Fasting is associated with a series of physiological responses that protect body tissues from degradation by efficiently using expendable energy reserves while sparing protein. Lactation requires the mobilization of maternal nutrients for milk synthesis. The rare life history trait of fasting simultaneous with lactation results in the conflicting demands of provisioning offspring while meeting maternal metabolic costs and preserving maternal tissues for her own survival and future reproduction. Certain tissues continue to require glucose for operation during fasting and might constrain tissue mobilization for lactogenesis due to a need for gluconeogenic substrates. This study investigated glucose flux, glucose cycle activity, and the influence of regulatory hormones in fasting lactating northern elephant seals. Measurements were taken early (5 days) and late (21 days) during the lactation period and, as a nonlactating comparison, after the completion of molting. Glucose cycle activity was highly variable in all study groups and did not change over lactation (P > 0.3), whereas endogenous glucose production decreased during lactation (t = −3.41, P = 0.008). Insulin and insulin-to-glucagon molar ratio decreased across lactation (t = 6.48, 4.28; P = 0.0001, 0.002), while plasma cortisol level increased (t = 4.15, P = 0.002). There were no relationships between glucose production and hormone levels. The glucose production values measured exceeded that predicted from available gluconeogenic substrate, indicating substantial glucose recycling in this species.
- glucose cycle
- Cori cycle
fasting brings about physiological responses that protect body tissues from degradation (11). At the onset of fasting metabolic rate decreases, reducing the amount of nutrients that must be catabolized to meet maintenance costs. With the progression of fasting, metabolic rate stabilizes at a suppressed level, carbohydrate and protein catabolism decline, and lipid oxidation increases to meet energetic requirements. These shifts deplete lipid reserves while sparing protein (33). In contrast, lactation requires key alterations in metabolic processes, including variation in hormone production and action, altered cardiac output, elevated metabolic rate, and mobilization of nutrients to mammary tissue for milk synthesis (1, 4, 24, 58). Lactation has been shown to be the most energetically expensive component of reproduction in many mammalian species (31). Most mammals increase food intake to compensate for these increased energy demands, and the timing of lactation is often correlated with food availability. Some domesticated animals have demonstrated a two- to threefold increase in food intake during lactation (84).
Some bears, mysticete whales, and phocid seals, as part of their natural life histories, combine long-duration fasting with the energetically costly activity of lactation. Many species of phocid seal separate marine feeding from terrestrial parturition. In these species, all nutrients necessary to nurse offspring are obtained on foraging trips before parturition (20). Milk is derived from body stores without replacement during lactation, and females experience considerable losses to blubber reserves and lean tissue (21, 66). This results in a conflict between the typical physiological response to fasting, which conserves body stores, and lactogenesis, which mobilizes body stores for investment in offspring.
Northern elephant seals (Mirounga angustirostris) provide an ideal system to examine the physiological adaptations that allow fasting adapted animals to resolve the conflicting demands of fasting and lactation. They have a long lactation period (∼27 days) compared with other phocids, and suckling pups consume large quantities of milk, ∼4.5 kg milk/day (67). The energetic content of the milk increases over lactation due to increasing lipid concentration (26), and the proportion of maternal energy reserves transferred to milk is one of the highest measured in nature (66). Lactating females may lose 40% of body mass (26), and pups may triple their mass by weaning (72). It has been suggested that reductions in nutrient reserves not only limit the duration of lactation in northern elephant seals, but also act as a threshold to energy expenditure during lactation (25). It follows that the efficiency by which nutrient reserves are utilized and partitioned regulates energy expenditure during lactation.
Previous studies in elephant seals have shown that plasma glucose levels remain high throughout fasting (22). Crocker et al. (25) estimated that protein catabolism increased over lactation, suggesting that elephant seals have a reduced ability to spare protein as lipid reserves are depleted with the progression of lactation. Increased protein utilization may be associated with an increased use of amino acids as gluconeogenic precursors. Fasting animals, however, typically exhibit decreased gluconeogenesis as peripheral tissues reduce glucose utilization and switch to fat-based metabolism (11), thus allowing glucose to be directed toward glucose-dependent tissues.
The regulation of glucose metabolism has been a point of interest in phocids for some time (22). Kirby and Ortiz (49) demonstrated that, when challenged with an injection of glucose, weaned elephant seal pups lacked an insulin response. They proposed that elephant seals may not regulate blood glucose by the normal mammalian insulin-glucagon push-pull model. High levels of circulating free fatty acids (FFAs), present in this species (10), are known to interfere with the regulation of glucose metabolism, probably by promoting insulin resistance (6, 77). Substrate cycles provide a mechanism by which total substrate flux can be altered in response to regulating factors (44) or substrate concentration (39). The enzymes of the glucose cycle (glucokinase and glucose-6-phosphatase) are sites potentially responsible for FFA-induced increases in hepatic insulin resistance and glucose production (53, 65). Variation in glucose cycle activity (GCA) may influence glucose metabolism in naturally hypoinsulinemic, hyperlipidemic northern elephant seals.
This study attempted to quantify the relationship between glucose kinetics and the duration of simultaneous fasting and lactation in the northern elephant seal. The goal of the study was to determine how the regulation of glucose availability relates to fuel partitioning under the physiological constraints of fasting and lactogenesis. Differentially labeled glucose was used to measure glucose production and GCA in a fasting-adapted species undergoing seasonal natural fasts. Ultrasound measurements of body composition were used to relate shifts in glucose production to the depletion of nutrient reserves and previously estimated rates of protein catabolism (25). Additionally, relationships between glucose production and plasma levels of regulatory hormones were explored.
Study site and subjects.
The study was conducted at Año Nuevo state reserve, San Mateo County, CA, during the 2003 elephant seal breeding season, January to March, and subsequent summer molt, May to June. All procedures were approved by the Sonoma State University Institutional Animal Care and Use Committee. During the breeding season, adult females were marked with hair dye (Lady Clairol, Stamford, CT) upon arrival at the rookery to facilitate identification throughout the study period. Experimental females were observed after parturition to ensure appropriate mother-pup bonding was established. During the lactation period, measurements were made in 10 females at 5 days postpartum (early lactation) and repeated 21 days postpartum (late lactation). All study animals exhibited normal lactation behavior and successfully weaned pups. In addition to the reproductive haul out, northern elephant seals haul out on land in the summer for 3–5 wk to replace their pelage (molt). As a fasting, nonlactating comparison, measurements were taken in 10 females after the completion of molting (postmolt).
Mass and body composition measurements.
Body composition was calculated using the truncated cones method (30). A portable ultrasound scanner (Ithaca Scanoprobe, Ithaca, NY) was used to measure blubber thickness. Dorsal, lateral, and ventral measurements were taken at each of six rings along the seal. Girth and distance from the tail were taken at each ring, and the total curved length of the study animal was measured. This technique has been validated against isotopic methods of measuring body composition and gave a mean error of 3% in estimating adipose tissue reserves (87). Mass was determined using a tripod and scale (MSI tension dynamometer, ±2 kg; Seattle, WA) in conjunction with a canvas sling (26). In two cases (seals M83 early and 61KT late lactation), mass was calculated using the morphometric measures taken for body composition due to mechanical failure of the scale or tripod. This method has been previously validated, and calculated masses were shown to be within 4% of measured masses (26). In this study, mass was calculated using morphometric measurements for weighed animals, and calculated masses were within 4.3 ± 3.6% of measured masses.
Sample collection and processing.
Seals were immobilized using an initial intramuscular injection of Telazol (teletamine/zolazepam HCl) at a dose of 1.0 mg/kg and administered intravenous doses of 100 mg ketamine and 5 mg diazepam as needed to maintain immobilization (all drugs from Fort Dodge Laboratories, Ft. Dodge, IA). Seals were kept awake and eupneic throughout procedures.
Using a bolus injection technique, a noncompartmental model was used to describe the glucose kinetics of each seal (91). To measure endogenous glucose production (EGP = gluconeogenesis + glycogenolysis) and GCA simultaneously, a differentially labeled glucose tracer was used. Each animal was administered 0.5 mCi each of [2-3H]glucose and [6-3H]glucose via the extradural vein at the onset of the flux measurement. After injection, blood samples were drawn at 5-min intervals for 30 min, then every 15 min thereafter until 3 h postinjection. Blood samples were collected in chilled, heparinized vacutainers, stored on ice, transported to the laboratory, and centrifuged for 15 min at 2,000 rpm and 4°C. The plasma was collected, and protein was precipitated from plasma by adding 1.5 ml each of barium hydroxide and zinc sulfate (0.3 N, Sigma-Aldrich, St. Louis, MO) to 1.0 ml plasma with subsequent vortexing and chilling for 20 min in an ice-water bath. Samples were then centrifuged at 3,000 rpm for 20 min, and the supernatant was decanted and stored at −80°C until further analysis.
To distinguish 3H at the second carbon (C-2) from the sixth carbon (C-6) of the glucose molecule, an enzymatic detritiation developed by Issekutz (40) and modified by Rooney et al. (75) was utilized to selectively remove the 3H from C-2. Deproteinated samples were thawed and passed through an ion exchange column containing cation resin (AG 50W-X8 200–400 mesh hydrogen form) and anion resin (AG 1-X8 200–400 mesh formate form; both resins from Bio-Rad Laboratories, Hercules, CA). The eluate was collected and lyophilized for 36 h to remove any 3H that had exchanged with the plasma water. Dried samples were reconstituted in 1.0 ml of 133 mM phosphate buffer (pH = 7.4) and aliquoted into four 200-μl fractions. Two fractions were detritiated, two were nondetritiated, and the remaining portion was used to determine glucose concentration of the sample. To each fraction that was to be detritiated, 500 μl of a detritiation solution were added. The detritiation solution consisted of 133 mM phosphate buffer, 8.4 mM ATP, 9.0 mM MgCl2, 2.4 U/ml hexokinase, and 10 U/ml phosphoglucose isomerase (all reagents from Sigma, St. Louis, MO). The pH of the final solution was adjusted to 7.4 with 1 M NaOH. To the nondetritiated aliquots, 500 μl of 133 mM phosphate buffer were added. All samples were incubated in a shaker water bath at 37°C for 2 h and were subsequently lyophilized overnight. Samples were reconstituted in 500 μl of 1.0 N H2SO4. Scintillation cocktail (6.5 ml; Econolite, Fisher, Pittsburgh, PA) was added to each sample, and the sample was then agitated for 1 min. Sample activity was determined by liquid scintillation on a Beckman LS 3801 scintillation spectrophotometer (Beckman, Fullerton, CA) using standard scintillation techniques. A quench correction factor was established for each sample based on a calculated H number using a series of 3H standards with variable degrees of quench. Glucose concentration was measured in duplicate on an YSI 2300 glucose autoanalyzer (YSI, Yellow Springs, OH), and the specific activity of counted samples was determined. Since the nondetritiated aliquots contain [2-3H]glucose and [6-3H]glucose, and the detritiated fractions contain only [6-3H]glucose, the specific activity of [2-3H]glucose (SA2) was determined by the equation where SAtotal is the sum of the specific activities resulting from the dual label, and SA6 is the specific activity attributed only to [6-3H]glucose.
Single-label tritiated glucose standards were run in parallel with samples to determine the degree of detritiation of each isotope and correct for detritiation efficiency. Average detritiation of [2-3H]glucose was 97.9 ± 0.5%. Within each assay, detritiation efficiency was corrected for by multiplying sample SA6 by the [2-3H]glucose standard detritiation efficiency. Average detritiation of [6-3H]glucose was <1.0%. Detritiation of [6-3H]glucose standards ranged from −2.5 to 3.8%; therefore, detritiation efficiency of less than 3 % was not corrected for, while efficiencies of 3.0–3.8% were adjusted for as above.
Plasma glucose concentration remained stable throughout the procedures, and steady-state equations were used, assuming that the rate of disappearance (Rd) of substrate matches its rate of appearance (Ra). The Ra of glucose, measured by the dilution of isotopically labeled glucose by unlabeled glucose produced over time, was determined by dividing the dose injected by the area under the clearance curve where Ra is of unlabeled glucose, Dosedpm is the radioactivity of the injected tracer in disintegrations per minute (dpm), and y(t) is the exponential function describing the decay of tracer-specific activity with respect to time (91). Two exponential functions were fit to the clearance curve by maximizing the r2 value for each function. Curve-fitting and integration were performed using the software program Mathematica (Wolfram Research, Champaign, IL). The volume of administered tracer was determined by gravimetric calibration of the injection syringe for each study animal.
In vivo, hydrogen at C-2 of glucose is removed in the conversion of glucose-6-phosphate to fructose-6-phosphate. Therefore, endogenous glucose output (EGO = gluconeogenesis + glycogenolysis + GCA) was measured as the Ra calculated with respect to [2-3H]glucose (Ra2). Hydrogen at C-6 is not removed until the phosphoenolpyruvate and citric acid cycles, so EGP (gluconeogenesis + glycogenolysis) was measured as the Ra calculated with respect to [6-3H]glucose (Ra6). GCA was calculated as the difference between EGO and EGP (GCA = Ra2 − Ra6). The contribution of glycogenolysis to glucose release during a prolonged fast was assumed negligible, since glycogen stores are rapidly used during the first days of fasting. Therefore, by removing glycogenolysis from the definitions of EGO and EGP presented, EGO can be redefined for the measurement periods made in this study as gluconeogenesis plus GCA, whereas EGP will represent exclusively gluconeogenesis.
Hormone and metabolite analysis.
Plasma samples drawn before tracer injection were thawed for use in assays of insulin, glucagon, cortisol, β-hydroxybutyrate (β-HBA), glycerol, lactate, and glucose. Insulin was assayed using a Sensitive Rat Insulin RIA kit (cat. no. SRI-13K), glucagon was assayed with a Glucagon RIA kit (cat. no. GL-32K; both kits from Linco Research Inc., St. Charles, MO), and cortisol levels were measured using a Cortisol RIA kit (cat. no. TKCO2, Diagnostic Products, Los Angeles, CA). All of these kits have been validated previously for this species (14, 69). The mean intra-assay coefficient of variation was 9.7, 8.9, and 8.1% for insulin, glucagon, and cortisol, respectively. Since insulin and glucagon are antagonistic and it is believed that their molar ratios determine their metabolic effect rather than absolute plasma concentrations (50), values of insulin and glucagon were used to calculate the insulin-to-glucagon molar ratio (I/G, assuming molecular weights of 5,800 and 3,483 for insulin and glucagon respectively). β-HBA, glycerol, and lactate were assayed in duplicate using a GM-7 Micro-Stat autoanalyzer (Analox Instruments, Lunenburg, MA). Plasma glucose was measured in triplicate using an YSI 2300 glucose autoanalyzer (YSI, Yellow Springs, OH).
Unexpectedly, there was no relationship between glucose production and mass within any study group (Fig. 1). Because mass does not affect glucose production, no attempt was made to control for body mass (i.e., use mass-specific values) in statistical comparisons. There is concern regarding the improper use of ratios in reporting metabolic data (71); however, we also report mass-specific values for comparisons with other species and previous research. GCA is reported as mg glucose cycled/min, as well as a proportion of glucose production. GCA proportional to glucose production (pGCA = GCA/EGP) has been used previously to represent the activity of this substrate cycle relative to glucose production (88). Early-lactation and late-lactation measurements were compared using paired t-tests on matched early- and late-lactation measurements. Comparisons of lactation means to postmolt means were made using separate t-tests. Regression analyses were performed within the groups for early-lactation, late-lactation, and postmolt samples, or across lactation for differences calculated between early- and late-lactation values.
Mass loss and body composition changes.
Female mass decreased from 489 ± 60 kg during early lactation to 368 ± 57 kg at the end of lactation (paired t = 15.7, P = 0.0001). Lactating females lost 24.8 ± 4.9% of initial mass at a rate of 7.5 ± 1.5 kg/day. The mean mass of molted females was 306 ± 31 kg. There was a significant mass difference between molted females and both early- and late-lactation females (t = 8.6, 3.1; P = 0.0001, 0.007; respectively). Lactation was associated with significant alterations in body composition. Percent adipose tissue decreased from 34.5 ± 1.3% during early lactation to 29.1 ± 1.3% late in lactation (paired t = 9.5, P = 0.0001). In molted females, adipose tissue comprised 30.1 ± 1.3% of body mass. Early-lactation females had significantly higher proportions of adipose tissue than molted females (t = 7.6, P = 0.0001), but there was no significant difference in body composition between late-lactation and molted females (P > 0.1).
Total EGO (EGOtotal) during early lactation was 750 ± 134 mg glucose/min, during late lactation was 657 ± 207 mg glucose/min, and was 765 ± 128 mg glucose/min in molted animals. There was a trend toward decreased EGOtotal across lactation (paired t = −2.10, P = 0.066), while EGOtotal of molted animals was not significantly different from early or late lactation (P > 0.2).
EGP values are summarized in Fig. 2. Total EGP (EGPtotal) during early lactation was 707 ± 124 mg glucose/min, during late lactation was 598 ± 147 mg glucose/min, and was 688 ± 101 mg glucose/min in molted animals. Total glucose production significantly decreased across lactation (paired t = −3.41, P = 0.008). EGPtotal of molted animals was not different from early or late lactation (P > 0.1). There was no correlation between body composition (%adipose) and EGPtotal within any study group (P > 0.2). There was, however, a strong relationship between the change in glucose production and the change in body composition (ΔEGPtotal = 4,545.2 Δ% body composition − 352.2; r2 = 0.61, P = 0.008, Fig. 3).
Mass-specific EGP (EGPMS) during early lactation was 1.47 ± 0.33 mg·kg−1·min−1, during late lactation was 1.65 ± 0.44 mg·kg−1·min−1, and was 2.28 ± 0.48 mg·kg−1·min−1 in molted animals. There was a statistically significant increase in EGPMS across lactation (paired t = 2.39, P = 0.04). Molted animals had higher EGPMS than lactating animals (t = −4.40, −3.05; P = 0.0004, 0.007, early and late lactation, respectively).
GCA was highly variable in fasting seals. Mean GCA during early lactation was 43.8 ± 31.4 mg glucose/min, late-lactation GCA was 61.5 ± 79.2 mg glucose/min, and was 76.4 ± 37.3 mg glucose/min in molted animals. Mean GCA was 6.1 ± 4.8% of EGP during early lactation, 9.0 ± 11.1% during late lactation, and was 10.8 ± 4.6% after molting. There was no difference between the absolute or proportional GCA of early- and late-lactation or molted seals (P > 0.3).
Early in lactation, pGCA decreased with plasma glucagon (pGCA = −0.003 glucagon + 0.192; r2 = 0.52, P = 0.019; Fig. 4) and increased with I/G (pGCA = 0.06 I/G − 0.03; r2 = 0.41, P = 0.04). Proportional GCA decreased with glucagon in molted animals (pGCA = −0.0022 glucagon + 0.213; r2 = 0.43, P = 0.04). In summary, GCA decreased with glucagon and increased with I/G during early lactation and decreased with glucagon postmolting, but these relationships were not evident late in lactation.
Hormones and metabolites.
Levels of plasma glucose, β-HBA, glycerol, and lactate are summarized in Table 1. Plasma glucose level increased during lactation (paired t = 3.15, P = 0.012). There was a strong inverse relationship between plasma glucose and lactate levels among molted animals (r = −0.91, P < 0.001), but not during lactation (P > 0.2). Mean plasma β-HBA level increased during lactation (paired t = 6.7, P = 0.0001).
Mean plasma insulin, glucagon, I/G, and cortisol levels are displayed in Table 1. There was a significant decrease in insulin concentration across lactation (paired t = 6.48, P = 0.0001), while glucagon showed no significant change (paired t = 2.14, P = 0.06). I/G significantly decreased during lactation (paired t = 4.28, P = 0.002), while plasma cortisol level doubled (paired t = 4.15, P = 0.003). Surprisingly, plasma glucose concentrations were not related to insulin, glucagon, or I/G within any group (P > 0.1); however, plasma glucose levels increased with cortisol concentration early in lactation (r = 0.65, P = 0.041) but not late in lactation or in molted animals (P > 0.3).
Rates of glucose production were high in elephant seals compared with nonfasting adapted species and decreased significantly across lactation. While there was a statistically significant increase in mass-specific glucose production over lactation, the lack of a relationship between mass and glucose production in any of the sampling periods strongly suggests that ratios are an inappropriate way to control for effects of body mass. Decreased glucose production is common in long-duration fasting (14, 29, 35, 42, 45, 64, 79, 81), but few studies have explored fasting concomitant with lactation. Protein catabolism increases with lactation duration in northern elephant seals (25), while the protein content of milk declines (26), presumably increasing the availability of gluconeogenic amino acids. This was not reflected, however, in increased EGPtotal late in lactation. The magnitude of the reduction in glucose production was correlated with adipose loss (Fig. 3), a relationship that was also exhibited by pups during the postweaning fast (14). Females maintaining adipose reserves had the largest reductions in EGPtotal, suggesting that glucose production is related to rates of lipid catabolism. FFAs have been shown to increase gluconeogenesis (15, 16), but not necessarily hepatic glucose production (gluconeogenesis + glycogenolysis) (16, 52). The compensatory changes in gluconeogenesis or glycogenolysis to maintain a static glucose production rate is termed hepatic autoregulation. Disruption of hepatic autoregulation has been demonstrated during 1) glycogen depletion (80), 2) FFA-induced insulin resistance, either by translocation of protein kinase C (54) or allosteric stimulation of glucose-6-phosphatase (53), and 3) in type II diabetics (5), possibly due to impaired insulin secretion or reduced liver glycogen. Thus the potential depletion of glycogen reserves, high-circulating FFA, and hypoinsulemia, characteristic of fasting elephant seals, may disrupt hepatic autoregulation and influence glucose production in elephant seals.
Mass-specific glucose production values of lactating females were 1.5–1.7 mg glucose·kg−1·min−1; this is markedly lower than the 2.3 mg·kg−1·min−1 in molted females, as well as the 2.2–2.8 mg·kg−1·min−1 we reported previously in fasting while developing elephant seal pups (14). Molted females and weanlings exhibited similar rates of EGPMS; rates of molted females were not significantly different from late-fasting weanlings (P > 0.05). Few studies have measured glucose production in truly prolonged fasting (see Table 2); however, in all physiological states measured, grey seals and northern elephant seals have rates of glucose production similar to or higher than rates of other species undertaking similar duration fasts. This was unexpected since reduced gluconeogenesis facilitates efficient protein sparing, a hallmark of fasting adaptation. Katz and Tayek (45) measured glucose production rates of 1.99–2.35, 1.81, and 1.77 mg·kg−1·min−1 in 12-, 20-, and 40-h fasted humans, respectively. After 3–5 wk of fasting, Streja and colleagues (81) measured 0.88 mg·kg−1·min−1 glucose production; after 1-mo fasting in dogs, glucose production was suppressed to ∼1.4 mg·kg−1·min−1. Thus the northern elephant seal, in a state of highly efficient protein sparing and undertaking a long duration fast, did not suppress EGP to the levels seen in humans or dogs.
Most mammalian milk contains significant amounts of carbohydrate, much of which is lactose and the primary source is plasma glucose (82). Elevated glucose production is likely necessary to support lactogenesis in many species. Burnol and colleagues (7) measured increased gluconeogenesis during lactation in the rat, and Tigas et al. (86) showed glucose production rates in lactating women were 33.5% higher than in nonlactating women after a 22- to 24-h fast. Additionally, in ruminants, and presumably most mammals, the increased glucose output during lactation is accompanied by reductions in glucose use by peripheral tissues (2, 3). However, the milk of elephant seals, like most pinnipeds, is largely absent of carbohydrate (26, 66). Despite the lack of carbohydrate found in milk, lactating elephant seals demonstrated more than twice the mass-specific rates of gluconeogenesis compared with lactating women (0.66 mg glucose·kg−1·min−1) undergoing a 24-h fast (86).
During the early-lactation measurement, the average Ra of glucose was 707 mg/min; this results in 5.65 mol (almost exactly 1 kg) of glucose produced per seal each day. This exceeded our expectations for an animal meeting 90% of its energy needs from fat catabolism (26). The Ra of urea was measured by Crocker et al. (25) in lactating elephant seals and used to estimate that 245-g protein were catabolized per day during midlactation. If we assume that all of this catabolized protein is directed toward gluconeogenesis, and 0.57 g glucose are produced per 1.0 g protein (51), then amino acids can account for ∼14% of EGP early in the lactation period. Later in lactation, glucose production decreases while protein catabolism increases; thus the amino acid contribution may increase during lactation, up to 24%. These are maximal contributions, and actual contribution is likely significantly lower. Recent studies of the contribution of glycerol toward gluconeogenesis suggest that <5% of glucose production derives from glycerol (38). Together, these estimates suggest a high degree of glucose recycling.
Erythrocytes are a significant glucose-dependent tissue in this species. Adult females have, on average, 212 ml blood/kg and maximal hematocrit levels >60% (85). Assuming this mass-specific blood volume and a hematocrit of 60%, a red blood cell (RBC) volume of 61 liters was calculated. Using rates of glucose consumption measured in RBCs by Castellini et al. (9), and correcting for the temperature dependence of enzymatic activity, it is estimated that RBCs require ∼700 nmol glucose·ml RBC−1·h−1; consequently, 1.02 mol glucose are consumed by RBCs per day. Estimates of glucose consumption would be lower if a lesser hematocrit value were assumed, which is a possibility if some portion of the RBC supply were sequestered by the spleen during the glucose kinetics measurement.
The primary mechanism by which erythrocytes meet metabolic costs is the breakdown of glucose to lactate. Produced lactate may subsequently act as a gluconeogenic precursor, with resulting glucose released back into circulation. This glucose-lactate-glucose cycling was first hypothesized by Cori (19) as a process of importance in the regulation of blood glucose. Subsequent tracer studies have established that equilibration of lactate, pyruvate, and alanine occurs (47). Fifty years after his original formulation, Cori (18) expanded his concept of the glucose-lactate cycle into a pyruvate cycle, and the term “Cori cycle” has been retained for all glucose recycling through pyruvate (45). Owen et al. (70) proposed that Cori cycle activity may be important in fasting humans, and glucose oxidation findings of Keith and Ortiz (48) in weaned elephant seal pups led them to suggest that the Cori cycle was the primary mechanism of recycling of radioactive carbon in fasting elephant seal pups. Davis (27) found that Cori cycling accounted for all of the glucose recycling in postabsorptive harbor seals. The effect of fasting on the Cori cycle varies, and the correct calculation of such has been controversial. Early studies utilized [14C]glucose and [3H]glucose and let the difference in Ra equal glucose recycling (e.g., R− = Cori cycle). Estimates of Cori cycle in rodents were 19–35% of glucose production (28, 29, 79) and 12–20% in humans (73). More recent studies use stable isotopes and mass isotopomer distribution analysis (46, 57). Measures of Cori cycle activity in fasted rats remained ∼15% of glucose production (47) and in humans after an overnight fast ∼20% (45, 83), increasing to ∼36% of glucose production after 40-h fasting (45). Landau and colleagues' (57) calculations of gluconeogenesis and Cori cycle values were lower than others, and they identified some potential inherent methodological concerns in their estimation of Cori cycle activity as 20% of glucose production after 60-h fasting (55). In most cases, the increase in proportional Cori cycle activity was facilitated by the reduction in glucose production with time fasting.
In the fasting state, the energy-to-fuel gluconeogenesis is likely supplied through fatty acid oxidation. Glucose that is released following hepatic gluconeogenesis could be utilized by erythrocytes and renal medulla via the Cori cycle. This allows glucose to act as an ATP shuttle between fat oxidation in liver and glycolysis in glucose-dependent tissue. By recycling glucose in this manner, elephant seals may provision erythrocytes and the renal medulla using the energy from fat. Measured lactate values were slightly higher than that observed in other phocids (43, 89); this is, however, the first time that plasma lactate levels have been reported for lactating elephant seals. The circulating lactate levels presented are consistent with our suggestion of increased Cori cycle activity. The degree of cycling suggested by the glucose flux measurements made here implies an energetically inefficient system and suggests that protection of lean tissue is more important than energetic efficiency. Goodman et al. (32) as well as Henry et al. (36) suggested that suppression of metabolic rate during long-term fasting is essential to facilitate protein sparing. Sparing of amino acid precursors may take priority over energy efficiency during extended fasts, provided extensive lipid energy reserves are available. Further research is needed to identify the substrates contributing to gluconeogenesis to determine the extent of recycling and the contribution of amino acids.
This glucose recycling model is reminiscent of early diving physiology hypotheses, notably that of Hochachka et al. (37) (also see Ref. 8), who postulated that a modified Cori cycle operated within central circulation between brain, heart, and lung during diving. Elevated levels of lactate dehydrogenase are common in pinnipeds (12), and the requirements of clearing products of glycolysis during dives exceeding aerobic capacity may contribute to elevated levels of glucose recycling in the fasting state.
It is possible to increase glucose availability in the face of static glucose output, if GCA were decreased. The increase in plasma glucose concentration across lactation observed in this study, however, was not due to alterations in GCA. GCA did not change across lactation or between lactation and postmolt groups. GCA during early lactation ranged from 0 to 15% of glucose production. This is very similar to GCA rates of other species studied (34, 40, 61, 75, 78). Previous studies are conflicting on the glucose cycle's response to alterations in hormone levels. Several studies have found increases in GCA in response to glucocorticoids (40), as well as both glucagon (59) and insulin (75). This study found a decline in pGCA with glucagon. A reduction in pGCA may increase glucose availability to tissues if glucose output were held constant, thereby magnifying the effect of glucagon via this substrate cycle. Glucagon varied from 31 to 64 pg/ml during early lactation; using the regression equation from Fig. 4, the relative effect of glucagon on GCA can be examined. A glucagon concentration of 40 pg/ml corresponded to a pGCA of 0.07; increasing the glucagon level by 50% to 60 pg/ml would result in a pGCA of 0.01, increasing EGO by 42.4 mg/min, ∼5% of EGO. Thus it would require large changes in glucagon to elicit relatively small changes in glucose flux via GCA. This does not fit with the model of substrate cycles responding to small alterations in hormone levels to have large effects on overall substrate flux (63). In a previous study of weaned elephant seals, we found no relationship between hormone levels and pGCA (14), and Weber et al. (88) argued that no changes in proportional GCA had been reported in response to regulatory factors. This study reports a relationship between proportional GCA and glucagon, as well as I/G. There was no difference, however, in glucagon levels or GCA across lactation or in postmolt animals, and there remains no clear evidence of the glucose cycle actively regulating glucose availability in fasting elephant seals.
Metabolites and hormonal regulation.
This study found no correlation between plasma concentrations of insulin and glucagon. Additionally, neither hormone nor I/G were related to plasma glucose level or rates of gluconeogenesis. Previous work found decreased plasma glucose level and I/G (14, 68) with time fasting in weaned elephant seal pups, while in the present study plasma glucose concentration increased during lactation. The decrease in I/G of pups was primarily due to increased glucagon, whereas the decreased I/G found in lactating seals in this study was caused by decreased insulin levels. Thus these different age classes affect a similar change in I/G by varying different hormones with the progression of fasting. EGP decreased during the postweaning fast and lactation, but plasma glucose increased in lactating females. Reduced glucose production concomitant with increased plasma glucose concentration implies reduced uptake by peripheral tissues, even during increased glucose availability. During lactation, tissue-specific alterations in hormone receptor numbers have been documented in dairy cattle (24). If significant changes in receptor concentrations occur over the study period, then I/G may become less important and individual hormone levels may predominantly influence metabolic changes. Additionally, alterations in receptor concentrations may affect glucose uptake by nontarget tissues. Insulin decreased substantially across lactation, while plasma glucose levels increased, and, paradoxically, decreased insulin levels were accompanied by decreased glucose production. Insulin, glucagon, I/G, or changes in each of these variables across fasting were not related to plasma glucose within any study group (early- or late-lactation or postmolt groups). Previous researchers have speculated that elephant seals do not regulate blood glucose via the standard insulin-glucagon model. This study explored the glucose cycle for possible regulatory properties within glucose metabolism and found none. Few relationships between hormone level and plasma glucose concentration or the rate of EGP were found. The mechanism by which elephant seals regulate blood glucose levels remains unresolved.
Increases in β-HBA with time fasting have been measured previously in this species (13), but the levels remain well below the clinical definition of ketoacidosis. Possible explanations for the lack of ketone accumulation in fasting northern elephant seals include 1) ketone production has been suppressed, 2) β-HBA and acetoacetate are efficiently utilized by tissues, or 3) acetoacetate is decarboxylated to acetone and enters the methylglyoxyl pathway. Given the high rate of lypolysis (10), extreme suppression of ketone production seems unlikely. Nordøy and Blix (64) argued that ketone utilization facilitates protein sparing in fasting grey seals, a mechanism that is observed in other fasting mammal models but that requires a state of ketoacidosis (62). Alternatively, acetone may be converted to lactate and fuel gluconeogenesis, as suggested by Schweigert in grey seals (76). Acetone has been shown to contribute to gluconeogenesis in mice (17) and may be responsible for as much as 11% of plasma glucose production in fasting humans (74). In a fasting animal with abundant lipid reserves, conversion of beta-oxidation by-products to glucose may be advantageous. The conversion of acetoacetate to acetone and subsequent contribution to gluconeogenesis via the methylglyoxylate pathway may be a mechanism to avoid ketoacidosis.
Possible confounding factors.
Placing the results of the glucose flux measurements into proper context cannot be done without some consideration of potential confounding factors. One such factor is the use of immobilizing drugs during the measurement period. There was a large degree of variation in the total amount of drugs required to maintain immobilization, from 10.4 to 31.5 ml of ketamine; initial immobilization with tiletamine/zolazepam was standardized to animal mass. There was no significant relationship of ketamine administered with EGPtotal (r2 = 0.01–0.05; P > 0.5) or EGPMS (r2 = 0.05–0.10, P > 0.3). No correlation was observed under a similar study in weaned elephant seal pups (14), suggesting that any effect from drugging had minimal impact on the study results. Another potential confound may be the incomplete removal of tritium from the administered tracers via the predicted pathways, or loss via unaccounted for pathways. Errors due to these processes may cause EGP to be underestimated by 3% and EGO by as much as 20% (56). There are no data on in vivo loss of label in seals, and we report uncorrected data in this study with the understanding that these values represent conservative estimates of both GCA and EGP.
In this study, high levels of glucose production were observed with levels decreasing across the fast. GCA varied with glucagon and I/G early in lactation but did not appear to play an important role in regulating glucose availability during fasting. The mechanisms of plasma glucose regulation require further investigation. β-HBA increased slightly during lactation, but remained well below concentrations that would be considered ketoacidotic. Although the glucose production rates during lactation were lower than after molting or compared with those of weaned pups, values were high relative to glucose utilization requirements and available substrates. This suggests elevated rates of glucose recycling, possibly by increased Cori cycle activity or other recycling pathways, in fasting while lactating elephant seals.
This material is based upon work supported by the National Science Foundation under Grant 0213095. Work was conducted under National Marine Fisheries Service Marine Mammal permit no. 786–1463.
We thank Dr. P. M. Bell for generous assistance with laboratory protocol of the detritiation process, A. Kosztowny for help in laboratory and field procedures, Año Nuevo Rangers for logistical support, and Clairol, Inc. for providing marking solutions. This manuscript was greatly improved by the suggestions of two anonymous reviewers.
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. Section 1734 solely to indicate this fact.
- Copyright © 2006 the American Physiological Society