the Special Diagnostic Treatment Unit and the staff of the blood drawing room at the Minneapolis Veterans Affairs Medical Center.
Footnotes
This research was supported by Department of Veterans Affairs funds.
Abbreviations: Ala, Alanine; apo, apolipoprotein; AUC, area under the curve; BMI, body mass index; FABP2, fatty acid binding protein 2; HDL, high-density lipoprotein; LDL, low-density lipoprotein; Thr, threonine; VA, Veterans Affairs; VLDL, very low-density lipoprotein.
Received December 5, 2001.
Accepted April 26, 2002.
References
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Endocrinology | Endocrine Reviews | J. Clin. End. & Metab. |
| Molecular Endocrinology | Recent Prog. Horm. Res. | All Endocrine Journals |
