Transport and Metabolism of Thiamin in Isolated Rat Hepatocytes*

This study examines thiamin transport in isolated rat hepatocytes and its relationship to thiamin phosphorylation.

This study examines thiamin transport in isolated rat hepatocytes and its relationship to thiamin phosphorylation.
In an Na+ medium, [35S]thiamin, 3 PM, was accumulated rapidly by the cells, and a near steady state intra-/extracellular distribution ratio of 3 was attained in 1 min. [35S]thiamin and its phosphoesters as a function of time was observed. However, the distribution ratio of intra-/extracellular ["'Slthiamin was reduced to 2.0, and the formation of ["?S]TPP was decreased to 9 nmol/ml of cell water in 1 h. Equilibrium dialysis indicated that thiamin is not bound to either bovine serum albumin or to cytosolic proteins (data not shown). It is concluded, therefore, that thiamin transport by the hepatocytes is concentrative.
In order to demonstrate the dependence of [35S]thiamin transport on medium Na+, liver cells were incubated in a medium containing choline (Fig. 2). Cellular [35S]thiamin concentration increased slowly, and the distribution ratio of intra-/extracellular [""Slthiamin did not exceed 1.1 even in 2 h of incubation.
Presumably owing to the low intracellular level of [""Slthiamin, the formation of labeled TPP, TMP, and TTP was also reduced in the absence of Na+ in the medium. Similar data were obtained with isolated liver cells incubated in either an Li+ or a K' medium.
The addition of ouabain (2.5 mM) to isolated liver cells incubated in an Na+ medium reduced the distribution ratio of intra-/extracellular r3?S]thiamin from 3.3, in the control, to 1.5 (data not shown). Thus, the concentrative process of thiamin transport in liver is coupled to Na' flux.
Energy Dependence of p'S/Thiamin Transport-Uncouplers of oxidative phosphorylation, carbonyl cyanidep-trifluoromethoxyphenylhydrazone and dinitrophenol, decreased the intracellular levels of [35S]thiamin and [35S]TPP in a dosedependent manner (Fig. 3). These data indicate that both the transport and the phosphorylation of thiamin in liver are dependent on biological energy.
Effect of Thiamin Analogs on PSJThiamin Transport and Metabolism-o-Benzoylthiamin disulfide is a thiamin analog that permeates plasma membranes very rapidly by simple diffusion and can be converted intracellularly to thiamin (10). On the other hand, pyrithiamin is a potent inhibitor of thiamin pyrophosphokinase (ll), but its effect on thiamin Thiamin Transport and Metabolism in Liver 7267 mM concentration it decreased the phosphorylation of ["%Ithiamin, and at 0.3 IIIM and higher concentrations, it completely abolished this reaction. These data, therefore, indicate that thiamin transport in liver is dissociable from thiamin phosphorylation.
Additionally, they strongly suggest the existence of a thiamin channel (or carrier) in the liver cell membrane which can be distinguished from thiamin pyrophosphokinase by the action of pyrithiamin. Amprolium is a thiamin analog that lacks a hydroxyethyl group and is not a substrate of thiamin pyrophosphokinase. It has been shown to inhibit thiamin absorption in ligated duodenal loops of hens (12). Amprolium differs from pyrithiamin and o-benzoylthiamin disulfide in that it decreased the distribution ratio of intra-/extracellular [""Slthiamin concentration (Fig. 5). It also decreased the rates of [""SITPP formation Liver cells, 55 mg wet weight per ml, were incubated as in Fig. 1. o-Benzoylthiamin disulfide dissolved in 0.01 ml of N,N'-dimethylformamide or the solvent alone was added. Similar data were obtained after 15.min incubation. Liver cells, 50 mg wet weight per ml, were incubated as described in Fig. 1. Carbonyl cyanide ptrifluromethoxyphenylhydrazone (FCCP) dissolved in N,N'-dimethylformamide, 0.01 ml, was added. This solvent per se had no effect on thiamin uptake and metabolism. Similar data were obtained after 30-min incubation. Cell viability based on trypan blue exclusion remained >80% with these concentrations of uncouplers. DNP, dinitrophenol. Hepatocytes, 62 mg min across the liver cell membrane (Fig. 4). However, at 0.03 wet weight per ml, were incubated as described in Fig. 1. It also did not exert any significant effect on [,'"S]thiamin phosphorylation.

DISCUSSION
In agreement with a recent report (5), we find that thiamin transport in isolated liver cells is concentrative, Na'-dependent, and dependent on biological energy. However, our results indicate that ["'Slthiamin is extensively phosphorylated to TPP, TMP, and TTP after its uptake into the cells. This finding differs significantly from that of Chen (5), who found that [""Slthiamin remained largely unmetabolized for as long as 1 h of incubation. This difference may be explained by his use of sonication and methanol for terminating the incubation. It has been speculated that under these conditions the phosphoesters of thiamin were hydrolyzed to thiamin by cellular phosphohydrolases (5). Thiamin transport in mammalian tissue has been studied most extensively in the small intestine (13-20). Because it has not been possible to reliably dissociate thiamin uptake and phosphorylation in such studies, some investigators have expressed the view that, as in E. coli, these processes in mammalian cells are tightly coupled (4). Although several earlier studies have shown that the distribution ratio of intra-/extracellular thiamin concentration exceeds unity when mammalian tissues are incubated in the presence of thiamin (18,(20)(21)(22), it was observed that TPP accumulated more rapidly than thiamin in tissue. Thus, it appeared possible that the intracellular thiamin might have arisen, not from concentrative transmembrane entry but, rather, from dephosphorylation of TPP. To our knowledge, the data in Figs. 1 and 4 provide the first evidence which delineates thiamin transport in mammalian cells as a metabolic step that is separable from and precedes thiamin phosphorylation.
The plasma concentration of thiamin is in the range of 0.1 to 0.2 pM (23, 24), and the intrahepatic level is about 1 PM (25). Since the K, of thiamine pyrophosphokinase from mammalian tissues for thiamin is 0.10 to 7.5 pM (21,(26)(27)(28)(29), it appears likely that the concentration of intracellular thiamin is a critical rate-limiting factor in TPP synthesis. This contention is well illustrated by the data in Figs. 1 and 2. At this point in time, no positive modulator for the thiamin pyrophosphokinase reaction has yet been identified. Similarly, the knowledge about the "coarse" control of this enzyme activity is meager, with only a recent demonstration that thiamin deficiency can significantly lower this activity in heart and in liver (29).
In these experiments, the concentrations of [""Slthiamin in the medium are about 1 order of magnitude higher than those present in blood plasma. However, based on the known amounts of endogenous thiamin and its phosphoesters present in liver (25), the levels of intracellular thiamin compounds as determined herein by radioactivity measurements have probably underestimated the true concentrations owing to dilution of radiolabel. Studies are now in progress to measure both the total amount and the radioactivity of all the thiamin compounds in the medium and in hepatocytes. In this manner, the relative rates of turnover of different intracellular pools of thiamin, TMP, TPP, and TTP, if they exist, can also be assessed. The finding that amprolium significantly lowers the distribution ratio of intra-/extracellular thiamin concentrations (Fig. 5) suggests that amprolium may compete with thiamin in its uptake. A less likely possibility is that amprolium may accelerate thiamin efflux. Since amprolium lacks a hydroxyethyl group and is not phosphorylated, it appears that amprolium can be used as a nonmetabolizable thiamin analog in the further characterization of thiamin transport. In accord with this expectation, our preliminary data indicate the [%]amprolium transport in isolated hepatocytes is, indeed, highly concentrative and is Na+-dependent. Chen (5) has also reported that ethanol, in the range of 206 to 824 mM, decreased the rate of thiamine transport 36 to 55%. Since these enormously high ethanol concentrations never occur in uiuo, we have examined the effect of 87 mM ethanol, already a near lethal concentration.
This level of ethanol did not produce any discernible effect of thiamin transport or metabolism in isolated hepatocytes.