Biotin uptake by human colonic epithelial NCM460 cells: a carrier-mediated process shared with pantothenic acid

Hamid M. Said, Alvaro Ortiz, Eric McCloud, David Dyer, Mary Pat Moyer, Stanley Rubin

Abstract

Previous studies showed that the normal microflora of the large intestine synthesizes biotin and that the colon is capable of absorbing intraluminally introduced free biotin. Nothing, however, is known about the mechanism of biotin absorption in the large intestine and its regulation. To address these issues, we used the human-derived, nontransformed colonic epithelial cell line NCM460. The initial rate of biotin uptake was found to be1) temperature and energy dependent,2) Na+ dependent (coupling ratio of 1:1), 3) saturable as a function of concentration [apparent Michaelis constant (K m) of 19.7 μM], 4) inhibited by structural analogs with a free carboxyl group at the valeric acid moiety, and 5) competitively inhibited by the vitamin pantothenic acid (inhibition constant of 14.4 μM). Pretreatment with the protein kinase C (PKC) activators phorbol 12-myristate 13-acetate (PMA) and 1,2-dioctanoyl-sn-glycerol significantly inhibited biotin uptake. In contrast, pretreatment with the PKC inhibitors staurosporine and chelerythrine led to a slight, but significant, increase in biotin uptake. The effect of PMA was mediated via a marked decrease in maximal uptake velocity and a slight increase in apparentK m. Pretreatment of cells with modulators of the protein kinase A-mediated pathway, on the other hand, showed no significant effect on biotin uptake. These results demonstrate, for the first time, the functional existence of a Na+-dependent, specialized carrier-mediated system for biotin uptake in colonic epithelial cells. This system is shared with pantothenic acid and appears to be under the regulation of an intracellular PKC-mediated pathway.

  • biotin transport
  • human colonic epithelial cells
  • membrane transport
  • transport regulation

biotin is an essential micronutrient for normal cellular functions, growth, and development (3, 9, 36). It acts as a coenzyme for four carboxylases that catalyze essential steps in critical cellular metabolic pathways, including fatty acid biosynthesis, gluconeogenesis, and the catabolism of several branched-chain amino acids and odd-chain fatty acids (3, 9, 36). Biotin deficiency in humans leads to a range of clinical abnormalities, including neurological disorders, growth retardation, and skin abnormalities (3, 9, 36, 39).

Humans and other mammals cannot synthesize biotin and thus must obtain the vitamin from exogenous sources via intestinal absorption. Biotin is presented to the intestine from two exogenous sources: from the diet and as a product of bacterial synthesis by the normal microflora of the large intestine. Dietary biotin exists in free and protein-bound forms (14), with the latter requiring conversion to free biotin before absorption (27, 40). Absorption of free biotin then takes place mainly in the proximal part of the small intestine via a specialized Na+-dependent, carrier-mediated system (5, 21, 23, 26, 28-31). As to the second source of biotin, previous studies have shown that a substantial portion of the biotin synthesized by the normal microflora of the large intestine is in the form of free unbound biotin, i.e., available for absorption (6, 7, 12,41). Furthermore, in vivo studies in humans, rats, and minipigs have shown that the colon is capable of absorbing significant amounts of luminally introduced biotin (1, 4, 33). Nothing, however, is known about the absorption mechanism involved or its cellular regulation. Addressing this issue is important from a physiological and nutritional perspective because this latter source of biotin may play a role in the localized nutrition of colonocytes, in addition to its contribution to the overall biotin body homeostasis. In this study, we investigated the mechanism and regulation of biotin transport in the large intestine using the human-derived, nontransformed, colonic epithelial cell line NCM460 (18) as an in vitro model system. We chose these cells because they possess characteristics that are similar to those of normal colonic epithelial cells, including similar transport processes (18,20). For example, recent studies in our laboratory have shown that these cells possess a folate uptake mechanism that is similar to that found in human native colonic apical membrane vesicles (10, 13).

METHODS

[3H]biotin (specific activity 58.2 Ci/mmol; radiochemical purity >97%) was obtained from DuPont NEN (Boston, MA). The culture medium M3:10 was a gift from INCELL (San Antonio, TX). Other cell culture ingredients were obtained from Sigma (St. Louis, MO). All other chemicals were of analytical grade and were obtained from commercial sources.

The human-derived normal colonic epithelial cell line NCM460 was propagated in the culture medium M3:10 to maintain its colonocyte features (18). The M3:10 medium is M3 base medium supplemented with 10% (vol/vol) fetal bovine serum and antibiotics and contains many growth factors and nutrients (16, 17, 35). NCM460 cells were used between passages 32 and42 in this study. The cells were grown in 75-cm2 plastic flasks (Costar) at 37°C in a 5% CO2-95% air atmosphere, with medium changes every 4 days. NCM460 cells were subcultured by trypsinization with 0.05% trypsin and 0.9 nM EDTA in Ca2+- and Mg2+-free phosphate-buffered solution and plated onto 12-well plates at a concentration of 5 × 105 cells/well. Uptake of biotin was studied 2–4 days after confluence. Cell growth was observed by periodic monitoring with an inverted microscope. Cell viability was tested by the trypan blue dye exclusion method and found to be >95%.

Uptake experiments were performed at 37°C, unless otherwise stated. Incubation was performed in Krebs-Ringer buffer containing (in mM) 123 NaCl, 4.93 KCl, 1.23 MgSO4, 0.85 CaCl2, 5 glucose, 5 glutamine, 10 HEPES, and 10 MES (pH 7.4), unless otherwise stated. [3H]biotin was added to the incubation buffer at the onset of experiments, and uptake was terminated after 3 min of incubation (unless otherwise specified) by the addition of 1 ml of ice-cold buffer followed by immediate aspiration. The monolayers were rinsed twice with ice-cold buffer and digested with 1 ml of 1 N NaOH, neutralized by HCl, and then counted for radioactivity in a liquid scintillation counter. Protein contents of cell digests were estimated on parallel wells by the method of Lowry et al. (15), using BSA as the standard. Uptake data are means ± SE of measurements on multiple separate monolayers performed on at least two different occasions and are expressed in femtomoles or picomoles per milligram protein per unit time. Pvalues for experimental vs. simultaneously performed control groups were calculated using the Student’st-test. Kinetic parameters of biotin uptake, i.e., maximal velocity (V max) and the apparent Michaelis constant (K m), were calculated using a computerized nonlinear regression analysis program of the Michaelis-Menten equation as described previously (38).

RESULTS

General characteristics of biotin uptake by the human-derived colonic epithelial cell line NCM460.

Figure 1 shows the uptake of low (6.4 nM) and high (100 μM) concentrations of biotin by confluent monolayers of NCM460 cells. At both concentrations, uptake was found to be linear with time for up to 15 min, and it occurred at rates of 3.53 fmol ⋅ mg protein−1 ⋅ min−1and 9.39 pmol ⋅ mg protein−1 ⋅ min−1, respectively. Thus, in all subsequent studies, 3 min was used as the standard incubation time (i.e., for initial rate). In another study, we examined the metabolic form of the radioactivity taken up by the cells following a 5-min incubation with 21 nM [3H]biotin. Cellulose-precoated TLC plates and a solvent system of butanol-acetic acid-water (4:1:1) were used. The results showed that 95% of the transported radioactivity was in the form of intact biotin.

Fig. 1.

Uptake of biotin by NCM460 cells as a function of time. Cells were incubated at 37°C in Krebs-Ringer buffer (pH 7.4) for different periods with low (6.4 nM; A) or high (100 μM; B) concentrations of biotin. Data are means ± SE of 3–6 separate uptake determinations performed on 2 separate occasions. When not shown, error bar is smaller than symbol. For A,y = 3.53x + 3.82 (r = 0.99). ForB, y= 9.39x + 37.49, (r = 0.99).

We also examined the effect of temperature on the uptake of biotin (6.4 nM) by NCM460 cells. Uptake was found to significantly decrease when the incubation temperature was lowered [15.3 ± 0.6 (n = 8), 6.1 ± 0.13 (n = 4;P < 0.01), and 2.9 ± 0.4 (n = 8;P < 0.01) fmol ⋅ mg protein−1 ⋅ 3 min−1 at 37, 22, and 4°C, respectively].

The dependence on Na+ of biotin uptake was also examined by replacing Na+ in the incubation medium with an equimolar concentration of Li+, K+, choline, or mannitol. This manipulation resulted in a significant (P < 0.01 for each) inhibition in biotin (6.4 nM) uptake when Na+was removed, regardless of what was used to replace it (Table1).

View this table:
Table 1.

Na+ dependence of biotin uptake by NCM460 cells

We also examined the effect of buffer pH on the uptake of biotin (6.4 nM) by NCM460 cells. The study was performed both in the presence and in the absence of Na+ in the incubation medium (Na+ was isosmotically replaced by K+). The results showed that decreasing the incubation buffer pH from 8.5 to 5.5 caused a modest increase in biotin uptake both in the presence and absence of Na+ (Table2).

View this table:
Table 2.

Effect of buffer pH on biotin uptake by NCM460 cells

Existence of a carrier-mediated system for biotin uptake by NCM460 cells.

The initial rate of biotin uptake as a function of the substrate concentration in the incubation medium (0.006–250 μM) was examined. The results showed that biotin uptake includes a saturable component. Uptake by this component was calculated by subtracting diffusion [calculated from the slope of the line between the point of origin and uptake at a high pharmacological concentration (1,000 μM) of biotin] from total uptake (Fig.2). Kinetic parameters of the saturable component were then calculated as described inmethods and found to be 19.7 ± 3.1 μM and 38.8 ± 1.9 pmol ⋅ mg protein−1 ⋅ 3 min−1 for the apparentK m andV max, respectively.

Fig. 2.

Uptake of biotin by NCM460 cells as a function of concentration. Cells were incubated at 37°C in Krebs-Ringer buffer (pH 7.4) for 3 min (i.e., initial rate) in presence of different concentrations of biotin (0.006–250 μM). Uptake shown is that of carrier-mediated component calculated as described in text. Data are means ± SE of 3–8 separate uptake determinations performed on 2 or 3 separate occasions. When not shown, error bar is smaller than symbol.

We also examined the effect of the biotin structural analogs thioctic acid, desthiobiotin, biocytin, and biotin methyl ester on the uptake of [3H]biotin (6.4 nM; Table 3). Thioctic acid and desthiobiotin caused a marked and concentration-dependent inhibition in [3H]biotin uptake, whereas biocytin and biotin methyl ester had no (or minimal) effect.

View this table:
Table 3.

Effect of biotin structural analogs on uptake of [3H]biotin by NCM460 cells

The effect of unlabeled biotin in the incubation medium on the efflux of [3H]biotin from preloaded NCM460 cells was also examined. Preloading with [3H]biotin was performed by incubating the cells for 10 min in the presence of 32 nM [3H]biotin, followed by washing of the cells and incubation for 10 min with or without 100 μM unlabeled biotin. The cell content of [3H]biotin was significantly lower in cells incubated in the presence of unlabeled biotin in the incubation medium than in cells incubated in the absence of unlabeled biotin [50.61 ± 3.65 and 63.95 ± 1.3 fmol ⋅ mg protein−1 ⋅ 3 min−1, respectively (n = 3;P < 0.01) ].

Stoichiometry of biotin-Na+carrier-mediated uptake.

The coupling ratio of biotin to Na+ was investigated using the “activation method” of Turner and Moran (37) as described by us before (24). In this method, stimulation in initial rate of biotin (6.4 μM) uptake was determined as a function of increasing the concentration of the activator, i.e., Na+, in the incubation medium (Fig.3 A). The K+ ionophore valinomycin (30 μg/ml) was added to the incubation medium as indicated (37). Uptake data were then applied to the Hill plot [log Na+ concentration vs. log (V/V maxV), whereV is initial rate of biotin uptake; Fig. 3 B]. The results showed a linear relationship (r = 0.99) with a Hill coefficient (i.e., slope) of 1.01, suggesting a Na+-to-biotin coupling ratio of 1:1.

Fig. 3.

A: biotin uptake by NCM460 cells as a function of Na+ concentration in incubation medium. Cells were incubated for 3 min (i.e., initial rate) at 37°C in modified Krebs-Ringer buffer (pH 7.4) containing biotin (6.4 μM) and different Na+concentrations ([Na+]; NaCl was isosmotically replaced with choline chloride). Valinomycin was added to incubation medium at 30 μg/ml. Results are means ± SE of 3–6 separate uptake determinations performed on 2 separate occasions. When not shown, error bar is smaller than symbol.B: Hill plot of data inA; log Q = log (V/V maxV), whereV is initial rate of biotin uptake in presence of a given Na+ concentration ([Na+]) andV max is maximum uptake velocity. For Hill plot, y = 1.01x − 1.69 (r = 0.99).

Effect of pantothenic acid and short-chain fatty acids on biotin uptake by NCM460 cells.

The effect of different concentrations of the anion pantothenate, (another water-soluble vitamin that is synthesized by the normal microflora of the large intestine) on the uptake of the anion biotin was investigated in this experiment. Pantothenic acid produced a concentration-dependent inhibition of biotin uptake that was found, by the Dixon method, to be competitive with an inhibition constant (K i) of 14.4 μM (Fig. 4).

Fig. 4.

Dixon plot of effect of pantothenic acid on biotin uptake by NCM460 cells. Cells were incubated at 37°C in Krebs-Ringer buffer (pH 7.4) containing 0.5 (○) and 5 (•) μM [3H]biotin and different concentrations of pantothenic acid. [3H]biotin uptake was measured over a 3-min incubation (i.e., initial rate), and data were applied to Dixon plot (i.e., concentration of inhibitor vs. 1/V, whereV is initial rate of [3H]biotin uptake). Data are means of 3–6 separate uptake determinations performed on 2 separate occasions.

We also examined the effect of pantothenic acid (20 μM) on the kinetic parameters of biotin uptake by NCM460 cells. Pantothenic acid caused a marked increase in the apparentK m of biotin uptake with no (or minimal) change in theV max of the uptake process (apparentK m of 16.64 ± 1.5 and 59.86 ± 1.9 μM;V max of 35.23 ± 0.93 and 38.9 ± 0.52 pmol ⋅ mg protein−1 ⋅ 3 min−1 for control and presence of pantothenic acid, respectively; Fig.5).

Fig. 5.

Effect of pantothenic acid on biotin uptake by NCM460 cells as a function of biotin concentration. Cells were incubated at 37°C in Krebs-Ringer buffer (pH 7.4) containing different concentrations of biotin (1–200 μM) and with (•) or without (○) 20 μM pantothenic acid. Data are means ± SE of 3–6 separate uptake determinations performed on 2 separate occasions. When not shown, error bar is smaller than symbol.

We also tested the effect of the anionic short-chain fatty acids acetate and butyrate (both at 1 mM) on biotin (6.4 nM) uptake. Neither of these compounds affected biotin uptake [16.61 ± 0.45, 16.11 ± 0.43, and 17.31 ± 0.63 fmol ⋅ mg protein−1 ⋅ 3 min−1 for control and presence of acetate and of butyrate, respectively (n = 6)].

Effect of metabolic inhibitors on biotin uptake by NCM460 cells.

The effect of preincubating NCM460 cells for 30 min with the metabolic inhibitors iodoacetate (10 mM), dinitrophenol (DNP; 0.5 mM), and ouabain (10 mM) on subsequent uptake of [3H]biotin was investigated in this experiment. These compounds significantly (P < 0.01 for each) inhibited biotin uptake [14.96 ± 0.42 (n = 7), 10.91 ± 0.22 (n = 7), 6.76 ± 0.7 (n = 7), and 8.14 ± 0.12 (n = 7) fmol ⋅ mg protein−1 ⋅ 3 min−1 for control and presence of DNP, iodoacetate, and ouabain, respectively].

Intracellular regulation of biotin uptake.

We tested in these experiments whether protein kinase C (PKC)- and protein kinase A (PKA)-mediated pathways are involved in the regulation of biotin uptake by NCM460 cells. One-hour pretreatment of cells with the PKC activator phorbol 12-myristate 13-acetate (PMA) led to a concentration-dependent inhibition in biotin uptake (Table4), whereas its negative control, i.e., 4α-PMA, had no effect. Similarly, pretreatment of cells with 1,2-dioctanoyl-sn-glycerol, another activator of PKC, also led to inhibition in biotin uptake (Table 4). In contrast, pretreatment of cells with staurosporine and chelerythrine, inhibitors of PKC, led to a slight but significant stimulation in biotin uptake (Table 4). When NCM460 cells were pretreated with PMA (1 μM) in the presence of staurosporine (1 μM), the PMA-induced inhibitory effect on biotin uptake was significantly (P < 0.01) reduced [16.49 ± 0.07, 12 ± 0.25, and 15.7 ± 0.48 fmol ⋅ mg protein−1 ⋅ 3 min−1 for control and after pretreatment with PMA and PMA plus staurosporine, respectively (n = 3)].

View this table:
Table 4.

Effect of modulators of PKC- and PKA-mediated pathways on biotin uptake by NCM460 cells

We also examined the effect of PMA on the kinetic parameters of biotin uptake by NCM460 cells. This was done by examining the effect of pretreatment of cells with 1 μM PMA on the uptake of biotin as a function of concentration. The results (Fig.6) showed that PMA cause a marked decrease in the V max of biotin uptake with a slight increase in the apparentK m of the uptake process (V max of 40.76 ± 1.58 and 25.21 ± 0.63 pmol ⋅ mg protein−1 ⋅ 3 min−1 and apparentK m of 18.68 ± 2.36 and 23.85 ± 2.19 μM for control and pretreatment with PMA, respectively).

Fig. 6.

Effect of pretreatment of NCM460 cells with phorbol 12-myristate 13-acetate (PMA) on biotin uptake as a function of concentration. Cells were pretreated for 1 h with 1 μM PMA at 37°C in Krebs-Ringer buffer (pH 7.4). •, PMA; ○, control. Different concentrations of [3H]biotin (0.006–250 μM) were then added to incubation medium, and uptake was determined after 3 min of incubation. Data are means ± SE of 3–6 separate uptake determinations performed on 2 separate occasions. When not shown, error bar is smaller than symbol.

In another experiment, we examined the effect of pretreatment of NCM460 cells with modulators of the PKA-mediated pathway on biotin uptake. None of these modulators significantly affected biotin uptake (Table4).

DISCUSSION

The normal microflora of the large intestine synthesizes biotin (6, 7,12, 41), and human, rat, and minipig colon are capable of absorbing significant amounts of luminally introduced biotin (1, 4, 33). The mechanism of biotin absorption in the large intestine and the intracellular regulation of that process, however, are not known. In the present study, we used the human-derived, normal colonic epithelial cell line NCM460 to address these issues. Uptake of biotin by these cells was found to be appreciable and linear with time for up to 15 min of incubation and was temperature dependent. No metabolic alteration in the transported substrate was observed following 5 min of incubation with [3H]biotin. The uptake process of biotin by NCM460 cells was dependent on the presence of Na+ in the incubation medium, as indicated by the drastic inhibition in the vitamin uptake observed when Na+ was replaced in the incubation medium with other monovalent cations or with mannitol. This suggestion was further supported by the observation of a significant inhibition of the vitamin uptake when cells were treated with the Na+-K+-ATPase inhibitor ouabain. Incubation buffer pH was found to have some effect on biotin uptake by these cells, as suggested by the modest increase in the substrate uptake when the incubation buffer pH was decreased from 8.5 to 5.5. This trend of pH effect was observed both in the presence and in the absence of Na+ in the incubation medium, suggesting that the effect is not mediated through an effect of pH on the Na+-dependent component of the vitamin uptake process. A similar type of effect of pH on biotin uptake has been seen previously in brush-border membrane vesicles of human jejunum and was attributed to the possible effect of pH on the ionic status of biotin (pK a of biotin is 4.51) (26).

Biotin uptake by NCM460 cells apparently involves a specialized, carrier-mediated system, as indicated by the saturation in the vitamin uptake as a function of increasing the substrate concentration in the incubation medium. This suggestion was further supported by the markedcis-inhibition in [3H]biotin uptake by certain structural analogs (thioctic acid and desthiobiotin) and by the stimulation of [3H]biotin efflux from preloaded cells by unlabeled biotin in the incubation medium. The contribution of this carrier-mediated system to the overall absorption process of biotin in the colon depends on the prevailing vitamin concentration in the colonic lumen, being higher at low physiological concentrations and lower at high concentrations. It is worth noting here that structural analogs with a free carboxyl group at the valeric acid moiety of the biotin molecule, such as desthiobiotin and thioctic acid, were potent inhibitors of [3H]biotin uptake compared with analogs with a blocked carboxyl group at this moiety, such as biocytin and biotin methyl ester, which showed no (or minimal) effect on [3H]biotin uptake. This demonstrates the importance of this carboxyl moiety of the biotin molecule in the recognition of and interaction with the substrate uptake carrier in these cells. Similar structural requirements in the biotin molecule have been reported for biotin transport in other cellular systems (8, 25, 34).

In a separate study, we investigated the uptake coupling ratio of biotin to Na+ using the activation method of Turner and Moran (37). The ratio was found to be 1:1, suggesting that one Na+ is transported with one biotin molecule, and that the event is electroneutral in nature. In another study, the uptake process of biotin was found to be dependent on cellular energy, as indicated by the significant inhibition in the vitamin uptake by pretreatment of cells with the metabolic inhibitors iodoacetate and DNP.

The normal microflora of the large intestine synthesizes significant quantities of not only biotin but also other substrates, such as the water-soluble vitamin pantothenic acid and the short-chain fatty acids acetate and butyrate. These substrates, like biotin, also exist as anions at the physiological pH of the large intestinal lumen, and thus might be expected to interact or interfere with uptake of the anionic biotin. Because of this, and because pantothenic and fatty acids have been shown to inhibit biotin uptake in other systems (2, 11, 19, 22,34), we examined their effect on biotin uptake by NCM460 cells. Pantothenic acid caused a concentration-dependent competitive inhibition in biotin uptake. The fact that pantothenic acid also caused a marked increase in the apparentK m of biotin uptake (from 16.64 μM for control to 59.86 μM in the presence of pantothenic acid), with no (or minimal) effect on theV max of the uptake process, further confirms the competitive nature of the inhibition. It is also worth mentioning here that theK i for pantothenic acid (14.4 μM) is close to the apparentK m of biotin uptake by these cells (range between 16.64 and 19.7 μM). These findings suggest that biotin and pantothenic acid share the same uptake system in these cells. Similar findings have been reported in the heart (2), the placenta (11, 19), and the small intestine (22). In contrast, pantothenic acid did not appear to affect the transport of biotin across the brain microvessel endothelial cells (32). The physiological and nutritional implications of interactions between these vitamins deserves further investigation. As to the effects of the short-chain fatty acids acetate and butyrate on biotin uptake, no inhibition was observed with either compound. This is unlike biotin transport through the blood-brain barrier, which was inhibited by nonanoic acid, a straight-chain fatty acid (34).

In separate studies, we investigated the potential for intracellular regulation of the biotin uptake process of NCM460 cells by PKC- and PKA-mediated pathways. Our findings showed that pretreatment of cells with PMA (but not with its negative control, 4α-PMA) and with 1,2-dioctanoyl-sn-glycerol, activators of PKC, caused significant inhibition in biotin uptake. On the other hand, pretreatment of cells with staurosporine and chelerythrine, inhibitors of PKC, produced a slight but significant stimulation in biotin uptake. Furthermore, the inhibition by PMA was significantly reduced when staurosporine was added to the pretreatment buffer. These findings point toward the involvement of a PKC-mediated pathway in the regulation of biotin uptake by these cells. The effect of PKC activation by PMA on biotin uptake was found to be mediated through marked inhibition in theV max of the uptake process and a slight increase in the apparentK m. This suggests that PKC activation leads to a marked decrease in the activity and/or number of the functional biotin uptake carriers and a slight decrease in their affinity, respectively. In contrast to the role of PKC, no role for a PKA-mediated pathway in the regulation of biotin uptake by NCM460 cells was found. This conclusion is based on the observations that specific modulators of this pathway did not significantly affect biotin uptake in these cells.

The above-described findings on the mechanism of biotin uptake by the human-derived colonic epithelial cells NCM460, showing involvement of a specialized, carrier-meditated and Na+-dependent process that is also shared by the vitamin pantothenic acid and appears to be under the regulation of a PKC-mediated pathway, are similar to those reported elsewhere for the vitamin uptake in small intestinal epithelial cells (22, 23, 28, 30, 31). Thus it is reasonable to suggest that absorption of dietary biotin in the small intestine and that of the bacterially synthesized vitamin in the large intestine occur via a similar cellular mechanism. In summary, our findings show for the first time the functional existence of a specialized carrier-mediated, Na+-dependent system for biotin uptake in colonic epithelial cells. This system is shared by the vitamin pantothenic acid and appears to be under the regulation of an intracellular PKC-mediated pathway.

Acknowledgments

This study was supported by grants from the Department of Veterans Affairs and by National Institute of Diabetes and Digestive and Kidney Diseases Grants DK-47203 and DK-02357.

Footnotes

  • Address for reprint requests: H. M. Said, VA Medical Center 151, Long Beach, CA 90822.

  • The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. §1734 solely to indicate this fact.

REFERENCES

View Abstract