Transport and Metabolism of Vitamin B6 in SalmoneLLa typhimurium LTZ*

Salmonella typhimurium LT2 concentrates radioactivity intracellularly from [3H]pyridoxal or [3H]pyridoxine up to 25 times the external concentration. After 1 min of uptake intracellular radioactivity is found as phosphorylated vitamin B6. The process is sensitive to temperature and is maximally active at pH 8.1, but under the conditions tested it is insensitive to monovalent cations or metabolic inhibitors, and does not require an exogenous energy source. The Km values for uptake of pyridoxine and pyridoxal are 2.0 x 10(-7) M and 1.2 x 10(-7) M, respectively; [3H]pyridoxamine is not transported. Evidence is presented for an uptake mechanism involving facilitated diffusion followed by trapping by pyridoxal kinase. S. typhimurium also appears to lack a periplasmic binding protein for vitamin B6.

Transport of amino acids (l-3) and sugars (4,5) by Salmonella typhimurium LT2 has been studied extensively. Although the immediate source of energy for amino acid transport has not been established, the requirement for a binding protein (3) and an additional protein required for transport (2) has been shown for histidine.
Sugar transport via the phosphotransferase system requires several proteins, as well as phosphoenolpyruvate as a source of energy (5). On the other hand nothing is known about the transport of vitamins in S. typhimurium. Griffith and Leach (6) have purified a thiaminbinding protein from Escherichia coli and have investigated briefly the transport of several other vitamins by this organism. They found that osmotic shock greatly reduced the uptake of thiamin, had a relatively minor effect on the transport of nicotinamide, pyridoxine, and lipoic acid, and did not alter biotin accumulation (6). Each of these vitamins except nicotinamide was bound by components of the shock fluid; however, binding of thiamin was greater than that of the other vitamins by 1 order of magnitude.
In addition, Oya has investigated the uptake of pyridoxine by E. coli (7 were counted to determine the elution profile; subsequently appropriate fractions were combined and counted to give the total radioactivity shown above each peak. Shane (15). The material used in these studies was greater than 95% pure as judged by radioautography following thin layer chromatogra-   (20), and the shock fluid was brought to a protein concentration of 1 mg/ml by ultrafiltration through a Diaflo PM-10 membrane.
Assays for binding protein employed the ultrafiltration apparatus described by Paulus (21). (3H]Pyridoxine (700 cpm/ pmol) was employed as substrate to minimize problems with nonspecitic binding of contaminating pyridoxal. Even with this precaution binding was observed in the presence of bovine serum albumin, and could only be reduced by pretreatment of the pyridoxine with sodium borohydride or cysteine (Table I).

Demonstration
and General Characteristics of Vitamin B6 Transport-As shown in Fig. 2, Salmonella typhimurium accumulates pyridoxal and pyridoxine, but not pyridoxamine. Uptake is linear for 10 min, and the intracellular concentration of radioactivity reaches a level 25 times that in the medium. The uptake is optimal at pH 8.1 (Fig. 3), is not significantly stimulated by addition of 50 mM NH,+, K+, Nat, or Mgz+, and has a Q1, of about 2.0 between 20 and 40". The apparent K, for transport of pyridoxine is 2.0 x lo-' M, and for pyridoxal, 1.2 x lo-' M (Fig. 4).
The most striking characteristic of active transport systems is their ability to concentrate free metabolites intracellularly. Thus it was of interest to determine whether the intracellular radioactivity was free transport substrate, or a metabolically trapped derivative. Fig. 5 shows that after 1 min of accumulation the bulk of intracellular radioactivity is found as vitamin B6 phosphates and pyridoxal; almost none is found as pyridoxine, which is present at 0.1 flM extracellularly.
Since there is no net concentration of free pyridoxine intracellularly at this or subsequent times, this criterion may not be used to show the presence of an active transport system. Transport of 5'-Deoxypyridowal and 5'-Deoxypyridoxine-To test the possibility that pyridoxine is actively transported but also rapidly phosphorylated, two nonmetabolizable analogs, 5'-deoxypyridoxine and 5'-deoxypyridoxal, were checked as transport substrates. The former was not concentrated above the level in the medium, while Et'-deoxypyridoxal was concentrated lo-fold. However, because of Uptake Time (mln) PH   FIG. 2 (left). Uptake of 3H-labeled vitamin B6 by Salmonella typhimurium.
Transport was assayed by the standard method at an extracellular vitamin concentration of 0.5 pM. The arrow represents the expected level of vitamin uptake in the absence of any concentration gradient across the membrane.
Effect of pH on [8H]pyridoxine uptake. Cells were centrifuged from growth medium (Medium A plus 0.5% glucose) and aliquots resuspended in 20 mM potassium phosphate/20 mM sodium citrate buffer adjusted to the indicated pH with 6 M HCl or 10 M NaOH. Transport was then assayed in the usual manner. The data shown were obtained after 10 min of uptake.

FIG. 4. Determination
of kinetic parameters for transport of pyridoxal and pyridoxine. Transport was assayed by the standard method, but with varying concentrations of *H-labeled vitamin B6. The data shown were obtained after 10 min of accumulation.
its free aldehyde group, 5'-deoxypyridoxal binds strongly and relatively nonspecifically to protein amino groups, and thus the possibility existed that the gradient observed resulted from such intracellular binding of this compound, rather than from active transport. Fig. 6A shows this to be the case, since the apparent concentration gradient rises in the presence of toluene, a treatment which destroys the permeability barrier provided by the cell membrane (22). In contrast, the accumulation of pyridoxine was markedly reduced in the presence of toluene (Fig. 6B).
Inhibition of Pyridoxine Transport by 5'-Deoxypyridoxine and 5'-Deoxypyridoxal-In conjunction with the above study, 5'-deoxypyridoxine and 5'-deoxypyridoxal were checked as inhibitors of pyridoxine uptake. 5'-Deoxypyridoxine had no effect at concentrations up to 10 NM (100 times higher than the concentration of pyridoxine present), while inhibition by 5'-deoxypyridoxal appeared to be noncompetitive (Fig. 7). Since 5'-deoxypyridoxa1 is a potent inhibitor of pyridoxal kinase from Escherichia coli (23), it seemed possible that the inhibition of pyridoxine uptake might result from inability of the cells to phosphorylate accumulated pyridoxine in the presence of the analog. This hypothesis was tested by preloading cells with 5'-deoxypyridoxal, as detailed in Fig. 8 8 (r&t).
As a control, cells were first incubated with 1.0 &tM 5'-deoxy[SH]pyridoxal and the total uptake was determined by the usual method. A parallel set of samples was filtered after 10 min, then, while still on the filters, the cells were layered with medium containmg 0.5% glucose, and after 10 more min, filtered and washed. By this method it was determined that 70% of the initial amount of 5'-deoxypyridoxal taken up remained in the cells after 10 min incubation on a filter with uptake medium.
In the experiment shown, cells were preloaded with 5'-deoxypyridoxal (A) and filtered as above, but the layering medium contained 0.3 PM [SH]pyridoxine. Control cells (0) were treated in the same way except that the first incubation lacked 5'-deoxypyridoxal. in intact cells of Salmonella typhimurium, and of cell-free pyridoxal kinase from the same organism. Pyridoxal kinase was determined in a crude cell extract prepared by sonic oscillation of S. typhimurium, using 0.12 mM pyridoxal as substrate. Assay was by an unpublished spectrophotometric method employing apotryptophanase.' Transport was assayed by the standard method (using 1.2 MM [3H]pyridoxine) except that sodium arsenite was added 10 min before substrate.
Binding Protein Assays-In an effort to account for the specificity of transport (i.e. for the lack of pyridoxamine uptake) several attempts were made to detect a vitamin B6 binding protein. Although there appeared to be a small binding capacity for pyridoxine, it was abolished by treating the substrate with sodium borohydride to remove [3H]pyridoxal. In addition, boiling the shock fluid for 15 min did not decrease the apparent binding of untreated [3H]pyridoxine (Table I) typhimurium primarily as phosphorylated derivatives. The uptake is sensitive to pH and temperature changes, is not stimulated by monovalent cations, and is not inhibited by metabolic poisons. Transport shows saturation kinetics with K, values of 0.20 pM and 0.12 fiM for pyridoxine and pyridoxal, respectively. The lack of effect of metabolic inhibitors, the rapid appearance of phosphorylated vitamin intracellularly, and the parallel sensitivity of pyridoxal kinase and uptake to inhibition by arsenite argue against active transport, and suggest some sort of diffusion process followed by trapping of the vitamin by phosphorylation.
Three pieces of evidence indicate that vitamin B6 does not enter S. typhimurium by simple diffusion: (a) pyridoxamine, unlike pyridoxine or pyridoxal, is not concentrated intracellularly to any significant extent. The measured K, value for phosphorylation of pyridoxal by the kinase from S. typhimurium was 40 PM; pyridoxamine inhibits this process (at low concentrations of pyridoxal and pyridoxamine) with a K, value of about 100 PM. Technical difficulties prevented a more accurate measurement of its K, value, or of its K, value when present alone. However, crude extracts of S. typhimurium phosphorylate pyridoxamine at twice the rate of pyridoxal when given high (1 mM) levels of these substrates. It is thus likely that the failure of S. typhimurium to concentrate pyridoxamine is due to its inability to cross the cell membrane, rather than lack of trapping. This lack of transport probably explains the observation (24) that most vitamin BG-dependent mutants of S. typhimurium grow equally well on pyridoxal or pyridoxine, but not on pyridoxamine.
(b) 5'-Deoxypyridoxal apparently also encounters a diffusion barrier, since its binding to cellular components is increased almost 3-fold in the presence of toluene (Fig. 6A); nevertheless this compound inhibits uptake of both pyridoxine and pyridoxal. This effect is not intracellular, judging from the lack of inhibition of pyridoxine transport observed when cells were preloaded with 5'-deoxypyridoxal.
The observation that this pyridoxal analog is a noncompetitive inhibitor of transport leads one to conclude that it acts by binding to membrane proteins important for transport. (c) The measured K, for transport of pyridoxal (0.12 WM) is radically different than the K, of pyridoxal kinase for pyridoxal (40 FM). Although there are other possibilities, it seems most likely that the Michaelis-Menten constant for transport reflects the affinity toward vitamin B6 of some intermediate carrier. Efforts to detect a specific vitamin BG-binding protein were completely unsuccessful. Although the supernatant fluids from osmotically shocked cells did bind radioactivity from [9H]pyridoxine, the binding was not competitive with unlabeled pyridoxine, was not abolished by boiling or digestion with pronase, and was eliminated by treatments which destroy or complex pyridoxal. Since the bound radioactivity represents less than 0.5% of the total present, and traces of pyridoxal are readily formed by nonenzymatic oxidative processes during manipulation of pyridoxine solutions (25) we believe the material bound is probably pyridoxal or possibly some other undetected radioimpurity, and that the binding is nonspecific. In summary, it appears that pyridoxine and pyridoxal are accumulated by S. typhimurium via a facilitated diffusion mechanism, followed by trapping through the action of pyridoxal kinase. No periplasmic factors are involved in binding, although a factor exposed to the medium and sensitive to of Vitamin B6 inhibition by 5'-deoxypyridoxal is implicated in, and lends specificity to, the transport process. The lack of stimulation of pyridoxine uptake by added glucose, or of inhibition by dinitrophenol was unexpected in light of Oya's results (7), but suggests a difference in the ability of S. typhimurium and E. coli to utilize endogenous energy stores for phosphorylation of vitamin B6.
Bassham for arranging our use of facilities at the Lawrence Laboratory for handling large amounts of tritiated materials.

REFERENCES
It should be noted that the uptake mechanism described here for S. typhimurium differs in most respects from that found in Saccharomyces carlsbergensis (13). Both organisms accumulate vitamin B6 from the medium, but in the yeast this accumulation is via an active transport mechanism producing free intracellular vitamin, rather than via the facilitated diffusion plus trapping mechanism which results in accumulation of vitamin B6 phosphates in S. typkimurium.
In S. carlsbergensis control of active transport results in an overshoot phenomenon, with intracellular vitamin B6 being released into the medium, while in S. typhimurium accumulation proceeds monotonically to a steady state plateau. 5'-Deoxypyridoxine, which cannot be phosphorylated by pyridoxal kinase, is concentrated by the yeast in the same way as pyridoxine, but is inert both as a substrate and inhibitor of the transport system in S. typhimurium.
Finally, the net accumulation of vitamin B6 in cells of S. carlsbergensis is about 10 times that observed in S. typhimurium.
These marked differences in properties of vitamin B6 transport systems from the two organisms studied emphasize the need for studies in a variety of different organisms before generalizations concerning the nature of vitamin transport are drawn.