Context:

Ketosis-prone diabetes (KPD) is mostly observed in males of West African descent and is characterized by phasic or permanent insulin dependence without apparent autoimmune process.

Objective:

KPD subjects display a propensity to hyperglycemia-induced acute insulin deficiency, suggesting that they exhibit a propensity to oxidative stressinβ-cells.The enzyme glucose-6-phosphate dehydrogenase (G6PD) is a defense mechanism against oxidative stress, and G6PD deficiency, an X-linked genetic disorder with male predominance, is frequent in West Africans. We hypothesized that mutations in theG6PD gene could predispose to KPD.

Design:

We studied G6PD erythrocyte enzyme activity and the insulin secretory reserve (glucagon-stimulated C peptide) in a cohort of hospitalized West Africans with KPD (n = 59) or type 2 diabetes (T2DM; n = 59) and in normoglycemic controls (n = 55). We also studied theG6PD gene in an extended population of KPD patients (n = 100), T2DM patients (n = 59), and controls (n=85).

Results:

The prevalence of G6PD deficiency was higher in KPD than in T2DM and controls (42.3%; 16.9%; 16.4%;P = 0.01). In KPD, but not in T2DM, insulin deficiency was proportional to the decreased G6PD activity (r = 0.33;P = 0.04). We found no increase in the prevalence ofG6PD gene mutations in KPD compared with T2DM and controls.Rather,we found a 20.3% prevalence of G6PD deficiency in KPD without gene mutation.

Conclusions:

This study suggests that 1) G6PD deficiency alone is not causative of KPD; and 2) alterations in genes controlling both insulin secretion and G6PD-mediated antioxidant defenses may contribute to the predisposition to KPD in West Africans. (J Clin Endocrinol Metab 90: 4446–4451, 2005)

Patient and control populations

For the clinical study, we examined 59 consecutive KPD patients (40 men and 17 women), 59 patients with type 2 diabetes (T2D; 32 men and 27 women), and 55 normoglycemic controls (30 men and 25 women). The criteria used to define ketosis-prone diabetes have been described previously (1). Briefly, KPD was defined as new-onset diabetes without precipitating illness (infection or stress), with the presence of strong ketosis (urinary ketones,>80 mg/dl) or DKA, and in the absence of autoantibodies to islet cells and to glutamic acid decarboxylase 65. T2DM was defined as diabetes treated by diet or oral hypoglycemic agents, with no episode of ketosis and in the absence of autoantibodies to islet cells and glutamic acid decarboxylase 65. The control population was composed of subjects with normal fasting blood glucose and without a known family history of diabetes. Diabetic and control populations were comparable with regard to age.

The extended KPD and control populations used for the genetic study have been described and consisted of 100 KPD patients (71 men and 29 women) and 85 normoglycemic control subjects (41 men and 44 women) (2). All patients and controls originated from West Africa. The local ethical committee approved the study.

Measurement of erythrocyte G6PD enzyme activity

We measured the erythrocyte G6PD enzyme activity 48 h after normalization of blood glucose using the spectrophotometric method (15). Erythrocyte homogenate was saturated with G6PD and NADPH, followed by spectrophotometric detection of NADPH appearance at the absorbance of 340 nm in the kinetic mode. The results are calculated by evaluating the increase in OD per minute (slope) for unknowns against the slope for a standard with a known G6PD activity. Total erythrocyte enzyme activity was determined (units per gram of hemoglobin) and residual erythrocyte enzyme activity (REA) was calculated by dividing the observed value for each patient by the value of standard with known normal G6PD activity. G6PD deficiency was defined as an REA less than 40% of the normal value (16). The presence of G6PD deficiency was confirmed by a second measurement after at least 3 months of metabolic stabilization.

Insulin secretion

Pancreatic insulin secretory reserve was assessed 48 h after resolution of ketosis and normalization of blood glucose by measuring the C peptide level before and 8 min after the iv injection of 1 mg glucagon as previously described (1). The test was performed 12 h after an overnight fast and before the morning insulin injection in a subset of patients who agreed to participate and whose baseline characteristics were representative of the rest of the group. C Peptide was measured by RIA (IRMA-C-PEP, CIS Bio International, Gif-sur-Yvette, France).

G6PD gene screening

Genomic DNA was extracted from peripheral leukocytes in 100 unrelated KPD patients. The coding region of the humanG6PD gene was amplified in five different reactions: exon 1: primers G6PD1F2 (CACAAGGAGTGATTTGGGCA) and G6PD-1R (GCACTTCCTGGCTTTTAAGA); exons 2 and 3: primers G6PDAF (CAGCCACTTCTAACCACACACCT) and G6PDAR (CCGAAGTTGGCCATGCTGGG); exon 4: primers G6PD-4F (CGGACTCAAAGAGAGGGGCT) and G6PD-4R (GCACTGCCTGGGCCAGCCT); exons 5–7: primers G6PD5F (GCAGCTGTGATCCTCACTCC) and G6PD7R (CTGCGACAGGGCATGCTCCT); exons 8–10: primers G6PD8F (GCACCCCAACTCAACACCCA) and G6PD-9R (CGCCCACCCTCCACACTGCT); exons 11 and 12: primers G6PD11F (GAG AAG GAG CAG TGT GGA GG) and G6PD12R (GAG GAG AGG CAT GAG GTA GC); and exon 13: primers G6PD13F (ATG GCA GGT GAG GAA AGG GTG) and G6PD13R (GAC CCA GTG GCC AAT AAG CTC). Sequencing reactions were performed with the BigDye terminator kit (Applied Biosystems, Foster City, CA). Sequencing was performed on an ABI PRISM 3700 automated DNA sequencer (Applied Biosystems). Sequences were analyzed using Sequence Navigator (Applied Biosystems).

Single nucleotide polymorphism (SNP) genotyping

G6PD gene variants were genotyped in T2DM and controls using a TaqMan allelic discrimination assay system after designing flanking primers and fluorogenic probes, both of which were specific to target alleles, as previously described (2).

Statistical analysis

Results are expressed as the frequency or as the mean ± sd unless otherwise stated. Group comparison was performed byx² tests, a test for proportions was performed with continuity correction when necessary, and ANOVA and Mann-Whitney U test for means were conducted. The relation between insulin secretion and residual enzyme activity was analyzed by Spearman’s rank-order correlation. The haplotypes for the females in each of the three groups were constructed for SNPs A202G, A376G, A542T, and T968C using the SNPHAP program (http://www.gene.cimr.cam.ac.uk/clayton/software/snphap.txt). The haplotypes for the males were constructed based upon the observed of their hemizygous genotypes. The Fisher-Freeman-Halton exact test was used to test for differences in haplotype frequencies among the three groups (controls, T2DM, and KPD), because some of the expected values in the contingency table were less than 2.

Prevalence of G6PD deficiency is increased in KPD

We assessed the G6PD REA in a cohort of KPD subjects, T2D subjects, and normoglycemic controls. Although T2D subjects and controls had similar REAs, the REA was 30% lower in KPD subjects (Fig. 1A). The prevalence of G6PD deficiency (REA, 40% of the normal value) was similar in T2D patients and controls, but was 2-fold higher in KPD subjects (Fig. 1B).

G6PD deficiency is associated with the severity of pancreatic β-cell failure

We examined the hypothesized relationship between G6PD deficiency and insulin secretory reserve during an iv glucagon stimulation test in all diabetic subjects. The C peptide response to glucagon in each group is presented inFig. 1C. In T2D patients, there was no significant correlation between G6PD deficiency and insulin deficiency (r = 0.27;P = 0.3). Conversely, KPD patients presenting with G6PD deficiency had lower insulin secretory reserve compared with those with G6PD REA greater than 40% (increase in C peptide between 0 and 8 min, 0.90 ± 0.60vs. 1.33 ± 0.53µg/liter;P = 0.03). Moreover, in KPD patients, the degree of decreased G6PD activity was correlated to the severity of insulin deficiency (Fig. 1D). This correlation was stronger in insulin-dependent KPD patients (r = 0.62;P = 0.03).

Analysis of coding variants of the G6PD gene in KPD

G6PD deficiency could result from 1)G6PD gene mutations, or 2) alterations in factors controlling G6PD expression/activity (17).G6PD gene variations are common in West Africans, and the most common mutations associated with biochemical G6PD deficiency through enzyme instability consist of the amino acid change Asn¹²⁶Asp (A376G) associated with either amino acid change Met⁶⁸Val (G202A) or, less frequently, Asp¹⁸¹Val (A542T) or Leu³²³Pro (T968C). The last three mutations are not found alone; they are always associated with Asn¹²⁶Asp, the double mutation creating enzyme instability and G6PD deficiency (G6PD A⁻). Asn¹²⁶Asp is frequently found alone, but does not cause G6PD deficiency (G6PD A⁺). The wild-type gene without mutation is the form G6PD B (14,18).

We screened the coding region of theG6PD gene in 100 unrelated KPD subjects, and we genotyped the observed variants in a common type 2 diabetic and normoglycemic control population (Table 1). All the mutations described above were detected in KPD subjects at a frequency not different from those in the T2D and control populations (Table 1). Furthermore, we analyzed the haplotypes for the SNP A202G, A376G, A542T, and T968C in KPD, T2D, and control subjects. There was no statistical difference in haplotype frequencies among the three groups (P = 0.76).

Prevalence of G6PD mutations in subjects with G6PD deficiency

TheG6PD gene mutations described inTable 1 explained half of the G6PD deficiencies observed in KPD (G6PD A⁻). The frequency of A⁻ functional gene mutations associated with G6PD deficiency was not significantly increased in the KPD patients (22%) compared with T2DM (16.9%;P = 0.5) and control subjects (14.5%;P = 0.3;Table 2). This may be related to our small sample size. However, in the remaining KPD patients with G6PD deficiency, no other functional mutation was found (G6PD B). Rather, we observed a 20.3% prevalence of G6PD deficiency with no mutation, which was not observed in the T2DM subjects and controls. In these G6PD-deficient subjects, we did not find variants in intronic regulatory sequences (data not shown;Table 2).

• 1.Mauvais-Jarvis F, Sobngwi E, Porcher R, Riveline JP, Kevorkian JP, Vaisse C, Guillausseau PJ, Charpentier G, Vexiau P, Gautier JF2004. Ketosis-prone type 2 diabetes in patients of sub-Saharan African origin: clinical pathophysiology and natural history ofβ-cell dysfunction and insulin resistance.Diabetes53:645–653 [DOI] [PubMed] [Google Scholar]
• 2.Mauvais-Jarvis F, Smith SB, Le May C, Leal SM, Gautier JF, Molokhia M, Riveline JP, Rajan AS, Kevorkian JP, Zhang S, Vexiau P, German MS, Vaisse C2004. PAX4 gene variations predispose to ketosis-prone diabetes.Hum Mol Genet13:3151–3159 [DOI] [PMC free article] [PubMed] [Google Scholar]
• 3.Umpierrez GE, Casals MM, Gebhart SP, Mixon PS, Clark WS, Phillips LS1995. Diabetic ketoacidosis in obese African-Americans.Diabetes44:790–795 [DOI] [PubMed] [Google Scholar]
• 4.Poitout V, Robertson RP2002. Minireview: secondaryβ-cell failure in type 2 diabetes: a convergence of glucotoxicity and lipotoxicity.Endocrinology143: 339–342 [DOI] [PubMed] [Google Scholar]
• 5.Prentki M, Joly E, El-Assaad W, Roduit R2002. Malonyl-CoA signaling, lipid partitioning, and glucolipotoxicity: role inβ-cell adaptation and failure in the etiology of diabetes.Diabetes51(Suppl 3):S405–S413 [DOI] [PubMed] [Google Scholar]
• 6.Evans JL, Goldfine ID, Maddux BA, Grodsky GM2002. Oxidative stress and stress-activated signaling pathways: a unifying hypothesis of type 2 diabetes.Endocr Rev23:599–622 [DOI] [PubMed] [Google Scholar]
• 7.Robertson RP, Harmon J, Tran PO, Tanaka Y, Takahashi H2003. Glucose toxicity inβ-cells: type 2 diabetes, good radicals gone bad, and the glutathione connection.Diabetes52:581–587 [DOI] [PubMed] [Google Scholar]
• 8.Grankvist K, Marklund SL, Taljedal IB1981. CuZn-superoxide dismutase, Mn-superoxide dismutase, catalase and glutathione peroxidase in pancreatic islets and other tissues in the mouse.Biochem J199:393–398 [DOI] [PMC free article] [PubMed] [Google Scholar]
• 9.Tiedge M, Lortz S, Drinkgern J, Lenzen S1997. Relation between antioxidant enzyme gene expression and antioxidative defense status of insulin-producing cells.Diabetes46:1733–1742 [DOI] [PubMed] [Google Scholar]
• 10.Ihara Y, Toyokuni S, Uchida K, Odaka H, Tanaka T, Ikeda H, Hiai H, Seino Y, Yamada Y1999. Hyperglycemia causes oxidative stress in pancreaticβ-cells of GK rats, a model of type 2 diabetes.Diabetes48:927–932 [DOI] [PubMed] [Google Scholar]
• 11.Tanaka Y, Tran PO, Harmon J, Robertson RP2002. A role for glutathione peroxidase in protecting pancreatic cells against oxidative stress in a model of glucose toxicity.Proc Natl Acad Sci USA99:12363–12368 [DOI] [PMC free article] [PubMed] [Google Scholar]
• 12.Pandolfi PP, Sonati F, Rivi R, Mason P, Grosveld F, Luzzatto L1995. Targeted disruption of the housekeeping gene encoding glucose 6-phosphate dehydrogenase (G6PD): G6PD is dispensable for pentose synthesis but essential for defense against oxidative stress.EMBO J14:5209–5215 [DOI] [PMC free article] [PubMed] [Google Scholar]
• 13.Tian WN, Braunstein LD, Apse K, Pang J, Rose M, Tian X, Stanton RC1999. Importance of glucose-6-phosphate dehydrogenase activity in cell death.Am J Physiol276:C1121–C1131 [DOI] [PubMed] [Google Scholar]
• 14.Beutler E1994. G6PD deficiency.Blood84:3613–3636 [PubMed] [Google Scholar]
• 15.WHO1967. Standardization of procedure for the study of glucose-6-phosphate dehydrogenaseGeneva: World Health Organization; World Health Organization technical report 366; [PubMed] [Google Scholar]
• 16.May J, Meyer CG, Grossterlinden L, Ademowo OG, Mockenhaupt FP, Olumese PE, Falusi AG, Luzzatto L, Bienzle U2000. Red cell glucose-6-phosphate dehydrogenase status and pyruvate kinase activity in a Nigerian population.Trop Med Int Health5:119–123 [DOI] [PubMed] [Google Scholar]
• 17.Kletzien RF, Harris PK, Foellmi LA1994. Glucose-6-phosphate dehydrogenase: a “housekeeping” enzyme subject to tissue-specific regulation by hormones, nutrients, and oxidant stress.FASEB J8:174–181 [DOI] [PubMed] [Google Scholar]
• 18.Gomez-Gallego F, Garrido-Pertierra A, Bautista JM2000. Structural defects underlying protein dysfunction in human glucose-6-phosphate dehydrogenase A(−) deficiency.J Biol Chem275:9256–9262 [DOI] [PubMed] [Google Scholar]
• 19.Luzzatto L, Usanga FA, Reddy S1969. Glucose-6-phosphate dehydrogenase deficient red cells: resistance to infection by malarial parasites.Science164: 839–842 [DOI] [PubMed] [Google Scholar]
• 20.Eppes RB, Lawrence AM, McNamara JV, Powell RD, Carson PE1969. Intravenous glucose tolerance in negro men deficient in glucose-6-phosphate dehydrogenase.N Engl J Med281:60–63 [DOI] [PubMed] [Google Scholar]
• 21.Billis AG,Xefteris ED,Ioannides PJ,Papastamatis SC1970. Abnormal glucose tolerance in favism.Diabetologia6:425–429 [DOI] [PubMed] [Google Scholar]
• 22.Saha N1979. Association of glucose-6-phosphate dehydrogenase deficiency with diabetes mellitus in ethnic groups of Singapore.J Med Genet16:431–434 [DOI] [PMC free article] [PubMed] [Google Scholar]
• 23.Meloni T, Pacifico A, Forteleoni G, Meloni GF1992. G6PD deficiency and diabetes mellitus in northern Sardinian subjects.Haematologica77:94–95 [PubMed] [Google Scholar]
• 24.Wan GH, Tsai SC, Chiu DT2002. Decreased blood activity of glucose-6-phosphate dehydrogenase associates with increased risk for diabetes mellitus.Endocrine19:191–195 [DOI] [PubMed] [Google Scholar]
• 25.Niazi GA1991. Glucose-6-phosphate dehydrogenase deficiency and diabetes mellitus.Int J Hematol54:295–298 [PubMed] [Google Scholar]
• 26.MonteAlegre S,Saad ST,Delatre E,Saad MJ1991. Insulin secretion in patients deficient in glucose-6-phosphate dehydrogenase.Horm Metab Res23:171–173 [DOI] [PubMed] [Google Scholar]
• 27.Laybutt DR, Kaneto H, Hasenkamp W, Grey S, Jonas JC, Sgroi DC, Groff A, Ferran C, Bonner-Weir S, Sharma A, Weir GC2002. Increased expression of antioxidant and antiapoptotic genes in islets that may contribute toβ-cell survival during chronic hyperglycemia.Diabetes51:413–423 [DOI] [PubMed] [Google Scholar]