Cell Surface Expression of LDL Receptor Is Decreased in Type 2 Diabetic Patients and Is Normalized by Insulin Therapy

The study population consisted of 21 type 2 diabetic patients who were admitted to the metabolic disease department of our hospital for the introduction of insulin treatment because of failure of oral antidiabetic therapy and 21 healthy normolipidemic subjects. Control subjects did not take any medication. At study entry, type 2 diabetic patients were treated with oral antidiabetic therapy (both sulfonylureas [15 mg/day glibenclamide] and metformin [2,550 mg/day] in all patients). They were not taking other drugs that could affect lipid metabolism. After the first measurement of LDL receptors on blood mononuclear cells, oral antidiabetic therapy was stopped and replaced with insulin treatment consisting of two daily injections of intermediate-acting human insulin at a dose of 0.30–1.5 units · kg⁻¹ · day⁻¹. The patients performed capillary glucose monitoring three times per day. They were educated in order to adapt insulin doses according to capillary blood glucose level. The goal was to obtain a fasting blood glucose level between 4.5 and 8.25 mmol/l. The patients were asked to increase their insulin dose by 2 units when fasting glycemia was >8.25 mmol/l and to reduce it by 2 units when fasting glycemia was <4.5 mmol/l.

LDL receptors were quantified at the surface of mononuclear cells in fresh fasting blood samples taken at 8:00 a.m. This analysis was performed twice in type 2 diabetic patients, first before the introduction of insulin therapy and then 3 months later. Blood samples were drawn in tubes with EDTA (Becton Dickinson, Meylan, France). At the same time, blood was taken for the measurement of lipid parameters, glycemia, insulinemia, and glycated HbA_1c.

LDL receptor expression was quantified on peripheral mononuclear cells isolated from blood samples immediately after blood drawing. Erythrocytes were lyzed by mixing 1 ml blood and 4 ml of an hypotonic solution containing 0.83% NH₄Cl, 0.004% EDTA, and 0.1% KHCO₃. After a 30-min incubation at 4°C, samples were centrifugated at 300g for 5 min and the pellet was washed twice with the lysis solution. The pellet was then suspended in ice-cold PBS with 1% BSA and 0.1% NaN₃. A volume containing 10⁶ cells was then centrifuged, supernatant was removed, and cells incubated at 4°C for 30 min with 50 μl of a pure mouse monoclonal anti-LDL receptor IgG2b (clone 15C8) (Oncogene Research Products, Boston, MA). Subsequently, the samples were washed twice with PBS/1% BSA/0.1% NaN₃ and incubated at 4°C for 30 min in the dark with 50 μl fluorescein isothiocyanate (FITC)-labeled goat anti-mouse IgG diluted 1:50 (QIFIKIT; Dako, Glostrup, Denmark). Again, cells were washed twice with PBS/1% BSA/0.1% NaN₃ and resuspended in 1 ml of the same solution. Blanks were processed in parallel with the samples and in the same manner, except that anti-LDL receptor antibody was replaced with PBS/1% BSA/0.1% NaN₃. A calibration curve linking the intensity of fluorescence and the number of antigenic sites was established using QIFIKIT. SetUp and Calibration beads (50 μl each) were washed as recommended by the manufacturer and incubated at 4°C for 30 min in the dark with 50 μl goat FITC-labeled anti-mouse IgG diluted 1:50. Then, they were washed and suspended in PBS before flow cytometer analysis.

Measurements were performed on a GALAXY flow cytometer (Dako, Trappes, France). The monocyte population was selected in a forward- versus side-scatter window. For each sample, the FL1 chanel fluorescence (FITC) was recorded. Data were analyzed with FluoMax software (Dako). The fluorescence related to specific binding to LDL receptors was obtained by subtracting the mean fluorescence of the blank from the mean fluorescence of the sample. The number of antigenic sites per cell was calculated from the calibration curve derived from QIFIKIT beads. The coefficient of variation evaluating the repeatability of the assay was 5%.

Plasma glucose concentrations were measured by an enzymatic method (glucose oxidase) on a Vitros 750 analyzer (Johnson & Johnson Clinical Diagnostics, Rochester, NY). HbA_1c was measured by ion-exchange high-performance liquid chromatography (Bio-Rad, Richmond, CA). Serum insulin concentrations were determined by radioimmunoassay (CIS Bio International, Gif sur Yvette, France). Total and HDL cholesterol and triglyceride concentrations were measured in a Cobas Integra analyzer (Roche, Basel, Switzerland) with dedicated reagents. LDL cholesterol concentrations were calculated using the Friedewald formula (12).

Results are expressed as means ± SD. Statistical calculations were performed using the SPSS software package. Means in control and type 2 diabetic groups were compared using the Student’s t test. Means in type 2 diabetic patients before and after insulin therapy were compared using the paired Student’s t test. Correlation coefficients were calculated by the Spearman test. For type 2 diabetic patient subgroup analysis, the nonparametric Mann-Whitney U test was used. P values <0.05 were considered statistically significant.

Clinical and glucose metabolism characteristics are presented in Table 1. The type 2 diabetic patients were significantly overweight compared with control subjects (BMI 28.4 ± 3.5 vs. 21.2 ± 2.3 kg/m², P < 0.001) and did not significantly change their weight on insulin therapy (BMI 29.2 ± 3.9 vs. 28.4 ± 3.5 kg/m²). Insulin therapy significantly improved glycemic control in type 2 diabetic patients, as assessed by fasting blood glucose concentration (10.1 ± 2.5 vs. 13.0 ± 3.6 mmol/l, P < 0.01) and HbA_1c (8.7 ± 1.5 vs. 10.1 ± 1.9%, P < 0.01). Fasting plasma insulin level was comparable in type 2 diabetic patients at their entry in the study and in control subjects (8.0 ± 6.0 vs. 5.3 ± 1.7 mU/l) and was significantly higher in insulin-treated patients than in control subjects (15.6 ± 10.9 vs. 5.3 ± 1.7 mU/l, P < 0.01).

Before the introduction of insulin therapy, type 2 diabetic patients presented a 2.5-fold increase in fasting plasma triglycerides compared with control subjects (2.38 ± 1.48 vs. 0.92 ± 0.51 mmol/l, P < 0.001) (Table 1). On insulin therapy, plasma triglycerides fell by 25% (1.79 ± 0.82 vs. 2.38 ± 1.48 mmol/l, P < 0.05). Nevertheless, plasma triglycerides remained 1.9-fold higher in type 2 diabetic subjects than in control subjects (P < 0.001). Total and LDL plasma cholesterol concentrations were initially comparable between patients and control subjects and were not modified by insulin treatment in type 2 diabetic patients. HDL cholesterol concentration was reduced by 26% in type 2 diabetic patients before insulin therapy (1.19 ± 0.24 vs. 1.63 ± 0.42 mmol/l, P < 0.001). It increased on insulin therapy (1.38 ± 0.37 vs. 1.19 ± 0.24 mmol/l, P < 0.05) but remained 14% lower than that in control subjects (P < 0.05).

The mean number of LDL receptors at the surface of blood mononuclear cells was reduced on average by 41% (6,439 ± 2,310 vs. 10,846 ± 2,764, P < 0.001) in type 2 diabetic patients with poor metabolic control before insulin treatment compared with control subjects (Fig. 1). After 3 months of insulin therapy, LDL receptor expression in type 2 diabetic patients increased by 57% (10,096 ± 5,657 vs. 6,439 ± 2,310, P < 0.01) and was similar to that observed in control subjects. Considering type 2 diabetic patients before insulin and control subjects together, LDL receptor expression was negatively correlated with age, BMI, glycemia, and triglycerides (data not shown) and positively correlated with HDL cholesterol. However, none of these correlations remained significant when we analyzed each group separately. LDL receptor expression was correlated with neither total or LDL cholesterol nor insulinemia when considering the two groups together or individually. Moreover, in type 2 diabetic patients, LDL receptor expression was not correlated with glycated HbA_1c. A step-by-step multiple regression analysis performed on the data of control subjects and type 2 diabetic patients before insulin treatment, including the presence of diabetes, age, sex, BMI, fasting blood glucose, triglyceridemia, and LDL and HDL cholesterol levels, demonstrated that the number of LDL receptors per monocyte was only dependent on the presence or absence of type 2 diabetes (P < 0.001). This factor explained 57% of the variations of LDL receptor expression. The difference of LDL receptor expression in type 2 diabetic patients before insulin treatment with that after was not significantly correlated with any of the differences for BMI, fasting blood glucose, HbA_1c, insulinemia, triglycerides, or total, HDL, or LDL cholesterol. Moreover, the increase in LDL receptors in insulin-treated patients was not correlated with the amount of injected insulin, as expressed in units per kilograms of body weight.

We wondered whether LDL receptor expression was parallel to glycemia in type 2 diabetic patients. To try to answer this question, we divided the type 2 diabetic group into two subgroups: patients with fasting glycemia before insulin treatment <12 mmol/l (n = 10) and patients with fasting glycemia >12 mmol/l (n = 11). Mean glycemia was 10.1 ± 1.4 and 16.1 ± 2.4 mmol/l (P < 0.001) in each subgroup, respectively. Before the introduction of insulin therapy, LDL receptor expression was comparable between these two subgroups (6,612 ± 2,932 vs. 6,283 ± 1,697). After 3 months of insulin therapy, LDL receptor expression was still comparable between these two subgroups (9,696 ± 4,079 vs. 10,459 ± 6,481). The difference of LDL receptor expression after insulin treatment with that before was similar in the two subgroups (3,084 ± 4,313 vs. 4,176 ± 7,219), whereas the difference of fasting glycemia after insulin treatment with that before was significantly lesser in patients with an initial fasting glycemia <12 mmol/l than in patients with an initial fasting glycemia >12 mmol/l (−0.83 ± 4.05 vs. −5.13 ± 3.60 mmol/l, P < 0.01).

This study was designed to test the hypothesis that changes in the expression of LDL receptors in type 2 diabetic patients before and after the introduction of insulin therapy could explain both the decrease of LDL FCR observed in poorly controlled type 2 diabetic subjects and its normalization in insulin-treated patients. By quantifying LDL receptor expression on blood mononuclear cells with a flow cytometry method, we demonstrated for the first time that LDL receptor expression is decreased in type 2 diabetic patients with a poor metabolic control on oral antidiabetic drugs compared with nondiabetic subjects and, moreover, that it is normalized after a 3-month insulin treatment period.

Liver but not mononuclear cell LDL receptors play a key role in LDL catabolism in vivo. For ethical reasons, receptor quantification on liver biopsies is not currently conceivable. Several studies have shown that LDL receptor measurement on mononuclear cells is physiologically relevant for the determination of LDL receptor status. Indeed, in human mononuclear cells, LDL receptor gene expression has been demonstrated to parallel and be coordinately regulated to gene expression in human liver (13,14). Moreover, the number of LDL receptors on mononuclear cells has been shown to be susceptible to the effect of well-recognized modulators of LDL receptor number such as saturated fatty acids and dietary cholesterol (14,15).

It is noteworthy that we performed LDL receptor measurements on freshly isolated mononuclear cells; therefore, our results are likely to reflect the in vivo situation.

Insulin upregulates both the mRNA expression of LDL receptor and its expression at the surface of cultured cells (9,11). Thus, the variations of the LDL receptor expression we observed both before and after the introduction of insulin therapy in poorly controlled type 2 diabetic patients are likely to be related to insulin and insulin action. We found that plasma insulin level was in a normal range in type 2 diabetic patients before insulin therapy. Nevertheless, type 2 diabetes is characterized by a resistance of peripheral tissues to the action of insulin, and a normal insulinemia in regard to elevated plasma glucose levels indicates a relative insulin deficiency (16). Thus, the decreased expression of LDL receptors we observed in type 2 diabetic patients before the introduction of insulin therapy is likely to be explained by a relative insulin deficiency. This hypothesis is highly strengthened by the second part of our study.

The other major finding of our work is the normalization of the expression of LDL receptors at the mononuclear cell surface in type 2 diabetic patients treated by insulin for 3 months. To our knowledge, only one previous study (17) reported a positive effect of insulin therapy on LDL receptor expression in diabetic patients, but this work differed from ours on several points: only seven patients were recruited, the type of diabetes (type 1 or 2 diabetes) was not specified, and the effect of insulin was studied only in the short term (1 week after the introduction of insulin therapy). In the present study, the normalization of LDL receptor expression is probably directly linked to the administration of insulin, which tended to correct the relative insulin deficiency. At first glance, the lack of correlation between the amount of injected insulin and the difference of the expression of LDL receptors between the two quantifications might be surprising. In fact, insulin action depends on several factors and is not correlated with plasma insulin level, which explains the lack of correlation between either the dose of administrated insulin or plasma insulin level and their effect on the expression of LDL receptors (18,19).

Based on plasma concentration or urinary excretion of mevalonate, cholesterol synthesis has been demonstrated to be both increased in type 2 diabetic patients and normalized by intensive insulin treatment (20,21). Therefore, the decreased expression of LDL receptors in these patients might also be regarded as the consequence of an increased intracellular concentration of cholesterol, related to an enhanced cholesterol synthesis (22). However, in cultured cells, insulin upregulates cholesterol synthesis (23). In light of this observation, the elevated cholesterol synthesis demonstrated in type 2 diabetes, which is characterized by a relative insulin deficiency, cannot be interpreted as a primary event likely to induce a negative feedback on LDL receptor expression. Rather, the increased cholesterol synthesis seems to be the consequence of reduced LDL receptor expression in poorly controlled type 2 diabetic patients. The decrease of cholesterol synthesis in insulin-treated type 2 diabetic patients also seems to be the consequence of increased LDL receptor expression.

The increase in LDL receptor expression we observed in type 2 diabetic patients is, in the strict sense, the result of both removal of oral antidiabetic drugs and addition of insulin. It is highly unlikely that the 41% increase in LDL receptor expression 3 months after the introduction of insulin therapy in type 2 diabetic patients was due to removal of oral antidiabetic drugs. Indeed, sulfonylureas stimulate insulin secretion and metformin stimulates insulin sensitivity. If these two drugs had an effect on LDL receptor expression, they would increase LDL receptor expression. Moreover, in the present study, we think that the influence of antidiabetic drugs on LDL receptor expression is weak. Indeed, in our patients, the glycemic control was poor, indicating that antidiabetic drugs do not effectively improve metabolic disturbances in type 2 diabetes.

A very interesting finding of this work is that the variations of LDL receptors at the cell surface in type 2 diabetic patients parallel the fluctuations of LDL FCR we and others observed in in vivo kinetic studies. We recently demonstrated that LDL catabolism was slowed down by 25% on average in type 2 diabetic patients with poor metabolic control, whereas it was normalized after 3 months of insulin treatment (5,8). In this work, LDL receptor expression was diminished by 41% in poorly controlled patients, as well as normalized on insulin treatment. We wondered which responsibility could be assigned to the decrease of LDL receptors for the decrease of LDL catabolism in patients before insulin treatment. In heterezygous familial hypercholesterolemia, characterized by only 50% functional and efficient LDL receptors, LDL FCR has been demonstrated to be significantly slowed down (24,25). So, the hypothesis that a 41% decrease of the expression of LDL receptors in type 2 diabetic patients is highly implied in the decrease of LDL FCR seems to be relevant. In the same manner, the normalization of the expression of LDL receptors after the introduction of insulin therapy probably plays an important role in the normalization of LDL FCR in these patients.

In type 2 diabetic patients, LDLs are triglyceride rich and glycated, which is likely to decrease their affinity to the LDL receptor (26). The correction of these abnormalities by insulin therapy may also contribute to the normalization of LDL FCR in insulin-treated type 2 diabetic patients, parallel to the increase of LDL receptor expression. However, the correction of LDL qualitative abnormalities by insulin therapy is far from being constant in these patients (27,28). It is noteworthy that in our group of type 2 diabetic patients, we observed no significant change of the triglyceride-to-apoB, cholesteryl ester–to–apoB, and triglyceride–to–cholesteryl ester ratios and of LDL glycation 3 months after the introduction of insulin therapy (data not shown).

Another apparently intriguing result is the lack of decrease in plasma LDL cholesterol concentration in insulin-treated patients, whereas LDL receptor expression is increased, suggesting that LDL clearance rate is probably also increased. In fact, the in vivo kinetic study we recently performed showed that 3 months after the introduction of insulin therapy in patients with poor metabolic control, the increase in LDL FCR was accompanied by an increase in LDL synthetic rate, explaining why LDL concentration can be unchanged after the introduction of insulin therapy despite an increase in FCR.

In conclusion, we demonstrate that despite oral antidiabetic treatment, LDL receptor expression is decreased in type 2 diabetic patients with a poor metabolic control compared with nondiabetic subjects and that it is normalized after 3 months of insulin therapy. This study therefore favors an important in vivo role of insulin in the expression of LDL receptors. Moreover, the number of LDL receptors at the cell surface evolves in the same way as LDL FCR in type 2 diabetes, suggesting that a low LDL receptor expression may play an important role in the decrease of LDL FCR observed in type 2 diabetic patients with poor metabolic control, whereas the normalization of this expression may contribute to the normalization of LDL FCR in insulin-treated patients.

We are indebted to Dominique Battaut, Elisabeth Niot, and Liliane Princep for their invaluable technical assistance.