Disposition of a Glucose Load into Hepatic Glycogen by Direct and Indirect Pathways in Juvenile Seabass and Seabream

In carnivorous fish, conversion of a glucose load to hepatic glycogen is widely used to assess their metabolic flexibility towards carbohydrate utilization, but the activities of direct and indirect pathways in this setting are unclear. We assessed the conversion of an intraperitoneal glucose load (2 g.kg−1) enriched with [U-13C6]glucose to hepatic glycogen in juvenile seabass and seabream. 13C-NMR analysis of glycogen was used to determine the contribution of the load to glycogen synthesis via direct and indirect pathways at 48-hr post-injection. For seabass, [U-13C6]glucose was accompanied by deuterated water and 2H-NMR analysis of glycogen 2H-enrichment, allowing endogenous substrate contributions to be assessed as well. For fasted seabass and seabream, 47 ± 5% and 64 ± 10% of glycogen was synthesized from the load, respectively. Direct and indirect pathways contributed equally (25 ± 3% direct, 21 ± 1% indirect for seabass; 35 ± 7% direct, 29 ± 4% indirect for seabream). In fasted seabass, integration of 2H- and 13C-NMR analysis indicated that endogenous glycerol and anaplerotic substrates contributed an additional 7 ± 2% and 7 ± 1%, respectively. In fed seabass, glucose load contributions were residual and endogenous contributions were negligible. Concluding, direct and indirect pathways contributed equally and substantially to fasting hepatic glycogen repletion from a glucose load in juvenile seabream and seabass.

glycogen from non-glucose precursors such as glycerol and gluconeogenic amino acids, which are abundant precursors in the typical high protein diets of carnivorous fish.
In seabass, analysis of hepatic glycogen position 5 enrichment from deuterated water ( 2 H 2 O) was used to estimate indirect pathway contributions to glycogen synthesis 13 . With this approach, it was shown that essentially all hepatic glycogen synthesis during feeding with conventional fishmeal diet occurred via indirect pathway. However, this contribution could drop to ~70% if the fishmeal was partially replaced with digestible starch.
To our knowledge, direct and indirect pathway contributions to hepatic glycogen synthesis following a glucose load have not been reported for any fish species. This setting is expected to maximize direct pathway activity due to the high availability of glucose that in turn promotes insulin secretion 8 and activation of hepatic glycogen synthesis 9 . A glucose load can be enriched with [U- 13 C 6 ]glucose whose incorporation into hepatic glycogen can be used to independently measure direct and indirect pathway contributions of the glucose load to glycogen synthesis 14 . When [U- 13 C 6 ]glucose is used in combination with 2 H 2 O, indirect pathway contributions from non-glucose precursors can also be accounted for. NMR methodology can resolve glycogen 2 H-and 13 C-enrichment from 2 H 2 O and [U- 13 C 6 ]glucose allowing both tracers to be given simultaneously [14][15][16] . Taking this approach into consideration, we aimed at quantifying hepatic glycogen synthesis from a glucose load via both direct and indirect pathways in seabass and seabream by using an enriched glucose load with [U- 13 C 6 ]glucose and quantifying the 13 C-isotopomer distributions of glycogen. Also, we aimed at determining the contributions of both glucose and non-glucose precursors to hepatic glycogen synthesis in seabass following a glucose load by integrating [U-13 C 6 ] glucose and 2 H 2 O tracers. The results generated from 13 C-isotopomer and 2 H-enrichment data were incorporated into a comprehensive metabolic model describing hepatic glycogen synthesis from all major substrate precursors, represented by Fig. 1.

Materials and Methods
Fish handling. Twenty four juvenile gilthead seabream (S. aurata L.) were maintained in two 200 L aquariums with recirculating water filtered by an external filter. Fish had a mean body weight of 27.6 ± 1.4 g. Fish were acclimated in a recirculating aquaculture system (RAS) equipped with a central filter and UV unit at 20 °C, 30‰ salinity and well aerated seawater (>90% O 2 ). Water temperature, pH, dissolved oxygen and salinity were continuously monitored and nitrogenous waste products were measured twice weekly and were found to be within optimal ranges. During acclimation fish were fed daily until apparent satiety with a commercial diet (Dourasoja Ultra Energia 2, SORGAL, S.A.; 45.2% crude protein, 20.0% fat, 9.5% ash, 1.5% fiber, 1.0% phosphorus). Therefore, seabream were then fasted for 21 days.
Eighty juvenile European seabass (D. labrax L.), with body weight of 44.5 ± 3.1 g were randomly distributed into four 200 L tanks (n = 20 each). Fish were acclimated under the same conditions as previously described for seabream. Seabass were fed daily to apparent satiety with Aquagold 5 fish-feed, SORGAL, S.A. The diet was composed of 45.0% crude protein, 18.0% fat, 10.0% ash, 2.0% fiber, 2.0% calcium, 1.0% phosphorus, 0.4% sodium. After acclimation, two groups were fasted for 21 days while the others continued with the described feeding regime. Tracer delivery and sampling. At the end of the 21-day feeding/fasting period, both seabream and seabass were anesthetized in small seawater tanks containing 0.1 g.L −1 of MS-222. They were then measured and weighed and fish from half of the groups for each species were injected intraperitoneally with 2 g.kg −1 glucose enriched to 10.0-12.5% with [U-13 C 6 ]glucose and transferred back to the main rearing tanks for 48 hours without feeding. For the seabass, the seawater was also enriched to 3.5% with 2 H 2 O during the 48 hour post-injection period. Fish from the other half of the groups were injected with saline to assess hepatic glycogen levels in the absence of a glucose load. After this period, the fish were again anesthetized in seawater tanks containing 0.1 g.L −1 of MS-222 and also enriched to 3.5% with 2 H 2 O in the case of seabass. Fish were measured, weighed and livers were extracted and freeze clamped in liquid nitrogen and then stored at −80 °C until further processing. After sacrifice by cervical section perivisceral fat was collected and weighed. Enriched seawater with 2 H 2 O from the seabass was collected to confirm its enrichment. All the above procedures were in compliance with the Guidelines of the European Union Council (2010/63/EU) after approval by the Animal Care Committee of the University of Coimbra.
Hepatic glycogen extraction and derivatization. In order to insure glycogen samples that yielded 2 H-and 13 C-NMR spectra with high signal-to-noise ratios, livers from 2-3 fish were pooled. Glycogen extraction, quantification, hydrolysis to glucose and derivatization to monoacetone glucose (MAG) was as previously described 13 . MAG was further purified by solid phase extraction using Discovery ® DSC-18 3 mL/500 mg disposable columns (Sigma-Aldrich). Columns were activated with 3 mL acetonitrile followed by 10 mL distilled water. The MAG preparation was dissolved in 1 mL distilled water and added to the column. The column was then washed with 0.5 mL distilled water to remove impurities and the MAG was eluted with 2.5 mL of 10/90 v/v acetonitrile/water and the solvent then evaporated.
NMR spectroscopy and analysis. Proton-decoupled 13 C-NMR spectra of MAG derived from seabream were acquired at 14.1 T with a Varian VNMRS 600 MHz NMR (Agilent, Santa Clara, CA, USA) spectrometer equipped with a 3 mm broadband probe with z-gradient. Spectra were obtained at a temperature of 25 °C, using a 70° pulse, 2.5 s acquisition time, 0.5 s pulse delay and 5,000 to 15,000 free-induction decays (f.i.d.). 13 C-NMR spectra of MAG derived from seabass were obtained with a Bruker Avance III HD 500 (Bruker Biospin, Billerica, MA, USA) spectrometer equipped with a BBFO 5 mm broadband probe using the same pulse parameters and f.i.d. numbers. Seabass 13 C NMR spectra were obtained following 2 H-NMR acquisition and evaporation of the acetonitrile/water solvent mixture (see below) and resuspension of the MAG residue in 0.5 ml 99.9% 2 H 2 O. Quantification of the glycogen 13 C-isotopomers representing direct pathway and indirect pathway conversion of [U-13 C 6 ]glucose to glycogen was performed by deconvolution of the MAG carbon 1 resonance multiplet as previously described 14,17 . The two methyl natural-abundance carbon signals of MAG were used as intramolecular standards 18 for converting the carbon 1 13 C isotopomer multiplet areas into 13 C-enrichment values.
Proton-decoupled 2 H-NMR spectra of MAG samples obtained from seabass were also obtained with a Bruker Avance III HD 500 spectrometer using a 2 H-selective 5 mm probe incorporating a 19 F-lock channel. Samples were resuspended in 0.5 ml 90% acetonitrile/10% 2 H-depleted water to which 50 μl of hexafluorobenzene were added. 2 H-NMR spectra were obtained with a 90° pulse angle, 1.6 s of acquisition time and a 0.1 s interpulse delay. The number of f.i.d. collected ranged from 5,000-20,000. Positional 2 H-enrichments were determined using the MAG methyl signals as an intramolecular standard 19 . To quantify fish plasma body water and seawater 2 H-enrichments, triplicate 10 µL samples of plasma were analyzed by 2 H NMR as previously described 20 but with 50 μl of hexafluorobenzene added to the NMR sample. 13 C-and 2 H-NMR spectra were analyzed with ACD/NMR Processor Academic Edition software (ACD/Labs, Advanced Chemistry Development, Inc.).
Theory and metabolic flux calculations. 13 C Glycogen enrichment from [U- 13 C 6 ]glucose. Analysis of glycogen 13 C-isotopomer distributions after an intraperitoneal injection of [U-13 C 6 ]glucose informs the fractional rates of glycogen production from the glucose load via both direct and indirect pathways 14,21 . Direct pathway conversion of glucose to glycogen via G6P does not cleave the hexose carbon skeleton thus [U-13 C 6 ]glucose is converted into [U-13 C 6 ]glucosyl units of glycogen. The direct pathway fractional contribution is calculated according to the following equation: 13 6 g F g represents the percent [U-13 C 6 ]glucose enrichment of the glucose load. The indirect pathway metabolism of glucose to glycogen involves its initial conversion to pyruvate via glycolysis followed by the re-formation of G6P via anaplerosis and gluconeogenesis and subsequent conversion to glycogen. During this sequence, the [U-13 C 6 ]glucose skeleton is cleaved and the 13 C-label also undergoes both dilution and randomization to form partially-labeled glycogen 13 C-isotopomers distributed within each triose half (see Fig. 1). The most prevalent are [1,2-13 C 2 ]-and [1,2,3-13 C 3 ]glycogen, with corresponding [5,6-13 C 2 ]-and [4,5,6-13 C 3 ]glucosyl isotopomers from the other triose half 22 . Assuming that the two triose halves have equivalent 13 C-isotopomer distributions, the fractional contribution of the indirect pathway to the glycogen pool can be calculated from the C123 triose moiety -whose component 13 C-isotopomers have better-resolved 13 C-NMR signals -as follows. The factor of 1.5 accounts for dilution of 13 C enrichment at the level of the Krebs cycle and is based on the assumption that net anaplerotic outflow is ~2-fold that of Krebs cycle oxidative flux, as has been measured in both rodents 23 and humans 24 . The fraction of hepatic glycogen that was synthesized during the period that the fish were administered with [U-13 C 6 ]glucose and 2 H 2 O tracers, defined as newly synthesized glycogen, was calculated as follows:

= +
Newly synthesized glycogen (%) Direct pathway fraction all indirect pathway fractions (7) Finally, the fraction of pre-existing hepatic glycogen was estimated as the balance between total and newly synthesized glycogen fractions.
Newly synthesized fraction (8) Fractional pathway contributions were converted to absolute values by multiplication with liver glycogen amounts, expressed as grams per 100 grams liver wet weight.

Statistical analysis.
Values are expressed as means ± standard error of the mean (S.E.M.). Statistical differences between different groups with P < 0.05 were assessed using Student's two-tailed unpaired t test.
Data availability. The datasets generated and analysed during the current are available from the corresponding author on reasonable request.

Results
Hepatic metabolite levels. There were no differences in weights between seabass groups (see Table 1).
The hepatosomatic index (HSI) was significantly different between fed and fasted seabass as well as between fasted seabass and fasted seabream (P < 0.0001 for both). Fed and fasted seabass also had significantly different for perivisceral fat somatic index (PFSI) (P < 0.05). This index could not be calculated for seabream since they had no observable perivisceral fat. Similar differences were also observed for hepatic glycogen levels with significantly lower amounts in fasted compared to fed seabass (P < 0.05). Hepatic glycogen levels were also significantly different between fasted seabream and fasted seabass (P < 0.0005, respectively). Compared to our previous study with 90 g fasted seabass 9 , liver glycogen levels were somewhat higher: 8.4 ± 0.5 g.100 g −1 liver in this study versus 6.6 ± 0.6 g.100 g −1 liver previously. From a comparison of hepatic glycogen levels in saline-injected versus glucose-injected seabass, the glucose load resulted in a significant increase in hepatic glycogen levels of fasted fish while it had no significant effects on glycogen levels of fed fish (Table 1).
Hepatic glycogen 13 C-enrichment from the [U-13 C 6 ]glucose load and estimates of direct and indirect pathway contributions to glycogen synthesis in fasted seabream and seabass.
Enrichment of the load with [U-13 C 6 ]glucose generates glycogen 13 C-isotopomers that can be attributed to direct and indirect pathway activities, as shown in Fig. 1. Figure 2 illustrates the identification of direct and indirect pathway 13 C-isotopomers by 13 C-NMR spectroscopy of glycogen following its derivatization to MAG. Each of the six hexose signals is composed of a central singlet arising from the background 13  ]glucosyls -as would be expected from extensive exchange of the triose-P precursors catalyzed by triose phosphate isomerase (data not shown). The prominence of the D signals in both seabream and seabass spectra indicate that the indirect pathway had a significant role in the conversion of the glucose load to hepatic glycogen for both species. Table 2 shows estimates of 13 C-isotopomer enrichments obtained from the 13 C-NMR spectra corresponding to direct and indirect pathway activities. From these data, the fraction of hepatic glycogen derived from direct and indirect pathway metabolism of the glucose load were estimated using Equations (1) and (2) from the materials and methods section and the results are represented by Fig. 3. The data demonstrate that a substantial portion of hepatic glycogen (~50% for seabass and ~65% for seabream) was Final liver glycogen (g.100 g −1 liver) 10.0 ± 0.5 * (n = 6) 8.4 ± 0.5 (n = 6) 1.6 ± 0.7 ** (n = 3) Saline load final liver glycogen (g.100 g −1 liver) 10.4 ± 0.2 ** (n = 6) 3.4 ± 1.1 ‡ ‡ (n = 6) 0.2 ± 0.01 (n = 3) Table 1. Physiological parameters and hepatic glycogen levels for fed and fasted seabass and fasted seabream 48 hours after the administration of a glucose load. Also shown are hepatic glycogen levels for seabass that received a saline injection instead of a glucose load. Each n for liver glycogen represents a pooled sample from livers of 2-3 fish. Values are means ± S.E.M. Significant differences between different seabass feeding conditions and between different species in fasted state (except weight) are indicated by asterisks (t-test, *P < 0.05; **P < 0.001). Significant differences between Saline load and Final liver glycogen are indicated with ‡ (t-test, ‡ ‡ P < 0.005). 1 Hepatosomatic index = 100 × (liver weight/body weight). 2 Perivisceral fat somatic index = 100 × (perivisceral fat weight/body weight). 3 Absence of perivisceral fat. derived from the glucose load with near-equivalent contributions from direct and indirect pathways. The remaining unlabeled fraction represents glycogen that was synthesized from unlabeled precursors and/or pre-existing glycogen.

Hepatic glycogen synthesis from all sources in seabass using [U-13 C 6 ]glucose and 2 H 2 O. When
[U-13 C 6 ]glucose is combined with 2 H 2 O, glycogen synthesis from the glucose load and from non-glucose precursors can be resolved. Also, the fraction of pre-existing glycogen can be estimated. This is particularly relevant in carnivorous fish under fed conditions where endogenous glycogenic precursors such as gluconeogenic amino acids may be abundant, and pre-existing hepatic glycogen stores are also plentiful. Figure 4 shows 13 C-and 2 H-NMR spectra taken from liver samples of fasted and fed seabass. In the 2 H NMR spectrum, the position 2 intensity appears to be less than that of position 5 -consistent with a previous seabass study of glycogen enrichment from 2 H 2 O 13 . This is due to a significant kinetic isotope effect that limits the exchange between the hydrogen 2 of G6P and body water 9 thereby decreasing the enrichment of this specific site from 2 H 2 O. The magnitude of this effect can be quantified from the analysis of glycogen 2 H-enrichment from [U-2 H 7 ]glucose allowing the position 2 enrichment from 2 H 2 O to be corrected for incomplete exchange 9 (see Table 2). The 13 C-multiplet intensities of the 13 C-NMR spectra of the fed fish were substantially lower compared to fasted fish indicating a much smaller enrichment of glycogen from [U-13 C 6 ]glucose in fed compared to fasted fish      Table 2). The 2 H NMR signals and estimated 2 H-enrichment levels, which inform glycogen synthesis from both [U-13 C 6 ]glucose and endogenous substrates, were also much less in fed versus fasted fish. Integration of 2 Hand 13 C-NMR data informs the fraction of glycogen synthesized from all sources via direct and indirect pathways, and these data are presented in Table 3. These results show that under fed conditions, the fraction of hepatic glycogen derived from the glucose load was very small. Of this small fraction, the indirect pathway had a substantial contribution, being roughly equivalent to that of the direct pathway. The majority of indirect pathway carbons and essentially all Krebs cycle sources were accounted for by the glucose load. Under fasted conditions, the fraction of newly synthesized glycogen was far more substantial, accounting for 60% of the total glycogen sampled at 48 hours post load. While about 2/3 rds of the newly synthesized glycogen was derived from the glucose load -with approximately equal contributions from direct and indirect pathways -there were also meaningful contributions from endogenous Krebs cycle and glycerol sources. Absolute values of glycogen synthesis from the glucose load and endogenous sources, obtained by multiplying fractional rates by the total glycogen concentration, are shown in Table 4. These data reveal a ~10-15-fold higher uptake of glucose into glycogen under fasted compared to fed conditions. Glycogen synthesis from endogenous sources was also significantly higher in the fasted state.
These results are also in good agreement with our biochemical estimates of hepatic glycogen excursions based on hepatic glycogen quantification of fish administered with glucose and saline loads (Table 1). For the fed seabass, the glucose load did not significantly alter the levels of hepatic glycogen compared to saline. For the fasted seabass, the difference in glycogen levels between saline and glucose loads (~5.6 g.100 g −1 liver) corresponds to ~67% of the total glycogen sampled. This is consistent with the 60% fraction of newly synthesized glycogen calculated from the tracer studies. On the left-hand side are shown the carbon 1 multiplet from the 13 C NMR spectra for each condition with S, D and Q notations as described in Fig. 2. On the right-hand side are shown the corresponding 2 H NMR spectra with the number above each signal representing its position in the glycogen hexose unit. The 13 C and 2 H NMR glycogen signals of fasted and fed fish were scaled by matching the intensities of the 13 C and 2 H MAG methyl signals (not shown) which represent constant 2 H and 13 C enrichment levels across the entire set of samples.

Discussion
Rationale of our study design and setting. The conversion of glucose to glycogen by the liver is a key process to define the capacity of an organism to utilize dietary glucose. Thus, the rate of hepatic glycogen synthesis is widely used as a proxy for glucose utilization in many different organisms including fish 3,4,9,10,13,28 . Resolution of hepatic glycogen synthesis into direct and indirect pathway contributions gives further insight on the roles and capacities of these pathways in mediating the conversion of glucose to glycogen. Many studies have shown that hepatic glycogen repletion in fish is most active when they are presented with food or with a glucose bolus after fasting 13,[29][30][31] . Quantifying direct and indirect pathway activities in this setting, where glucose availability is not a limiting factor, provides valuable insight on the capacity of each pathway in sustaining hepatic glycogen synthesis.

Comparison of direct and indirect pathway activities of seabass and seabream with other species.
In omnivorous species such as humans and rats that are considered to be well adapted to assimilate and metabolize glucose, the indirect pathway contributes significantly to hepatic glycogen re-synthesis after fasting induced by a carbohydrate-containing meal or a glucose load. In rats, the indirect pathway accounts for 50-70% of glycogen incorporation from a glucose load 14,21 and ~54% of hepatic glycogen repletion during nocturnal feeding with standard chow 26 . In a model of insulin-dependent (Type 1) diabetes induced by streptozotocin, the indirect pathway contribution was significantly higher (68%) but could be restored to the levels found in non-diabetic controls by insulin administration 32 . In healthy humans, the indirect pathway accounts for 30-50% of hepatic glycogen synthesis following a meal or a glucose load [33][34][35] . As with rats, this contribution is significantly increased in Type 1 diabetes 34,36 but it does not revert to normal following insulin therapy 33,36 . Given that carnivorous fish have been previously compared to glucose-intolerant or diabetic mammalian species in their poor capacity to clear a glucose load 37 , and based on our previous measurements of high indirect pathway activities in seabass fed with either standard or starch-supplemented aquafeeds 7,13 we anticipated that the indirect pathway would dominate hepatic glycogen synthesis from a glucose load for seabass and seabream. Instead, we found that the direct pathway contributed between one-third and one-half of hepatic glycogen synthesis from glucose -similar to that of a healthy rat. These results suggest that the capacity for hepatic glucose metabolism in seabass and seabream is comparable to that of omnivorous mammals and that their glucose intolerance may be more related to peripheral glucose uptake and metabolism or/and inability to shut down gluconeogenesis even if glycolysis and glycogenesis seem well regulated after glucose load. Also, the fact that a glucose load appears to sustain higher direct pathway contributions compared to that measured after feeding with aquafeed enriched in dietary starch 7 suggests that polysaccharide digestion may be a limiting step in the supply of glucose for direct pathway flux in carnivorous fish.
Role of extrahepatic glucose metabolism in sustaining indirect pathway flux. The indirect pathway of glucose conversion to hepatic glycogen involves the initial glycolytic conversion of glucose to pyruvate followed by the gluconeogenic conversion of pyruvate to G6P. In the whole body, glucose can be converted to pyruvate in peripheral tissues such as muscle and this pyruvate can be then exported to the liver (usually in the form of lactate or alanine) to be converted into G6P. Conversion of this G6P to glucose completes the Cori and/ or glucose-alanine cycle. If this G6P is recruited for glycogen synthesis, it is tantamount to the conversion of glucose to glycogen via the indirect pathway. Tracer methods cannot distinguish this form of indirect pathway flux involving extrahepatic glucose metabolism from that conducted solely within the liver. Therefore, it is not known to what extent peripheral tissues contribute to the indirect pathway formation of hepatic glycogen from a glucose load. We do know that there is extensive Cori cycling of a glucose load in seabass 11 hence it is likely that a significant portion of indirect pathway flux involves extrahepatic glucose metabolism.
Glycogen cycling and its impact on 2 H 2 O measurements of direct pathway flux. Futile cycling between glycogen and G6P has been described in humans under certain nutritional conditions or disease states [38][39][40] and was also found to be active in rats following a glucose load 41 . With 2 H 2 O, futile cycling contributes to the enrichment of glycogen position 2 (H2) independently of net glycogen synthesis thereby confounding direct pathway estimates 27,34,39 . Glycogen cycling does not influence the enrichment of position 5 (H5) hence estimates of indirect pathway contributions based on the H5/body water enrichment ratio are unaffected. For determining the extent of glycogen cycling, the ratio of glycogen position 2 to tank water enrichment (H2/TW), representing the fraction of newly synthesized glycogen plus the fraction of glycogen that underwent cycling, is compared with the estimate of newly synthesized glycogen obtained by mass balance 39 . Based on our biochemical measurements of hepatic glycogen in glucose-and sham-injected fish, we estimated that 67% of the post-load glycogen was newly-synthesized. From H2/TW (see Table 2), the fraction of newly-synthesized plus cycled glycogen was 88%.  Thus the difference, 21%, can be attributed to glycogen cycling. This additional labeling of position 2 represents the conversion of unlabeled pre-existing glycogen to [2-2 H]glycogen via glycogen cycling (see Fig. 1). Since pre-existing glycogen constituted 33% of the total, the fraction of glycogen that underwent cycling can be estimated as 21/33 = 64%. Thus in seabass administered with a glucose load, hepatic glycogen cycling appears to be very extensive. As a result, analysis of glycogen would have yielded substantial overestimates of direct pathway contributions to hepatic glycogen synthesis (53% versus 25% calculated using [U-13 C 6 ]glucose).
Effects of pentose phosphate pathway activity on flux estimates. The oxidative portion of the pentose phosphate pathway (PPP) results in the loss of carbon 1 while the non-oxidative portion results in carbon rearrangements through transaldolase and transketolase activities. Therefore, both oxidative and non-oxidative PPP activity will alter glycogen enrichment distributions from 2 H-and 13 C-tracers. Such activity is revealed most clearly in the 13 C-isotopomer data 42,43 . In the absence of PPP flux, the 13 C-isotopomers derived from indirect pathway metabolism have identical distributions in the 123-and 456-carbons of glycogen. If PPP flux is active, the 123-carbons undergo significant rearrangement while the 456-moiety remains intact 42 . Analysis of the carbon 2 and carbon 5 13 C signals from our experiments (data not shown) showed near-identical 123-and 456-isotopomer distributions indicating that PPP flux was low in relation to G6P flux into glycogen. These observations are consistent with our previous studies of seabass glycogen enrichment from a glucose load enriched with [U-2 H 7 ] glucose where the effects of transaldolase exchange activity on glycogen 2 H-enrichment distributions were found to be insignificant 9 .
Conclusions. There is high interest in understanding the degree to which carnivorous fish such as seabass and seabream can adapt to dietary carbohydrate. Since the hepatic conversion of glucose to glycogen is considered a central pathway in carbohydrate assimilation, we studied this process using stable-isotope tracers that allowed glycogen synthesis from both an exogenous glucose load as well as endogenous non-glucose substrates to be measured. Moreover, the overall conversion of glucose into hepatic glycogen was resolved into direct and indirect pathway contributions. We found that for both fasted seabream and seabass, hepatic glycogen was efficiently replenished by a glucose load. This repletion involved approximately equal contributions of direct and indirect pathways, which is similar to that observed in rodents and in humans. In our study of hepatic glycogen repletion from glucose in fasted seabass using a combination of [U-13 C 6 ]glucose and 2 H 2 O, we found evidence of substantial substrate cycling between glycogen and G6P. Since this activity can result in the enrichment of glycogen position 2 from 2 H 2 O without the net synthesis of new glycogen, analysis of direct pathway contributions to hepatic glycogen synthesis based on position 2 enrichment can result in significant overestimates. These uncertainties may be avoided by using a combination of 2 H 2 O and [U-13 C 6 ]glucose tracers.