The expression and activities of glucokinase (GK) and hexokinase (HK) were assessed in different tissues of rainbow trout (Oncorhynchus mykiss) under different feeding conditions (fed, fasted for 14 days, and refed for 7 days). Two different HK-I cDNAs were identified with different tissue distributions. One transcript named heart or H-HK-I was observed in the four brain regions assessed, white muscle, kidney, and gills but not in liver or erythrocytes. A second transcript named liver or L-HK-I was found in all tissues surveyed. GK mRNA was identified only in liver and the four brain regions. GK expression was altered by feeding conditions, especially in liver and hypothalamus where food deprivation decreased and re-feeding increased expression; changes in expression reflected activity changes and changes in tissue glycogen levels. In contrast, feeding conditions did not alter expression of either HK-I transcript but did alter tissue HK activities. The reduced phosphorylating capacity noted with food deprivation correlates primarily with changes in tissue HK, whereas increased capacity, as with refeeding, was associated with changes in GK; these changes fit with the different Km values of the GK and HK enzymes. These results provide evidence for the hypothalamus acting as a glucosensor in trout, as hyperglycemia produced increased GK expression and activity, as well as increased glycogen levels. Thus, even though trout use glucose poorly, none of the parameters tested here relate to this inability to use glucose and suggest that, at least, rainbow trout, if given an appropriate carbohydrate diet, could metabolically adjust to such a diet.
- glucose use
hexokinases (hk) catalyze the first reaction of glycolysis, thus playing a key role in tissue intermediary metabolism. Four HKs are found in vertebrates (HK-I, HK-II, HK-III, and HK-IV), where they form a family of closely related enzymes encoded by unique genes (25, 42).
HK-I is ubiquitously distributed in vertebrate tissues with particularly high expression levels in brain and kidney, where it generally exists in a mitochondrially bound form. This enzyme is characterized by a relatively high affinity for glucose (low Km), inhibition by glucose 6-phosphate and inorganic phosphate antagonism of glucose 6-phosphate inhibition (42, 43). HK-I is particularly active in glucose-dependent tissues (heart, brain) of mammals (43) and fish (18) and is influenced by dietary change (21). HK-I mRNA is reported in numerous tissues (liver, heart, brain, white muscle, and kidney) of carp (45) and in the liver of gilthead sea bream (4).
HK-IV or glucokinase (GK) is characterized by a low affinity for glucose (Km 5–12 mM), no inhibition by glucose 6-phosphate and a molecular weight of 50 kDa rather than the 100 kDa of HK-I. GK was first found in the mammalian liver but a high Km glucose phosphorylating activity is also reported in the brain, intestine, kidney, mammary glands, and islets of Langerhans (25, 42). Studies in humans, animal models, and isolated hepatocytes established that hepatic GK exerts a strong influence on glucose utilization and glycogen synthesis (25). In mammals, the factors regulating GK expression in the brain are different from those in liver (46). Finally, GK in the pancreatic β-cells plays a central role in blood glucose sensing and insulin release (15, 25).
Fish are generally considered to possess poor control over blood glucose levels (44), and one explanation for this was thought to be the absence of inducible hepatic GK activity (17, 22). However, GK enzyme activities are reported in several fish species (21, 22, 23, 39, 40) and are induced by feeding (13) and dietary carbohydrates (21, 22, 39). Hepatic GK transcripts were isolated and characterized from the liver of carp, rainbow trout, and gilthead sea bream (5, 16, 21, 22), but GK expression was not found in white muscle, heart, or kidney of rainbow trout or carp (4, 22). Whole brain GK expression was reported in rainbow trout but not in carp or gilthead sea bream (22). Although direct evidence that these teleost GK transcripts generate functional GK enzymes is lacking, the nucleotide and amino acid sequences, and the observation of a high hepatic GK activity in some fish fed with carbohydrates supports a functional GK enzyme in these species (22).
Mammalian feeding behavior and general energy homeostasis are regulated by circulating levels of nutrients (e.g., glucose) and peptide hormones (e.g., leptin, insulin). Sensors to detect levels of glucose reside within specific nuclei of the hypothalamus and in pancreatic β-cells (15). These glucosensors consist of the low-affinity glucose transporter GLUT2 and the low-affinity glucose-phosphorylating enzyme GK (38). Several regions of the mammalian brain contain specialized neurons that utilize glucose as a signaling molecule rather than as an energy substrate. Thus glucose-excitory neurons increase firing, whereas glucose-inhibitory neurons decrease firing rates as ambient glucose levels rise in a manner analogous to glucosensing in pancreatic β-cells (9). Because GK is selectively expressed in brain areas where glucosensing neurons reside, GK has the physiological properties that would make it an ideal glucosensor in neurons (9). An increase either in glycemia or insulinemia in mammals induces GK gene expression, enhances activities, of GK and promotes glucose uptake in those brain regions thought to contain glucosensing neurons (11).
Experimentally induced hyperglycemia in some fish produces an increased glucose uptake in most tissues (3) but only a small inhibitory effect on food intake, suggesting a low sensitivity of the brain to detect hyperglycemia (33). However, a fast counter-regulatory response (increased food intake) after intracerebroventricular treatment of rainbow trout with 2-deoxyglucose demonstrated indirectly the presence of brain glucosensors sensitive to hypoglycemia (33). Whether glucosensors patterned after those in the mammalian brain exist in the fish brain remains unresolved. There are no studies assessing how changes in food availability change the expression and activities of GK and HK in different fish tissues. Thus the purpose of this study was twofold: 1) to determine whether feeding condition alters GK and HK activities and transcript expression in different tissues of the rainbow trout, and 2) to determine the expression of GK in different brain regions, and, in particular, the hypothalamus, as a clue for the presence of a brain glucosensor in this species. Accordingly, rainbow trout were fed, food-deprived, and refed over a 3-wk period, and GK and HK expression and enzyme activities in various tissues were assessed.
MATERIALS AND METHODS
Rainbow trout (Oncorhynchus mykiss Walbaum) were obtained from a local fish hatchery (Soutorredondo, Noia, Spain). Fish were maintained for 1 mo in 150-liter tanks under laboratory conditions and a natural photoperiod in dechlorinated tap water at 14°C. Fish mass was 102 ± 3 g (n = 35). Fish were fed once daily (0900>) to satiety with commercial dry fish pellets (Dibaq-Diprotg SA, Segovia, Spain; proximate food analysis was 48% crude protein, 6% carbohydrates, 25% crude fat, and 11.5% ash; 20.2 MJ/kg of feed). The experiments described comply with the Guidelines of the European Union Council (86/609/EU) and was approved by the Ethics Committee of the University of Vigo (Spain) for the use of animals in research. Experiments were undertaken in October 2004.
Fish were randomly assigned to 150-liter experimental tanks after acclimation, and each tank was randomly assigned to one of three treatments (two replicates per treatment): fed, fasted, and refed. From that time forward, one group of fish in the two replicate tanks (fed) remained on the feeding regime described above for 14 days, another group of fish in two tanks (fasted) was deprived of all food for 14 days, and another group of fish in two tanks (refed) was deprived of food for 14 days followed by the feeding regime (to satiety) described above for a subsequent 7 days. The period of food deprivation was chosen based on other studies showing significant changes in metabolic status after 2 wk of food deprivation (20).
On the sampling day, 5 or 6 fish were removed from each replicate holding tank at 1500, 6 h after feeding in the fed and refed fish, resulting in 11 or 12 fish per group (fed, fasted, and refed) and anesthetized in MS-222 (50 mg/l) buffered to pH 7.4 with sodium bicarbonate. This time period was chosen to provide optimal induction of GK expression in trout (21). Samples obtained from four fish per group (two per tank) were used to assess GK and HK mRNA expression in tissues. Samples obtained from the remaining 7 or 8 fish per group (3 or 4 per tank) were used to assess enzyme activities and metabolite levels.
Blood was obtained by caudal puncture using ammonium-heparinized syringes. Plasma samples were obtained after centrifugation of the blood (1 min at 10,000 g; Eppendorf 5415R), and were immediately deproteinized (6% perchloric acid, PCA) and neutralized (1 mol/l potassium bicarbonate) before freezing in liquid nitrogen and further storage at −80°C until assayed. The remaining erythrocyte pellet was frozen and stored in a similar fashion. The liver, a portion of white muscle (hypaxial muscle taken anterior of the dorsal fin), kidney, gill filaments, intestinal mucosa, and heart were removed, freeze-clamped in liquid nitrogen, and stored at -80°C until assayed. The brain, without the pituitary, was removed, placed on a chilled Petri dish, and dissected into four regions (hypothalamus, midbrain, hindbrain, and telencephalon), as described previously (8). Then it was freeze-clamped in liquid nitrogen and stored at −80°C until assayed.
Enzyme and Metabolite Assays
Frozen tissue samples were minced on a chilled petri dish into small pieces that were vigorously mixed and divided into two different (but relatively homogeneous) aliquots to assess enzyme activities and metabolite levels. The aliquots of still frozen tissue used for the assessment of metabolite levels were homogenized immediately by ultrasonic disruption in 7.5 vol ice-cold 6% PCA, neutralized (using 1 mol/l potassium bicarbonate), centrifuged (2 min at 13,000 g, Eppendorf 5415R), and the supernatant was used to assay tissue metabolites. Tissue glycogen levels were assessed using the amyloglucosidase method (12). Glucose obtained after glycogen breakdown in tissues (after subtracting free glucose levels), as well as plasma glucose levels, were determined with a commercial kit (Biomérieux, Madrid, Spain). Tissue and plasma total α-amino acids were assessed colorimetrically using the nynhidrin method of Moore (19) with modifications to adapt the assay to a microplate format. Plasma lactate levels were determined spectrophotometrically using a commercial kit (Spinreact, Gerona, Spain).
Tissue aliquots used to assess enzyme activities were homogenized by ultrasonic disruption with 9 vol ice-cold-buffer consisting of 50 mmol/l Tris (pH 7.6), 5 mmol/l EDTA, 2 mmol/l 1,4-dithiothreitol, and a protease inhibitor cocktail (Sigma Chemical, St. Louis, MO; P-2714). The homogenate was centrifuged (5 min at 900 g, Eppendorf 5810R), and the supernatant was used immediately for enzyme assays.
Hexokinase Enzymatic Code ([EC) 22.214.171.124] and glucokinase (EC 126.96.36.199) activities were estimated at 37°C in a plate reader (SpectraFluor, Tecan, Zurich, Switzerland) as described by Panserat et al. (21) by coupling ribulose-5-phosphate formation from glucose 6-phosphate to the reduction of NADP+ at 340 nm. GK activities were corrected for glucose dehydrogenase (GlDH; EC 188.8.131.52) activities. Thus GK activity is calculated as the total HK activity minus low Km HK activity minus 1/3 of GlDH activity, as described (21). Pyruvate kinase (PK; EC 184.108.40.206) and fructose 1,6-bisphosphatase (FBPase; EC 220.127.116.11) were estimated using a Unicam UV-2 spectrophotometer (Thermo Unicam, Waltham, MA). The reactions were started by the addition of homogenate (0.05 ml) at a preestablished protein concentration, omitting the substrate in control cuvettes (final volume 1.35 ml), and allowing the reactions to proceed at 15°C for preestablished times (5–15 min). Protein was assayed in triplicate in homogenates following the method of Bradford (7) with bovine serum albumin (Sigma) as a standard. Enzyme analyses were assessed at maximum rates in each tissue, with the reaction mixtures established in preliminary tests to render optimal activities. The activity ratio of pyruvate kinase is defined as activity at low/high substrate phosphoenolpyruvate concentrations and expressed as a percentage. Low-substrate concentration was 0.05 mmol/l for the brain regions (hypothalamus, midbrain, hindbrain, and telencephalon) and 0.09 mmol/l for the remaining tissues. High substrate concentration was 1 mmol/l for the brain regions and 2 mmol/l for the remaining tissues. The specific conditions for PK and FBPase assays were described previously (28, 29).
Total RNA was extracted from frozen tissues using TRI reagent, as recommended by the manufacturer (Sigma). The quality and quantity of the isolated RNA was assessed spectrophotometrically. Total RNA (2 μg) was reverse transcribed into first-strand cDNA when primed with 5′-pd(T)12–18-3′ (Amersham Biosciences, Barcelona, Spain) using Moloney murine leukemia virus reverse transcriptase for 1 h at 37°C using methods recommended by the manufacturer (Promega, Madison, WI).
GK cDNA was PCR-amplified using specific primers described for rainbow trout by Panserat et al. (21): 5′-TGATGTTGGTGAAGGTGGGG-3′ (forward, F) and 5′-TTCAGTAGGATGCCCTTGTC-3′ (reverse, R). This gave an expected sequence size of ∼240 bases. Primers for HK-I were designed using the CODEHOP program (http://blocks.fhcrc.org/codehop.html) following alignment with the Clustal-W multiple alignment algorithm (10) of the HK-I sequences from Homo sapiens (GenBank accession no. NM_033500), Cyprinus carpio (AF 119837), Danio rerio (BCO67330), chicken (NM_204101), rat (NM_012734), and mouse (BCO27316), and the preliminary sequencing data from Oncorhynchus mykiss were kindly provided by Dr. Stephane Panserat (INRA-IFREMER, St-Pée-sur-Nivelle, France). Two degenerate primers were chosen corresponding to the most conserved regions of the sequence. Degenerate primer set 1 consisted of 5′-CCTGGACCTGGGCggnacnaaytt-3′ (F) and 5′-GGGCCAGCCGGTCGswytcdatytg-3′ (R). Degenerate primer set 2 consisted of 5′-GCATCTCCGACTTCCTGgaytayatggg-3′ (F) and 5′-GCAGCTTGTACAGGGTGccrtcnacncc-3′ (R). Using these primers and RNA isolated from trout heart, we found a single band after PCR analysis with primer set 2; no band was detected using primer set 1. After sequencing of this PCR product (see below), gene-specific primers (GSPs) were designed to amplify a 245-base product with high-sequence identity to HK-I. Rainbow trout tissues were probed with this GSP: 5′-CGGCATGTACCTGGGAGAAA-3′ (F) and 5′-AGGGACACGGTGCTGCATAC-3′ (R). As noted in the results, this GSP was unable to detect a product in either liver or erythrocytes; the product of this GSP was therefore designated heart or H-HK-I. Using degenerate primer set 1 (above) a single band was found in rainbow trout liver. On the basis of the sequence of this PCR product we designed a liver-specific or L-HK-I primer: 5′-GGCTTCTCAGAGAGGGGATT (F) and 5′-TTTCGAAAATGCCTCTGGTC (R). This primer set amplified a 447-base product. Trout β-actin was used to assess the relative cDNA levels of GK and HK. β-Actin cDNA was amplified by PCR using the GSP 5′-ACCCTCAGCTCGTTGTAGA-3′ (F) and 5′-ACGGATCCGGTATGTGCAA-3′ (R) designed from the sequence of rainbow trout β-actin (GenBank accession no. AJ438158). Amplification with these primers produced a band for β-actin of 253 bases.
The PCR reaction for GK was carried out using a PTC-200 Peltier thermal cycler (MJ Research, Waltham, MA) in a final volume of 20 μl containing 2 μl (liver) or 8 μl (remaining tissues) cDNA template for GK and 2 μl cDNA template for β-actin, 1 × PCR buffer (50 mM KCl, 20 mM Tris-HCl, and 0.1% Triton X-100), 0.2 mM dNTPs, 1.5 mM MgCl2, 10 μM of each primer (forward and reverse), and 1 U Taq polymerase (Ecogen, Barcelona, Spain). Semi-quantitative RT-PCR was used to assess the relative expression of HK and GK using β-actin as the comparator gene. Cycle numbers were determined for each gene and were within the linear phase of the PCR reaction. Amplification of cDNA was achieved with an initial denaturation at 94°C followed by 20 (β-actin) or 35 (GK) cycles of denaturation (94°C for 1 min), annealing (60°C for 1 min) and extension (72°C for 1 min) followed by a final extension period of 10 min at 72°C before termination. HK cDNA synthesis was carried out with the same RNA generated for the GK experiment using the first-strand cDNA Synthesis Kit for RT-PCR (Roche Molecular Biologicals, Laval, QC, Canada). PCR was carried out in a 20-μl total volume and included 1 × PCR buffer, 0.2 mM dNTP, 2 mM MgCl2, 1 U Taq polymerase (Invitrogen), 0.2 μM forward and reverse primers, and 0.5 μl cDNA. Amplification occurred using the protocol noted above for 32 cycles using a Eppendorf Mastercycler gradient thermal cycler. For both GK and HK, negative controls without reverse transcriptase or cDNA were performed to ensure observed bands were not simply contamination. The PCR products were subjected to electrophoresis in a 1.5% agarose gel. The size of each PCR product was established by comparing with a 50-base step DNA ladder (Promega). Quantification of PCR products was performed by densitometric analysis of ethidium bromide-stained gels using a gel documentation system and analysis software. For each sample treatment, the integrated density value (IDV) obtained for the target gene-specific (GK or HK) band was divided by the signal obtained for the β-actin band producing a relative mRNA abundance value. The β-actin band IDV did not change significantly with treatment. As the same cDNA was used to assess the two HK-I forms and β-actin, comparison between relative abundance of the two HK-I forms is possible. Results are presented as relative intensities ± SE of n values (fish).
To ensure the bands of interest were, in fact, trout GK and HK-I, the bands were gel purified using the GFX PCR DNA and gel band purification kit (Amersham Biosciences) and cloned using Pgem-T Vector Systems II (Promega). White colonies were amplified by PCR using primers T7 and M13 (flanking the insert) and sequenced in both directions using the dRhodamine terminator cycle sequencing kit (Applied Biosystems). The reactions were run on an Applied Biosystems automated sequencer model ABI PRISM 310. To confirm the liver- and heart-HK-I sequences, sequencing was done at the CORE DNA facility center (Ottawa, Canada). The resulting sequences were compared with known sequences on GenBank using BLASTn (http://www.ncbi.nlm.nih.gov/BLAST/). Also, to ensure that the products of each GSP were GK, heart- and liver-HK-I, each product was cloned, sequenced, and compared with their original sequences.
Data are presented as means ± SE. Comparisons among groups were performed with one-way ANOVA using the SigmaStat (SPSS, Chicago, IL) statistical package. Post hoc comparisons were made using a Student-Newman-Keuls test, and differences were considered statistically significant at P < 0.05.
GK and HK Tissue Expression
The PCR reaction for GK using GSPs produced, as expected, a single band that after sequencing resulted in a 241-base nucleotide sequence (not shown) with a 95% homology with that already published for rainbow trout GK (GenBank accession no. AF053331) and 86% homology with gilthead sea bream GK (AF053330) supporting the presence of GK transcripts. Using this GSP, 11 trout tissues were screened for GK transcripts. Samples were taken 6 h after feeding to maximize GK liver expression (21). Figure 1 demonstrates significant relative GK mRNA expression in the liver and hypothalamus; a lower but measurable expression is also observed in midbrain and hindbrain. Expression in the telencephalon is very low at least in these fed fish (see below). In none of the other 11 tissues could transcripts be detected using our procedures.
After testing, both HK degenerate primers with RNA isolated from the trout heart, only primer set 2 produced a clear band of the expected size (∼850 bases). Sequencing this band provided an 852-base cDNA product (Fig. 2), which was compared with existing sequences using the BLAST algorithm. The comparison showed strong homologies with predicted amino acid sequences for HK-I in other fish species (90% identity with carp, AAF28854 and zebrafish, AAH67330) and other vertebrates [81–86% homology with human (BAD92763), rat (NP_036866), mouse (AAH72628), chicken BAC20932)], providing evidence that this particular sequence is, indeed, HK-I from rainbow trout heart (Gen Bank accession no. AY864082). GSPs were designed from this sequence and used to identify similar cDNAs in other trout tissue samples. HK-I transcripts are found in 8 of the 11 tissues studied (Fig. 3). This GSP could not detect HK-I in either liver or erythrocytes; the RNA sample from the intestine was degraded and unusable for samples tested with this primer. However, in another intestine sample taken from a separate group of rainbow trout, HK-I expression was observed (data not shown). As this primer set was designed from trout heart RNA, this particular transcript is called H-HK-I for heart HK-I. Using degenerate primer set 1, we detected a band using liver RNA, so a liver-specific primer was designed to detect liver-specific HK-I transcripts (see materials and methods), which is called L-HK-I for liver-HK-I; this GSP demonstrated L-HK-I in all 11 tissues assayed with the highest relative levels found in erythrocytes (Fig. 4). This particular transcript is also found in heart. Given that the same cDNA was probed to detect H-HK-I, L-HK-I, and β-actin, generally, transcript levels of H-HK-I were greater than L-HK-I. An amino acid alignment for H- and L-HK-I noted that over the 285 amino acids, there are 29 differences with only 10 of these being significant derivations providing an identity of 89% between these forms (Fig. 5).
Effects of Feeding on GK and HK Tissue Expression
Fed trout showed GK mRNA expression primarily in the liver and hypothalamus (Fig. 1). The expression of liver GK mRNA was affected by feeding conditions with food deprivation decreasing expression, whereas refeeding significantly increased expression, even when compared with the fed fish (Fig. 6). A similar expression pattern was also observed in both the hypothalamus and the hindbrain (Fig. 7). In midbrain, no significant change occurred between groups. Finally, in telencephalon GK expression was low in the fed fish but significantly increased in the refed fish; expression was not detected in fasted fish (Fig. 7).
The relative abundance of H- and L-HK-I mRNA was assessed in the tissues from fed, food-deprived, and refed trout. None of the conditions assessed modified the relative expression of either HK-I transcript in these tissues (Tables 1 and 2).
Tissue GK and HK Enzyme Activities
GK activities in intestine and erythrocytes were very low and probably not physiologically relevant, and no activity could be detected in the heart (Table 3). GK activities in liver, hindbrain, and kidney significantly decreased with fasting, whereas activities in liver, hypothalamus, hindbrain, and telencephalon of refed fish were higher than those of fed fish. Activities of GK were high in white muscle and declined with both fasting and refeeding. No GK transcript was detected in white muscle, so the significance of measurable activity in this tissue is not known.
HK enzyme activities in liver decreased in fasted fish and increased in refed fish compared with fed fish (Table 4), and activities were 5- to 10-fold lower than GK activities in this tissue (Tables 3 and 4). Brain sections with the exception of hypothalamus had comparable HK and GK activities; hypothalamic activities were higher (three- to seven-fold) for GK. All other tissues had higher HK than GK activities, and this was particularly apparent in the heart. Activities were generally not affected by nutrient condition with a few exceptions. HK activities decreased in kidney, gills, intestine and heart comparing the fasted with the fed fish, whereas activities actually increased in white muscle. In refed fish, HK activities significantly increased in intestine and erythrocytes above the fed and fasted group, whereas activities returned to fed levels for gills, kidney, white muscle, and midbrain.
Other Metabolic Parameters
Inadequate tissue was available to assess PK activities in intestine, white muscle, and heart. Liver PK activities increased in refed fish compared with fed and fasted fish, as did the activity ratios of the enzyme (Table 5). Hypothalamic PK activities were unchanged, although activity ratios did decrease in the fasted and refed fish compared with the fed group. In midbrain, optimal activity was higher in fed compared with refed fish in the absence of changes in the activity ratio. Hindbrain PK activity increased in refed fish compared with fasted and fed fish, whereas the activity ratio was higher in fed than in fasted fish. Similar changes in activity were noted in telencephalon, but the activity ratio was lower for the fasted than either the fed or refed fish. Kidney, gill, and erythrocytes showed no changes in PK activities, although the activity ratio was generally higher in refed compared with fed and fasted fish.
Hepatic FBPase activities were higher in fasted than in either fed or refed fish, whereas no activity changes were found in the kidney (Table 6).
Tissue glycogen levels changed in a number of tissues as predicted (Table 7); inadequate sample was available to assess telencephalon glycogen. Liver glycogen levels decreased in fasted and increased in refed fish compared with fed fish. A similar trend was noted in hypothalamus and hindbrain, although here refeeding did not increase glycogen levels above that of the fed group. No significant changes were noted for midbrain. White muscle and kidney glycogen levels of refed fish were higher than those of fed and fasted fish. In contrast, gill, kidney, and erythrocyte glycogen levels of refed fish were lower than those of fed and fasted fish. Finally, heart glycogen levels of refed fish were lower than those of fasted fish.
Total α-amino acids could only be assessed in liver, white muscle, and kidney (Table 8). Levels in liver were higher in fasted than in fed and refed fish. No changes were noticed in white muscle. In kidney, levels were lower in fasted than in fed fish.
Finally, plasma glucose levels generally decreased in fasted compared with fed fish, whereas levels in refed fish were higher than in the other two groups (Table 9). Plasma lactate was higher in fasted compared with either the fed or refed fish. An opposite pattern was noted for plasma α-amino acids that were lower in the fasted than either of the other two groups.
HK-IV or GK is expressed in liver of rainbow trout in agreement with that reported in the livers of a number of fish species, including rainbow trout, carp, and gilthead sea bream (5, 16, 21, 23). The expression of GK in the liver is thought to be related to plasma glycemia and glycogen synthesis (25), and, as noted, GK expression did rise in trout 6 h after a meal (see Liver). The expression of GK was also evident in the four brain regions assessed with the highest expression noted in the hypothalamus, although measurable expression was noted in the remaining regions. In fish, GK expression was described in rainbow trout whole brain (22) but not carp whole brain (4, 22). In mammals, high GK expression is reported in the hypothalamus (9, 14, 15, 26) with measurable expression noted in other regions, including the thalamus, olfactory nucleus, preoptic area, mesencephalon, and cortex (14, 15, 26). The absence of GK expression in the other trout tissues examined is consistent with previous fish studies (22).
Two partial cDNAs were sequenced from rainbow trout with high identity to the HK-I sequences available on GenBank from human (accession no. BAD92763), rat (NP_036866), mouse (AAH72628), bovine (AAQ11378), chicken (BAC20932), carp (Cyprinus carpio, AAF28854) and zebrafish (Danio rerio, AAH67330). The partial cDNAs identified are 284 (heart, H-HK-I) and 286 (liver, L-HK-I) bases in length and represent bases 583 to 867 in the human and zebrafish full-length sequences. The two amino acid sequences are 89% identical and the L-HK-I sequence is 86, 82, 83, 83, 88, 91, and 93% identical with respect to the species noted above whereas the H-HK-I is 83, 81, 81, 82, 86, 90, and 90% identical to the same group. These high-identity values are consistent with the highly conserved nature of this enzyme across species where it has been sequenced (2, 43) and support our contention that these sequences are indeed rainbow trout HK-I. In fact, the key amino acids necessary for catalysis are identical in these two sequences as they are in all HK-Is sequenced to date (2).
The L-HK-I is ubiquitously expressed in all tissues studied, but the H-HK-I was not detected in either liver or erythrocytes. Given that HK-I is generally considered to be ubiquitously distributed in mammals (43), we speculate that the L-HK-I may be closest to the ancestral HK-I form in trout as further supported by the higher-sequence identity values to other HK-Is for the L- than the H-HK-I form noted above. A full-length sequence and gene structure information are necessary to validate this hypothesis. The question of the role of the H-HK-I cannot be addressed by our study, but given that relative levels of this transcript generally exceed those found for the L-HK-I, presumably, it plays a specific role in those tissues in which it is found. An erythrocyte and testis HK-I form is reported in mammals that differ with respect to other somatic HK-Is by the absence of a porin-binding domain at the C-terminus that targets HK-I to the mitochondria (27, 36). As the partial sequences reported here are at the catalytic N-terminal end, we can only speculate at this time that changes may occur within the C-terminal or regulatory end of these two sequences.
Influence of Feeding Conditions
Metabolic parameters assessed in plasma revealed a nonsignificant decrease in glucose but a significant decrease in amino acids and an increase in lactate in the fasted compared with fed trout. This result is contrary to the hypoglycemia reported in many studies in fish (see Ref. 20), although in a more recent study, rainbow trout food deprived for 14 days did not show significant changes in plasma glucose (13). Refeeding fish after food deprivation produced a substantial increase in plasma glucose levels, whereas amino acids and lactate levels returned to levels similar to those observed in fed fish. Changes in plasma glucose levels generally induce metabolic changes in different tissues that act as glucosensors. Thus, in mammals, increased plasma glucose stimulates GK gene expression in β-cells of the pancreas (38) and in the hypothalamus (45), whereas fasting and insulin deficiency are associated with inhibition of GK transcription and with a gradual decline of total GK activity in liver (11, 38). Accordingly, we could expect changes in GK/HK expression and activities in different tissues of fish submitted to different feeding conditions.
GK expression in liver of fasted fish was reduced compared with fed fish, whereas refeeding increased expression above that of fed fish. This dual effect on expression levels is similar to that in liver of mammals (46) and in gilthead sea bream (16); Kirchner et al. (13) reported that food deprivation decreased GK expression in rainbow trout. The reduced GK expression coincides in time with a nonsignificant decrease in plasma glucose of fasted fish. It does appear that in trout, hepatic GK expression generally parallels glycemic levels (16), suggesting that glucose is an inducer of this enzyme at least in trout. Presumably, the refed trout were hyperphagic. Whether total calories or glucose specifically resulted in increased GK transcript and activity levels is not known, although the literature supports glucose as the primary trigger. This induction of GK activity was linked to higher glycogen levels, as reported when rainbow trout (21), perch (6), and rainbow trout, carp, and gilthead sea bream (21) were fed a high carbohydrate diet. GK activities declined with fasting in trout (13) and gilthead sea bream (16). Precisely whether this increase in activity results from increases in GK gene transcription or simply translocation from a regulatory protein (1) is unknown, although no GK regulatory protein is characterized in any fish species.
Considering that HK activity is nutritionally regulated in fish liver (6, 21, 37), it is not surprising that fasting and refeeding altered HK activities in liver similar to changes noted for GK. The decreased activity observed in fasted fish is in contrast to the absence of changes reported in liver of food-deprived rainbow trout in another experiment (13), and with the increase noted in Atlantic salmon (37), though a decrease was also observed in liver of fasted cod (37) and gilthead sea bream (24).
These results linking hepatic glucose phosphorylating capacity with nutrient status are also reflected by changes in the glycolytic and gluconeogenic pathways in the same tissues as previously noted in food-deprived fish (16, 24, 30, 34). The increased glycolytic capacity with refeeding is consistent with studies in mammals showing that GK overexpression increases glycolysis and glycogen synthesis in liver (1).
Decreased liver glycogen levels were noted in fasted trout and an overshoot occurred with refeeding. These changes are, in general, in agreement with the literature in food-deprived (24, 30, 34) and refed fish (16). Certainly, the mammalian literature demonstrates that increased plasma glucose levels induce increased GK activities in several tissues that act as glucosensors (pancreatic β-cells and hypothalamus), thus stimulating GSase activity leading ultimately to increased glycogen synthesis (31). A similar series of events appears to be operating in the trout even though the vertebrate liver is not thought to act as a glucosensor; the increased glycogen synthesis may be linked to the actions of insulin on the liver.
No changes were observed in the expression of either HK transcript in the brain regions of trout under our feeding conditions. In contrast, a marked decrease of GK expression was noted in the hypothalamus, hindbrain, and telencephalon (a nonsignificant decrease was also observed in the midbrain) of fasted fish compared with fed fish. A similar decrease in the expression of GK has been reported in mammalian β cells (38), though not in brain (9, 46). Thus this is the first report of changes in GK expression in the brain of a fasted animal. Refeeding fasted trout induced a spectacular increase in GK expression compared with the fed trout in the hypothalamus and a smaller increase in hindbrain and telencephalon. Again, midbrain did not produce any significant change. No similar changes with refeeding have been reported previously in any fish species.
Changes in GK expression agree, in general, with the pattern of GK activity changes in the same brain regions. HK and GK activities also displayed different behaviors in various brain regions. In the hypothalamus, hindbrain, and telencephalon, no significant changes were observed in HK activities, suggesting that changes in the activity of GK, not HK, may produce changes in the capacity of the brain regions to detect changes in plasma glucose levels. The absence of changes in HK activity in fasted fish are inconsistent with data obtained in whole brain of fasted rainbow trout and Atlantic salmon (34, 35), though consistent with the lack of changes reported in food deprived gilthead sea bream (30) and brown trout (3). These species differences may relate to specific differences in brain metabolic patterns. The different behaviors of HK and GK in the brain compared with those found in liver lend further support for a specific role of GK in fish brain metabolism.
Pyruvate kinase activities in the various brain regions demonstrate an interesting change with feeding conditions. Activities (either total or activity ratio) decreased in brain regions of fasted fish and recovered to fed values in the refed fish. This implies a decreased glycolytic capacity in the brain regions with fasting. Other studies that used whole brain showed either a decrease (24, 34, 35) or an increase (41) in the glycolytic capacity during food deprivation, which was recovered under refeeding conditions. Thus it does appear that individual brain regions have distinct metabolic patterns that may not be detected using the whole brain.
The general decrease observed in glycogen levels in the brain regions of fasted fish agree with observations during food deprivation in salmonids using the whole brain (34, 35), although not in gilthead sea bream (24, 30). The increased glycogen levels observed in brain regions of refed fish have not been previously reported in any fish species and must relate to the importance of carbohydrate as metabolic substrate in this organ, as previously suggested (32).
The mammalian hypothalamus is postulated to be a glucosensor much like the pancreatic β-cells (9, 14). Our results support a role for the hypothalamus at least in rainbow trout as a glucosensor since one of the requisites to be a glucosensor, possessing GK activity and expression, is fulfilled by this region. Another requisite is that changes in glycogen levels parallel changes in plasma glucose since in mammals glucose phosphorylation is a key step in the activation of glycogen synthesis through increases in intracellular glucose 6-phosphate, leading to allosteric activation of glycogen synthase (GSase) or covalent activation or translocation of GSase (27, 31, 36). Of the brain regions assessed in the present study, a very significant increase in glycogen levels was observed in the hypothalamus of refed fish reinforcing the possible role of this brain region as a glucosensor.
Increased GK activity and expression and glycogen levels in response to increased plasma levels of glucose in refed fish suggest that the rainbow trout hypothalamus is acting as a glucosensor with metabolic potential, and capacities similar to those described not only in mammalian hypothalamus but also in β-cells (15, 45). Other brain regions of rainbow trout also displayed changes in GK expression and activity, although considering the lower magnitude of change, these regions may play only a minor role. However, considering that glucosensitive sites controlling food intake and blood glucose are also reported in hindbrain (medulla oblongata and mesencephalon) of the rat (26), and changes described in glycogen levels and GK activity in trout hindbrain are similar to those observed in the hypothalamus, we cannot exclude that the fish hindbrain may also act as a glucosensor, like the hypothalamus.
In tissues other than liver and brain no detectable expression of GK or significant changes in either the H- or L-HK expression were noted. In contrast, changes in activity occur for both HK and GK in several cases.
Food deprivation resulted in decreased capacity for phosphorylating glucose in kidney, intestine, gills, heart, and erythrocytes. In some cases, this decrease can be attributed to changes in HK (kidney, gills, intestine, and heart) and in others also to GK (kidney) activities. This reduced capacity agrees with the few studies available in the literature assessing this pathway in tissues other than liver in food-deprived fish such as in gills of gilthead sea bream (24) and rainbow trout intestine (13), although not with the absence of changes observed in gilthead sea bream kidney (24, 30). The decreased capacity for phosphorylating glucose is not reflected in changes in glycolytic or gluconeogenic enzymes in the same tissues. Only Kirchner et al. (13) described no changes in kidney gluconeogenic capacity of rainbow trout. The absence of changes in glycogen levels in the same tissues in food-deprived fish is in agreement with the results obtained in gills and kidney of food-deprived gilthead sea bream (24, 30).
Blasco et al. (3) reported that hyperglycemia induced by a glucose load to previously food-deprived brown trout enhanced glucose uptake and utilization in many different tissues. This situation is comparable with that observed in the present experiment in refed fish showing hyperglycemia. Refed fish showed increased phosphorylating capacity for glucose in intestine and erythrocytes that can be attributed to changes in HK and GK activities. No significant changes were noted in the other tissues assessed. The activities, however, were very low for both erythrocytes and intestine, except HK activity in intestine, suggesting that these changes are marginal compared with those observed in liver and brain.
The present study provides for the first time evidence for the expression of two different transcripts of HK-I in different tissues of rainbow trout. Further studies are required to understand whether these two transcripts produce protein with different physiological roles. The expression pattern of GK and the fact that GK and HK activities change with nutrient status are not unlike changes observed in mammals under similar conditions. Food deprivation results in a general decrease, whereas refeeding results in a general increase, in tissue glucose-phosphorylating capacity. In some tissues, the decrease can be attributed to HK (liver, kidney, gills, intestine, and heart), and in others, the decrease can be attributed to GK (liver, hindbrain, kidney) activities. The reduced glucose-phosphorylating capacity appears to be attributable more to changes in HK, whereas increased capacity to GK, which fits with the different Km values of the GK and HK enzymes. Our results also implicate the hypothalamus as a glucosensor in fish, as suggested in mammals, though a possible role for other brain regions like hindbrain cannot be excluded. Thus, even though the rainbow trout uses glucose poorly, none of the parameters tested here relate to this inability to use glucose. This implies that this species could adjust to a higher carbohydrate diet, assuming that the proper carbohydrate is added to the diet. Studies in this regard are ongoing.
This study was supported by research grants from Ministerio de Ciencia y Tecnología and European Fund for Regional Development (VEM2003–20062), and Xunta de Galicia (PGIDT004PXIC31208PN and PGIDT05PXIC31202PN) to J. L. Soengas, and a Natural Sciences and Engineering Research Council of Canada Discovery Grant to T. W. Moon. S. Polakof and S. Sangiao-Alvarellos were recipients of a predoctoral fellowship from the Universidade de Vigo and Xunta de Galicia, respectively.
The authors wish to thank Dr. Paloma Morán (Universidade de Vigo) for assisting in the sequencing of HK and Dr. Stephane Panserat (IFREMER, St-Pée-sur-Nivelle, France) for providing preliminary trout HK sequence.
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