In the intestinal lumen thiamine is in free form and very low concentrations. Absorption takes place primarily in the proximal part of the small intestine by means of a dual mechanism, which is saturable at low (physiological) concentrations and diffusive at higher. Thiamine undergoes intracellular phosphorylation mainly to thiamine pyrophosphate, while at the serosal side only free thiamine is present. Thiamine uptake is enhanced by thiamine deficiency, and reduced by thyroid hormone and diabetes. The entry of thiamine into the enterocyte, as evaluated in brush border membrane vesicles of rat small intestine in the absence of H+ gradient, is Na+- and biotransformation-independent, completely inhibited by thiamine analogs and reduced by ethanol administration and aging. The transport involves a saturable mechanism at low concentrations of vitamin and simple diffusion at higher. Outwardly oriented H+ gradients enhance thiamine transport, whose saturable component is a Na+-independent electroneutral uphill process utilizing energy supplied by the H+ gradient, and involving a thiamine/ H+ 1:1 stoichiometric exchange. The exit of thiamine from the enterocyte, as evaluated in basolateral membrane vesicles, is Na+-dependent, directly coupled to ATP hydrolysis by Na+-K+-ATPase, and inhibited by thiamine analogs. Transport of thiamine by renal brush border membrane vesicles is similar to the intestinal as far as both H+ gradient influence and specificity are concerned. In the erythrocyte thiamine transport is a Na+-independent, electroneutral process yet with two components: saturable, prevailing at low thiamine concentrations, and diffusive at higher. The saturable (specific) component is missing in patients of the rare disease known as thiamine-responsive megaloblastic anaemia (TRMA), producing a general disturbance of thiamine transport up to thiamine deficiency. The TRMA gene is located in chromosome 1q23.3. Recently, the thiamine transporter has been cloned: it is a protein of 497 amino acid residues with high homology with the reduced-folate transporter.

Erythrocyte thiamin metabolism and transport were investigated in 7 patients from Brazil, Israel and Italy suffering from thiamin-responsive megaloblastic anaemia (TRMA) associated with diabetes mellitus and sensorineural deafness. All patients discontinued thiamin therapy for 4-7 days before the investigation. TRMA patients showed invariably reduced total thiamin levels in erythrocytes (percentage reduction compared with healthy controls, -46.8+3%; mean+SEM). The proportions of individual thiamin compounds, expressed as a percentage of total thiamin content, were within the normal range, whereas their absolute amounts were significantly decreased in the following order: thiamin monophosphate>thiamin pyrophosphate>thiamin. Thiamin pyrophosphokinase activity was also reduced as compared with controls (mean reduction+SEM, -25.9+1%). The saturable, specific component of thiamin uptake, which normally prevails at physiological concentrations of thiamin (< 2/~mol/L), was absent in erythrocytes obtained from TRMA patients, while the non-saturable (diffusive) component of uptake was normally present.These results confirm observations made previously in two patients and demonstrate that TRMA is consistently associated with a state of thiamin deficiency, which is presumably secondary to reduced thiamin cellular transport and absorption (caused by lack of a membrane-specific carrier), and to impaired intracellular pyrophosphorylation.Investigators from our group recently showed that erythrocytes obtained from patients with thiamin-responsive megaloblastic anaemia (TRMA, McKusick 249270), exhibit no saturable, specific component of thiamin transport as well as a reduced thiamin pyrophosphokinase (EC 2.7.6.2; TPK) activity, resulting in decreased red cell total MS received 22.3.94 Accepted 16.5.94

Mucosal biopsy specimens obtained by routine endoscopy from 108 human subjects, including one patient with thiamine deficiency, were incubated at 37 degrees C in oxygenated calcium-free Krebs-Ringer solution (pH 7.5) containing tritiated thiamine and [14C]dextran as a marker of adherent mucosal water. The amount of labeled thiamine taken up was measured radiometrically. In subjects with no clinical evidence of thiamine deficiency, 1) thiamine uptake by duodenal mucosa had a hyperbolic time course, reaching equilibrium at 10 min; 2) thiamine concentrations < 2.5 mumol/L were taken up predominantly by a saturable mechanism displaying Michaelis-Menten kinetics (K(m) 4.4 mumol/L and Jmax 2.3 pmol.mg wet tissue-1.6 min-1), whereas higher concentrations were taken up by passive diffusion; 3) thiamine transport had different capacities along the gastrointestinal tract (duodenum >> colon > stomach); and 4) thiamine uptake was competitively inhibited in the duodenum by thiamine analogs, albeit with a different order of potency compared with rats, and was blocked by 2,4-dinitrophenol. In the thiamine-deficient patient, the duodenal saturable uptake was increased, with higher K(m) and Jmax values. In conclusion, physiologic concentrations of thiamine were transported in human small intestine by a specific mechanism dependent on cellular metabolism, whose transporters appear to be down-regulated.

Thiamin transport in human erythrocytes and resealed pink ghosts was evaluated by incubating both preparations at 37 or 20 degrees C in the presence of [3H]-thiamin of high specific activity. The rate of uptake was consistently higher in erythrocytes than in ghosts. In both preparations, the time course of uptake was independent from the presence of Na+ and did not reach equilibrium after 60 min incubation. At concentrations below 0.5 microM and at 37 degrees C, thiamin was taken up predominantly by a saturable mechanism in both erythrocytes and ghosts. Apparent kinetic constants were: for erythrocytes, Km = 0.12, 0.11 and 0.10 microM and Jmax = 0.01, 0.02 and 0.03 pmol.microliter-1 intracellular water after 3, 15, and 30 min incubation times, respectively; for ghosts, Km = 0.16 and 0.51 microM and Jmax = 0.01 and 0.04 pmol.microliter-1 intracellular water after 15 and 30 min incubation times, respectively. At 20 degrees C, the saturable component disappeared in both preparations. Erythrocyte thiamin transport was not influenced by the presence of D-glucose or metabolic inhibitors. In both preparations, thiamin transport was inhibited competitively by unlabeled thiamin, pyrithiamin, amprolium and, to a lesser extent, oxythiamin, the inhibiting effect being always more marked in erythrocytes than in ghosts. Only approximately 20% of the thiamin taken up by erythrocytes was protein- (probably membrane-) bound. A similar proportion was esterified to thiamin pyrophosphate. Separate experiments using valinomycin and SCN- showed that the transport of thiamin, which is a cation at pH 7.4, is unaffected by changes in membrane potential in both preparations.

The results provide strong evidence for the presence in renal BBMVs of a thiamin/H+ antiport having a 1:1 stoichiometric ratio.