INTRODUCTION

Top Abstract Introduction Materials & Methods Results Discussion References

HIGHER PRIMATES differ from most other mammals in that red blood cells from adult individuals show a very high capacity for equilibrative glucose transport, a feature typical of fetal and neonatal erythrocytes in many mammalian species (12, 30). Glucose transport in human red blood cells has been shown to be mediated by the GLUT-1 glucose transporter isoform (nomenclature from Ref. 14) of the equilibrative hexose transporter family of integral membrane glycoproteins. Glucose transporter polypeptides are abundant [~250,000-500,000 copies per cell (1, 17, 25)] in the plasma membrane of adult human red blood cells. The GLUT-1 transporter isoform is widely expressed in mammalian cells and tissues and shows strong conservation of sequence in diverse species (for review see Refs. 3 and 13). The functional characteristics of the GLUT-1-mediated transport pathway in human erythrocytes have been the subject of intensive investigation; however, a definitive kinetic characterization of this transport system remains elusive (reviewed by Carruthers in Ref. 4). This uncertainty may be due, in part, to technical difficulties in the accurate estimation of very rapid initial rates of D-glucose transport in human red blood cells. Even at sub-physiological temperatures, transmembrane glucose fluxes are very rapid, with a half time for D-glucose entry at low substrate concentrations of less than 1 s at 20°C (20, 21). However, kinetic complexity of glucose transport in human red blood cells may also reflect physical complexities of the transport pathway, for example, the influence of cytosolic hexose binding sites on sugar transport (6). Study of glucose transport in a nonprimate species showing rapid glucose permeation into adult red blood cells could assist in the identification of structural and functional features of the transporter that may be important for physiological activity and could clarify biological properties of the glucose transport pathway in human erythrocytes.

Erythrocytes from an adult beluga whale (Delphinapterus leucas; see Ref. 9) and bottle-nosed dolphins (Tursiops truncatus; see Ref. 7) show a high permeability toward D-glucose, supporting a conclusion originally drawn from measurements of sugars in a blood sample taken from a dying porpoise (2). Small odontocetes present an opportunity to examine rapid glucose transport in red blood cells from adult mammals other than the higher primates. Paleontological evidence suggests that cetacean lineages diverged from terrestrial mammals ~55 million years ago and that the artiodacyls (pigs, camels, ruminants) represent the closest modern terrestrial relatives of the whales (27). Erythrocytes from adults of these terrestrial species exhibit low permeability toward D-glucose (reviewed by Whittam in Ref. 28), an exception to this general finding being red blood cells from the pig, which are unable to utilize extracellular glucose (18) because they possess no mediated pathway for D-glucose entry.

The biological advantages of retention of a very high glucose transport capacity in erythrocytes of adult humans and dolphins have not been clearly ascertained. This physiological phenotype is probably not determined by metabolic requirements of the red blood cells themselves, since in humans the hexose transport capacity appears to be well in excess of any possible metabolic demands of the erythrocyte (29). Dolphin red blood cells incubated in vitro have been shown to catabolize D-glucose more slowly than human red blood cells (15). One rationalization for a high glucose transport capacity is that there is a requirement to maximize glucose delivery to specific regions of the brain under conditions of physiological stress (for example under hypoxic conditions in aquatic mammals), and calculations of rates of glucose consumption by the brain and rates of glucose delivery to the central nervous system support this conjecture (5). Both primates and odontocetes are notable for large and complex central nervous systems, with high brain mass/body mass ratios (23).

In this study, D-glucose entry into dolphin erythrocytes was examined at a low temperature so that initial rates of entry could be estimated directly from progress curves for uptake. A simple kinetic analysis of the functional properties of the D-glucose transport pathway was undertaken. Zero-trans and equilibrium exchange measurements [nomenclature of Stein (26)] showed that, at 6°C, the dolphin erythrocyte glucose transport pathway displays marked trans-stimulation effects, similar to those observed for glucose transport in human red blood cells. Transporter site density in dolphin red blood cell membranes, estimated from analysis of D-glucose-displaceable cytochalasin B binding, appears to be lower than that found in human erythrocyte membranes. Molecular studies demonstrated the presence of GLUT-1 equilibrative glucose transporter polypeptides in dolphin red blood cell membranes. This protein was readily fragmented by exposure to trypsin and showed a markedly greater mobility on sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) than human red cell GLUT-1, probably because of differences in glycosylation.

	MATERIALS AND METHODS

Top Abstract Introduction Materials & Methods Results Discussion References

Reagents and Solutions

Luminescent detection of antibodies on immunoblots was accomplished using an ECL kit purchased from Amersham Canada. Glycopeptidase F, Ponceau S dye, and Tween 20 were from Sigma Chemical (St. Louis, MO). Trypsin (from bovine pancreas; ~40 U/mg) was from Boehringer Mannheim Canada (Laval, Quebec, Canada).

Ascites fluid (stored at 70°C) was used as a source of monoclonal antibody 65D4. This antibody has been shown to recognize GLUT-1 polypeptides in erythrocyte membranes from adult humans, neonatal pigs, and adult mice (8). A polyclonal monospecific antiserum, raised against a 12-amino acid COOH-terminal peptide sequence for human GLUT-1, was purchased from East Acres Biologicals (Southbridge, MA). The composition of phosphate-buffered saline (PBS) was (in mM) 136.8 NaCl, 2.68 KCl, 4.29 Na₂HPO₄, and 1.47 KH₂PO₄, pH 7.4, and red blood cell lysis buffer was ice-cold 5 mM phosphate buffer, pH 8.0, including 0.1 mM phenylmethylsulfonyl fluoride (PMSF; Sigma Chemical) added as a stock solution of PMSF in anhydrous isopropanol immediately before use.

Venous blood samples used in this study, obtained from captive dolphins without the use of physical restraint, were excess samples left over after blood had been taken for routine veterinary tests.

Monosaccharide Flux Measurements

Equilibrium exchange entry. Red blood cells were isolated from whole blood as described in the zero-trans entry protocol and then washed with ~10 volumes of a solution containing the desired concentration of D-glucose in phosphate saline. Cells were incubated at room temperature for 15-45 min (longer times for higher glucose concentrations) and then washed again with a fresh portion of the appropriate glucose solution. After this washing step, the cell suspension (at ~20% hematocrit) was equilibrated to the temperature appropriate for the transport assay. An experimental protocol identical to that described for zero-trans entry measurements was then followed. Preliminary experiments showed that at 6°C equilibrium exchange fluxes of D-glucose were much more rapid than zero-trans entry fluxes, so shorter incubation times were used to define progress curves for equilibrium exchange entry. Initial rates of tracer entry and transport parameters were estimated as described for zero-trans entry.

Preparation of erythrocyte plasma membranes (ghost membranes) from human, pig, and dolphin erythrocytes. Erythrocytes were collected into 7-ml evacuated sample tubes containing 100 USP units of lithium heparin (Vacutainer, Becton Dickinson, Rutherford, NJ), mixed by inversion, and chilled on ice. Cells were stored at 6°C until used; preparation of red blood cell membranes was completed within 24 h of the blood samples being taken. Blood samples were subjected to gentle centrifugation (300 g, 5 min, 6°C) to pellet the erythrocytes. Blood plasma and white blood cells were removed by aspiration, and the red blood cells were resuspended in ~10 volumes of PBS, pH 7.4. This centrifugation and washing procedure was repeated twice to remove blood plasma proteins and platelets. The loose red blood cell pellet was added to ~100 volumes of ice-cold lysis buffer and gently stirred on ice for ~3 min. The lysed cell suspension was subjected to centrifugation (10,000 revolutions per minute, 15 min, 6°C, Sorvall SA-600 rotor), and supernatant fluid was removed by aspiration. The loose red blood cell membrane pellet was resuspended in ice-cold 5 mM phosphate buffer, and the small dense "button" pellet was discarded. The centrifugation and washing procedure was repeated (three or four times) to give an off-white membrane pellet that was resuspended in 5 mM phosphate buffer, pH 8.0, and frozen in small portions for storage at 70°C. Protein content of the suspension was estimated by a dye-binding method using protocols and reagents supplied by Bio-Rad.

Cytochalasin B binding assays. Reversible binding of tritiated cytochalasin B to dolphin erythrocyte membranes was assayed in the presence of 400 mM D-glucose or the same concentration of sorbose (a sugar known to show little interaction with erythrocyte glucose transport systems). Portions (40 µl) of erythrocyte suspension (0.93 mg protein/ml) were placed in 5 mm × 20 mm polyallomer ultracentrifuge tubes (Beckman Airfuge, Beckman Instruments, Palo Alto, CA) together with 140 µl 1 M D-glucose or L-glucose solution. Graded amounts of tritiated cytochalasin B were added in 2 µl of ethanol. The mixture was incubated for 40 min at room temperature before centrifugation (Beckman Airfuge, 28 lb/in.² at gauge, 15 min). Supernatant fluids were sampled and the concentration of free tritiated cytochalasin B estimated by scintillation counting (Ecolite scintillation fluid, ICN Biochemicals). Quench corrections were performed using an external standard method. The membrane pellet was dispersed by addition of 50 µl of 0.5% (vol/vol) Triton X-100 detergent in distilled water, and the amount of bound tritiated cytochalasin B was estimated by scintillation counting. Nonspecific binding of tritiated cytochalasin B was estimated from tubes in which 68 µM unlabeled cytochalasin B was included in addition to the tritiated ligand. Binding data were analyzed using Enzfitter nonlinear regression data analysis program (Biosoft, Cambridge, UK)

Photolabeling of erythrocyte membranes with tritiated cytochalasin B. Red blood cell membrane suspensions (final concn 0.75-1 mg protein/ml) in medium containing 2.5 mM phosphate, pH 8.0, 10 µM cytochalasin E, 500 mM D-glucose or L-glucose, and 0.5-1 µM tritiated cytochalasin B were incubated on ice for 2 h. Samples were briefly purged with nitrogen and were then irradiated in 1 mm light path quartz cuvettes at a distance of 7 cm from a water-jacketed 450 W mercury lamp (Conrad Hanovia) for 20 s. Following irradiation, dithiothreitol solution (0.5 M) was added to give a final concentration of 10 mM, and the red blood cell membranes was recovered by centrifugation (90 s, microcentrifuge), solubilized in SDS-PAGE sample buffer (incubation at 37°C for 10 min), and then subjected to electrophoresis (9% polyacrylamide gel). For fluorography, gels were treated with Enhance fluorography cocktail (NEN Research Products, DuPont Canada) according to the manufacturer's directions. Exposures of the treated and dried gels were performed at 70°C, using Kodak XAR-5 X-ray film.

SDS-PAGE and immunoblot protocols. Polyacrylamide slab gels (8, 9, or 10% acrylamide) were prepared and run using standard protocols and reagents supplied by Bio-Rad (Mississauga, Ontario, Canada) using the Laemmli buffer system. Prestained molecular weight standards (Rainbow Markers, Amersham Canada) and unstained molecular weight standards (high molecular weight, Sigma Chemical) were included in separate lanes.

For immunoblot analysis, proteins were transferred to nitrocellulose (0.45-µm pore size, Bio-Rad) or polyvinylidine difluoride (Immobilon, Millipore Canada, Napean, Ontario, Canada) membranes using the Towbin buffer system [25 mM tris(hydroxymethyl)aminomethane (Tris), 192 mM glycine, pH 8.8], with 5% (vol/vol) methanol. When appropriate, transferred proteins were visualized using Ponceau S or Amido Black stains. In immunodetection protocols, blots were treated with PBS containing 0.05% (vol/vol) polyoxyethylene sorbitan monolaurate (Tween 20) and 3% (wt/vol) dried nonfat milk powder to block nonspecific antibody binding. After incubation with primary and secondary antibodies (diluted in blocking solution) and appropriate washing steps (PBS containing 0.05% Tween 20), following the protocols given by the manufacturer of the ECL Western blot kit (with the substitution of PBS for Tris-buffered saline), antibody binding was determined by exposure of X-ray film (Fuji RX) to luminescence generated by peroxidase activity bound to the blot (exposure times 5 s to 8 min).

When it was necessary to reprobe blots (with a different antibody preparation), the membranes were first washed twice with 100-ml volumes of PBS-Tween solution (10 min with gentle agitation). The blot was then placed in stripping buffer (100 mM 2-mercaptoethanol, 2% wt/vol SDS, 62.5 mM Tris · HCl, pH 6.8) warmed to 50°C and incubated for 30 min with occasional gentle agitation. This procedure removed both primary and secondary antibodies. Blots were washed three times at room temperature (100 ml PBS-Tween, 10 min, gentle agitation) to remove stripping buffer and then incubated for 1 h in blocking medium, before the immunodetection protocol was repeated.

Digestion of erythrocyte membrane polypeptides with trypsin. Digestions were performed by incubation of 30 µl dolphin or human erythrocyte membranes (1.5 mg/ml protein concentration) suspended in 5 mM phosphate buffer, pH 8.0, on ice for 30 min in the presence of 0.25 or 0.025 mg/ml trypsin. The digestion was terminated by addition of 2 µl 100 mM PMSF (freshly dissolved in dimethyl sulfoxide) and incubation at room temperature for 5 min. Samples were dissolved by addition of 35 µl electrophoresis sample buffer, with incubation at room temperature for 5 min before SDS-PAGE analysis.

Digestion of erythrocyte membrane glycoproteins with glycopeptidase F. Digestions were performed by solubilization of 30 µg membrane protein in a volume of 10 µl of 1% SDS. The solubilized sample was prepared for digestion by addition of 30 µl digestion buffer (2.6% vol/vol Nonidet P-40; 40 mM Tris buffer, pH 8.0; 5 mM 1,10 o-phenanthroline; and 33 mM 2-mercaptoethanol). Glycopeptidase F (0.5 U) was added, and the sample was incubated at 35°C for 6 h, when additional glycopeptidase F (0.5 U) was added. The incubation was continued for a further 16 h at 35°C. The digestion was terminated by addition of 5 µl 10% (wt/vol) SDS and subsequent addition of 45 µl SDS-PAGE sample buffer. The mixture was incubated at room temperature for 10 min before samples were applied to the polyacrylamide gel and subjected to SDS-PAGE.

	RESULTS

Top Abstract Introduction Materials & Methods Results Discussion References

Kinetic Results

1

1

1

1

View larger version (12K): [in this window] [in a new window]   Fig. 1.   Progress curve for the entry of D-glucose into washed dolphin red blood cells (zero-trans entry) at 6°C. Extracellular D-glucose concentration was 5.32 mM at time 0. Solid curve shows a 2nd-order polynomial curve fitted to the experimental points; the straight line represents an initial rate of entry (0.276 mmol · l cell water1 · min1) of D-glucose, calculated from the polynomial curve fit.

View larger version (18K): [in this window] [in a new window]   Fig. 2.   Concentration dependence of initial rate of entry of D-glucose into dolphin erythrocytes at 6°C in phosphate-buffered saline solution under zero-trans (A) and equilibrium exchange (B) conditions. Solid curves illustrate Michaelis-Menten parameters [zero-trans entry (A): Michaelis constant (Km) = 0.78 mM and maximal velocity (Vmax) = 300 µmol · l cell water (lcw)1 · min1; and equilibrium exchange entry (B): Km = 17.5 mM and Vmax = 8,675 µmol · lcw1 · min1]. Eadie-Hoftsee replots of the data are included to show the fit of these parameters to the data points.

Inhibition of D-Glucose Entry by Cytochalasin B

Glucose-Displaceable Binding of Tritiated Cytochalasin B

View larger version (21K): [in this window] [in a new window]   Fig. 3.   D-Glucose displaceable binding of tritiated cytochalasin B to isolated dolphin erythrocyte membranes under equilibrium conditions as described in MATERIALS AND METHODS. Inset: Scatchard plot of the data. Solid lines show maximal binding = 0.37 nmol/mg membrane protein and Kd = 0.48 µM.

Photolabeling of Dolphin Red Blood Cell Membranes With Tritiated Cytochalasin B

View larger version (77K): [in this window] [in a new window]   Fig. 4.   Fluorogram of SDS-polyacrylamide gel electrophoresis (PAGE) gel of dolphin red blood cell membranes subjected to photoirradiation in the presence of tritiated cytochalasin B, as described in MATERIALS AND METHODS. Lane 1, irradiation in the presence of 500 mM L-glucose; lane 2, irradiation in the presence of 500 mM D-glucose.

Identification of Dolphin Erythrocyte Glucose Transporter Polypeptides on Western Blots

View larger version (82K): [in this window] [in a new window]   Fig. 5.   Immunoblot of red blood cell membrane proteins probed with polyclonal antiserum directed against the COOH-terminal region of human GLUT-1 protein. Lane A, human erythrocyte membranes, lane B, erythrocyte membranes from adult domestic pig; lane C, dolphin erythrocyte membranes. Protein loading of the SDS-PAGE gel was 10 µg/lane.

View larger version (64K): [in this window] [in a new window]   Fig. 6.   Digestion of GLUT-1 in dolphin and human erythrocyte membranes by trypsin. Red blood cell membranes were digested and subjected to SDS-PAGE and immunoblotting as described in MATERIALS AND METHODS. Red blood cell membranes were incubated on ice for 30 min in the presence or absence of trypsin (dolphin, lanes 1-3; human, lanes 4-6) before analysis. All lanes were loaded with 10 µg protein. A: signal detected after probing blot with polyclonal antiserum directed against the COOH-terminal peptide of human GLUT-1. B: signal from the same blot after stripping and subsequent reprobing with monoclonal antibody 65D4. Dolphin erythrocyte membranes: lane 1, no trypsin; lane 2, 0.025 mg/ml trypsin; and lane 3, 0.25 mg/ml trypsin. Human erythrocyte membranes: lane 4, no trypsin; lane 5, 0.025 mg/ml trypsin; and lane 6, 0.25 mg/ml trypsin. The lane adjacent to lane 4 contained molecular weight standards, one component of which appeared to cross-react with the polyclonal antiserum (A) but which did not give a signal when the blot was reprobed with the 65D4 monoclonal antibody (B). , Position of a faint band not visible after photography.

Digestion of Erythrocyte Membrane Polypeptides With Trypsin

Digestion of GLUT-1 with Glycopeptidase F

View larger version (102K): [in this window] [in a new window]   Fig. 7.   Immunoblot of red blood cell membranes (dolphin, lanes A and B; human, lanes C and D) after incubation at 35°C with (lanes B and D) or without (lanes A and C) glycopeptidase F, probed with polyclonal antibody directed against COOH-terminal peptide of human GLUT-1, as described in MATERIALS AND METHODS.

	DISCUSSION

Top Abstract Introduction Materials & Methods Results Discussion References

The pathway for rapid mediated entry of D-glucose into erythrocytes from adult dolphins shows remarkable functional similarity to the GLUT-1-mediated pathway in human red blood cells; in particular, a very marked trans-stimulation of hexose entry was observed at low temperatures. Mass law analysis of D-glucose-displaceable cytochalasin B binding to dolphin red cell membranes suggests that the site density of equilibrative glucose transporters is lower than that observed in equivalent preparations of human erythrocyte membranes, consistent with observations of D-glucose transport rates reported in this study, and previously (7), that indicated that entry of D-glucose into dolphin red blood cells was less rapid than entry into adult human red blood cells. A trans-stimulation effect (29-fold, calculated from V_max) is less than that reported for human red blood cells at 0°C [~100 fold (21)]. However, trans-stimulation of glucose transport in human red blood cells declines with increasing temperature (19, 21), so the value estimated for dolphin red blood cells at 6°C is close to that expected for human red blood cells. K_m values for equilibrium exchange for the dolphin system are within the range of values reported for human red blood cells [13-25 mM, (20)]; however, a higher K_d value for glucose-displaceable cytochalasin B binding, suggests that GLUT-1 in T. truncatus erythrocytes may differ slightly in its functional properties from the human red cell glucose transporter. Plasma glucose levels reported for peripheral venous blood samples of T. truncatus [6.3 ± 1.3 mM (5), 7.2 ± 0.2 mM (24), 10.1 ± 0.2 mM (22)] are higher than those observed for fasting humans [3.9-5.6 mM (11)], so a difference in K_d values between humans and dolphins could reflect a minor functional adaptation of the dolphin GLUT-1 protein.

Production of a sharp band of M_r 46,000 following treatment with glycopeptidase F suggests that dolphin red blood cell membranes contain GLUT-1 polypeptides very similar in molecular weight, but not in glycosylation, to the human GLUT-1 transporter polypeptide. Tryptic digestion of unsealed red blood cell membranes demonstrates that tryptic sites are present in dolphin GLUT-1 and that at least one of the tryptic sites is positioned within the molecule at a site(s) analagous to tryptic sites associated with a large cytoplasmic loop connecting two halves of the molecule in the human polypeptide. This finding is consistent with molecular studies of equilibrative glucose transporter polypeptides in other species, which have demonstrated that the GLUT-1 transporter isoform has been subject to rigorous conservation through mammalian evolution. The amino acid sequence of GLUT-1 from rat, pig, mouse, cow, and rabbit all show over 96% identity with the human GLUT-1 sequence (3).

At 6°C, rates of D-glucose entry in the presence of 40 µM cytochalasin B were too slow for quantitative estimation over periods of up to 5 min. At higher temperatures, some influx was apparent even in the presence of high (20-40 µM) concentrations of cytochalasin B. At 21°C, the rate of D-glucose entry from 12 mM D-glucose in the extracellular medium in the presence of 40 µM cytochalasin B was ~2% of the uninhibited rate (data not shown). At 37°C, entry of D-glucose into dolphin red blood cells treated with 40 µM cytochalasin B was faster than the entry of L-glucose, and D-glucose entry was reduced by addition of 2 mM HgCl₂, suggesting that at least part of the entry flux was via a mediated pathway. This stereospecific glucose transport may be a consequence of incomplete inhibition of dolphin GLUT-1 by cytochalasin B, some of which may be sequestered within the red blood cell. However, the presence of an additional mediated hexose transport pathway, with very low activity, cannot be ruled out. Taken together, these findings point to a remarkable functional similarity in the glucose transport properties of red blood cells from adult dolphins and adult humans. The existence of this physiological phenotype in an advanced cetacean species implies that the extraordinarily high glucose permeability of erythrocytes from adult humans and apes serves a physiological role and that it is not a quirk peculiar to the primate lineage.

Perspectives