FAD-glycerophosphate dehydrogenase activity in lymphocytes of type-2 diabetic patients and their relatives

DiabetesResearchand Clinical Practice31 (1996)17-25

FAD-glycerophosphate dehydrogenase activity in lymphocytes of type-2 diabetic patients and their relatives Josep Vidalb, Joanne RasschaerP, Abdullah Sener”, Ramon Gomisb, Willy J. Malaisse”3* “Laboratory

of Experimental

Medicine, Erasmus Medical School, Brussels Free University, B- 1070 Brussels, Belgium ‘Hospital Clinic, Endocrinology Unit, 08036 Barcelona, Spain

808 Route de Lennik,

Revised4 January 1996;accepted9 January 1996

Abstract The activities of FAD-linked glycerophosphate dehydrogenase (m-GDH), glutamate dehydrogenase (GlDH), glutamate-pyruvate transaminase (GFT) and glutamate-oxalacetate transaminase (GOT) were measured in purified populations of CD3 + lymphocytes from 55 control subjects, 62 type-2 diabetics and 50 non-diabetic relatives of the latter patients. The activity of m-GDH was measured by both a radioisotopic procedure and colourimetric technique. As judged from these measurements and relative to the paired value for GlDH, the incidence of abnormally low m-GDH activity was significantly higher in type-2 diabetics than in control subjects. Moreover, the paired ratio in reaction velocity between the colourimetric and radioisotopic assay of m-GDH was abnormally high in patients with low m-GDH activity. Low m-GDH activity often coincided with increased GPT activity in plasma or high GPT/GOT ratio in lymphocytes. No obvious clustering of these anomalies was found in relatives of diabetic patients. These findings suggest that an inherited or acquired genomic defect of m-GDH in lymphocytes, and possibly in pancreatic B-ceils, may participate to the pathogenesis of non-insulin-dependent diabetes mellitus.

1. Introduction A low activity of the mitochondrial FADlinked glycerophosphate dehydrogenase (mGDH) in pancreatic islet homogenates was recently documented in several animal models of type-2 diabetes mellitus (see ref. [I] for review) and a few patients affected by this disease {Z]. It * Correspondingauthor.

was further proposed that this enzymatic anomaly may account for a preferential impairment of the pancreatic B-cell secretory response to D-glucose [3]. The extension of this concept to type-2 diabetic patients is obviously hampered by the limited access to freshly isolated islets from such subjects, As an indirect approach, however, the activity of m-GDH was measured in lymphocytes prepared from these subjects. A low activity of m-GDH was found to be far from exceptional in

0021.9150/96/$15.00 0 1996ElsevierScienceIreland Ltd. All rights reserved PI2 SOO21-9150(96)01202-8

18

J. Vidal et al. 1 Diabetes Research and Clinical Practice 31 (1996) 17-25

the lymphocytes of patients with non-insulin-dependent diabetes [4,5]. The major aim of the present study is to investigate the possible familial clustering of such an enzymatic defect. 2. Materials and methods The present study concerns 55 control subjects (age: 28 to 91 years), 62 type-2 diabetic patients (36 to 95 years) and 50 first degree non-diabetic relatives of type-2 diabetics (16 to 62 years). The control group includes both normal subjects and non-diabetic patients with normal glycemia and HbAlc percentage and no first degree familial history of diabetes. The type-2 diabetic patients were defined according to the National Diabetes Data Group criteria [6]. They were treated by acarbose (2 cases), oral hypoglycemic agents (22 cases), insulin (28 cases), both insulin and oral hypoglycemic agents (4 cases) or diet only (2 cases).The relatives of diabetic patients were considered as non-diabetic on the basis of normal glycemia and HbAlc percentage [7]. For 37 of such relatives, a first degree diabetic proband was available for inclusion in this study. None of the subjects suffered from a hematological disease or displayed evidence of infection. An abnormal blood cell count was considered as an exclusion criterion. The experimental procedure was comparable to that described elsewhere [4]. Briefly, venous blood samples (about 20 ml) were collected in heparinized tubes after an overnight fast. Peripheral blood mononuclear cells were isolated using standard Ficoll-Isopaque gradient centrifugation. Isothrice in cells were washed lated phosphate-buffered saline and cultured at 0.5 x lo6 cells/ml in RPM1 medium (Gibco, Karlsruhe, Germany) mixed with 10% (v/v) inactivated calf serum and containing L-glutamine (2.0 mM), penicillin (100 U/ml), gentamicin (50 @g/ml), phytohaemagglutinin (50 pg/ml) and, from the second day onwards, recombinant interleukin 2 (10 U/ml). The culture medium was replaced every 3-4 days. After culture for 7 days, the cells were sonicated (20 x lo6 cells/ml) in a Hepes-NaOH buffer (20 mM, pH 7.4) containing sucrose (250

mM), EDTA (2.5 mM), and L-cysteine (2.0 mM). After centrifugation (1000 x g x 10 min), the supernatant was lyophilized. The lymphocyte extract was reconstituted in H,O and examined for its activity in m-GDH (EC 1.1.99.5) [8,9], glutamate dehydrogenase (GlDH, EC 1.4.1.3) [lo], glutamate pyruvate transaminase (GPT, EC 2.6.1.2.) [ll], and glutamate-oxalacetate transaminase (GOT, EC 2.6.1.1) [l l] by methods described in the cited references. The final concentrations of substrates, cofactors and activators were as follows: L-glycerol-3-phosphate 5 mM and 2-(4-iodophenyl)-3-(4nitrophenol)-5phenyltetrazolium 2 mM in the colourimetric assay of m-GDH, L-[2-3H]glycerol-3-phosphate 0.5 mM and FAD 0.05 mM in the radioisotopic assay of m-GDH, [U-14C]2-ketoglutarate 0.35 mM, ammonium acetate 25 mM, NADPH 0.3 mM and ADP 0.5 mM in the assay of GlDH, [U-14C]2-ketoglutarate 1.8 mM, pyridoxal phosphate 0.8 mM and either L-alanine or L-aspartate 20 mM in the assay of GPT and GOT. All enzymatic assays were conducted, in triplicate, over 30 min incubation at 37°C. The results were corrected for blank values found in the absence of either tissue homogenate (radioisotopic assay of m-GDH) or a suitable substrate (i.e. L-glycerol-3-phosphate in the colourimetric assay of m-GDH, NADPH in the assay of GlDH, and amino acid in the assay of GPT and GOT). The activity of GlDH, GPT and GOT was expressedas pmol/min per lo6 cells. In the case of m-GDH, the results were expressed relative to the paired value of GlDH in order to refer specifically to mitochondrial enzymes. The activity of mGDH was considered as low whenever it represented less than 50% of the mean control value. The latter percentage indeed corresponded to the lower limit of individual values in control subjects, after exclusion of 3 cases with readings below the 95% confidence interval in the radioisotopic assay and/or colourimetric procedure. All results were expressed as mean values ( + S.E.M.), together with the number of individual observations (n). A geometric mean was used whenever the individual data refer to a paired ratio in enzymatic activities. In such cases, the quoted SEM takes into account the upper and

19

J. Vidal et al. / Diabetes Research and Clinical Practice 31 (1996) 17-25

Table 1 Clinical data in control subjects, diabetic patients, and their relatives Controls

Diabetics

Male/female Age (years) Body mass index Glycemia (mM) HbAlc (%) Plasma insulin (PM) C-peptide (nM) Duration of diabetes (years) Complications (yes/no) Triglycerides (mM) Cholesterol (mM) Systolic blood pressure (mmHg) Diastolic blood pressure (mmHg) Hypertensive/normotensive Plasma GOT (IUjl) Plasma GPT (IUjl)

11145 60.0 rt 28.1 + 13.68 k 8.59 If: 67.2 k 0.83 + 12.2 * 27,l28 1.92 + 6.55 k 130 & 75 + 15145 22 k 30 k

11/44 45.5 If 25.5 k 5.03 + 4.53 + 58.8 + N.D. 1.13 f 5.14 k 124 + 76 f 7140 21 + 20 k

3.1 (54) 1.0 (47) 0.10 (43) 0.09 (41) 4.2 (44)

0.13 (46) 0.15 (46) 2 (47) 1 (47) 1 (46) 1 (46)

Relatives 1.7 (62) 0.9 (59) 2.34 (59)b 0.25 (58) 8.4 (21) 0.13 (29) 1.1 (54) 0.17 (59)s 0.76 (59) 2 (59)b 1 (59) 2 (58) 3 (58)b

21129 31.6 k 24.4 + 5.05 + 4.24 k 61.5 f N.D. 1.05 * 5.13 + 118 f 72 k l/498 20 i 24 +

1.7 (5Oyy 0.6 (5O)fl 0.08 (5O)B 0.06 (48)Y 5.4 (49)

0.09 (49)Y 0.12 (49) 1 (5O)y 1 (50) 1 (50) 1 (50)

aP < 0.05; bP < 0.01 and “P < 0.001 versus controls. “P < 0.005; BP < 0.01 and ?P < 0.001 versus diabetics.

N.D.: not determined.

lower deviation from the mean value [12]. The statistical significance of differences between mean values was assessed,as required, by either covariante analysis or Student’s t-test. 3. Results Table 1 provides information on the metabolic status of the three groups of subjects. Except for a younger mean age of diabetic relatives than control subjects, there was no significant difference between these two groups. Diabetic patients displayed, on average, a higher age, body mass index, glycemia, HbAlc percentage, plasma triglyceride concentration and systolic blood pressure than either their relatives or control subjects. The plasma GPT activity was higher in diabetic than control subjects. The mean plasma cholesterol concentration and diastolic blood pressure were also somewhat higher in diabetic patients than in their relatives. In those patients that were not treated with insulin, the mean plasma insulin concentration was comparable to that found in either control subjects or diabetic relatives. Likewise, the mean plasma C-peptide concentration in diabetic patients was comparable to that otherwise found in normal subjects (0.79 Ifi 0.18 nM).

Since the mean age of control subjects was significantly lower than that of diabetic patients, a group of old control subjects (n = 9; 86.7 f 1.3 years) was compared to another group including all other control subjects (n = 46; 37.1 + 2.1 years). There was no significant difference between these two groups in terms of glutamate dehydrogenase, GOT, GPT or m-GDH activity (radioisotopic and colourimetric assay). In lymphocytes, the activity of glutamate dehydrogenase, expressed as pmol/min per lo6 cells, was not significantly different in control subjects (551 + 29; n = 54) and either diabetics (546 + 24; n = 62) or their relatives (550 f 23; n = 47). In the latter two groups, there was no significant difference between subjects with either normal or low m-GDH activity (Table 2). Likewise, the activity of either GOT or GPT, expressed as pmol/min per lo6 cells, was not significantly different in control subjects and diabetic patients or their relatives. There was also no obvious difference of GOT and GPT activities in diabetic subjects and relatives with either low or normal m-GDH activity (Table 2). There was a highly significant correlation (P < 0.001; n = 138) between measurements of mGDH activity in the radioisotopic assay and

20

J. Vidal et al. / Diabetes Research and Clinical Practice 31 (1996) 17-25

Table 2 Enzymatic data in control subjects, diabetic patients and their relatives Diabetic

Control

Glutamate 551 * dehydrogenase (54) (pmol/min per lo6 cells) GOT 1278 + (pmol/min per (51) lo6 cells) GPT (pmol/min 44.9 f per lo6 cells) (43) m-GDH 1.31 f (radioisotopic)/ (52) GlDH (%a) m-GDH 262.3 k (colourimetric)/ (47) GlDH (%o)

Relatives

Normal m-GDH

Low m-GDH

Normal m-GDH

Low m-GDH

29

565 k 30 (42)

506 + 40 (20)

544 k 25 (40)

584 * 51 (7)

53

1309 f 61 (40)

1196 k 112 (17)

1308 + 55 (37)

1241 + 169 (5)

6.5

52.3 + 8.0 (35)

43.8 f 5.3 (15)

57.3 i

64.9 + 24.8 (5)

0.11

1.41 f 0.12 (42)

0.42 k 0.06 (20)

1.75 k 0.16 (40)

0.47 + 0.16 (7)

19.6

321.8 + 24.6 (39)

137.5 f 28.6 (13)

309.5 i

74.2 i

colourimetric procedure, both measurements being expressed relative to the paired activity of glutamate dehydrogenase. As shown in Fig. 1 and as already indicated above in the Materials and methods section, only 3 out of 52 control subjects displayed a low m-GDH activity in the radioisotopic assay (26.0 + 5.5% of mean control value; n = 3) and/or colourimetric procedure (33.8 + 7.4% of mean control value; n = 3), all results referring to the paired ratio between m-GDH and GlDH activity. One of these subjects underwent pancreatectomy for chronic pancreatitis, whilst a familial history of type-2 diabetes (maternal diabetic grandfather) was identified in another of these three control subjects. The mean activity of m-GDH was not significantly different in control subjects and either diabetic patients or their relatives, whether in the radioisotopic assay or colourimetric procedure. However, the incidence of low m-GDH activity in one or both assay procedures was higher (Pearson x2: P < 0.001) in diabetic patients than control subjects (Fig. 1). As documented in Table 2, low activities of m-GDH coincided, as a rule, in both

8.7 (34)

22.3 (35)

12.3 (4)

the radioisotopic and colourimetric assay. The diabetic patients with low m-GDH activity could not be distinguished from those with normal m-GDH activity in terms of sex ratio, age, body mass index, glycemia, HbAlc percentage, plasma concentration of insulin or C-peptide, duration of diabetes, cholesterolemia, systolic or diastolic blood pressure, or plasma GOT activity (Table 3). Likewise, the percent of patients requiring insulin treatment was not significantly different in diabetic patients with either low (12/20: 60%) or normal (19/37: 51%) m-GDH activity. However, the diabetic patients with low m-GDH activity were found to display a higher plasma triglyceride concentration, higher plasma GPT activity and higher incidence of hypertension than the diabetic patients with normal m-GDH activity (Table 3). Among relatives of diabetic patients, only 7 out of 47 subjects displayed low m-GDH activity in the radioisotopic assay and/or colourimetric procedure (Fig. 1). The incidence of low m-GDH activity was not significantly higher in relatives of diabetic patients than in control subjects, when tested by Fisher’s exact test. The information on

J. ViaM et al. /Diabetes

Radioisotopic

21

Research and Clinical Practice 31 (1996) 17-25

Cotourimet

assay

r ic 8ssay

0 0 0

0

00 o”o %og 000

0 8’00 oo”o oooo 00 n

00 08

$00

08+ 0 00 oQ9 +OO

OOOO

0a 80 0

08 0000

0” 00 “0” --

0 “p-

-w&A

-

. .

. .

l.

: .

. .

l

Control

.

.

..

.

Relative + Diabat ic

. Control

Diabttic

Relative

Fig. 1. Individual values for the paired ratio between m-GDH and GIDH activity (logarithmic scale) in control subjects, diabetic patients and their relatives. Results are expressed relative to the mean value found in control subjects in either the radioisotopic assay (left panel) or colourimetric procedure (right panel). Closed circles refer to subjects with low m-GDH activity in one or both assay procedures. The horizontal dotted lines refer to a relative value of 0.5, taken as the lower limit of the normal range.

the diabetic subjects and their relatives is summarized in Table 4. Among the 20 families in which the m-GDH activity was measured in the diabetic proband( ten families were identified in which normal m-GDH activity was present in both the diabetic patient(s) and their relatives. Among 14 non-diabetic relatives of patients with low mGDH activity, only two subjects also displayed low m-GDH activity.

Fig. 2 illustrates the pedigree of two families in which 5-6 subjects were tested. In one case, the diabetic female proband and her four children all displayed normal m-GDH activity. In the other case, three diabetic sisters with low m-GDH activity had each a non-diabetic child with normal m-GDH activity. As expected [8], the reaction velocity was much higher in the colourimetric assay than in the

22

J. Vidal et al. / Diabetes Research and Clinical Practice 31 (1996) 17-25

radioisotopic procedure. The paired ratio in reaction velocity averaged in diabetic patients 301.9 & 24.4 (n = 52) as compared (P < 0.001) to 229.5 f 12.3 (n = 86) in either control subjects or relatives of diabetic patients. In this respect, no significant difference (P > 0.2) was found between control subjects (paired ratio: 242.6 k 17.7; n = 47) and relatives of diabetic patients (paired ratio: 213.6 f 17.7; n = 39). The paired ratio in reaction velocity was also higher (P < 0.005) in diabetic patients with low m-GDH activity (444.3 + 78.2; IZ = 13) than in the patients with normal m-GDH activity (260.7 f 25.1; n = 39). The latter value was virtually identical (P > 0.5) to that found in control subjects (Fig. 3). Table 3 Clinical data in diabetic patients with either normal or low m-GDH activity m-GDH activity

Normal

Male/female Age (years)

11/31 58.7 (42) Body mass index 27.6 (40) Glycemia (mM) 14.39 (41) HbAlc (%) 8.45 (42) Plasma insulin (pM) 79.02 (13) Plasma C-peptide (nM) 0.94

(21)

Low

+ 2.3

6114 62.8 k 2.4

(20) & 1.1 * 3.38 k 0.30 + 12.36 f 0.16

Duration of diabetes 13.2 f (years) (37) Complications (yes/no) 19/l 8 Triglycerides (mM) 1.63 k (41) Cholesterol (mM) 6.69 k (41) Systolic blood pressure 128 + (~I-&) Diastolic blood 74 & pressure (mmHg) Hypertensive/normotens 7135 ive Plasma GOT (IU/l) 19 * Plasma GPT (II-I/l) 25 f

1.2

29.0 + 1.3 (19) 12.08 * 0.59 (18) 8.94 + 0.49

(16) 48.48 + 6.90 (8) 0.54 * 0.09

(8)

0.15

10.2 k 2.3 (17) S/l0 2.58 k 0.38

1.10

6.25 k 0.27

Wb

(18) 2 (41)

134 + 5 (18)

1 (41)

76 + 3 (18) sj1oa

1 (40) 2 (40)

26 k 6 (18) 40 + 8

Wb “P < 0.05. bP < 0.01.

Among 19 diabetic patients or relatives with low m-GDH activity in which the paired GPT/ GOT ratio was also measured, 14 subjects were found to display a high GPT/GOT ratio, exceeding by 50% or more the mean corresponding control value (27.6 + 3.7%“; n = 43). In these 14 subjects, the GPT/GOT ratio indeed averaged 223 f 20% (P < 0.001) of such a mean control value. 4. Discussion The procedure used in the present study for isolation and culture of lymphocytes offers two major advantages. First, it allows to obtain from a single blood sample sufficient biological material for the measurement of several enzymatic activities. Second, the lymphocytes obtained in this procedure represent a homogenous population of CD3 + T cells that multiplied during culture at a standardized D-glucose concentration (11.1 mM). Hence, any enzymatic anomaly in such cells is likely to represent a genomic defect, rather than reflecting interference of circulating factors such as hyperglycemia. Our results confirm that a low activity of mGDH is not exceptional in lymphocytes of type-2 diabetics and is often associated with an abnormally high GPT/GOT ratio. The activity of GPT in plasma was also abnormally high in the diabetic patients with low m-GDH activity. A genomic link may exist between these two enzymatic anomalies [3]. The incidence of low m-GDH activity in type-2 diabetics in the present report (32.3%) was close to that found (37.5%) in our first study [4]. In the diabetic patients with low m-GDH activity, the paired ratio of reaction velocity in the colourimetric/radioisotopic assay was significantly higher than in either control subjects or diabetic patients with normal m-GDH activity. A comparable observation was already made in freshly isolated islets from two type-2 diabetic patients [2]. Since this dissociated behaviour does not appear attributable to a deficient transfer of 3H from FAD3H to 3HOH in the electron transfer chain, as could conceivably occur in patients with the

23

J. Viaizl et al. / Diabetes Research and Clinical Practice 31 (1996) 17-25

Table 4 Normal (N) and low (L) m-GDH activity in diabetic patients and their relatives m-GDH activity in diabetic proband Normal activity

Low activity

Activity not determined

Diabetics

Relatives

Diabetics

Relatives

Diabetics

Relatives

N1” N! N! N1.N2 N eu Nl El N! N! N

N2 N2,N2 N2,N2 N2 N2,N2,N2,N2 Nl N2,N2 N2,N2 N2,N2 N2,L2,L2 N2,N2

!J LI.Ll.Ll Li Ll L! !J Ll !A L!

L2 N2,N2,N2 N2 Nl N2,N2,Ll N2,N2 N2 Nl N2

1 1 1 1 2 2 2 2 2 2 2

N2 N2 N2 N2 N2 N2 N2 N2 L2 L2 N2,L2

The information collected in the diabetic probands is underlined. The numbers 1 and 2 refer to the first and second generation, respectively.

MELAS syndrome [13], it points to a defect in the intrinsic properties of m-GDH [2,S]. Our effort to explore the possible familial clustering of these enzymatic anomalies in relatives of type-2 diabetic patients failed to produce conclusive information. The data collected in relatives of

Fig. 2. Pedigree of low (L) and normal (N) m-GDH activity in two families of diabetic patients. Squares and circles refer to males and females, and open and shaded symbols to normal and diabetic subjects. The oblique lines refer to dead relatives. The present age of subjects is also indicated.

diabetic patients with low m-GDH activity strongly argue against a dominant genetic defect with complete penetrance, but do not rule out another mode of inheritance. Alternatively, our data are also compatible with an acquired genomic defect, as could result for instance from protein malnutrition in foetal or neonatal life [14-161. The latter mechanism could also account for clustering of the enzyme defect within a given generation, as was apparently the case in one familial pedigree (Fig. 2). In a recent oral presentation, Bell reported that a mutation of either the m-GDH or glucokinase gene does not represent a statistically significant feature in the overall population of type-2 diabetics’. It is known, however, that a mutation of the glucokinase gene is not uncommon in patients with maturity onset diabetes of the young [17]. It cannot be ruled out, therefore, that some genetic defect affecting m-GDH activity prevails in a given subgroup of type-2 diabetic patients. In this respect, it would obviously be most relevant to reinvestigate with the tools of molecular biology those patients yielding a low m-GDH activity in lymphocytes. ’ G.I. Bell: Molecular biology of insulin secretion. Presented at the symposium Pancreatic Beta-Cell 1994 (Kyoto; November 1994).

24

J. Vidal et al. I Diabetes Research and Clinical Practice 31 (1996) 17-25

References r, .Z .E ii I

500-

400-

E ; i

300-

.-0 F? *i 5 4

200-

0

i < 2L 3 E ‘C 0’ 3

loo-

OControl

Diabetic

Relative

Fig. 3. Mean values ( f S.E.M.) for the paired colourimetric/ radioisotpic ratio for m-GDH activity in lymphocytes of control subjects, diabetic patients and their relatives. In the diabetic group, the results refer to patients with either normal (open column) or low (shaded column) m-GDH activity. The number of subjects in each group is shown at the bottom of each column.

The present study may set the scene, therefore, for further investigations on the possible participation of an inherited or acquired islet m-GDH defect in the pathogenesis of type-2 diabetes.

Acknowledgements This work was supported by grant 92/0791 from the Fondo de Investigaciones Sanitarias (Madrid, Spain) and grant 3.4513.94 from the Foundation for Scientific Medical Research (Brussels, Belgium). J.V. is supported by a research fellowship from the Hospital Clinic (Barcelona, Spain). We are grateful to C. Demesmaeker for secretarial help.

[1] Malaisse, W.J. (1994) FAD-linked glycerophosphate dehydrogenase activity in pancreatic islets of diabetic rodents. In: V.E. Shafrir (Ed.), Lessons from Animal Diabetes. Smith-Gordon, London (in press). [2] Fernandez-Alvarez, J., Conget, Rasschaert, J., Sener, A., Gomis, R. and Malaisse, W.J. (1994) Enzymatic, metabolic and secretory patterns in human islets of Type 2 (non-insulin-dependent) diabetic patients. Diabetologia 37, 177-181. [3] Malaisse, W.J. (1993) Is type 2 diabetes due to a deficiency of FAD-linked glycerophosphate dehydrogenase in pancreatic islets? Acta Diabetol. 30, 1-5. [4] Malaisse, W.J., Malaisse-Lagae, F., Kukel, S., Reinhold, U. and Sener, A. (1993) Could non-insulin-dependent diabetes mellitus be attributable to a deficiency of FADlinked glycerophosphate dehydrogenase? Biochem. Med. Metab. Biol. 50, 226-232. [5] Malaisse, W.J., Zlhner, D., Malaisse-Lagae, F. and Sener, A. (1993) Absence of FAD-glycerophosphate dehydrogenase deficiency in lymphocytes of insulin-dependent diabetic subjects. Med. Sci. Res. 21, 269-270. [6] National Diabetes Data Group. (1979) Classification and diagnosis of diabetes mellitus and other categories of glucose intolerance. Diabetes 28, 1038-1057. [7] Cook, J.T.E., Hattersley, A.T., Levy, J.C., Patel, P., Waierscoat, J.S., Hoackaday, D.R. and Turner, D. (1993) Distribution of Type 2 diabetes in nuclear families. Diabetes 42, 106-l 13. [8] Sener, A., Malaisse-Lagae, F. and Malaisse, W.J. (1992) Pancreatic islet FAD-linked glycerophosphate dehydrogenase activity in a model of B cell glucotoxicity. Med. Sci. Res. 20, 701-703. [9] Rasschaert, J. and Malaisse, W.J. (1991) Hexose metabolism in pancreatic islets. Glucose-induced and Ca* + -dependent activation of FAD-glycerophosphate dehydrogenase. Biochem. J. 278, 335-340. [lo] Sener, A. and Malaisse, W.J. (1990) A sensitive radioisotopic method for the measurement of NAD(P)H: its application to the assay of metabolites and enzymatic activities. Anal. Biochem. 186, 236-242. [11] Perales, M.A., Sener, A. and Malaisse, W.J. (1992) Radioisotopic assay of aspartate and alanine aminotransferase. Clin. Biochem. 25, 105-107. [12] Leclercq-Meyer, V., Malaisse-Lagae, F., Coulic, V., Akkan, A.G., Marchand, J. and Malaisse, W.J. (1992) Preservation of the anomeric specificity of glucose-induced insulin release in partially pancreatectomized rats. Diabetologia 35, 505-509. [13] Rasschaert, J., Pueyo, M.E., Velho, G., Froguel, P. and Malaisse, W.J. (1995) FAD-glycerophosphate dehydrogenase activity in lymphocytes of patients with mitochondrial mutation of the tRNALeU(UUR)gene. Med. Sci. Res. 23, 143-144. [14] Hales, C.N. and Barker, D.J.P. (1992) Type 2 (non-insulin-dependent) diabetes mellitus. The thrifty phenotype hypothesis. Diabetologia 35, 595-601.

J. Vidul et al. /Diabetes

Research and Clinical Practice 31 (1996) 17-25

[15] Sener, A., Ma&se-Lagae, F. and Malaisse, W.J. (1993) Decreased activity of mitochondrial glycerophosphate dehydrogenase in islets of rat fed a low protein diet. Med. Sci. Res. 21, 625-626. [16] Rasschaert, J., Reusens, B., Dahri, S., Sener, A., Remacle, C., Hoet, J.J. and Malaisse, W.J. (1995) Impaired activity of rat pancreatic islet mitochondrial glyc-

25

erophosphate dehydrogenase in protein malnutrition. Endocrinology 136, 2631-2634. [17] Vionnet, N., Stoffel, M., Takeda, J., et al. (1992) Nonsensemutation in the glucokinase gene causesearly-onset non-insulin-dependent diabetes mellitus. Nature 356, 721-722.