Metabolic and Structural Role of Thiamine in Nervous Tissues

Abstract

In the literature, previous descriptions of the role of thiamine (B1 vitamin) focused mostly on its biochemical functions as a coenzyme precursor of some key enzymes of the carbohydrate metabolism. This report reviews recent developments on the metabolic and structural role of thiamine, e.g., the coenzyme and noncoenzyme functions of the vitamin. Taking into account analysis of our experimental data relating to the effects of thiamine deficiency on developing central nervous system (CNS) and data available in literature, we seek to establish a clear difference between the metabolic and structural role of thiamine. Our experimental data indicate that the specific and nonspecific effects express two diametrically diverse functions of thiamine in development: the nonspecific effects show up the metabolic consequences of thiamine deficiency resulting in apoptosis and severe cellular deficit; inversely, the specific effects announced the structural consequences of thiamine deficiency, described as cellular membrane damage, irregular and ectopic cells. The review highlights the existence of noncoenzyme functions of this vitamin through its interactions with biological membranes.

Introduction

The brain uses glucose as a primary fuel for energy generation. Glucose gains entry into the brain by facilitated diffusion across the blood–brain barrier (Rao and Seaquist 2006). Roughly, 30% of the glucose absorbed by the brain undergoes a complete oxidation through the tricarboxylic acid cycle (Siebert et al. 1986). Three enzymatic systems, essential for the cerebral metabolism of glucose, depend on thiamine (B1 vitamin): The mitochondrial pyruvate and α-ketoglutarate dehydrogenases complexes and the cytosolic transketolase (Martin et al. 2003). These three enzymes use, as cofactor, the thiamine pyrophosphate (TPP) which accounts for 80% of total thiamine present in nervous tissues (Ishii et al. 1979). In the literature, the major part of actions of the thiamine described in nervous tissues is limited to those metabolic aspects which show up the co-enzymatic function of the vitamin (Butterworth 1987). However, literature that takes an interest in the structural role of thiamine is growing. So, thiamine interferes with the membrane structure and function (Tanaka and Cooper 1968; Itokawa et al. 1972; Goldberg et al. 2004), acts against agents-induced cytotoxicity (Bâ et al. 1996; Aberle et al. 2004) and fixes membrane sites (Czerniecki et al. 2004; Spector and Johanson 2007). Those observations illustrate the noncoenzymatic function of the vitamin. We reported the metabolic to be different from the structural role of thiamine in nervous tissues, constantly (Bâ et al. 1996, 2005; Bâ 2005). This review attempts through discussion to differentiate the metabolic from structural role of thiamine, in other words, to separate the coenzyme and noncoenzyme functions of the vitamin.

Studies of the thiamine metabolism variations are often achieved according to chronic alcoholism. Chronic alcoholism weakens the metabolism of thiamine by a reduction of the rate of enzymes dependent on the vitamin (Butterworth 1989). In addition, the Fetal Alcohol Syndrome (FAS) was commonly described in children born to severely alcoholic mothers (Jones and Smith 1973). The Intrauterine Growth Retardation (IUGR), a frequent concomitant of FAS, was suggested to result from the effects of ethanol-induced thiamine deficiency (Rœcklin et al. 1985).

In our studies, we investigated in the rat, experimental models of FAS (Bâ et al. 1996, 1999) and IUGR (Bâ 2005; Bâ et al. 2005). For further understanding of the role of the thiamine in nervous tissues, we have studied the effects of thiamine deficiency on the developing central nervous system (CNS), (Bâ 2005; Bâ et al. 2005). Thus, thiamine deficiency was induced during three main periods of the rat CNS ontogenesis. Females were fed with a thiamine deficient diet during the periods of gestation and lactation; fetuses were exposed alternatively to pre-, peri- or postnatal thiamine deficiency. Thereafter, the effects of maternal thiamine deprivation were assessed on the development of psychomotor and sensory functions in the offspring: seven different developmental abilities (exploratory activity, emotional reaction, hind paws lifting reflex, wire grasping times, crawling and leap execution latencies, and nociception) were recorded in the offspring from the 10th to the 45th postnatal day (Bâ and Seri 1993, 1995), to assess the specific and nonspecific effects of maternal thiamine deficiency on fetal brain (Bâ 2005).

Control dams (C) received regular diet containing thiamine during gestation and lactation. B1 lack-induced anorexia in thiamine deficient dams (TD) was reproduced in pair-fed dams (PF) which received a reduced quantity of regular diet matched to the TD dams’ consumption. The malnutrition generated by every pattern of thiamine deficiency was controlled by its own PF group. For each pattern of thiamine deficiency, we made multiple comparisons between C-TD-PF data recorded on the corresponding offspring (Bâ 2005): Specific effect, e.g., effect caused by thiamine lack per se was diagnosed if C ≠ TD ≠ PF; nonspecific effect, e.g., effect of malnutrition that comes with a thiamine deprivation was identified when C ≠ TD = PF. On the 45th postnatal day, histologic studies were done on the brains of the offspring and the structure of the hippocampus was examined (Fig. 2a). Average reduction rate in nuclear size and density was assessed following each pattern of thiamine deficiency, within the dentate gyrus and the fields CA4, CA3 and CA1 of the hippocampus (Fig. 2a), (Bâ et al. 2005).

Fig. 2.

(a) Parasagittal sections of the left ventral hippocampus in 45-day-old rats. Paraffin 10 μm thick sections were stained with a combination of hematoxylin–eosin and indigo carmine; sections of the midtemporal hippocampus were assessed. (a) The arrowhead demarcates the CA3 region and the arrow denotes the CA1 region; field CA4 is enclosed by broken lines; bar = 100 μm. P: pyramidal cell layer; G: granule cell layer; sm: stratum moleculare; sr: stratum radiatum; so: stratum oriens. (b–d) Effects of developmental thiamine deficiencies on the hilar CA3 pyramidal cells of the hippocampus. (b) Controls; (c) Prenatal exposed pups; (d) Postnatal exposed; bar = 20 μm

Developmental Thiamine Deficiency

Figure 1 summarizes profiles from four typical effects of developmental thiamine deficiencies on the CNS, gathered through our previous studies (Bâ 2005; Bâ et al. 1999, 2005), e.g., specific and nonspecific effects, cellular death and atrophy, which were statistically assessed from prenatal to postnatal periods.

Fig. 1.

Profiles of four typical effects of developmental thiamine deficiency on the CNS. Extract from data published elsewhere (Bâ 2005; Bâ et al. 1999, 2005)

Our results show that the percentage of functions altered by the specific lack of B1 vitamin increases from prenatal (28.57%) to perinatal (42.85%) and to postnatal periods (57.14%). Inversely, the percentage of functions altered by the nonspecific effects of thiamine deficiency decreases from prenatal (60.14%) to perinatal (35.31%) and postnatal periods (14.32%), (Fig. 1), (Bâ 2005).

It appears also that the rate of cellular death, resulting from thiamine deficiency, decreases progressively from prenatal (39.97%) to perinatal (26.75%) and postnatal periods (14.92%), whereas cellular atrophy does not seem to show any perceptible variations from prenatal (7.97%) to peri- (12.95%) and postnatal (10.77%) periods (Fig. 1), (Bâ et al. 2005).

Statistical analysis shows that the specific and nonspecific effects of thiamine deficiency, on the functional development of the CNS, are negatively correlated (R = −0.9988, t = 20.39, P < 0.025). Cellular death exhibits a positive correlation with the nonspecific effects of thiamine deficiency (R = 0.9998, t = 49.99, P < 0.01) and negative with the specific effects of that deficiency (R = −0.9994, t = 28.85, P < 0.025), during development. Cellular atrophy does not show any significant correlation with either the nonspecific effects (R = −0.5997, t = 0.75, P > 0.1) or the specific effects of thiamine deficiency (R = 0.5606, t = 0.67, P > 0.1).

For a better illustration of our goal, we compare in Fig. 2c and d, the effects of pre- and postnatal thiamine deficiencies on qualitative morphological alterations assessed from the hilar CA3 pyramidal cells of the hippocampus (Fig. 2a). Figure 2c shows that prenatal thiamine deficiency was characterized by a singular deficit of CA3 pyramidal cells comparative to regular diet (Fig. 2b), without any direct impact on nucleus or cytoplasm shape. On the contrary, postnatal thiamine deficiency exhibited more cornered, irregular and sparse pyramidal cells in the hippocampal CA3 field (Fig. 2d), comparative to regular diet (Fig. 2b). Figure 2d shows typical features of severe postnatal thiamine deficiency with characteristic breakings on cellular membranes illustrated by more cornered than pyramidal shaped cells in the hippocampal CA3 field. Moreover, postnatal thiamine deficiency resulted in a significant increase in pro-apoptotic-related morphological and biochemical changes characterized by cells containing condensed chromatin (Fig. 2d), (Kerr et al. 1994). It is obvious that prenatal thiamine deficiency kills cells metabolically, while postnatal thiamine deficiency damages cells structurally.

How to interpret the specific and nonspecific effects of thiamine lack on the developing CNS?

Metabolic Role of Thiamine

Our studies indicate that specific and nonspecific effects express two diametrically diverse functions of thiamine in development (Fig. 1). The percentage of functions altered by the nonspecific effects of thiamine deficiency decreases from prenatal (60.14%) to perinatal (35.31%) and postnatal periods (14.32%), (Bâ 2005); these periods overlap the moments of cellular migration, proliferation and developmental apoptosis that is active shortly after that (Bâ et al. 2005). Concomitantly, the rate of cellular death decreases progressively from prenatal (39.97%) to perinatal (26.75%) and postnatal periods (14.92%), (Fig. 1), (Bâ et al. 2005). Our present results confirmed a significant positive correlation between cellular death and the nonspecific effects of developmental thiamine deficiency. The nonspecific effects would be assigned to the undernourishment that comes with the thiamine deficiency. Consequently, developmental thiamine deficiency-induced cellular death would be caused by the nonspecific effects of that deficiency, e.g., the malnutrition that comes with the vitamin lack (Bâ et al. 2005). The nonspecific effects should be expressed by the metabolic role of thiamine. How thiamine (B1 vitamin) deficiency induces cellular death?

To explain mechanisms of cellular damage and death, most of investigations have focused on general metabolic effects of thiamine deficiency like ultimate actions of B1 vitamin lack. For instance, metabolic reduction of thiamine-dependent enzymes should be the starting box of any peripheral and central neuropathies (Butterworth 1989). However, following thiamine deficiency, thiamine-dependent enzymes, i.e., α-ketoglutarate and pyruvate dehydrogenases complexes, as well as non-thiamine-dependent enzymes, i.e., succinate and malate dehydrogenases of the tricarboxylic acid cycle are reduced in the brains of mice (Bubber et al. 2004). Another example is the Wernicke’s encephalopathy diagnosed in chronic alcoholics, described as a constellation of selective neuropathological lesions occurring in the brain of patients in which thiamine deficiency was indexed as the causal factor of selective neuronal loss (Navarro et al. 2005). However, the cause of Wernicke’s encephalopathy is thiamine deficiency as a result of any nutritionally deficient state, and the disease could not be confined only to alcoholics (Donnino et al. 2007). In addition, thiamine deficiency induces quantitative, distinct inflammatory responses and oxidative stress in vulnerable and nonvulnerable regions that lead to cellular loss (Karuppagounder et al. 2007).

Several basic mechanisms underlying thiamine deficiency-induced apoptosis and neurodegeneration have been reported recently. These mechanisms include: (i) Compromised energy production and lactic acidosis (Martin et al. 2003; Navarro et al. 2005); (ii) excess levels of free radicals and oxidative stress (Gibson and Blass 2007; Frank et al. 2008); (iii) changes in microglia making a start on neurodegeneration (Ke and Gibson 2004); (iv) the voltage-dependent K + membrane conductance alteration (Oliveira et al. 2007); (v) cellular membrane breaking (Bâ et al. 1996); (vi) the mitochondrial caspase 3-mediated apoptosis (Chornyy et al. 2007); (vii) translocation of amyloid precursor protein C-terminal fragments into the nucleus (Karuppagounder et al. 2008). Seemingly, most of these mechanisms are membrane-mediated effects of thiamine deficiency. In addition, experimental thiamine deficiency is a model of impaired oxidative metabolism (Ke and Gibson 2004). Thiamine-dependent mitochondrial dehydrogenase complexes produce oxygen free radicals (Gibson and Blass 2007). Oxygen-dependent free radical is generated during substrate turnover in the krebs cycle and its reaction with molecular oxygen results in the continuous production of reactive oxygen species (ROS) under aerobic conditions (Frank et al. 2008). Overproduction of ROS (arising from mitochondrial electron-transport chain) results in oxidative stress which damages cellular structures, including lipids and membranes, proteins, and DNA (Valko et al. 2007). Indeed, thiamine can act as a free radical scavenger (Gibson and Blass 2007). Free-radicals derive from mitochondrial dysfunction (Mancuso et al. 2007). Alteration of mitochondrial TPP transporter causes neural tube closure defect and results in Amish Lethal Microcephaly (Lindhurst et al. 2006). Thiamine deficiency which provokes mitochondrial dysfunction with increased glycolysis, increased lactate dehydrogenase and focal lactate accumulation (Navarro et al. 2005), results in a significant release of free radicals which, in turn, diminishes reduced glutathione (GSH) and other defense systems against oxidative stress at the origin of major neurodegenerative diseases. For instance, Prion diseases or Transmissible Spongiform Encephalopathies (TSE) are connected with a congenital and dramatic loss of antioxidant defense proteins in the brain (Brown 2005). Another metabolic disease related to thiamine and oxidatif stress is diabetes. Complications in diabetes mellitus are partially mediated by enhanced formation of reactive oxygen species (Schmid et al. 2008), similarly to thiamine deficiency. Benfotiamine, a lipophilic derivative of thiamine with better bioavailability, alleviates diabetes-induced cerebral oxidative stress (Wu and Ren 2006). Benfotiamine prevents oxidative stress-induced DNA damage (Schmid et al. 2008) and corrects defective replication in human umbilical vein endothelial cells cultured in high glucose (Pomero et al. 2001). The thiamine derivative prodrug prevents micro- and macrovascular endothelial dysfunction and the oxidative stress accompanying that dysfunction in type 2 diabetes (Stirban et al. 2006). Both thiamine and benfotiamine correct increased apoptosis due to high glucose in cultured vascular cells (Beltramo et al. 2004). These observations suggest that oxidative damage is critical to the pathogenesis of thiamine deficiency (Gibson and Blass 2007). From the analysis of the implication of B1 vitamin lack in different metabolic diseases, we propose an integrated mechanism for the metabolically thiamine deficiency-induced cellular death which starts with extended free radicals damage in the brain. Increased free radicals production related to thiamine deficiency launch cellular membrane damage, including lipoperoxydation (Bâ et al. 1999; Valko et al. 2007), alteration of neuron ion channels and transporters and microglia changes. Persistent free radicals produced by severe thiamine deficiency break cellular membrane (Bâ et al. 1996, 1999) which signals a cascade of cellular death pathways via intracellular messengers, like intracellular caspase 3-mediated apoptosis, toward the nucleus. Caspase 3 had been identified among the first components of the programmed cellular death machinery (Zakeri and Lockshin 2008). Subsequent nucleolysis machinery includes, among others, translocation of proteins carboxy-terminal fragments into the nucleus of neurons (Karuppagounder et al. 2008). The cytosolic pH changes by focal lactate acidosis should be an endogenous factor to launch and/or to amplify the cascade of cellular death pathways.

Structural Role of Thiamine

Our results also indicate a negative correlation between cellular death and the specific effects of thiamine deficiency, which increase over postnatal days. These observations suggest that specific effects bribe alterations in membrane processes which develop postnatally, e.g., cellular differentiation, synapses formation, axonal growth, and myelinogenesis (Bâ 2005). In example, it appears in our studies that structural alterations constitute the main effects of postnatal thiamine deficiency, including more cornered, irregular and sparse pyramidal cells in the hippocampal field CA3, comparative to either regular diet or prenatal thiamine deficiency. These structural changes were accompanied by biochemical modifications characterized by cells containing condensed chromatin visible in Fig. 2d.

Indeed, a number of experimental issues reveal that thiamine interferes with the membrane structure and function, acts against agents-induced cytotoxicity and highlight the presence of thiamine-binding sites on biological membranes.

Membrane Structure and Function

The interference of thiamine with the structure and function of biological membrane becomes obvious. Some previous studies reported that thiamine would be an active component of axoplasmic, mitochondrial (Tanaka and Cooper 1968; Itokawa et al. 1972) and synaptosomal membranes (Matsuda and Cooper 1981). It undergoes axonal anterograde transportation and retrograde as well (Tanaka et al. 1973; Bergquist and Hanson 1983), and intervenes in the synaptic transmission (Siegel et al. 1989). For instance, Thiamine was reported to fix membrane ion channels and to modify their activity (Tallaksen and Tauboll 2000). Thus, thiamine deficiency provoked a significant decrease in the voltage-dependent K⁺ membrane conductance of cerebellar granule neurons, by suppression of A-type K⁺ channels mainly, that leads to neuronal cell damage and loss (Oliveira et al. 2007). The consequence is the significant reduction of nervous conduction speed (Goldberg et al. 2004), followed by the blocking of the nervous fiber action potential by the pyrithiamine, an antagonist of the thiamine (Goldberg and Cooper 1975). Thiamine deficiency should provoke further disturbances in nervous electrical activities by the alteration of myelinogenesis (Trostler et al. 1977; Reddy and Ramakrishnan 1982), resulting in reduction of the diameter of myelinic fibers (Claus et al. 1985).

Membrane Stability

It seems obvious that thiamine exerts stabilizing interactions on the biological membranes. Investigations on the role of thiamine during developmental processes confirm physiologic protecting actions of thiamine on nervous cells and others tissues. For instance, thiamine stabilizes the membrane of newly generated neuronal cells during embryogenesis and slows the programmed cell death that is the developmental apoptosis (Wang et al. 2000; Bâ et al. 2005). In addition, the study on differential alterations in the distribution of three phosphatase enzymes, during early pregnancy in the rat, has shown that thiamine contributes to the process of plasma membrane transformation of uterine epithelial cells (Bucci and Murphy 1999). However, most of the studies highlight the capacity of thiamine to protect cellular membrane against alcohol-induced cytotoxic effects. Thus, thiamine prevents or reduces alcohol-induced damages of hippocampal CA1 pyramidal cells in rat CNS (Wenisch et al. 1996). In addition, pyrithiamine and ethanol-induced cytotoxicity is prevented in cerebellar slice cultures co-exposed to thiamine (Mulholland et al. 2005). For the former explanations of the observed alcohol-thiamine antagonism on cytotoxicity, Bâ et al. (1996) hypothesized that B1 vitamin acts against the effects of ethanol on fluidity, and therefore increases membrane stability (Bâ et al. 1996). To explain the mechanisms of action on biological membranes, thiamine was suggested to exert a nonspecific stabilizing interaction on the axonal membrane (Goldberg et al. 2004). However, a recent study has reported the specificity of vitamin B1, and not B6 or B12, in protecting cardiac ventricular myocytes against the effects of alcohol metabolite, the acetaldehyde-induced cytotoxicity and apoptotic cell death (Aberle et al. 2004). The last hypothesis states the problem of the existence of thiamine-binding sites on biological membranes.

Membrane Sites

There is an increasing argumentation reporting the existence of thiamine-binding sites on biological membranes (thiamine triphosphate) different from its membrane carrier sites (thiamine diphosphate). Contrary to the thiamine triphosphate (TTP) that fixes on the cellular surface, thiamine diphosphate (TDP) was fixed by a specific intra membrane carrier and transported into the cytosol actively (Spector and Johanson 2007). The first allusion to the existence of thiamine-binding sites on biological membranes, different from intra membrane carrier, was suggested trough the reversal of the effects of thiamine deficiency by antioxidants, suggesting that thiamine may act as a site-directed antioxidant (Gibson and Zhang 2002). Recent disclosure in peroxisomes of TDP-dependent enzyme that catalyzes alpha-oxidation of straight chain fatty acids makes possible the existence of membrane surface sites fixing thiamine (Sniekers et al. 2006; Fraccascia et al. 2007). Through our developmental studies, we reported that the specific need of B1 vitamin increased in nervous tissues during ontogenesis and corresponded to the post-natal development of membrane processes (Bâ 2005). An assumption is that the developing membranes should generate an increasing surface distribution of thiamine-binding sites insuring their stability. That assumption was corroborated by previous studies reporting changes of thiamine metabolism in the rat brain during postnatal development (Matsuda et al. 1989). Indeed, from birth to 3 weeks, the activity of thiamine diphosphatase (TDPase) developped more strongly in the liver than in the brain, while that of thiamine triphosphatase (TTPase) increased more powerfully in the brain than in the liver (Matsuda et al. 1989). Thus, microsomal and soluble TTPase in the cerebral cortex and cerebellum increased after birth and plateaued at 3 weeks; the increase of the enzyme activity was more significant in the former tissue than in the latter (Matsuda et al. 1989). These observations suggest that in the nervous tissue, TTPase activity should interfere with the post-natal development of membrane processes. Makarchikov et al. (2003) reported that TTP was widely distributed from prokaryotes to mammals and may have a basic role in cell metabolism or cell signaling. For instance, the main electric organ of Electrophorus electricus is particularly rich in TTP, which represents 87% of the total thiamine content in this tissue (Bettendorff et al. 1987). Such an observation suggests TTPase implication in the restoration of membrane stability after intense electrical activity. Thus, TTP seems to be essentially associated with neurons: In rat brain, the amount of TTP is about five times higher in neurons than in astrocytes (Bettendorff et al. 1991). Indeed, the membrane-associated enzyme form (TTPase) may play a physiologic role other than TTP hydrolysis in mammalian tissues (Szyniarowski et al. 2005). In vertebrate tissues, TTPase may act as a phosphate donor for the phosphorylation of certain proteins; that may be part of a new signal transduction pathway (Czerniecki et al. 2004). Thiamine deficiency decreases membrane-associated TTPase activity (Iwata et al. 1974). Membrane-associated TTPase is affected by proteolytic enzymes (Bettendorff et al. 1987). Distribution of TTPase activity, typical of thiamine-binding proteins in cellular structures, was studied (Ianchii et al. 2003; Czerniecki et al. 2004). TPPase immunocytochemical labeling showed the strongest staining in hippocampal pyramidal neurons, as well as cerebellar granule cells and Purkinje cells. Immunoreactive cells were distributed throughout cerebral cortical gray matter and the thalamus. White matter was not significantly labeled. TTPase immunoreactivity seems to be located mainly in the cytoplasm and dendrites (Czerniecki et al. 2004). Subcellular structures precipitated as microsomal fractions, e.g., endoplasmic reticulum and vesicular parts of plasma membranes were encrusted with TTP-binding proteins (Ianchii et al. 2003). Thus, there are several evident lines allowing the membrane-associated thiamine triphosphatate-binding proteins to be indexed like a stabilizing factor of the structure and function of cellular membrane. TTP-binding proteins should exert cytoprotective signals that maintain tissue homeostasis and counteracts apoptotic agents like free radicals. The regional vulnerability of central structures should be based on the lack or failure of TTP-binding proteins systems protection against radicals irradiation or others cytotoxic agents.

Growth Factor?

Our results show that thiamine does not exert any main role on cellular growth. Cellular atrophy did not show any significant correlation neither with nonspecific, or specific effects of developmental thiamine deficiency. These results indicate that the lack of thiamine is not a causative factor of cellular atrophy. Thus, thiamine cannot be classified like a growth factor. In recent studies, determination of the impact of the dietary concentration of 5 B vitamins (riboflavin, niacin, pantothenic acid, cobalamin, and folacin) showed no influence on pig growth performance; the greater need for these vitamins is not associated with greater dietary energy intake or body energy accretion rate, but is potentially due to shifts in the predominant metabolic pathways (Stahly et al. 2007; Böhmer and Roth-Maier 2007).