Induction and Activities of Pyruvate Dehydrogenase and α -ketoglutarate Dehydrogenase in Type 2 Diabetic Patients and Therapy with Vitamin B1

This work carried out collaboration between both authors. Author SSA designed the study, wrote the protocol, enrolled the patients, arranged for the placebo and B1, conducted the research and wrote the first draft of the manuscript. Author SR played a supportive role throughout the research, assisted in the patient enrollment, sampling, the experimental process and manuscript revision. Both authors read and approved the final manuscript. ABSTRACT only for baseline estimation of the parameters. Results: The enrolled patients with type 2 diabetes showed decreased activities of mononuclear enzymes as compared to the healthy controls. Q-PCR study showed that the expression levels of the genes encoding PDE1 β and α KGDE1k were significantly reduced in the patients with type 2 diabetes as compared to the healthy controls. Thiamine therapy resulted in significant increases in the expression of PD E1 β and α KGD E1 genes, which persisted even 2 months after the washout. Thiamine therapy therefore resulted in significant increase in activities of these enzymes and incremental activity persisted into the washout period. Conclusion: These results indicate that the thiamine acts as an inducer in the expression of mononuclear PDH and α KGD thus enhancing their activities in the type 2 diabetes patients with incipient nephropathy.


INTRODUCTION
A variety of micronutrient approaches have been suggested as therapeutic interventions in patients with type 2 diabetes [1][2][3]. Thiamine (vitamin B1) is a water-soluble vitamin and an essential normal dietary component [4]. Thiamine deficiency exists in both type 1 and type 2 diabetes patients [5] and in patients with type 2 diabetes and microalbuminuria [6][7][8][9]. Thiamine and benfotiamine therapy prevents the development of microvascular complications in experimental diabetes [10][11] and could possibly also have a beneficial role to play in the treatment of type 2 diabetes mellitus [5].
The pyruvate dehydrogenase multienzyme complex consists of PDE1 (pyruvate dehydrogenase: PDH), PDE2 (dihydrolipoyl transacetylase) and PDE3 (dihydrolipoyl dehydrogenase) subunits. TPP on binding to PDH initiates a series of reactions resulting in oxidative decarboxylation of pyruvate to yield acetyl-CoA and NADH [12]. PDH serves as the gate keeper anzyme that strategically links glycolysis, Krebs cycle and lipogenic pathways [13].
The α-KGDE1k (OGDH: oxoglutarate dehydrogenase) also functions as a complex in association with E2k (dihydrolipoyl transsuccinylase) and E3k (dihydrolipoyl dehydrogenase) subunits. Binding of thiamine to OGDH is required for the complex to convert oxoglutarate, a key intermediate in the krebs cycle, to succinyl co A and generating NADH and CO 2 in an irreversible reaction [14].
A dysfunctional PDH complex, with an unchanged total amount, is found in animal tissues in experimental diabetes, obesity and in skeletal muscles of diabetic patients [15][16][17][18]. In diabetes, thiamine dependent megaloblastic anaemia and sensorineural deafness associated with deficient α-KGD activity has also been reported [19]. Reduction in activities of thiamine dependant enzymes transketolase, α-KGD and PDH are thought to be responsible for the tissue damage and impaired cell function that accompany thiamine deficiency [20]. Specially, depreciation in oxidative decarboxylation of the α-keto acids through loss of activities of the thiamine dependent enzymes would reduce ATP synthesis, leading to cellular acidosis [21]. Thiamine deficiency in vitro leads to enhanced degradation of apoenzymes leading to loss of activity of the TPP requiring enzyme. [22] Thiamine may also have a role in the expression of genes that encode the TPP requiring enzymes [23][24] and its replenishment by thiamine therapy improved both activity and expression [25].
Experimental diabetes is associated with thiamine deficiency characterised by a marked decrease in plasma thiamine concentration and decreased activity and expression of the thiamine-dependent enzyme transketolase in renal glomeruli [6]. In this model correction of decreased plasma thiamine concentration countered the adverse effects of hyperglycaemia by activation of the reduced pentosephosphate pathway, providing an alternative route for disposal of accumulating glycolytic intermediate (fructose-6-phosphate and triosephosphates) preventing metabolic dysfunction in renal glomeruli and development of early stage nephropathy [26].
Diabetic nephropathy develops progressively over 5-40 years of diabetes and research shows patients to have low thiamine levels [6]. Correction of low levels of plasma thiamine in diabetic patients and related metabolic responses may regress the microalbuminuria and prevent early decline in GFR [26]. Based on the above evidence a novel strategy to counter biochemical dysfunction linked to the development of diabetic nephropathy is highdose thiamine therapy. In our previous published pilot study, we evaluated the effect of oral high dose supplements of thiamine on urinary albumin excretion (UAE), a marker of early-stage diabetic nephropathy, in type 2 diabetic patients with microalbuminuria and an improvement in microalbuminuria and transketolase levels [6,9]. AKDH and PDH are both thiamine dependent enzymes as well, required for effective utilization of the glucose intracellularly in all tissues including the kidney and in generation of ATP for all processes including renal function. Their activities and expression could be reduced in type 2 diabetics based on the evidence stated previously leading to dysfunction of glomerular epithelial cells, podocytes and tubular epithelial cells causing loss of thiamine resulting in incipient nephropathy and microalbuminuria. [27] This in turn would further compound the loss of thiamine from the kidneys and a vicious cycle of lowering puruvate dehydrogenase and alphaketoglutarate dehydrogenase function and expression associated with worsening diabetic nephropathy would ensue.
This study, based on a double blinded placebo controlled clinical trial reports the effects of high dose thiamine therapy on the activities and expression of PDE1 and α-KGDE1 in patients having type 2 diabetes with microalbuminuria.  1,2006. Stringent inclusion and exclusion, and randomization and treatment procedure of the patients and controls was done as described previously [6]. The trial duration was three months therapy and two months washout period. The medicine was given for three months only. Out of the enrolled diabetic individuals and controls, 40 patients and 20 controls completed the 3 month thiamine/placebo administration and 2 month follow up period for this study. Baseline data of the recruited patients and controls has been presented in our previously published data. [6] At baseline and throughout the study, nine patients in the thiamine-treatment group and three in the placebo group were receiving insulin therapy (p<0.05). There were no other significant differences in the proportions of patients receiving therapeutic agents in the thiaminetreatment and placebo groups. One patient achieved glycaemic control by diet only; all others received therapy with hypoglycaemic agents (sulfonylureas, metformin and thiazolidinediones). The age and gender of healthy controls was comparable to the diabetic patients.

Sampling
Fasting Blood samples were obtained from the patients and healthy individuals at baseline, after 3 months therapy and 2 months washout. These were processed for the biochemical analysis as well as isolation of erythrocytes and peripheral blood mononuclear cells (PBCMs), which were used for pyruvate dehydrogenase (PDE1) and αketoglutarate dehydrogenase (α -KGD) assays and gene expression studies. 24 hr urine collections were also made for determination of microalbuminuria.

Mononuclear Cell Separation
The PBMCs were isolated from EDTA treated blood samples, which were equally distributed into 2 sets of 15 ml falcon tubes at room temperature on a ficoll-paque(GE Healthcare) gradient [24]. The cells were recovered from the interface and subjected to short washing steps with Hanks balanced salt solution to remove any platelets or plasma [28]. All samples were processed fresh within 2-3 hours after collection and maintained on ice until the enzyme assays the same day.

Enzyme Assays in Mononuclear Cells
All routine reagents used in the assay were of analytical grade and were either procured from Sigma (Germany), Fluka or Reidel-de-Haen(Germany) . Each assay was conducted in a microplate reader (Biotek 808IU) using a 96 well quartz plate (Hellma). All test samples were run in duplicate with (sample minus substrate) serving as controls.

PDE1 assay
The method used for the PDE1 assay was the same reported previously [29]. However, in order to adapt this method for the microplate reader the reaction mixture volumes were reduced as mentioned below. This protocol describes a coupled assay to measure the pyruvate dehydrogenase activity. In this citrate synthase reaction was applied for the assay of two acetyl CoA producing enzymes pyruvate dehydrogenase and acetyl-CoA synthetase. Reaction were initiated by the addition of 0.1 mg total protein to an otherwise complete reaction mix of 12.5 µl 0.2 M sodium pyruvate, 12.5 µl 4 mM sodium coenzyme A, 12.5 µl 40 mM NAD, 25 µl 10 mM MgCl 2 ,50 µl 0.25 M Tris-HCl Buffer(pH 8.0), 69 µl deionized water, 12.5 µl 200 mM DTT, 25 µl 25 mM oxaloacetate,12.5 µl 0.05 g 5,5-dithiobis(2-nitrobenzoic acid) (DTNB) in 10 mL 100% ethanol and 5 µl Citrate synthase (250 U/mL). Final volume was 300 µl.This method had a run time of 100 s at 30°C at a wavelength of 415 nm. Activity was measured in U/mg protein. One unit (U) of pyruvate dehydrogenase E1 activity is defined as the amount of enzyme required to produce 1.0 µmole of acetyl CoA in one minute.

α-KGD assay
In this assay as well the methodology remained the same as previously reported [22]. However adaptation was made for the microplate reader by reducing the reaction volume as given below. Reactions were initiated by the addition of 0.1 mg total protein to an otherwise complete reaction mix of 43 µl 50 mmol/L MOPS (pH 8.0), 72 µl 1.2 mmol/L MgCl 2 , 72 µl 1.2 mmol/L CaCl 2 , 10 µl 0.16 mmol/L coenzyme A, 37.5 µl 6 mmol/L αketoglutarate, 33 µl 0.1 mmol/L NAD, 1.6 µl 0.5 g/L Triton X-100 and 9.5 µl 0.04 mmol/L rotenone. Final volume was 300 µl. The formation of NADH, which is directly proportional to α-KGDH activity, was measured at 340 nm wavelength at 30°C. One unit of alphaketoglutarate dehydrogenase activity was defined as that which converts 1.0µmole of β NAD to βNADH per minute at pH 7.4 at 30°C in the presence of saturating levels of coenzyme A. Protein concentration was determined using the Bradford method [23].
The intra-CV of PDH assays was 1.42% and inter-CV assay was 2.5%. Intra-CV for AKDH assay was 1.05% and inter-CV was 2.0%. The lowest detection limit for assay of aKDH was 2 nmol/mg protein per minute and for assay of PDH was 39 nmol/mg protein per minute.

Quantitative RT-PCR Analysis
Total RNA was isolated from the whole blood using Tri reagent LS (TS 120) Gentra (USA) following standard protocol. DNase-treated total RNA (2 µg) was reverse transcribed with the Revertaid minus Strand cDNA synthesis Kit (Fermentas) and semiquantitative PCR was performed to verify the amplification of the thiamine dependent enzyme genes. Primers were designed using Primer 3 [30] and UCSC Genome Bioinformatics [31] as shown in Table 1. These were synthesized by Eurofins MWG/ Operon (USA). Amplification was done in ABI Biocycler using the conditions: initial denaturation at 94°C (5min), annealing at 62°C (30 s) and extension at72°C(40 s) for 35 cycles followed by final extension at 72°C for 7 min. The relative gene expression analysis was done by using SDS 3.1 software provided by ABI. Each real time PCR assay was performed in duplicate with glyceraldehyde 3 phosphate dehydrogenase (GAPDH) gene as a control for normalization. These procedures were performed as per standard referred protocols and methods mentioned in the kits of Fermantas company.

Power calculation
The primary endpoint was urinary albumin excretion. Group CV values and intervention effects of 30% were assumed, similar to previous studies. For power =0.8 and α<0.05, patient group size was 17. Patient groups of 20 were employed to allow for noncompliance to therapy.

Statistical analysis
The data collected was analysed for statistical significance and correlation testing using statistical package SPSS15 Chicago, (IL, U.S.A). A normality check was performed for the entire data using the Kolmogorov Smirnov test. Following that Anova'(Analysis of Variance) statistical package was used for Parametric data when groups are found to be normal. While Kruskall-wallis test was used when non-normality was found. Significance of difference between mean, median analytes of thiamine and placebo groups were determined by using students t test and Mann Whitney U test respectively. Nonparametric correlation analysis of study groups were calculated by Spearman's rho statistic. While Real Time PCR results were also analyzed for significance using Anova' Analysis of Variance test.

RESULTS
In the present research work, only two patients reported gastrointestinal distress and three complained of tachycardia in the thiamine treated group and had to discontinue intervention. Tachycardia has also been previously been reported and might also have a reason due to sensitivity to thiamine in these patients and lesser dose could have been more acceptable. However due to requirements of 300 mg daily dose of thiamine in this clinical trial these patients were omitted from the trial completely. Two patients infact reported improvement in skin, hair and nail strength.
The baseline urinary albumin excretion levels in mg /24 hrs was significantly higher in the type 2 diabetic patients in both thiamine and placebo assigned groups. The thiamine group showed 53.86±22.62 mg/24 hrs and placebo group showed 55.74±23.71 mg/24 hrs as compared to controls (7.96±5.07 mg/24 hrs). The baseline HbA1C levels of the thiamine allocated diabetics was 9.2±1.3%, while the patients randomized to the placebo arm had mean HbA1C levels of 8.82±1.8% as compared to control (5.6±0.39%).
Detailed biochemical profiles were maintained for the enrolled type 2 diabetic patients with microalbuminuria at baseline, after 3 months post therapy and at 2 month washout period and have been published previously [6]. These revealed certain beneficial effects of high dose 300 mg/day thiamine therapy and absence of adverse effects. Markedly lower baseline median plasma thiamine concentration of diabetic patients (7.5 nM) was present compared to normal range of normal healthy human subjects (944.6 -93.7 nM). Thiamine treatment for 3 months increased median plasma thiamine concentration 10 fold and urinary thiamine excretion 29 fold. It importantly caused a regression of microalbuminuria to normal albumin levels in 35% of the patients .A decrease in mean glycated haemoglobin levels by 1.4% was also observed two months after washout of thiamine.

Primer name
Sequence

Expression Analysis of PDE1 and α-KGDE1
The mean baseline expression levels of mononuclear PDE1β gene in the thiamine group was 0.65 fold and 0.85 fold in the placebo group. While baseline α-KGDE1 gene expression levels in thiamine treated group was 0.60 fold and 0.71 fold in the placebo treated diabetics as compared to normal healthy controls. (p=0.007) (Fig. 1). Statistically the expression levels of mononuclear PDE1 and α-KGDE1 genes were significantly reduced in thiamine and placebo treated diabetics as compared to normal healthy controls. High dose thiamine therapy for 3 months resulted in significantly higher PDE1 expression of 2.9 fold, p<0.001 than baseline, an increase of about 200%, which was maintained after washout at 3.02 fold, p<0.001. The PDE1 gene, in the placebo group showed non significant changes in expression of 1.15 fold, p=0.05 after 3 months therapy and similarly after 2 months washout to 0.96 fold, p=0.055 (Fig. 2). The results indicated that the expression levels of PDE1β gene were significantly high in thiamine treated diabetics as compared to placebo treated diabetics patients. High dose thiamine therapy for 3 months also resulted in a statistically significant increase in expression levels of the alphaketoglutarate dehydrogenase gene by 2.65 fold (p<0.001), which remained statistically significant, at 2.72 fold (p<0.01) even after washout. While in the placebo group statistically non significant changes in fold expression of the oxoglutarate dehydrogenase gene were observed at 0.91 fold after therapy, and 1.09 fold (p=0.065) after washout (Fig. 3). diabetic patients and less but visible (band9) (baseline) of placebo treated type 2 diabetics as compared to (band1 -normal controls). Following thiamine therapy expression of AKGDE1 increased significantly (band4) and even more in washout (band5) as compared to its baseline (band 3). In comparison there was slight decrease in AKGDE1 expression seen in placebo treated patients at post therapy (band10) as compared to baseline (band 9) and slightly increased expression at post washout (band 11) GAPDH was maintained as internal control in bands (2,(6)(7)(8)(12)(13)(14) Comparison of expression of AKGDE1 gene expressed as relative fold induction in type 2 thiamine treated versus placebo treated patient's quantified using Real time PCR with GAPDH as internal control.

DISCUSSION
In Type 2 diabetes, plasma thiamine levels are markedly reduced and there is enhanced washout of thiamine as well. The aKDH and PDH are both thiamine dependent enzymes and therefore their activities were found to be correspondingly reduced in type 2 diabetics because of lowered thiamine levels. Our study shows that thiamine reduction impacts both activities and expression of the enzymes adversely. These enzymes are located in all the cells and a part of energy cycles. Therefore, in diabetes dysfunction of a varying range of cells is observed such as diabetic retinopathy, neuropathy and nephropathy.
Our previous study also showed that thiamine therapy led to the reduction in HbA1C which was representative of the improvement in overall long term glucose regulation, as improved functioning of the thiamine dependent enzymes aKDE1 and PDH was observed. These enzymes are obviously vital for glucose handling.
Additionally thiamine intervention also resulted in decreased microalbuminuria. This might have been a consequence of increasing plasma concentration of thiamine reversing dysfunction of glomerular endothelial cells, podocytes and tubular epithelial cells improving glomerular and tubular structure and function and reducing low grade vascular inflammation observed as decreased urinary albumin excretion in type 2 diabetic patients of our study.

Thiamine and its Effect on PDE1
Thiamine may regulate the expression of genes that encode the enzymes that utilize ThDP [20].
As previously reported by us these diabetics with microalbuminuria were also thiamine depleted [7] and a possibility of a change in pyruvate dehydrogenase and α-ketoglutarate dehydrogenase activities and the expression levels of their genes also needed to be investigated. Activity of the mononuclear pyruvate dehydrogenase (PDE1) subunit in diabetic patient was <54% than that of normal healthy individuals. Corresponding expression analysis of the mononuclear PDHE1β gene in diabetic patients revealed it to be significantly lowered to 75% of that in normal controls. The result thus indicated that the activity of pyruvate dehydrogenase enzyme and expression levels of PDE1β gene was significantly increased in mononuclear cells of thiamine treated diabetic patients. The result is also supported by the previous data which showed an enhanced degradation of apoenzymes generated during thiamine deficiency may explain the loss of activity of the TPP requiring enzymes [23]. As suggested earlier thiamine deficiency in vitro decreased activities of TK and PDH and its replenishment by thiamine therapy improved both activity and expression [22]. Our previously published results indicated that high dose thiamine therapy for 3 months significantly enhanced both mononuclear transketolase activity and gene expression in type 2 diabetic patients with incipient nephropathy [20].

Thiamine and Its Effect on α-KGDE1
In our present study, alphaketoglutarate dehydrogenase activity was reduced by nearly 82% in diabetics respectively as compared to normal controls. While α-KGD E1 gene expression in diabetics was significantly reduced to 65% of that in normal controls. The impact of thiamine therapy on α-KGD E1 gene expression was also evident as a highly significant increase 2.65 fold increase in the mRNA levels was observed persisting into washout at levels of 2.72 fold. Therefore thiamine administration increased both enzyme activity and expression of α-KGD E1, in PBMCs of diabetic patients.
The activities of PDE1, and α-KGDE1 Subunit were even higher in wash period due to reason as thiamine has a long biological half life (9-18 days) so even though the plasma concentrations and urinary excretion of thiamine in the thiamine treatment arm returns to baseline after two months washout period [32]. It is likely that increased tissues levels of TPP (thiamine pyrophosphate), activities of thiamine dependent enzymes and related pharmacological responses remained above baseline for at least one biological half life of thiamine in to the washout period of the patients in the thiamine treatment group. No adverse effects of high activities of both PDE1, and α-KGDE1 were noted [33].
This result was different from that of the cultured thiamine responsive megaloblastic anaemia patients (TRMA) lymphoblasts cells and in the α-KGD E1 enzyme activity and mRNA levels of these cells, where no correlation between total protein level/or activity and mRNA could be found [22]. The reason for this difference could have been in vitro limiting conditions and that being a tissue culture study [22]. The current study was obviously different in that the mononuclear cells were extracted from the diabetic patients who were orally administered thiamine or placebo and an in vivo analysis of enzyme activity and expression conducted in freshly extracted samples.
Our research on activity and expression analysis of PDE1 and alphaketoglutarate dehydrogenase in mononuclear cells revealed continued increment of both activity and expression levels of these enzymes after 3 months of high dose thiamine therapy and its persistence even 2 months after stoppage of therapy. The mean baseline plasma thiamine levels of the diabetic patients were found to be directly proportional to their enzyme activities and expression levels i.e both showed improvement with thiamine therapy [6][7][8][9].

CONCLUSION
This clinical intervention trial on 300 mg/day B1 therapy was the first randomized, double blinded, placebo controlled and pilot scale project for a period of 5 months to study the effect of high dose thiamine therapy on biochemical profile and activities of thiamine dependent enzymes on diabetics in the Pakistani population. The trial was also pioneering on the subject of diabetic nephropathy and the effect of thiamine supplementation on it. Baseline activity and expression levels of mononuclear PDE1 and α-KGDE1 in diabetic patients are significantly lower in diabetics as compared to healthy individuals. High dose thiamine therapy of 300 mg/ day significantly improved PDE1 and α-KGDE1 activity and gene expression in diabetic patients and caused a regression of microalbuminuria to normal albumin levels in 35% of the patients. Thus, further exemplifying the advantages of thiamine therapy in its beneficial and cooperative role in enhancing Krebs cycle glucose metabolism through PDE1 and α-KGDE1. These findings however deserve further examination and extension in larger clinical intervention studies of diabetic patients. Thiamine therapy may prove to be a valuable adjunct in treatment of diabetes and its complications.