Inositol Transport by Hepatopancreatic Brush-Border Membrane Vesicles of the Lobster Homarus Americanus

Myo-inositol is a ubiquitous compound found in bacteria, fungi, plants and animals. It is a cyclitol with a structure similar to that of D-glucose. The structural relationship and widespread occurrence of these two compounds have been the cause of speculation on the biogenetic origin of myo-inositol from D-glucose. Interconversion between these two carbohydrates (Daughaday et al. 1955; Loewus et al. 1978) as well as the biosynthesis of inositol-l-phosphate from glucose-6-phosphate have been documented (Zsindely et al. 1983). Despite its common occurrence, myo-inositol has been recognized as a vitamin or dietary requirement for a variety of cells and organisms. Tomita et al. (1978) showed abnormal lipid metabolism in rats and yeasts deficient in myo-inositol and inositol deficiency resulted in increased mortality and depressed growth in juvenile trout (Kitamura et al. 1967).

Carrier-mediated transport of myo-inositol has been described in a variety of tissues including hamster and chicken intestines (Caspary & Crane, 1970; Lerner & Smagula, 1979), rat and rabbit renal brush-border membrane vesicles (Takenawa et al. 1977; Hammerman et al. 1980) and bovine retinal capillary pericytes (Weiye et al. 1986). Generally, inositol transport in these tissues was Na⁺-dependent and inhibited by D-glucose, but D-glucose transport was not affected by myo-inositol (Caspary & Crane, 1970; Hammerman et al. 1980).

The present study is a characterization of myo-inositol transport by lobster hepatopancreatic brush-border membrane vesicles, specifically examining the nature of the driving forces responsible for the uptake of this compound and the interaction between inositol and D-glucose in transport by this invertebrate epithelium.

Live Atlantic lobsters (Homarus americanus) were purchased from commercial dealers in Hawaii and maintained at 10 °C in filtered sea water. Hepatopancreatic brush-border membrane vesicles (BBMV) were prepared from fresh tissue removed from individuals. Vesicles were produced using a magnesium chloride precipitation technique described previously (Ahearn et al. 1985, 1986; Ahearn & Clay, 1987a).

Purity of the vesicle preparations was assessed by comparing the activities of marker enzymes of the final pellet with those of the original tissue homogenate. These comparisons showed final pellet enrichments of alkaline phosphatase, Na⁺/K⁺-ATPase and cytochrome c oxidase of 15·3-, 1·0-and 0·2-fold, respectively (Ahearn et al. 1985), suggesting that this method produced membranes that were rich in brush borders and reduced in contamination from the basolateral membrane or membranes from cellular organelles.

Transport studies were conducted at 15 °C using the Millipore filtration technique developed by Hopfer et al. (1973). Vesicles were preloaded with 200 mmol 1⁻¹ mannitol at pH 5-5 or 7-0 using 12mmoll⁻¹Mes-Tris or Hepes-Tris as buffering agents. The composition of each incubation medium is described in the text or in the appropriate figure legend. Two types of experiments were performed. In long-term experiments, 20μl of vesicles was added to 180 μl of radiolabelled medium containing [³H]myo-inositol. After incubation (15 s, 1, 2, 5, 10, 20, 120 or 180min), 20μl of the suspension was withdrawn and plunged into 2ml of ice-cold stop solution (100 mmoll⁻¹ choline chloride at pH5·5 or 7·0). The resulting mixture was rapidly filtered through Millipore filters (0-65μm), and washed with another 8 ml of stop solution. The filters were then added to Beckman Ready Solv HP scintillation cocktail for radioactivity estimation in a Beckman LS-8100 scintillation spectrometer.

In short-term incubations, 5 μl of membrane suspension was mixed for a predetermined time with 45 μl of incubation medium, using a rapid uptake apparatus (Inovativ Labor AG, Adliswil, Switzerland). After injection of ice-cold stop solution, the vesicles were treated as described above.

Inositol uptake was expressed in pmol mg protein⁻¹, using the specific activity of the incubation medium. Protein values were established using the Bio Rad protein assay. For a given experiment, three or four replicates were used, and mean values are presented with their standard errors.

[³H]myo-inositol and D-[³H]glucose were obtained from New England Nuclear, Boston, MA. The unlabelled myo-inositol, D-glucose, valinomycin and other reagent grade chemicals were obtained from Sigma Chemical Co., St Louis, MO.

To establish the cation dependency of inositol uptake, vesicles were loaded with 200 mmol 1⁻¹ mannitol at pH 7·0, and incubated in media at the same pH containing 100 mmol 1⁻¹ NaCl, 100 mmol 1⁻¹ KC1 or 200 mmol 1⁻¹ mannitol, and 0·1 mmol 1⁻¹ [³H]inositol. Incubation in NaCl resulted in uptake of myo-inositol into the vesicles with a maximum accumulation occurring between 10 and 20 min (Fig. 1). At this time the concentration of intravesicular inositol exceeded that at 120 min (near equilibrium) by a factor greater than two. This overshoot phenomenon was not observed in the KC1 and mannitol media which led to slow hyperbolic uptake curves. Although the 120 min incubation point for inositol uptake in NaCl medium was still greater than those in other media, longer exposure to the isotope (e.g. 180 min) resulted in equilibrium values for all treatments that were not significantly different. These results suggest that inositol uptake is Na⁺-dependent and that other cations, such as K⁺, do not substitute for this ion requirement.

To assess the effects of variable pH on inositol transport, the time course of inositol uptake in the presence of an inwardly directed NaCl gradient was determined with three sets of pH conditions. Vesicles were loaded with 200mmoll⁻¹ mannitol, at pH7·0 or pH5·5. In the presence of an extravesicular pH of 7·0, and an intravesicular pH of 7·0, an inositol uptake overshoot was observed (Fig. 2). Maximal apparent uptake was achieved by 20 min of incubation. With an intravesicular pH of 7·0 and an extravesicular pH of 5·5, no overshoot was observed. Similarly, with an intra-and extravesicular pH of 5·5, no overshoot was observed. The 120 min equilibrium values for all treatments were not significantly different. This experiment suggests that inositol transport was maximal at neutral pH, and declined as pH was lowered. As a result, all subsequent experiments were conducted at intra-and extravesicular pH of 7-0.

To examine the electrogenic nature of the transport system, two types of experiments were performed. In the first experiment, Na⁺ salts with anions of different permeability were used. Vesicles were loaded with 200 mmol 1⁻¹ mannitol, and incubated in media containing 100 mmoll⁻¹ NaSCN, NaCl or sodium gluconate, and 01 mmol 1⁻¹ [³H]inositol. Inositol uptake exhibited an overshoot by 20 min of incubation in the presence of the anions SCN⁻ and Cl⁻, but not in the presence of gluconate, a non-permeant anion (Fig. 3). In addition, in the presence of SCN⁻, the apparent initial rate of inositol uptake (197 ± 6·4pmolmg protein⁻¹ s⁻¹), based on the first 60s of transport, was significantly (P< 0·0001) greater than with Cl⁻ (127 ± 6·6 pmol mg protein⁻¹ s⁻¹). Since SCN⁻ is more permeant in these membranes (Ahearn et al. 1985) than Cl⁻, a stronger transmembrane potential (inside negative) was likely to be generated with the former anion than with the latter. These results suggest that Na⁺-dependent inositol uptake by these vesicles may be an electrogenic process.

To investigate the electrogenic nature of the transport system further, a different approach was used. Na⁺-dependent inositol uptake was examined by using vesicles with a transmembrane potential difference (inside negative) generated by loading vesicles with K⁺ and incubating them in a medium containing the ionophore valinomycin. Under these conditions a transmembrane potential was created by the diffusional outflow of K⁺ through the ionophore channel. One group of vesicles was loaded with 100 mmoll⁻¹ KC1 and the ionophore. In the other group (control) the ionophore was not added. Both groups were then incubated in media containing 100 mmol 1⁻¹ NaCl or 200 mmol 1⁻¹ mannitol, and 0·1 mmol 1⁻¹ [³H]inositol. Those incubated in NaCl + valinomycin exhibited a more rapid uptake time course with an overshoot of greater magnitude than those without valinomycin (Fig. 4). Vesicles incubated in mannitol showed no overshoot under either ionophore condition. At equilibrium (180min), the uptake values for all anions were essentially identical. These results suggest that Na⁺-dependent inositol transport is electrogenic and therefore regulated by the membrane potential as well as the Na⁺ concentration gradient.

The time course of uptake of 0·1 mmol 1⁻¹ [³H]inositol was examined at very short intervals (10, 20, 30, 45 and 60s) using a rapid uptake apparatus to assess the initial entry rate. Vesicles were loaded with 200 mmol 1⁻¹ mannitol and incubated in media containing 100 mmoll⁻¹ NaCl and 0·lmmoll⁻¹ [³H]inositol. Uptake of 0·1 mmol 1⁻¹ inositol was a linear function of time from 10 to 60s of incubation (Fig. 5). A similar curve was obtained for the time course of uptake of 1 mmol 1⁻¹ [³H]inositol. The straight lines in Fig. 5 were based on linear regression analysis and provide an estimate of myo-inositol influx (slope) and initial binding (extrapolated vertical intercept). Calculated influx rates were 2·5 ±0·4 and 16·5 ± 3·0 pmol mg protein⁻¹ s⁻¹ at 0·1 and 1·0mmoll⁻¹ inositol, respectively. Binding to vesicles represented approximately 44 and 54% of total uptake at 45 s for 0·1 and 1·0mmoll⁻¹ inositol, respectively. The average of these binding values was used as an estimate of inositol binding at all external inositol concentrations. Similar quantitative binding values for inositol to these vesicles at each substrate concentration were obtained using vesicle blanks in which samples of membranes and radiolabelled incubation medium were independently added to stop solutions without prior mixing.

Fig. 6 shows the influx of [³H]inositol as a function of increasing external inositol concentration. The uptake values were adjusted by subtracting the calculated binding component (48-7%, Fig. 5) as described above. Vesicles were loaded with 200 mmol 1⁻¹ mannitol and incubated for 45 s in media containing 100mmoll⁻¹ NaCl and either 0·1, 0·25, 0·5, 1·0, 2·0, 5·0 or 100mmoll⁻¹ [³H]inositol. These media were osmotically balanced with mannitol. Influx was curvilinear at low inositol concentrations and linear at higher concentrations. Under these conditions inositol influx could be described as the sum of at least two independent processes operating simultaneously: (1) a Michaelis-Menten carrier mechanism exhibiting saturation kinetics and (2) a nonsaturable linear entry system whose rate was proportional to the external inositol concentration. The dotted line in Fig. 6 is an estimate of the nonsaturable portion of myo-inositol influx. Nonsaturable inositol influx was estimated from the slope of the entry curve between 2 and 10 mmol 1⁻¹ inositol. This apparent diffusional influx, subtracted from total influx at each concentration, provided an estimate of the carrier transport component. Influx of 0-1 and 10 mmol 1⁻¹ inositol in KC1 medium did not differ significantly (P > 0·05) from estimated nonsaturable inositol influx in NaCl saline, corroborating this method of characterizing apparent diffusional entry of this compound. Estimated carrier-mediated inositol influx was then shown on an Eadie-Hofstee plot (Fig. 6, inset). The following results were obtained using an iterative curve-fitting program. Apparent K_t was 0·79 ± 0·18 mmol 1⁻¹ and maximal influx rate, J_max, was 6·3 ± 0·4 pmol mg protein⁻¹ s⁻¹. The apparent diffusional permeability coefficient, P, was estimated as 5·92 pmol mg protein⁻¹ s⁻¹ (mmol 1⁻¹)⁻¹.

To examine the specificity of the inositol transport system, inositol uptake was measured in the presence of various sugars. One group of vesicles was loaded with 200 mmol 1⁻¹ mannitol and incubated for 15 s in media containing 100mmoll⁻¹ NaCl, 0·1 mmol 1⁻¹ [³H]inositol and 1 mmol 1⁻¹ of either mannitol, inositol, D-glucose, L-glucose, D-galactose or phloridzin. To control for diffusion, another group of vesicles was incubated for 15 s in 200 mmol 1⁻¹ mannitol and 0·1 mmol 1⁻¹ [³H]inositol, without a Na⁺ gradient. The greatest amount of inhibition was obtained in the presence of phloridzin, followed by inositol, D-glucose, D-galactose, L-glucose and mannitol (Fig. 7). The lowest uptake value was obtained in the absence of a Na⁺ gradient (diffusion). The numbers inside the distribution bars on Fig. 7 represent the percentage inhibition of carrier-mediated inositol transport after subtracting diffusion (unshaded area). The results were compared using a Duncan-Waller multiple comparison test. The degrees of inhibition by inositol, D-glucose, D-galactose and phloridzin were not significantly different (P>0·05) from each other, but were significantly different (P< 0·001) from that in the presence of L-glucose. These results suggest that inositol, D-glucose and D-galactose may share a common carrier.

To investigate the nature of the inhibition of inositol uptake by D-glucose, two experiments were conducted. In the first, vesicles were incubated in [³H]inositol, and in the presence of increasing D-glucose concentrations for 45 s (Fig. 8). Two concentrations of [³H]inositol were used (0·1 and 1·0mmoll⁻¹). Two hyperbolic curves were obtained, with maximum inositol uptake occurring in the absence of D-glucose, and maximum inhibition at 0·5 mmol 1⁻¹ D-glucose. [³H]inositol uptake values at 1, 5, 10 and 20 mmol 1⁻¹ D-glucose were identical, so only those obtained at the highest inhibitor concentration are displayed in Fig. 8. These results were then analysed using a Dixon plot (inset; Dixon, 1952), after subtracting the non-inhibitable component of inositol uptake. The intersections of the resulting regression lines on the x-axis were not significantly different, suggesting noncom-petitive inhibition of inositol uptake by D-glucose.

The second experiment to clarify the nature of the inhibition between D-glucose and inositol involved the trans-stimulation of [³H]inositol uptake by preloaded carbohydrates to ascertain if both D-glucose and myo-inositol were transported by the same carrier process. In this experiment, vesicles were preloaded for 1h at room temperature (22°C) with mannitol (199 or 200 mmol 1⁻¹), NaCl (100mmoll⁻¹) and either myo-inositol or D-glucose (both at 1·0 mmol 1⁻¹). [³H]myo-inositol (0·lmmoll⁻¹) uptake by these preloaded vesicles for 15 s (apparent influx) and 120 min (equilibrium uptake) was measured in an incubation medium containing 200 mmol 1⁻¹ NaCl. Those vesicles preincubated with 1·0 mmol 1⁻¹ myo-inositol exhibited a 39% increase in initial uptake rate of the labelled substrate compared with that shown by the control group of membranes preloaded with an equal concentration of mannitol (Table 1). In contrast, vesicles preloaded with D-glucose had an influx rate that was 18% lower than that of the control, an inhibition probably resulting from cis interactions between the substrates as a result of D-glucose carryover from the preincubation medium to the incubation medium. Both groups of test vesicles had equilibrium uptake values that were not significantly different (P<0-05) from that of the control. These results clearly indicate that [³H]inositol uptake is significantly trans-stimulated by unlabelled inositol, but not by D-glucose, suggesting that the latter sugar is probably not a transported substrate of the inositol carrier.

The purpose of this experiment was to examine the effect of inositol on the D-glucose carrier. Uptake of D-glucose was measured in the presence of inositol and various sugars. One group of vesicles was loaded with 200 mmol 1⁻¹ mannitol and incubated for 15s in media containing 0·1mmoll⁻¹ D-[³H]glucose, 100mmoll⁻¹ NaCl, and 1 mmol 1⁻¹ of either mannitol, inositol, D-glucose, L-glucose, D-fructose, D-galactose or phloridzin. To control for diffusion, another group of vesicles was incubated for 15 s in 200 mmol 1⁻¹ mannitol and 0-lmmoll⁻¹ D-[³H]glucose, without a Na⁺ gradient. The lowest uptake values were obtained in the absence of a Na⁺ gradient (diffusion). In the presence of a Na⁺ gradient, the highest inhibition was obtained with phloridzin, followed by D-glucose, D-galactose, D-fructose, L-glucose, inositol and mannitol (Fig. 9). The results from the treatments were compared using a Duncan-Waller test. The addition of inositol, L-glucose or D-fructose did not significantly (P< 0·0001) affect the transport of D-glucose. D-Galactose, D-glucose and phloridzin inhibited carrier-mediated D-glucose uptake by 61·2, 85·3 and 99·1%, respectively. These results suggest that inositol does not share the D-glucose carrier system.

D-Glucose uptake in lobster hepatopancreatic membrane vesicles (BBMV) has been shown to be Na⁺-dependent (Ahearn et al. 1985). Similarly in the present investigation, myo-inositol influx was stimulated by an inward Na⁺ concentration gradient, but not by a K⁺ gradient (Fig. 1). Na⁺-dependent myo-inositol uptake has been reported in hamster and chicken intestines (Caspary & Crane, 1970; Lerner & Smagula, 1979), rat and rabbit renal brush-border membrane vesicles (Takenawa et al. 1977; Hammerman et al. 1980) and bovine retinal capillary pericytes (Weiye et al. 1986). In addition to a Na⁺ concentration gradient, a permeant anion (e.g. Cl⁻) concentration gradient was also required to stimulate inositol uptake as a sodium gradient alone was not effective (Fig. 3).

Na⁺-dependent inositol uptake by lobster BBMV was shown to be electrogenic using two different approaches. In one, anion gradients of increasing permeability (SCN⁻ >C1⁻ > gluconate) resulted in increased inositol uptake (Fig. 3). In the second, a valinomycin-induced outward K⁺ diffusion potential stimulated myo-inositol uptake in the presence of a NaCl gradient, but not in the presence of mannitol (Fig. 4). These results indicate that inositol uptake can be driven by a Na⁺ chemical gradient and is further enhanced by a transmembrane electrical gradient. Hammerman et al. (1980) showed Na⁺-inositol cotransport to be electrogenic in rabbit renal brush-border membrane vesicles. A similar electrogenic transport system has also been described for D-glucose in lobster BBMV (Ahearn et al. 1985).

Inositol transport was shown to be sensitive to pH. Fig. 2 shows that influx was optimal around neutral pH inside and outside the vesicles, and minimal at low pH or in the presence of a pH gradient (pH 5-5 outside and pH 7-0 inside). This result contrasts sharply with those obtained in earlier studies with D-glucose, alanine, lysine and glutamate in the same lobster species using similar vesicle preparation techniques. In these previous publications D-glucose and amino acid uptakes were shown to increase as pH decreased (Ahearn et al. 1985, 1986; Ahearn & Clay, 1987a,b). This result suggests that uptake of inositol and these other nutrients in the lobster hepatopancreas may occur at different times following digestion of food, as their respective optimal uptake rates occur at opposite ends of the pH range. The pH of gastric juice in Homarus americanus has been reported to be about 5·0 (Gibson & Barker, 1979). The sequential uptake of myo-inositol and other organic solutes as a function of gut pH is consistent with the observation of Gibson & Barker (1979) who reported an increase in gastric juice pH in several recently fed decapod crustacean species. A freshly fed lobster could have an initially more alkaline gut luminal pH because of the ingestion of sea water (pH8·2) with food. As food becomes assimilated, the pH may decline. Since inositol exhibited optimal transport near neutral pH, whereas that for D-glucose and several amino acids occurred under more acidic conditions, this model suggests that inositol would be absorbed first after a meal and D-glucose and amino acid absorption would follow as gut pH dropped.

To examine the specificity of the myo-inositol carrier, inhibition experiments were performed. Fig. 7 indicates that D-glucose significantly inhibited inositol uptake (71·2% of carrier-mediated transport), and Fig. 9 shows that inositol did not have any significant effect on D-glucose transport (P>0·05). That L-glucose had little effect on myo-inositol or D-glucose influx in either experiment confirms that inhibition is not simply due to structural similarity. Similar results have been reported by Caspary & Crane (1970) and Hammerman et al. (1980). The latter concluded that D-glucose inhibited inositol transport in rabbit kidney by sharing the inositol carrier (competitive inhibition) and also by depleting the Na⁺ gradient. The results from Fig. 7 are consistent with this interpretation. However, the fact that inositol did not inhibit D-glucose uptake implies that the mechanism of inositol inhibition by D-glucose cannot be explained by the depletion of a shared Na⁺ gradient alone, although this may still be a significant factor. Otherwise, at least partial inhibition should be observed in both experiments because of similar control uptake rates of both compounds (Figs 7,9). These two experiments suggest that D-glucose as well as D-galactose may bind to the inositol carrier, but that inositol does not interact with the D-glucose carrier.

To examine the mechanism of inhibition between inositol and D-glucose, two experiments were performed. First, Fig. 8 shows that the interaction between these two solutes was noncompetitive, ruling out competition for the active site. Second, Table 1 indicates that only preloaded myo-inositol trans-stimulated the uptake of [³H]myo-inositol; preloaded D-glucose was mildly inhibitory, probably because of cis interactions between the solutes as a result of the carryover of 0-05 mmol 1⁻¹ D-glucose from the preloading medium to the incubation medium during influx estimation. These two results clearly show that in lobster hepatopan-creatic brush-border membrane vesicles cis D-glucose may bind to a site other than the active site on the inositol carrier, possibly acting like a modulator. In this respect, it is striking that even 0·1 mmol 1⁻¹ D-glucose markedly reduced the influx of 1·0mmoll⁻¹ inositol (Fig. 8), suggesting that the affinity of the D-glucose ‘regulatory’ site may be much higher than the carrier active site. Caspary & Crane (1970), Weiye et al. (1986), and S. Vilella, S. J. Reshkin, G. A. Ahearn & C. Storelli (in preparation) also found that inhibition of myo-inositol influx by D-glucose in mammalian and teleost epithelia was noncompetitive. The present investigation therefore proposes that an inositol carrier modulated by D-glucose and D-galactose may occur in the crustacean hepatopancreatic epithelium and may be even more widespread.

We would like to thank Stephan Reshkin and Eric Titus for their valuable comments throughout this study, and the East-West Center for their financial support. This investigation was supported by National Science Foundation grant number DCB85-11272.