RSS Feeds
Polymorphonuclear Leukocytes in Non-Insulin-dependent Diabetes Mellitus: Abnormalities in Metabolism and Function
- Jadwiga M. Alexiewicz, MD;
- Dinesh Kumar, MD;
- Miroslaw Smogorzewski, MD;
- Mariusz Klin, MD; and
- Shaul G. Massry, MD
- From the University of Southern California School of Medicine, Los Angeles, California. Grant Support: By National Institute of Diabetes and Digestive and Kidney Diseases grant DK 29955. Dr. Klin was supported by the Fogarty International Fellowship. Request for Reprints: Shaul G. Massry, MD, Chief, Division of Nephrology, University of Southern California School of Medicine, 2025 Zonal Avenue, Los Angeles, CA 90033. Current Author Addresses: Drs. Alexiewicz, Smogorzewski, Klin, and Massry: Division of Nephrology, University of Southern California School of Medicine, 2025 Zonal Avenue, Los Angeles, CA 90033. Dr. Kumar: Division of Endocrinology, Diabetes and Hypertension, University of Southern California School of Medicine, 2025 Zonal Avenue, Los Angeles, CA 90033.
Abstract
Objective: To determine basal levels of cytosolic calcium ([Ca2+]i) and phagocytic activity in polymorphonuclear leukocytes (PMNLs) from patients with non–insulin-dependent diabetes (NIDDM).
Design: Prospective cohort study.
Setting: A university-county hospital.
Measurements: Cytosolic calcium levels, adenosine triphosphate (ATP) content, and phagocytosis of PMNLs from patients with NIDDM and from controls.
Intervention: In patients with NIDDM, we evaluated the effect of treatment with an oral hypoglycemic agent (glyburide) on [Ca2+]i levels, ATP content, and the phagocytosis of PMNLs.
Patients: 22 controls and 34 patients with NIDDM were examined. Fifteen patients were studied before and after 3 months of treatment with glyburide.
Results: Polymorphonuclear leukocytes from patients with NIDDM showed significantly elevated basal levels of [Ca2+]i (68 ± 9.6 compared with 43 ± 4.9 nmol/L; P < 0.01); reduced ATP content (1.30 ± 0.58 compared with 2.35 ± 0.45 nmol/106PMNLs; P < 0.01); and impaired phagocytosis (117 ± 21.0 compared with 145 ± 17.4 µg oil/107PMNLs per minute; P < 0.01) compared with controls. There was a direct and significant correlation (P < 0.01, r = 0.80) between [Ca2+]i levels in PMNLs and serum glucose levels and an inverse correlation between phagocytic ability and [Ca2+]i levels (P < 0.01; r = 0.62) as well as between phagocytic activity and fasting serum glucose levels (P < 0.01, r = 0.54) in patients with NIDDM. Glyburide therapy resulted in significant reduction in fasting serum glucose levels; in PMNLs, this treatment resulted in a significant reduction in [Ca2+]i levels, a significant increase in ATP content, and a significant improvement of phagocytosis.
Conclusions: Patients with NIDDM have elevated [Ca2+]i levels in PMNLs. This abnormality is probably induced by hyperglycemia and is primarily responsible for the impaired phagocytosis seen in these patients.
A large body of evidence indicates that the cytosolic calcium ([Ca2+]i) level is elevated in many cells in both patients with insulin-dependent diabetes mellitus (IDDM) and those with non–insulin-dependent diabetes mellitus (NIDDM) [1-6]. This phenomenon led Levy and Gavin [7] to propose that many of the complications of diabetes mellitus are, at least in part, related to an elevation in [Ca2+]i levels. This is plausible because other conditions in which [Ca2+]i levels are elevated, such as chronic renal failure [8] and phosphate depletion [9-12], are associated with organ dysfunctions similar to those seen with diabetes mellitus. However, a relation between the elevation in [Ca2+]i levels and a specific cell dysfunction in diabetes mellitus has not yet been documented.
The mechanisms responsible for the elevation in [Ca2+]i levels in patients with diabetes mellitus are not known. Hyperglycemia may be associated with increased calcium influx into cells and a corresponding increase in [Ca2+]i levels [13]. However, data on the relation among the degree of hyperglycemia, the magnitude of the increase in [Ca2+]i, and cell dysfunction are not available. Furthermore, it is not known whether the reversal of hyperglycemia by insulin therapy or by the use of oral hypoglycemic agents is followed by the return of [Ca2+]i levels and cell function to normal.
Phagocytosis by polymorphonuclear leukocytes (PMNLs) is impaired in patients with IDDM and NIDDM [14, 15]. However, data on [Ca2+]i levels in PMNLs and on the relations among hyperglycemia, [Ca2+]i levels in PMNLs, and phagocytosis are not available. The study of these accessible cells in patients with diabetes mellitus may provide an opportunity to explore these issues. We therefore examined PMNLs from patients newly diagnosed with NIDDM in order to delineate the interactions among hyperglycemia, [Ca2+]i levels, and phagocytosis.
Methods
We studied 22 controls and 34 patients with NIDDM to evaluate [Ca2+]i levels and phagocytosis in PMNLs. At the time of diagnosis, we obtained blood samples for the study of PMNL metabolism and function. All patients received the oral hyperglycemic agent glyburide (5 to 20 mg/d; mean dose, 7 ± 2.8 mg/d). Only 15 patients returned for follow-up after 3 months of therapy with this drug; no specific reasons were given for why the remaining 19 patients did not return for follow-up. All biochemical variables were assessed in the patients but not in the controls. The normal values for our laboratory are provided.
Separation of Polymorphonuclear Leukocytes
Peripheral venous blood was drawn under sterile conditions into vacutainers containing 20 U of preservative-free heparin (Gibco Laboratories, Grand Island, New York) per 1.0 mL of blood. Polymorphonuclear leukocytes were isolated from the whole blood according to the method described by Ferrante and Thong [16]. Fresh heparinized blood was layered in 3.5-mL aliquots over Ficoll-Hypaque solution with a density of 1.114 g/mL (Mono-Poly Resolving Medium, Flow Laboratories, McLean, Virginia) and centrifuged at 300 g for 35 min at room temperature. This procedure resulted in the separation of mononuclear and polymorphonuclear cells into two distinct bands, with the red blood cell pellet at the bottom of the tube. The PMNL layer was aspirated with a Pasteur pipette and washed twice in Hanks balanced salt solution. The purity of cells as determined by Wright stain was greater than 97%, and the viability of cells exceeded 98% as assayed by the trypan blue exclusion method.
Determination of Phagocytosis
We estimated the rate of ingestion by PMNLs spectrophotometrically according to the bioassay described by Southwick and Stossel [17]. Briefly, PMNLs were fed oil droplets containing oil red O and coated with Escherichia coli lipopolysaccharide (2 mg per experiment). These particles were previously opsonized in fresh autologous serum. After the PMNLs and the particles were incubated for 5 min at 37 °C, the reaction was rapidly stopped by adding ice-cold isotonic saline that contained 1 mmol/L N-ethylmaleimide, which poisons the cells and stops ingestion. Uningested particles were separated from cells that contained ingested particles by centrifugation at 250 g for 10 min. Oil red O was extracted from the cell pellet with dioxane, and the optical density of the extract was measured using a Perkin Elmer Lambda 2 UV/VIS spectrophotometer (Perkin Elmer, Norwalk, Connecticut) at a wavelength of 525 nm. The ingestion rate was expressed in micrograms of oil engulfed per 107PMNLs per minute. We purchased all reagents from Aldrich Chemical (Milwaukee, Wisconsin). Lipopolysaccharide E. coli 026:B6 was obtained from Boivin preparation (Difco Laboratories, Detroit, Michigan).
Assay of Adenosine Triphosphate in Polymorphonuclear Leukocytes
We measured the ATP content of PMNLs according to the method described by Lundin and colleagues [18]. We added an aliquot of 0.5 mL of 0.6 M ice-cold perchloric acid to 5 × 106PMNLs in 0.5 mL of RPMI (Gibco Laboratories). After 10 minutes of extraction on ice, we added 62.5 µL of 2 M potassium carbonate, and the mixture was centrifuged at 10 500 g for 10 min. The supernatant was removed and immediately frozen with liquid nitrogen and stored at −70 °C. On the day of the assay, we added 50 µL of the supernatant to 950 µL of 40 mmol/L Tris buffer (pH, 7.4) and diluted this mixture 20 times with distilled water. Samples of 50 µL were assayed for ATP by the firefly luciferase assay with LAD 535 luminometer (Turner Design, Sunnyvale, California). Adenosine triphosphate standards were prepared with 40 mmol/L Tris buffer, distilled water, and amounts of RPMI, perchloric acid, and potassium carbonate similar to those in the PMNL preparation.
Measurements of Cytosolic Calcium
We measured resting levels of [Ca2+]i in PMNLs with Fura 2-AM (Sigma, St. Louis, Missouri). A sample of 5 × 106PMNLs was washed with solution 1, which contained 132 mmol/L sodium chloride; 3 mmol/L potassium chloride; 1 mmol/L magnesium sulfate; 1.2 mmol/L monosodium acid phosphate; 10 mmol/L D-glucose; 10 mmol/L Hepes; and 0.02 mmol/L calcium chloride (pH was adjusted to 7.4 with Tris buffer) and was then spun at 300 g for 15 min. The pellet was resuspended in 490 µL of this solution and 10 µL of Fura 2-AM dissolved in dimethyl sulfoxide, giving a final concentration of 4 µmol/L of Fura 2. The mixture was then incubated in a water bath of 37 °C for 30 minutes. After this incubation, the cells were washed and resuspended in solution 1. Cytosolic calcium levels were measured with Perkin Elmer fluorescence spectrophotometer (model LS 513) at excitation wavelengths of 340 and 380 nm and an emission wavelength of 510 nm, with slits of 10 mm and 20 mm, respectively. We added a 100-µ L aliquot of the PMNL suspension to a spectrophotometer cuvette that contained 1.9 mL of solution 2, which was the same as solution 1 except that its calcium chloride concentration was 1 mmol/L. Autofluorescence from cells or added reagents (or both) was monitored during each experiment and was not found to be a significant factor. We evaluated maximal fluorescence and minimal fluorescence with 0.05% Triton and 5 mmol/L ethylene glycol tetra-acetic acid in Tris base buffer (pH, 8.0), respectively. Cytosolic calcium was calculated using the Grynkiewicz equation [19], and the dissociation constant for Ca2+-Fura 2 was assumed to be 225 nmol/L.
Biochemical Measurements
The concentration of calcium in serum was determined by Perkin Elmer spectrophotometer model 505 of phosphorus; that of creatinine by an autoanalyzer (Technicon, Tarrytown, New York); that of serum glucose by a multichannel analyzer (Beckman, Irvine, California); and that of hemoglobin A1c (HbA1c) by an ion exchange chromatography according to a modification of the method of Trivelli and colleagues [20]. Serum parathyroid hormone (PTH) levels were estimated with radioisotopic assay (Intact PTH-Nichols Institute Diagnostics). The normal value of PTH is 10 to 65 pg/mL.
Statistical analysis was done using a paired and an unpaired t-test, and the relation between two variables was determined by the correlation coefficient. Data are reported as mean ±SD unless otherwise specified.
Results
The study included 22 controls and 34 patients with NIDDM. The patients were examined at the time of their visit to the outpatient clinic and consisted of 20 women and 14 men (mean age, 45 ± 12.8 years; range, 30 to 71 years). No patients had infection, nor were they receiving any medications. The diagnosis of NIDDM was made by measurements of fasting hyperglycemia in the absence of ketonemia [21]. However, some of these patients may have had late-onset IDDM. All patients had normal blood pressure (mean, 127 ± 19.8/77 ± 11.1 mm Hg). The mean body mass index was 32.6 ± 9.3 kg/m2 for women and 30.0 ± 6.4 kg/m2 for men. The body mass index was normal in 2 women, in the overweight range in 4, and in the obese range in 11. It was normal in 2 men, in the over-weight range in 7, and in the obese range in 8.
Table 1 provides the clinical and biochemical data of the patients at the time of diagnosis. They had hyperglycemia, elevated blood levels of HbA1c, normal blood levels of creatinine and PTH, and blood cholesterol levels in the upper normal range. Ten patients had elevated blood levels of triglycerides, 19 patients had elevated low-density lipoprotein cholesterol levels, and 16 patients had decreased high-density lipoprotein cholesterol levels. The controls were 9 women and 13 men (mean age, 35 ± 4.9 years; range, 26 to 45 years).
In the patients with NIDDM, the basal [Ca2+]i levels in PMNLs (68 ± 9.6 nmol/L) were significantly higher than those in the controls (43 ± 4.9 nmol/L) (P < 0.01) (Figure 1). We observed a direct and significant correlation between [Ca2+]i levels in PMNLs and fasting serum glucose levels (r = 0.80; P < 0.01) (Figure 2). The PMNLs from the patients showed a significant reduction in phagocytic property (P < 0.01). The ingestion of opsonized oil droplets by these PMNLs was 117 ± 21.0 µg oil/107PMNLs per minute, a value significantly lower (P < 0.01) than that of PMNLs from the controls (145 ± 17.4 µg oil/107 PMNL per minute) (Figure 1). We observed a significant inverse correlation between phagocytosis and [Ca2+]i levels in PMNLs (r = 0.62; P < 0.01) (Figure 2) and between phagocytosis and fasting serum glucose levels (r = 0.54, P < 0.01) (Figure 2). We did not observe a significant correlation between phagocytosis and serum cholesterol or triglyceride levels. Among controls, no significant correlation was noted between [Ca2+]i levels in PMNLs and phagocytosis.
The ATP content of PMNLs in the patients with NIDDM (1.3 ± 0.58 nmol/106 PMNLs) was significantly lower (P < 0.01) than that of controls (2.35 ± 0.45 nmol/106 PMNLs) (Figure 1).
Table 2 shows the changes in the biochemical profiles of the 15 patients who received treatment with glyburide for 3 months. This treatment resulted in fairly good control of hyperglycemia. Fasting serum glucose levels were reduced significantly (P < 0.01); the values in 6 patients were less than 5.55 mmol/L (mean, 5.16 ± 0.42 mmol/L), and values in 9 patients were modestly elevated (mean, 7.44 ± 0.87 mmol/L). A significant (P < 0.01) reduction in the blood levels of HbA1c was also noted, with the values approaching normal levels. Furthermore, serum levels of cholesterol, low-density lipoprotein cholesterol, and triglycerides decreased significantly (P < 0.01) after patients received therapy with glyburide, and high-density lipoprotein cholesterol levels increased significantly (P < 0.01) after treatment.
After therapy with glyburide, the [Ca2+]i levels in PMNLs decreased significantly in all patients, from 71 ± 9.3 nmol/L to 53 ± 5.4 nmol/L (P < 0.01) (Figure 3). This change in [Ca2+]i was associated with significant improvement in the ATP content of the PMNLs (from 1.2 ± 0.50 nmol/107PMNLs to 2.0 ± 0.43 nmol/107 PMNL; P < 0.01) and with improvement in their phagocytic ability (from 116 ± 11.2 µg oil/107 PMNLs per minute to 125 ± 4.8 µg oil/107 PMNLs/minute; P < 0.01). The blood pressure of these 15 patients was normal before (127 ± 18.9/79 ± 10.1 mm Hg) and after (122 ± 13.9/73 ± 6.6 mm Hg) therapy.
Discussion
The results of our study show that, compared with controls, the basal levels of [Ca2+]i in PMNLs in patients with NIDDM are significantly elevated, whereas ATP content and phagocytic ability are reduced. These derangements were corrected after serum glucose levels returned to normal with glyburide therapy. The values of these variables in the PMNLs from controls did not differ from those previously determined in our laboratory [22].
Hyperglycemia may induce calcium influx into cells [13] and increase the activity [23] or the translocation of protein kinase C [24] or both. The activation of this enzyme may increase [Ca2+]i levels in many cells [25-27]. Hyperglycemia-induced stimulation of protein kinase C may therefore be at least partly responsible for elevated basal levels of [Ca2+]i in PMNLs. Two findings in our study support the notion that elevated [Ca2+]i levels in PMNLs are, at least in part, related to the hyperglycemia. First, we noted a direct and significant relation between [Ca2+]i levels of PMNLs and the degree of hyperglycemia. Second, the decrease in blood glucose levels toward normal values after therapy with glyburide was associated with the return of the [Ca2+]i levels in PMNLs toward normal values. However, it is possible that glyburide may directly affect [Ca2+]i levels of PMNLs. Data to confirm or refute such a possibility are not available.
Hypertension is associated with elevated [Ca2+]i levels in platelets [28], lymphocytes [29], and peripheral blood mononuclear leukocytes [30]. However, because our patients with diabetes were normotensive, hypertension could not have contributed to the elevated [Ca2+]i levels in PMNLs.
The age of the controls was significantly lower than that of the patients with NIDDM, and it is possible that, with increasing age, [Ca2+]i levels in PMNLs may increase and phagocytosis may decrease. Our data showed no difference in [Ca2+]i levels or phagocytosis with increasing age in either the controls or the patients. Furthermore, for any given age group (in decades), the patients had higher [Ca2+]i levels in PMNLs and lower phagocytic activity in PMNLs than did the controls.
An increase in calcium influx into cells by itself is not adequate to cause an increase in the basal levels of [Ca2+]i, because cells are endowed with powerful mechanisms that directly (Ca2+ATPase, Na+-Ca2+ exchanger) or indirectly (Na+-K+ATPase) extrude calcium from the cells [31]. Therefore, for basal levels of [Ca2+]i to increase, an impairment in the function of these pumps must also be present. Indeed, the activity of Ca2+ATPase, Na+-K+ATPase, and Na+-Ca2+ exchanger are lessened in conditions associated with elevated basal levels of [Ca2+]i, such as chronic renal failure [8, 32] and phosphate depletion [12]. We did not examine the activity of these pumps in the PMNLs from our patients, but others have found that Ca2+ ATPase activity is reduced in sarcolemmal preparations of heart from rats with streptozotocin-induced diabetes [33], and in erythrocytes of Zucker obese diabetic rats [34]. Also, Na+-K+ATPase activity is reduced in the hearts of rats with NIDDM [35], in the sciatic nerve of rats with streptozotocin-induced diabetes [36], and in the platelets of patients with NIDDM [37]. Furthermore, Na+-Ca2+ exchanger was found to be reduced in sarcolemmal preparations of heart from diabetic rats [33, 35]. Thus, it is reasonable to assume that the functions of Ca2+ATPase, Na+-K+ATPase, and Na+-Ca2+ exchanger are impaired in the PMNLs of patients with NIDDM.
The decrease in ATP content may be related to the hyperglycemia-induced calcium influx. Indeed, chronic and sustained entry of calcium into cells is associated with inhibition of mitochondrial oxidation and hence with reduced ATP content [38-40]. The decrease in ATP content may contribute to the potential decrease in the activity of Ca2+ ATPase and Na+-K+ATPase, because both enzymes require ATP [31].
On the basis of their studies of phosphate depletion, Craddock and colleagues [41] proposed that the reduction in the ATP content of PMNLs was responsible for the impairment of the phagocytosis in phosphate depletion. However, later studies by Kiersztejn and colleagues [11] clearly showed that the impaired phagocytosis in phosphate depletion is due to an increase in [Ca2+]i levels in PMNLs and not to a decrease in ATP content. They found that treatment of phosphate-depleted rats with verapamil, which prevented the increase in [Ca2+]i levels in PMNLs, was associated with the return of phagocytosis to normal despite sustained reduction in the ATP content of PMNLs. Furthermore, elevated [Ca2+]i levels in PMNLs in humans [27] or animals [42] with chronic renal failure was associated with impaired phagocytosis, and prevention of the increase in [Ca2+]i levels in PMNLs in humans [43] and animals [42] with chronic renal failure resulted in the return of phagocytosis to normal.
The results of our study are also consistent with the idea that the elevated [Ca2+]i levels in PMNLs in patients with NIDDM is at least partly responsible for the impairment in phagocytosis. Indeed, we found an inverse relation between phagocytosis and [Ca2+]i levels in PMNLs. Furthermore, the decrease in [Ca2+]i levels in PMNLs after glyburide therapy was associated with improvement in phagocytosis. Although the present data indicate that the increase in [Ca2+]i levels in PMNLs has an important role in the impairment of the phagocytosis of these cells, it is also possible that hyperglycemia itself may impair phagocytosis through an as yet unidentified mechanism.
Our data may also be consistent with the observation that hyperglycemia may initiate an increase in calcium influx into PMNLs and that this continued increase in calcium entry into the cells inhibits mitochondrial oxidation, leading to a decrease in ATP production and hence in ATP content of PMNLs. Under these circumstances, the decrease in ATP content would cause a decrease in the activity of the enzymes responsible for the extrusion of calcium from the cells. The combination of an increase of calcium influx into and a decrease in calcium efflux out of the PMNLs would lead to an increase in [Ca2+]i levels, which would impair phagocytosis by an as yet unidentified mechanism.
If hyperglycemia-induced calcium influx initiates the process described above, this increased influx may be prevented by calcium channel blockers, as in chronic renal failure [27] and phosphate depletion [11]. The use of such agents may therefore also correct the abnormality in PMNLs in patients with uncontrolled hyperglycemic NIDDM. Studies are needed to explore this possibility.
- Copyright ©2004 by the American College of Physicians
Related Clinical Content
