• We are sorry, but NCBI web applications do not support your browser and may not function properly. More information
Logo of physiolgenomicsPublished ArticleArchivesSubscriptionsSubmissionsContact UsPhysiological GenomicsAmerican Physiological Society
Physiol Genomics. Nov 2008; 35(3): 283–295.
Published online Aug 26, 2008. doi:  10.1152/physiolgenomics.90213.2008
PMCID: PMC2585019

Effect of starvation on transcriptomes of brain and liver in adult female zebrafish (Danio rerio)


We used microarray and quantitative real-time PCR (qRT-PCR) analyses in adult female zebrafish (Danio rerio) to identify metabolic pathways regulated by starvation in the liver and brain. The transcriptome of whole zebrafish brain showed little response to 21 days of starvation. Only agouti-related protein 1 (agrp1) significantly responded, with increased expression in brains of starved fish. In contrast, a 21-day period of starvation significantly downregulated 466 and upregulated 108 transcripts in the liver, indicating an overall decrease in metabolic activity, reduced lipid metabolism, protein biosynthesis, proteolysis, and cellular respiration, and increased gluconeogenesis. Starvation also regulated expression of many components of the unfolded protein response, the first such report in a species other than yeast (Saccharomyces cerevisiae) and mice (Mus musculus). The response of the zebrafish hepatic transcriptome to starvation was strikingly similar to that of rainbow trout (Oncorhynchus mykiss) and less similar to mouse, while the response of common carp (Cyprinus carpio) differed considerably from the other three species.

Keywords: microarray analysis, metabolic signaling, quantitative real-time polymerase chain reaction, neuropeptide

the adaptive physiological response to starvation conserves the health and function of key organs and redistributes resources toward essential biological functions. In mammals and birds, this reallocation of resources is part of a predictable and sequential transition of metabolic changes (reviewed in Ref. 72). These metabolic changes appear to be similar in fish and other ectotherms, although the transitions occur over a much longer time frame, largely the result of lower metabolic rates (72). Variation in metabolic rate in fish is in turn influenced by body size and body temperature (10, 29). As the most diverse group of vertebrates, fish species vary considerably in body size, thermal habitat, and metabolic rate, and many species differ in susceptibility or have unique adaptations to prolonged periods of fasting in their natural environment (37, 72). The considerable variation among fish species in the response to starvation provides a useful context for identifying mechanisms that are conserved among vertebrate species. Understanding these conserved mechanisms requires multiple investigations of a diverse array of species that identify the metabolic pathways that respond to starvation and genetic mechanisms that regulate them.

Here we used microarrays and quantitative real-time PCR (qRT-PCR) to examine the transcriptomic response to starvation in both brain and liver tissues of adult female zebrafish (Danio rerio). The zebrafish is an important model organism for the study of development and is now emerging as a model organism in other fields, including behavioral, physiological, and nutritional genomics (42, 49, 76). Despite its importance as a developmental and genomic model organism, our understanding of the nutritional physiology of the zebrafish remains very limited. Our goal was to identify the metabolic pathways impacted by starvation in zebrafish, contrasting two organs that 1) serve biosynthetic and energy-mobilizing functions (liver) and 2) consume energy and direct behavioral responses (brain). In addition, we offer an interpretation of these data in a comparative discussion of transcriptomic responses to starvation in other fish species (21, 52, 72) and mammals (2, 9, 26, 30, 79).


Experimental Organisms

We used a total of 96 adult female zebrafish of the Scientific Hatcheries strain, ranging in age from 156 to 225 days postfertilization (dpf) (mean ± SE: 191 ± 7 dpf). Only females were studied, to eliminate potentially confounding effects of sexual dimorphism in the transcriptomes of the liver (49) and brain (53). Groups of eight fish of the same age were randomly assigned to each of twelve 3-liter tanks (10 cm wide × 15 cm tall × 25 cm deep) arranged on the same shelf of a flow-through rearing system (Aquaneering, San Diego, CA). On the day of assignment to tanks and at 1-wk intervals throughout the study, individual fish were anesthetized in 0.017% MS222 (Western Chemical, Ferndale, WA), briefly blotted dry, and measured for standard length (tip of snout to end of caudal peduncle) and body mass (to the nearest 1 mg). Fish were maintained at 28.5°C on a 14:10-h light-dark photoperiod throughout the experiment.

Fish were fed a commercial flake food (Aquarian Tropical Fish food flake, Aquarium Pharmaceuticals, Chalfont, PA; min. 33.5% crude protein, min. 10.0% crude fat, and max. 2.0% crude fiber) to apparent satiety three times each day (0900, 1230, 1630). During feeding, a small amount of food was delivered by hand to the surface. Additional food was added when no food remained on the surface. When surface feeding ceased, the fish were judged to have reached apparent satiety and no additional food was added. Total mass of food delivered to each tank was monitored daily. Fish were acclimated to the test tanks and feeding regimen for 11 days before starvation treatments commenced. All procedures for animal handling were approved by the University of Idaho's Institutional Animal Care and Use Committee (www.uro.uidaho.edu/acuc).

Starvation Treatments

Test tanks were assigned to starvation or control treatments in an alternating manner so that treatments were uniformly distributed across fish ages and tank location. There was no significant difference in body mass or standard length between fish assigned to starvation treatment or control treatment on the day when treatments commenced. At this point, fish in the starvation treatment received no food, while fish in the control treatment continued to be fed to apparent satiety three times daily. The treatments were administered for 21 days. Fish were not individually marked; thus means of body mass and standard length were calculated for each tank to assess growth (see below).

At the end of the experiment, fish were anesthetized in MS222 and euthanized by decapitation. Six fish were sampled from each tank. Brain and liver were removed from each individual, placed in 1 ml of TRIzol (Invitrogen, Carlsbad, CA), and immediately homogenized. Brains were pooled at the time of homogenization, resulting in two pools per test tank, each pool containing samples from three individuals. Liver samples were pooled in the same manner (2 pools per tank, 3 individuals per pool) but after extraction of total RNA. Ovaries were removed and weighed to the nearest 1 mg to calculate a gonadosomatic index [GSI (%) = ovary mass/total body mass × 100]. To ensure that differences in transcription levels were the result of long-term physiological changes rather than acute periprandial responses, fish were not fed on the day of sampling, which equated to a 24-h fast for the control group.

Microarray Procedures and Analysis

Total RNA was extracted by the TRIzol method according to the manufacturer's protocol (Invitrogen). GeneChip Zebrafish Genome Arrays (Affymetrix, Santa Clara, CA) were used to measure the expression of 15,509 features representing at least 14,900 transcripts. A total of eight microarrays were run on brain samples, four biological replicates from each treatment. For liver, a total of 10 microarrays were hybridized, 5 biological replicates from each treatment. One RNA pool was hybridized to each microarray. For each tissue, RNA pools were selected for microarray analysis so that each biological replicate came from a different tank. Microarrays were hybridized at the Washington State University-University of Idaho Center for Reproductive Biology Genomics Core Facility (Pullman, WA) by methods described elsewhere (49). Briefly, for each array, up to 10 μg of total RNA was converted to cDNA and then to biotinylated cRNA in vitro with the GeneChip expression 3′ amplification one-cycle target labeling kit (Affymetrix). Microarrays were hybridized with fragmented biotinylated cRNA for 16 h at 45°C with constant rotation (45 rpm) and processed with the Affymetrix GeneChip Fluidic Station 450. Arrays were stained with streptavidin-conjugated phycoerythrin (SAPE), followed by amplification with a biotinylated anti-streptavidin antibody and another round of SAPE staining. The microarrays were then scanned with a GeneChip Scanner 3000 (Affymetrix). The microarray data from this study were deposited in the NCBI Gene Expression Omnibus (GEO) under accession number GSE11107.

All analyses of microarray data were conducted with Bioconductor (15) for R (www.R-project.org). Data for brain and liver were normalized and analyzed separately because variances of expression differed considerably between the tissues. The microarray data were filtered with two criteria (49). First, probe sets were eliminated if they were not classified as present in at least 75% of the microarrays from one treatment (control or starved) with PMA calls. Second, four separate normalizations were performed [MAS 5, robust multiarray average (RMA), GCRMA, and PLIER], and we retained only those probe sets with intensities above the 95% percentile of the background in all four normalizations. Filtered probe sets were then normalized with RMA (23), and the effect of starvation was tested with the Linear Models for Microarray Data (LIMMA) package (58). P values were corrected for a false discovery rate (FDR) of 0.05 with the Benjamini-Hochberg method (3). Gene Ontology (GO) analysis was performed with the Database for Annotation, Visualization, and Integrated Discovery (DAVID) (12) to identify GOs and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathways that were significantly overrepresented among differentially expressed genes.

Quantitative Real-Time PCR

qRT-PCR was performed to 1) validate the results of the microarray analyses, 2) verify that pooling procedures reflected patterns in individuals (possible in liver samples only), and 3) assay gene expression for loci that were not represented on the Zebrafish Genome Array. Expression of 10 genes was measured for the purpose of validating the liver microarray analysis. These genes were selected from various functional categories and included up- and downregulated genes with different magnitudes of expression. Differential expression was validated for β-actin (forward 5′-CTTCTTGGGTATGGAATCTTGC-3′, reverse 5′-GTACCACCAGACAATACAGTG-3′, from GenBank accession no. BC154531); hydroxyacyl-coenzyme A dehydrogenase/3-ketoacyl-coenzyme A thiolase/enoyl-coenzyme A hydratase (trifunctional protein), β-subunit (hadhb: forward 5′-TTCTCTGTCCCTCGGTCATC-3′, reverse 5′-CTTCGATCACCATTGCATGA-3′, from BC046041); fructose-1,6-bisphosphatase 1 (fbp1: forward 5′-CGCTATGTGGGGTCTATGGT-3′, reverse 5′-GGGGTTGCACTCATACAACA-3′, from NM_213132); pyruvate kinase, liver and red blood cells (RBC) (pklr: forward 5′-TAGAGGCCGTGGCCATGAT-3′, reverse 5′-GAGTCAAGCGCCGCAACT-3′, from NM_201289); proteasome (prosome, macropain) subunit, β-type, 5 (psmb5: forward 5′-CAACAAAGAGCGCATTTCAG-3′, reverse 5′-TTGCCCTCTGAATCCACATA-3′, from NM_131151); ELOVL family member 5, elongation of long chain fatty acids (elovl5: forward 5′-ACGCTCATCCTGCTCTTCTC-3′, reverse 5′-CGCTGGACATCACTCCATTA-3′, from NM_200453); claudin 7 (cldn7: forward 5′-GTTGCCTGTGGCTGGTTTAC-3′, reverse 5′-AACATGCCTCCACCCATTAT-3′, from NM_131637), peroxisomal biogenesis factor 19 (pex19: forward 5′-CCGACCAATTCACGAGGTAT-3′, reverse 5′-TTTGGAGGTTGACCCAAGTC-3′, from BC056815); growth hormone-inducible transmembrane protein (ghitm: forward 5′-GGGGATCATGATGAGCAACT-3′, reverse 5′-GAGCCCAGGACAGGTGTTTA-3′, from NM_200591); and eukaryotic translation initiation factor (eIF)3, subunit 6a (eif3s6a: forward 5′-CGTGCCTTGAGGACTTCATT-3′, reverse 5′-GGTTGACGATCCATCTTTCTG-3′, from NM_200839). We examined several potential housekeeping genes commonly used as reference genes for qRT-PCR analyses. Of these, 18S ribosomal RNA (18S rRNA) was the only gene that did not have a significant response to starvation [raw threshold cycle (CT)] and was therefore selected as the reference gene for liver (forward 5′-GAACGCCACTTGTCCCTCTA-3′, reverse 5′-GTTGGTGGAGCGATTTGTCT-3′, from BX296557:164887–166470 bp). qRT-PCR was performed on 29 of the individual RNA samples that had been used to construct the pooled samples used in the microarray hybridization.

In addition, to evaluate microarray results in the brain, expression of four genes was measured with qRT-PCR: matrix metalloproteinase 9 (mmp9: forward 5′-AGTCTCCGCAGGAATGTTGT-3′, reverse 5′-AAAATCTGCAATCCCCATCA-3′, from NM_213123.1), neuropeptide Y (npy: forward 5′-GGAGGAGCTCGCCAAGTAT-3′, reverse 5′-GGGACTCTGTTTCACCAATCA-3′, from NM_131074.1), a predicted gene similar to agouti-related peptide 2 (Dr.20586.1.A1_at: forward 5′-TCTCTCTGTGCACCATGACAC-3′, reverse 5′-AAACTTTGCCAGCGTGAATC-3′, from XM_001334910.1), and a transcript weakly similar to hemoglobin γ-A and γ-G chains (Dr.6751.1.S1_at: forward 5′-TCCTGTGCATCTGCAATCAT-3′, reverse 5′-GGGCGTAAGTGGTTTTGATG-3′, from NM_001003431). qRT-PCR was also performed to quantify the expression of four neuropeptide genes that were not represented on the microarray but are known to respond to starvation and regulate feeding behavior (16, 69, 70): agouti-related protein 1 (agrp1: forward 5′-CCTGAAATGGAGCACCTTGA-3′, reverse 5′-GGAGCGTGTGCCTCTTCTC-3′, from sequence reported in Ref. 61), cocaine- and amphetamine-regulated transcript protein type I (cart1: forward 5′-GCAGAGCAAACGGATCTCAC-3′, reverse 5′-TCCTCGATCCTTTCCTGATG-3′, from XM_680337), ghrelin preprohormone (forward 5′-AGCATGTTTCTGCTCCTGTGT-3′, reverse 5′-CTCTTCTGCCCACTCTTGGT-3′, from NM_001083872), and orexin preprohormone (forward 5′-TCTACGAGATGCTGTGCCGAG-3′, reverse 5′-CGTTTGCCAAGAGTGAGAATC-3′, from Ref. 38). We examined several housekeeping genes as potential reference genes for qRT-PCR analysis of brain. eIF-1α1 (eef1α1: forward 5′-GTACAGTTCCAATACCTCCA-3′, reverse 5′-GTACTACTCTTCTTGATGCCC-3′, from AY422992) was selected as the reference gene for analysis of gene expression in the brain because expression of this gene was not significantly affected by starvation (raw CT) and had the least variability in technical replicates. Because no comparisons were made between expression in brain and liver samples, normalization with different reference genes did not hamper the analysis. qRT-PCR was performed on the 8 brain RNA pools screened by the microarrays and the 16 remaining pools.

cDNA synthesis and qRT-PCR procedures were identical between the two tissues. Possible DNA contamination was removed by incubating 400 ng of total RNA at 37°C for 30 min with 1 unit of DNase I (Fermentas, Glen Burnie, MD) in a total volume of 8 μl. A chelating agent (1 μl of 25 mM EDTA) was added to each reaction to protect the RNA from hydrolysis, and the DNase I enzyme was denatured by incubation at 65°C for 10 min. First-strand cDNA synthesis was then performed. The entire DNase-treated sample was combined with 100 ng of random primers (Invitrogen) and 1 μl of dNTPs (10 mM) in a 12-μl total volume and incubated at 65°C for 5 min, 4°C for 10 min, and 25°C for 10 min. Next, the following reagents were added to the reactions for a 20-μl final volume: 4 μl of 5× First-Strand Buffer, 2 μl of DTT (0.1 M), and 100 U of Superscript II reverse transcriptase (Invitrogen). The cDNA reactions were then incubated at 42°C for 50 min, followed by heat inactivation of the reverse transcriptase (70°C for 15 min). The resulting cDNA was diluted 1:5 in Tris-EDTA (pH 8). qRT-PCR was performed separately for each gene in 20-μl total volume reactions consisting of 10 μl of 2× SYBR Green Master Mix (Applied Biosystems, Foster City, CA), 10 pmol of each primer, and 2 μl of diluted cDNA. The reactions were then run in an ABI Prism 7900HT Real-Time PCR System with the default PCR protocol (2 min at 50°C, 10 min at 95°C, followed by 40 cycles of 15 s at 95°C and 1 min at 60°C). Each 96-well plate included reactions for the reference gene and one target gene. A four-point dilution curve of a pool of all cDNA samples was included for each gene on each plate to monitor quality control, calculate amplification efficiencies, and aid comparisons among plates. Negative controls were included to test for contamination. Each sample was assayed with two technical replicates, each performed on separate plates. To verify the specificity of each primer pair, amplicons were separated by electrophoresis in agarose gels stained with ethidium bromide during primer optimization, and a melting curve analysis was performed at the completion of each qRT-PCR run.

Statistical Analysis

Phenotypic and transcriptional responses to starvation were assessed with analysis of covariance (ANCOVA) with PROC MIXED from SAS version 9 (SAS Institute, Cary, NC). Standard length and body mass were analyzed at each time point and GSI at the final time point. The model included Treatment (starved or control) as a main effect, Initial Age (dpf) as a covariate, Tank within Treatment as the experimental unit for testing the effect of Treatment, and Individuals as repeated measures. To satisfy assumptions of normality, mass was log-transformed and GSI was arcsine-transformed.

Specific growth rate (SGR, %) was calculated with the equation: 100 × (ln masst − ln mass0)/t, where t is the number of experimental days. Because fish were not individually marked, SGR was calculated from the mean body masses for each tank and therefore provided an index of the average growth rate for each tank. SGR was analyzed with ANCOVA as described above except that there were no repeated measures for each tank.

Finally, gene expression measures were analyzed in the form of CTs, using ANCOVA to test the main effect of Treatment, with the CTs of the reference gene as a covariate for normalization. This method of analysis gives results comparable to other methods of normalization and analysis (78). The ANCOVA approach has an advantage over other normalization techniques in that error variances around the target gene and reference gene are kept separate rather than compounded by normalization before statistical analysis. Also, the technique does not assume equal amplification efficiencies for the target and reference genes. The complete model tested Treatment as the main effect, Tank within Treatment as the experimental unit, Sample within Tank as the repeated measure for Tank, Technical Replicate within Sample as the repeated measure for Sample, and Reference gene expression as a covariate. Mean gene expression for each treatment was calculated with the LSMEANS statement, which adjusted for the reference gene covariate.


Phenotypic Response to Starvation

During the 11-day preexperiment period, average body mass (±SE) of female zebrafish increased from 330 mg (±78 mg) to 422 mg (±88 mg), a 27.7% (±6.3%) increase (Fig. 1). Average body length increased from 24.9 mm (±1.5 mm) to 26.5 mm (±1.6 mm), a 6.1% (±2.0%) increase. During this time, fish collectively consumed food at an average rate of 10.3% (±1.1%) of tank biomass per day. Throughout the 21-day starvation treatment, food consumption by fish in control tanks declined to an average of 4.7% (±0.6%) of tank biomass per day.

Fig. 1.
Mean body weight of fed and starved adult female zebrafish over the course of the experiment. All fish were fed to apparent satiety 3 times daily until day 0, after which fish in the starved group were no longer offered food. Error bars represent SE. ...

While control fish continued to grow, fish in the starvation treatment group showed a gradual decline in body mass, as would be expected (Fig. 1). After 7 days, the starved fish weighed significantly less than the control fish. By the end of the experiment, starved fish lost an average of 27.3% (±1.9%) of their initial body mass (day 0) while control fish gained 10.3% (±3.8%). SGR averaged −0.66%/day (±0.06%) for the starved fish and +0.20%/day (±0.01%) for the fed fish. Standard length did not change in starved fish, while it increased in control fish over the course of the experiment (data not shown). We noted that livers were dramatically smaller in the starved fish, but quantitative data were not collected. Declines of hepatosomatic index after starvation have been observed in another cyprinid, the ide (Leuciscus idus melanotus) (56). Starvation also reduced gonad growth and development, as shown by significantly lower GSI in the starved fish (6.6% ± 0.8%) compared with the fed fish (13.4% ± 0.8%) [F(1,9) = 15.85, P = 0.0032]. This agrees with findings that starvation dramatically decreased daily egg production by female zebrafish (21).

Effect of Starvation on Hepatic Transcriptome

In the liver, transcripts hybridizing to 3,797 probe sets were classified as expressed after application of our filter. Starvation significantly changed the expression of 574 of these probe sets, with 466 decreased and 108 increased in livers of starved fish (Supplemental Table S1).1

qRT-PCR successfully validated 9 of 10 genes identified by the microarray analysis, with a failure rate (10%) that was comparable to the FDR (5%) used in the microarray analysis. Validation of a subset of these genes is depicted in Fig. 2. Values for one-tailed t statistics were determined by taking the square root of the F statistic for the effect of starvation calculated in the ANCOVA. Successfully validated genes included β actin2 (1-tailed t = 2.7, df = 8, P = 0.0126), cldn7 (1-tailed t = 2.7, df = 8, P = 0.0136), eif3s6a (1-tailed t = 2.4, df = 8, P = 0.02), elovl5 (1-tailed t = −4.3, df = 8, P = 0.0013), fbp1 (1-tailed t = 2.2, df = 8, P = 0.0308), hadhb (1-tailed t = −1.9, df = 8, P = 0.0446), pex19 (1-tailed t = −2.0, df = 8, P = 0.0403), pklr (1-tailed t = −3.7, df = 8, P = 0.0031), and psmb5 (1-tailed t = −2.9, df = 8, P = 0.0103). qRT-PCR analysis of ghitm did not show a significant effect of starvation (1-tailed t = 0.8, df = 8, P = 0.7696), and this gene may represent a false positive. Comparisons between microarray and qRT-PCR measures of six of these genes are depicted in Fig. 2.

Fig. 2.
Validation of microarray results for 6 genes in the hepatic transcriptome that responded to starvation in adult female zebrafish. Log2 mean ratio of expression in livers of starved fish relative to controls (starved/control) measured from pooled liver ...

GO analysis using DAVID revealed several overrepresented GOs in our list of differentially expressed genes (Supplemental Table S2). The numbers of up- and downregulated genes in selected GOs are depicted in Fig. 3 and discussed in greater detail in the following sections.

Fig. 3.
Selected Gene Ontologies (GOs) that were significantly overrepresented (P < 0.05) among genes regulated by starvation in zebrafish liver. These GOs were identified with the Database for Annotation, Visualization, and Integrated Discovery (DAVID) ...

Lipid metabolism, biosynthesis, and transport.

Expression of five genes involved in fatty acid biosynthesis significantly decreased in livers of starved zebrafish (Table 1). These genes included stearoyl-CoA (Δ9) desaturase (scd), fatty acid (Δ6) desaturase 2 (fads2), elongation of very long-chain fatty acids-like 5 (elovl5), lysophospholipase 3 (lypla3), and phospholipase A2, group XIIB (pla2g12b). Both SCD and FADS2 catalyze insertions of double bonds into fatty acid chains and are rate-limiting steps in the synthesis of monounsaturated and polyunsaturated fatty acids, respectively (33, 45). Food deprivation has been shown previously to decrease SCD activity in mammals (22) and fish (24, 65). ELOVL5 is a polyunsaturated fatty acid elongase that lengthens monounsaturated and polyunsaturated fatty acids (73). In the present study, this gene was downregulated in the liver, in agreement with reduced mRNA expression and enzymatic activity previously observed in hepatocytes of fasted rats (73). Decreased expression of lypla3 and pla2g12b, both of which are involved in cell membrane remodeling functions (64), lends further evidence for an overall decrease in lipid synthesis within the liver of starved zebrafish.

Table 1.
Effect of starvation on expression of genes involved in lipid metabolism, binding, and transport

Generally, expression of genes involved in lipid binding and transport also decreased in livers of starved zebrafish. Downregulation of six genes, such as fatty acid binding protein 7a (fabp7a) and apolipoprotein A-IV (apoa4), suggests reductions in extracellular and intracellular lipid transport. Plasma levels of cholesterol, triglycerides, and total lipids decrease with food deprivation in European sea bass (Dicentrarchus labrax) (46), suggesting that decreased plasma lipids may coincide with decreased need for lipid transport. However, this effect was not universal, because expression of solute carrier family 27 (fatty acid transporter), member 2 (slc27a2) was significantly upregulated.

Starvation also decreased hepatic expression of genes involved in β-oxidation of fatty acids, including l-3-hydroxyacyl-CoA dehydrogenase, short chain (hadhsc), the α- and β-subunits of the hydroxyacyl-CoA dehydrogenase/3-ketoacyl-CoA thiolase/enoyl-CoA hydratase (trifunctional protein) (hadha and hadhb), and acyl-CoA oxidase, palmitoyl (acox1). HADHSC is required for β-oxidation of short-chain fatty acids and HADHA and HADHB for long-chain fatty acids. In the peroxisome, ACOX1 catalyzes an initial step of three reactions in peroxisomal β-oxidation, desaturating very long-chain acyl-CoAs to 2-trans-enoyl-CoAs. These observations suggest decreased β-oxidation in zebrafish liver after 21 days of starvation. In contrast, transcription and activities of genes associated with β-oxidation increased in mammals (2, 72).

Protein biosynthesis.

The patterns of expression of genes involved in protein metabolism suggest an overall decrease in protein biosynthesis. One way the liver may decrease protein biosynthesis is through regulation of eIFs. Expression of four subunits of the eIF3 complex increased in starved fish, while expression of subunits of eif1, eif2, and eif5 decreased (Table 2). eIFs are necessary to initiate translation and are highly conserved between yeast and mammals; however, in the absence of other eIFs, eIF3 has inhibitory effects, binding the 40S ribosomal subunit and preventing formation of the 80S ribosomal unit (17). In addition, during amino acid starvation in yeast, eIF2α is inactivated through phosphorylation by general control nonderepressible 2 (GCN2), inhibiting translation. This action requires activation of GCN2 kinase by a protein complex including GCN1 and GCN20 (28). Expression of GCN1-like 1 (gcn1l1), the vertebrate homolog of yeast gcn1, increased in livers of starved zebrafish, suggesting that this mechanism may be functioning similarly in fish. Thus transcriptional regulation may be an important mechanism underlying eIF-induced decreases in liver protein synthesis, in addition to phosphorylation and activation of specific proteins.

Table 2.
Effect of starvation on expression of genes affecting translation initiation and ribosomal proteins

Protein biosynthesis can also be decreased during starvation by reducing rates of ribosome biosynthesis, as observed in livers of starved rats (18) and freshwater catfish (Clarias batrachus) (66). In the present study, seven genes encoding ribosomal structural proteins were differentially expressed: five downregulated and two upregulated (Table 2). Overall, our findings suggest that zebrafish hepatocytes reduce protein biosynthesis as a response to reduced availability of ATP and nutrient substrates.


Genes involved in protein catabolism and the proteasome cell component were downregulated (Table 3) and significantly overrepresented in our list of responding genes (Supplemental Table S2). These genes included ubiquitin-conjugating enzyme E2A, which catalyzes the attachment of ubiquitin, marking proteins for degradation, and seven subunits of the proteasome that degrades ubiquitinated proteins (14). A transcript similar to TBP-interacting protein, also known as cullin-associated and neddylation-dissociated 1 (cand1), was also downregulated. In mammals, this protein interferes with formation of the ubiquitin ligase complex, inhibiting ubiquitination (43). Although the ubiquitin-proteasome system is the primary system responsible for proteolysis and amino acid recycling during starvation in mammalian skeletal muscle (14), observations in zebrafish (the present study) and rainbow trout (50–52) suggest this may not be the case in fish liver and skeletal muscle.

Table 3.
Effect of starvation on genes in ubiquitin-proteasome pathway

Expression of other genes involved in protein catabolism also decreased in livers of starved fish, including nothepsin, an aspartic proteinase homologous to cathepsin but with no known mammalian or avian ortholog (5). In zebrafish, expression of nothepsin is limited to females (47) and the specific function of the enzyme is not yet known. A transcript similar to ATP-dependent clp proteinase chain P (clpp), which hydrolyzes proteins into smaller peptides in an ATP-dependent reaction, was also downregulated.

Unfolded protein response.

Starvation also affected expression of another pathway associated with protein synthesis: the unfolded protein response (UPR) (reviewed in Ref. 55). This pathway is primarily associated with the response to excesses of unfolded proteins when translation exceeds the capacity of the endoplasmic reticulum (ER) to fold proteins into their final tertiary structures. The UPR also appears to have a role in nutrient sensing, as evidenced by its repression during nitrogen starvation in yeast (54, 74) and the activation of some of its components during glucose starvation in mice (55). The UPR is largely unexplored in organisms other than yeast and mice, so annotation of this system is sparse in zebrafish; however, a number of genes associated with the UPR in other species were differentially expressed in our data (Table 4), suggesting a similar involvement of the UPR with starvation in zebrafish. Genes with peptidylprolyl cis-trans isomerase activity (cyclophilin B and FK506 binding proteins) showed decreased expression, while members of the heat shock protein 90 family (hsp90ab1 and grp94) and the DnaJ family (hsp40) increased in expression. In particular, grp94 is an ER-specific chaperone that is upregulated by glucose starvation in chick fibroblasts (reviewed in Refs. 1, 11) and was induced by starvation in our study. Another ER chaperone of the UPR, heat shock 70-kDa protein 5 (hspa5), also known as grp78 or bip, also showed a trend of higher expression in livers of starved fish, but the increase was not significant after Benjamini-Hochberg correction (P = 0.07). Client-specific chaperone proteins, including lectin, mannose binding, 1 (lman1), and B cell receptor-associated protein 31 (bcap31) were also downregulated. In addition, cAMP-responsive element binding protein 3-like 3 (creb3l3) is a liver-specific transcription factor that is posttranslationally activated by ER stress in mammals (40, 80) but was downregulated in the present study. Finally, expression of Der1-like domain family, member 1 (derl1), involved in ER-associated degradation (ERAD) of misfolded proteins (55), also decreased. These results suggest downregulation of protein processing by the ER and ER stress-related components of the UPR but an upregulation of nutrient-sensing components (especially grp94). To our knowledge, this is the first reported evidence of involvement of components of the UPR in the response to starvation in a species other than yeast or mouse.

Table 4.
Effect of starvation on expression of genes involved in protein folding and the unfolded protein response

Glycolysis, tricarboxylic acid cycle, and oxidative phosphorylation.

Most genes associated with primary metabolism through glycolysis, the tricarboxylic acid (TCA) cycle, and oxidative phosphorylation were significantly downregulated in starved fish (Tables 5 and and6).6). For example, expression of hexokinase IV (hex4; aka glucokinase) significantly decreased in livers of starved zebrafish. This gene is inducible by feeding and available dietary carbohydrates in many fish species, including zebrafish (49). Similarly, hepatic expression and activity of HEX4 decreased in gilthead sea bream undergoing starvation or reduced rations (7). No studies of circulating levels of fatty acids, glucose, and insulin in zebrafish after starvation have yet been reported.

Table 5.
Effect of starvation on expression of genes involved in carbohydrate metabolism
Table 6.
Effect of starvation on expression of genes involved in oxidative phosphorylation


Several genes with gluconeogenic functions were upregulated in starved zebrafish (Table 7), including malate dehydrogenase (mdh1), fructose 1,6-bisphosphatase 1 (fbp1), and isocitrate dehydrogenase 2 (idh2). FBP1 has the opposite function of phosphofructokinase, converting fructose 1,6-bisphosphate to fructose 6-phosphate during gluconeogenesis. Expression of fbp1 increased with starvation in our study, consistent with increased activity of FBP in livers of starved gilthead sea bream (Sparus aurata) (4). Expression of mitochondrial isocitrate dehydrogenase (idh2) also increased with the starvation treatment. In addition to converting isocitrate to α-ketoglutarate during the TCA cycle, IDH2 produces NADH from the catalysis of isocitrate from other sources, such as fatty acid catabolism.

Table 7.
Effect of starvation on expression of genes involved in gluconeogenesis

Oxidative stress and cell redox homeostasis.

In mammals and fish, starvation leads to accumulation of reactive oxygen species (ROS), especially in the liver (13, 31, 44, 62). ROS can damage cell membranes, proteins, and DNA and are believed to be the main cause of cellular and organ damage resulting from starvation (48). Contrary to expectations, genes involved with neutralizing ROS and ameliorating oxidative stress were downregulated in livers of starved zebrafish (Table 8), including superoxide dismutase 2 (sod2), two glutathione peroxidases (gpx1a and gpx4b), and several other selenium-binding proteins with known or putative antioxidant functions (19). Downregulation of these genes disagrees with observations that activities of antioxidant enzymes increase in liver during starvation in other fish species (31, 34, 44) and suggests that prolonged starvation decreased the capacity of zebrafish liver to ameliorate oxidative stress.

Table 8.
Effect of starvation on expression of genes involved in oxidative stress and cell redox homeostasis

Despite the downregulation of some genes involved with neutralizing ROS and ameliorating oxidative stress, we did observe some indications of increased response to oxidative stress in livers of starved zebrafish. We observed increased expression of selenoprotein H (seph) in starved zebrafish, the only selenoprotein that exhibited this pattern. Selenoprotein H encodes a protein with DNA binding and oxidoreductase activities (39, 41). SEPH also appears to act as a transcription factor that is activated by oxidative stress. Overexpression of human seph in murine hippocampal cells increased the antioxidant capacity of transformed cells and upregulated transcription of endogenous seph and glutathione S-transferase genes that are involved in glutathione synthesis and phase II detoxification (41). In our study, increased expression of a transcript similar to microsomal glutathione S-transferase (Table 8) may have resulted from regulation by SEPH. Upregulation of seph contradicts the downregulation of other genes involved in the response to oxidative stress, suggesting that those genes are regulated by a different mechanism.

An increase in oxidative stress, coupled with decreased expression of genes that ameliorate this stress, may have caused inflammation and cellular damage as suggested by increased expression of high mobility group box 1 (hmgb1) in livers of starved zebrafish. HMGB1 is the most abundant nonhistone protein in the nucleus, where it maintains nucleosome structure, regulates transcription, and interacts with steroid hormone receptors (77). HMGB1 also acts as a cellular damage signal that is released to the extracellular fluid passively during cell necrosis or actively from damaged or dying cells and activated macrophages, stimulating the release of cytokines and promoting chronic inflammation (77). Modification of HMGB1 by oxidative stress stimulates both translocation of HMGB1 to the nucleus and release of this protein to the extracellular fluid from hepatocytes (20, 63, 67). It is not known whether oxidative stress also has a role in regulating transcription of hmgb1.

Together these findings suggest that, as in other species, starvation increased oxidative stress in the liver of zebrafish. However, this prolonged starvation appears to have impaired the capacity to ameliorate this stress and may have resulted in inflammation and cellular damage.

Comparative observations of effects of starvation on hepatic transcriptome.

The effect of starvation on the transcriptome-wide response of liver has been studied in mouse (2), rat (79), carp (21, 72), and rainbow trout (52). These studies differed in laboratory methodology, gene composition of microarray platforms, and method of statistical analysis. Also, given differences in metabolic rate and body size, it is unlikely that the study organisms were in the same physiological phase of starvation. The studies also differ in sex and maturity of the study organisms, which can also affect gene expression (49). Although these differences complicate comparative interpretations of transcriptome-wide responses to starvation, several interesting comparative observations can be made.

The response in zebrafish was strikingly similar to that of immature female rainbow trout (~200 g), which showed an overall decrease in transcription in response to starvation, particularly of genes associated with biosynthesis and metabolism (52). In fact, at least 14 orthologous genes were regulated in the same direction in both rainbow trout and zebrafish (notable examples are identified in Tables 18). The results of both studies suggest an overall decrease in cellular activity of rainbow trout and zebrafish hepatocytes. Studies in mammalian cells (6) and fish hepatocytes (27) have shown that protein synthesis and RNA/DNA synthesis are energetically expensive, together constituting ~50% of total respiration. Protein synthesis and RNA/DNA synthesis were also the most sensitive to ATP production, dramatically decreasing when ATP levels were low (6), as would be expected during starvation.

Transcriptome analyses in mouse (2) and rat (79) also showed decreased expression of genes involved in biosynthetic pathways but, in contrast, showed an upregulation of a larger number of genes involved in energy mobilization and amelioration of oxidative stress. These findings agree with physiological observations in mammals (72). It is not known whether these differences can be attributed to differences in the overall physiological response to starvation between mammals and fish, or are simply the result of measurements at different physiological time points or other experimental differences.

In contrast, carp showed a much different response after 28 days of starvation, paradoxically elevating expression of genes associated with glycolysis, the TCA cycle, the ubiquitin-proteasome system, and lipid metabolism and decreased expression of genes associated with gluconeogenesis (21, 72). The fact that zebrafish are in the same family as carp (Cyprinidae) but have a response more similar to that of rainbow trout (family Salmonidae) is intriguing; however, considering that the cyprinid family is the largest family of fish (32), species diversity in the response to starvation is not unexpected. In fact, carp are known to store large amounts of hepatic glycogen compared with other fishes and maintain high levels during starvation (37).

Effect of Starvation on Brain Transcriptome

In contrast to the liver, starvation had little or no detectable effect on the transcriptome of whole brain (Table 9). Although 8,640 probe sets passed the filter of detectable expression in the brain, analysis with LIMMA detected only two differentially expressed genes, both of which were downregulated in starved fish: a transcript with weak similarity to human hemoglobin γ-A and γ-G chains (Dr.6751.1.S1_at: log2 fold change = −2.20, FDR-corrected P value = 0.0425) and matrix metalloproteinase 9 (mmp9: log2 fold change = −0.97, FDR-corrected P value = 0.0239). Neither gene could be validated by qRT-PCR, suggesting that these genes may be false positives (Dr.6751.1.S1_at: 1-tailed t = 0.88, df = 10, P = 0.8002; mmp9: 1-tailed t = 0.81, df = 10, P = 0.7816). However, qRT-PCR agreed with the microarray analysis in detecting no effect of starvation on expression of neuropeptide Y (npy: 2-tailed t = −0.57, df = 10, P = 0.5824) and a predicted gene similar to agouti-related protein 2 (Dr.20586.1.A1_at: 2-tailed t = −0.80, df = 10, P = 0.4409). Quality control analyses of the microarray data, such as examination of variation among arrays and RNA degradation, did not indicate that the lack of response in the brain was a result of technical variability.

Table 9.
Transcriptional response to starvation in brain of adult female zebrafish analyzed with microarrays and qRT-PCR

This lack of a transcriptional response of whole brain to starvation was also evident from qRT-PCR analysis of neuropeptides that are known to influence feeding behavior (Table 9). Of these, agouti-related protein 1 (agrp1) was the only neuropeptide to change expression in response to food deprivation. AGRP has orexigenic effects in mammals and functions by blocking anorexigenic signaling through melanocortin receptors (75), and expression is known to increase in the brain during starvation in zebrafish (61) and goldfish (8) in response to food deprivation. Starvation did not have significant effects on expression of npy, ghrelin, cart1, and orex in the whole brains of the starved zebrafish in our study, in contradiction to the results of another study in zebrafish (38) and observations in other fish species (25, 36, 57, 68, 71).

The lack of a transcriptional response in the zebrafish brain is surprising considering that starvation is known to decrease metabolism in whole brain in other fish species (59) and several neuropeptide and metabolic genes are transcriptionally regulated in the brain during starvation (16, 60, 69, 70). The apparent absence of a transcriptomic response has several potential explanations. First, the most likely explanation is that responses to starvation are often isolated to specific regions of the brain, and this may render them undetectable by analyses using RNA from whole brain. Numerous studies have detected transcriptional responses to starvation that are localized to specific regions of the brain (35, 57, 60, 71), and in some cases the responding region differed among fish species (35, 36, 57). This does not explain the disparity between our study and that of Novak et al. (38), who detected effects of starvation on expression of orexin in whole zebrafish brain. In addition, the 24-h fast before sampling may have had a greater than expected impact on gene expression in the brains of control fish, resulting in expression profiles more similar to those in starved fish. This may be particularly important for npy, ghrelin, and cart1, which are regulated periprandially in other fish species (36, 68, 71). Finally, transcripts expressed in the brain may be poorly represented on the Zebrafish Genome Array. This possibility seems unlikely considering that the total number of transcripts detected as expressed in the zebrafish brain was quite high.

Comparative observations of effects of starvation on transcriptome of whole brain.

The low responsiveness of zebrafish brain to starvation agrees with the low numbers of genes responding in mouse and rat (9, 26, 30), despite focus on the hypothalamic region. There was no obvious overlap of responding genes among the studies (9, 26, 30). Genes that responded to starvation in the hypothalamus of rodents were involved in metabolic, regulatory, and structural functions, as well as neuroprotection.


Prolonged starvation had significant and widespread effects on the hepatic transcriptome of adult female zebrafish, influencing 574 of 3,797 expressed genes. The patterns of regulation corresponded with expectations of increased energy conservation, including overall reductions in biosynthesis, metabolic activity, and overall transcription. We also observed increased expression of genes involved in gluconeogenesis. A number of possible mechanisms for transcriptional regulation were suggested by the data, including nutrient levels (e.g., grp94) and transcription factors involved in the UPR (e.g., creb3l3) or response to oxidative stress (e.g., seph). In addition, our study is the first report of a response of the UPR to starvation in a species other than yeast and mice. The general pattern of the transcriptomic response in zebrafish liver was more similar to that in rainbow trout but differed from that in carp, mouse, and rat. In contrast, the transcriptome of the zebrafish brain showed very little response to starvation.


This research was supported by funding from the National Science Foundation and the Idaho Experimental Program to Stimulate Competitive Research (EPSCoR) under award number EPS0447689. The zebrafish facilities were constructed with funding from National Institutes of Health Grant P20-RR-016448 from the Centers of Biomedical Research Excellence (COBRE) Program of the National Center for Research Resources.

Supplementary Material

[Supplemental Tables]


The authors are grateful to M. Papasani, S. Lewis, M. Benner, and M. Oswald for assistance during dissections and tissue collections. D. Pouchnik and the Center for Reproductive Biology Genomics Core Lab at Washington State University performed the microarray hybridizations. Finally, the authors thank T. Cavileer for providing training and advice in the collection and analysis of qRT-PCR data.


Address for reprint requests and other correspondence: B. D. Robison, Dept. of Biology, PO Box 443051, Univ. of Idaho, Moscow, ID 83844-3051 (e-mail: ude.ohadiu@nosiborb).

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.


1The online version of this manuscript contains supplemental material.


1. Argon Y, Simen BB. GRP94, an ER chaperone with protein and peptide binding properties. Semin Cell Dev Biol 10: 495–505, 1999. [PubMed]
2. Bauer M, Hamm AC, Bonaus M, Jacob A, Jaekel J, Schorle H, Pankratz MJ, Katzenberger JD. Starvation response in mouse liver shows strong correlation with life-span-prolonging processes. Physiol Genomics 17: 230–244, 2004. [PubMed]
3. Benjamini Y, Hochberg Y. On the adaptive control of the false discovery rate in multiple testing with independent statistics. J Educ Behav Stat 25: 60–83, 2000.
4. Bonamusa L, García de Frutos P, Fernandez F, Baanante IV. Nutritional effects on key glycolytic-gluconeogenic enzyme activities and metabolic levels in liver of the teleost fish Sparus aurata. Mol Marine Biol Biotechnol 1: 113–125, 1992.
5. Borrelli L, De Stasio R, Filosa S, Parisi E, Riggio M, RS, Trinchella F. Evolutionary fate of duplicate genes encoding aspartic proteinases. Nothepsin case study. Gene 368: 101–109, 2006. [PubMed]
6. Buttgereit F, Brand MD. A hierarchy of ATP-consuming processes in mammalian cells. Biochem J 312: 163–167, 1995. [PMC free article] [PubMed]
7. Caseras A, Metón I, Fernández F, Baanante IV. Glucokinase gene expression is nutritionally regulated in liver of gilthead sea bream (Sparus aurata). Biochim Biophys Acta 1493: 135–141, 2000. [PubMed]
8. Cerdá-Reverter JM, Peter RE. Endogenous melanocortin antagonist in fish: structure, brain mapping, and regulation by fasting of the goldfish agouti-related protein gene. Endocrinology 144: 4552–4561, 2003. [PubMed]
9. Chiba T, Fujita S, Kubota H, Inoue D, Mizuno A, Komatsu T, Yamaza H, Higami Y, Shimokawa I. Identification of fasting-induced genes in the rat hypothalamus: relationship with neuroprotection. Ann NY Acad Sci 1119: 216–226, 2007. [PubMed]
10. Clarke A, Johnston NM. Scaling of metabolic rate with body mass and temperature in teleost fish. J Anim Ecol 68: 893–905, 1999.
11. Csermely P, Schnaider T, Sõti C, Prohászka Z, Nardai G. The 90-kDa molecular chaperone family: structure, function, and clinical applications. Pharmacol Ther 79: 129–168, 1998. [PubMed]
12. Dennis G, Sherman BT, Hosack DA, Yang J, Gao W, Lane HC, Lempicki RA. DAVID: Database for Annotation, Visualization, and Integrated Discovery. Genome Biol 4: R60, 2003. [PMC free article] [PubMed]
13. Domenicali M, Caraceni P, Vendemiale G, Grattagliano I, Nardo B, Dall'Agata M, Santoni B, Trevisani F, Cavallari A, Altomare E, Bernardi M. Food deprivation exacerbates mitochondrial oxidative stress in rat liver exposed to ischemia-reperfusion injury. J Nutr 131: 105–110, 2001. [PubMed]
14. Finn PF, Dice JF. Proteolytic and lipolytic responses to starvation. Nutrition 22: 830–844, 2006. [PubMed]
15. Gentleman RC, Carey VJ, Bates DM, Bolstad B, Dettling M, Dudoit S, Ellis B, Gautier L, Ge Y, Gentry J, Hornik K, Hothorn T, Huber W, Iacus S, Irizarry R, Leisch F, Li C, Maechler M, Rossini AJ, Sawitzki G, Smith C, Smyth G, Tierney L, Yang YH, Zhang J. Bioconductor: open software development for computational biology and bioinformatics. Genome Biol 5: R80, 2004. [PMC free article] [PubMed]
16. Gorissen MH, Flik G, Huising MO. Peptides and proteins regulating food intake: a comparative view. Anim Biol 56: 447–473, 2006.
17. Hinnebusch A eIF3: a versatile scaffold for translation initiation complexes. Trends Biochem Sci 31: 553–562, 2006. [PubMed]
18. Hirsch CA, Hiatt HH. Turnover of liver ribosomes in fed and in fasted rats. J Biol Chem 241: 5936–5940, 1966. [PubMed]
19. Holben DH, Smith AM. The diverse role of selenium within selenoproteins: a review. J Am Diet Assoc 99: 836–843, 1999. [PubMed]
20. Hoppe G, Talcott KE, Bhattacharya SK, Crabb JW, Sears JE. Molecular basis for the redox control of nuclear transport of the structural chromatin protein HMGB1. Exp Cell Res 312: 3526–3538, 2006. [PubMed]
21. Hung CY Survival Strategies of Common Carp, Cyprinus carpio, During Prolonged Starvation and Hypoxia (PhD thesis). Hong Kong: City University of Hong Kong, 2005, p. 289.
22. Inkpen CA, Harris RA, Quackenbush FW. Differential responses to fasting and subsequent feeding by microsomal systems of rat liver: 6- and 9-desaturation of fatty acids. J Lipid Res 10: 277–282, 1969. [PubMed]
23. Irizarry RA, Hobbs B, Collin F, Beazer-Barclay YD, Antonellis KJ, Scherf U, Speed TP. Exploration, normalization, and summaries of high density oligonucleotide array probe level data. Biostatistics 4: 249–264, 2003. [PubMed]
24. Kayama M, Hirata M, Hisai T. Effect of water temperature on the desaturation of fatty acids in the carp. Bull Jpn Soc Sci Fish 52: 853–857, 1986.
25. Kehoe AS, Volkoff H. Cloning and characterization of neuropeptide Y (NPY) and cocaine and amphetamine regulated transcript (CART) in Atlantic cod (Gadus morhua). Comp Biochem Physiol A Comp Physiol 146: 451–461, 2007. [PubMed]
26. Li JY, Lescure PA, Misek DE, Lai YM, Chai BX, Kuick R, Thompson RC, Demo RM, Kurnit DM, Michailidis G, Hanash SM, Gantz I. Food deprivation-induced expression of minoxidil sulfotransferase in the hypothalamus uncovered by microarray analysis. J Biol Chem 277: 9069–9076, 2002. [PubMed]
27. Mark FC, Hirse T, Pörtner HO. Thermal sensitivity of cellular energy budgets in some Antarctic fish hepatocytes. Polar Biol 28: 805–814, 2005.
28. Marton MJ, Vazquez de Aldana CR, Qiu H, Chakraburtty K, Hinnebusch AG. Evidence that GCN1 and GCN20, translational regulators of GCN4, function on elongating ribosomes in activation of eIF2a kinase GCN2. Mol Cell Biol 17: 4474–4489, 1997. [PMC free article] [PubMed]
29. McNab BK The Physiological Ecology of Vertebrates. A View from Energetics. Ithaca, NY: Cornell Univ. Press, 2002.
30. Mobbs CV, Yen K, Mastaitis J, Nguyen H, Watson E, Wurmbach E, Sealfon SC, Brooks A, Salton SRJ. Mining microarrays for metabolic meaning: nutritional regulation of hypothalamic gene expression. Neurochem Res 29: 1093–1103, 2004. [PubMed]
31. Morales AE, Pérez-Jiménez A, Hidalgo MC, Abellán E, Cardenete G. Oxidative stress and antioxidant defenses after prolonged starvation in Dentex dentex liver. Comp Biochem Physiol C Toxicol Pharmcol 139: 153–161, 2004. [PubMed]
32. Moyle PB, Cech JJ. Fishes: An Introduction to Ichthyology. Upper Saddle River, NJ: Prentice Hall, 2000.
33. Nakamura MT, Nara TY. Structure, function, and dietary regulation of Δ6, Δ5, and Δ9 desaturases. Annu Rev Nutr 24: 345–376, 2004. [PubMed]
34. Nam YK, Cho YS, Choi BN, Kim KH, Kim SK, Kim DS. Alteration of antioxidant enzymes at the mRNA level during short-term starvation of rockbream Oplegnathus fasciatus. Fish Sci 71: 1385–1387, 2005.
35. Narnaware YK, Peter RE. Effects of food deprivation and refeeding on neuropeptide Y (NPY) mRNA levels in goldfish. Comp Biochem Physiol B Biochem Mol Biol 129: 633–637, 2001. [PubMed]
36. Narnaware YK, Peyon PP, Lin X, Peter RE. Regulation of food intake by neuropeptide Y in goldfish. Am J Physiol Regul Integr Comp Physiol 279: R1025–R1034, 2000. [PubMed]
37. Navarro I, Gutiérrez J. Fasting and starvation. In: Biochemistry and Molecular Biology of Fishes, edited by Hochachka PW, Mommsen TP. Amsterdam: Elsevier Science, 1995.
38. Novak CM, Jiang X, Wang C, Teske JA, Kotz CM, Levine JA. Caloric restriction and physical activity in zebrafish (Danio rerio). Neurosci Lett 383: 99–104, 2005. [PubMed]
39. Novoselov SV, Kryukov GV, Xu XM, Carlson BA, Hatfield DL, Gladyshev VN. Selenoprotein H is a nucleolar thioredoxin-like protein with a unique expression pattern. J Biol Chem 282: 11960–11968, 2007. [PubMed]
40. Omori Y, Imai J, Watanabe M, Komatsu T, Suzuki Y, Kataoka K, Watanabe S, Tanigami A, Sugano S. CREB-H: a novel mammalian transcription factor belonging to the CREB/ATF family and functioning vie the box-B element with a liver-specific expression. Nucleic Acids Res 29: 2154–2162, 2001. [PMC free article] [PubMed]
41. Panee J, Stoytcheva ZR, Liu W, Berry MJ. Selenoprotein H is a redox-sensing high mobility group family DNA-binding protein that up-regulates genes involved in glutathione synthesis and phase II detoxification. J Biol Chem 282: 23759–23765, 2007. [PubMed]
42. Papasani MR, Robison BD, Hardy RW, Hill RA. Early developmental expression of two insulins in zebrafish (Danio rerio). Physiol Genomics 27: 79–85, 2006. [PubMed]
43. Parry G, Estelle M. Regulation of cullin-based ubiquitin ligases by the Nedd8/RUB ubiquitin-like proteins. Semin Cell Dev Biol 15: 221–229, 2004. [PubMed]
44. Pascual P, Pedrajas JR, Toribio F, López-Barea J, Peinado J. Effect of food deprivation on oxidative stress biomarkers in fish (Sparus aurata). Chem Biol Interact 145: 191–199, 2003. [PubMed]
45. Pereira SL, Leonard AE, Mukerji P. Recent advances in the study of fatty acid desaturases from animals and lower eukaryotes. Prostaglandins Leukot Essent Fatty Acids 68: 97–106, 2003. [PubMed]
46. Pérez-Jiménez A, Guedes MJ, Morales AE, Oliva-Teles A. Metabolic responses to short starvation and refeeding in Dicentrarchus labrax. Effect of dietary composition. Aquaculture 265: 325–335, 2007.
47. Riggio M, Scudiero R, Filosa S, Parisi E. Sex- and tissue-specific expression of aspartic proteinases in Danio rerio (zebrafish). Gene 260: 67–75, 2000. [PubMed]
48. Rios FS, Donatti L, Fernandes MN, Kalinin AL, Rantin FT. Liver histopathology and accumulation of melano-macrophage centres in Hoplias malabaricus after long-term food deprivation and re-feeding. J Fish Biol 71: 1393–1406, 2007.
49. Robison BD, Drew RE, Murdoch GK, Powell M, Rodnick KJ, Settles M, Stone D, Churchill E, Hill RA, Papasani MR, Lewis SS, Hardy RW. Sexual dimorphism in hepatic gene expression and the response to dietary carbohydrate manipulation in the zebrafish (Danio rerio). Comp Biochem Physiol D Genomics Proteomics 3: 141–154, 2008. [PMC free article] [PubMed]
50. Salem M, Kenney B, Rexroad CE III, Yao J. Microarray gene expression analysis in atrophying rainbow trout muscle: a unique non-mammalian muscle degradation model. Physiol Genomics 28: 33–45, 2006. [PubMed]
51. Salem M, Kenney PB, Rexroad CE III, Yao J. Molecular characterization of muscle atrophy and proteolysis associated with spawning in rainbow trout. Comp Biochem Physiol D Genomics Proteomics 1: 227–237, 2006. [PubMed]
52. Salem M, Silverstein J, Rexroad CE III, Yao J. Effect of starvation on global gene expression and proteolysis in rainbow trout (Oncorhynchus mykiss). BMC Genomics 8: 328, 2007. [PMC free article] [PubMed]
53. Santos EM, Kille P, Workman VL, Paull GC, Tyler CR. Sexually dimorphic gene expression in the brains of mature zebrafish. Comp Biochem Physiol A Mol Integr Physiol 149: 314–324, 2008. [PubMed]
54. Schröder M, Clark R, Liu CY, Kaufman RJ. The unfolded protein response represses differentiation through the RPD3-SIN3 histone deacetylase. EMBO J 23: 2281–2292, 2004. [PMC free article] [PubMed]
55. Schröder M, Kaufman RJ. The mammalian unfolded protein response. Annu Rev Biochem 74: 739–789, 2005. [PubMed]
56. Segner H, Braunbeck T. Hepatocellular adaptation to extreme nutritional conditions in ide, Leuciscus idus melanotus L. (Cyprinidae). A morphofunctional analysis. Fish Physiol Biochem 5: 79–97, 1988. [PubMed]
57. Silverstein JT, Breininger J, Baskin DG, Plisetskaya EM. Neuropeptide Y-like gene expression in the salmon brain increases with fasting. Gen Comp Endocrinol 110: 157–165, 1998. [PubMed]
58. Smyth GK Linear models and empirical Bayes methods for assessing differential expression in microarray experiments. Stat Appl Genet Mol Biol 3: Article 3, 2004. [PubMed]
59. Soengas JL, Aldegunde M. Energy metabolism of fish brain. Comp Biochem Physiol B Biochem Mol Biol 131: 271–296, 2002. [PubMed]
60. Soengas JL, Polakof S, Chen X, Sangiao-Alvarellos S, Moon TW. Glucokinase and hexokinase expression and activities in rainbow trout tissues: changes with food deprivation and refeeding. Am J Physiol Regul Integr Comp Physiol 291: R810–R821, 2006. [PubMed]
61. Song Y, Golling G, Thacker TL, Cone RD. Agouti-related protein (AGRP) is conserved and regulated by metabolic state in the zebrafish, Danio rerio. Endocrine 22: 257–265, 2003. [PubMed]
62. Stadler M, Nuyens V, Seidel L, Albert A, Boogaerts JG. Effect of nutritional status on oxidative stress in an ex vivo perfused rat liver. Anesthesiology 103: 978–986, 2005. [PubMed]
63. Tang D, Kang R, Xiao W, Jiang L, Liu M, Shi Y, Wang K, Wang H, Xiao X. Nuclear heat shock protein 72 as a negative regulator of oxidative stress (hydrogen peroxide)-induced HMGB1 cytoplasmic translocation and release. J Immunol 178: 7376–7384, 2007. [PMC free article] [PubMed]
64. Tocher DR Metabolism and functions of lipids and fatty acids in teleost fish. Rev Fish Sci 11: 107–184, 2003.
65. Tocher DR, Bell JG, Sargent JR. Induction of Δ9-fatty acyl desaturation in rainbow trout (Oncorhynchus mykiss) liver by dietary manipulation. Comp Biochem Physiol B Biochem Mol Biol 113: 205–212, 1996.
66. Tripathi G, Verma P. Starvation-induced impairment of metabolism in a freshwater catfish. Z Naturforsch [C] 58: 446–451, 2003. [PubMed]
67. Tsung A, Klune JR, Zhang X, Jeyabalan G, Cao Z, Peng X, Stolz DB, Geller DA, Rosengart MR, Billiar TR. HMGB1 release induced by liver ischemia involves Toll-like receptor 4-dependent reactive oxygen species production and calcium-mediated signaling. J Exp Med 204: 2913–2923, 2007. [PMC free article] [PubMed]
68. Unniappan S, Canosa LF, Peter RE. Orexigenic actions of ghrelin in goldfish: feeding-induced changes in brain and gut mRNA expression and serum levels, and responses to central and peripheral injections. Neuroendocrinology 79: 100–108, 2004. [PubMed]
69. Volkoff H The role of neuropeptide Y, orexins, cocaine and amphetamine-related transcript, cholecystokinin, amylin and leptin in the regulation of feeding in fish. Comp Biochem Physiol A Mol Integr Physiol 144: 325–331, 2006. [PubMed]
70. Volkoff H, Canosa LF, Unniappan S, Cerdá-Reverter JM, Bernier NJ, Kelly SP, Peter RE. Neuropeptides and the control of food intake in fish. Gen Comp Endocrinol 142: 3–19, 2005. [PubMed]
71. Volkoff H, Peter RE. Characterization of two forms of cocaine- and amphetamine-regulated transcript (CART) peptide precursors in goldfish: molecular cloning and distribution, modulation of expression by nutritional status, and interactions with leptin. Endocrinology 142: 5076–5088, 2001. [PubMed]
72. Wang T, Hung CCY, Randall DJ. The comparative physiology of food deprivation: from feast to famine. Annu Rev Physiol 68: 223–251, 2006. [PubMed]
73. Wang Y, Botolin D, Christian B, Busik J, Xu J, Jump DB. Tissue-specific, nutritional, and developmental regulation of rat fatty acid elongases. J Lipid Res 46: 706–715, 2005. [PMC free article] [PubMed]
74. Williams RM, Primig M, Washburn BK, Winzeler EA, Bellis M, de Menthiere CS, Davis RW, Esposito RE. The Ume6 regulon coordinates metabolic and meiotic gene expression in yeast. Proc Natl Acad Sci USA 99: 13431–13436, 2002. [PMC free article] [PubMed]
75. Wilson BD, Ollmann MM, Barsh GS. The role of agouti-related protein in regulating body weight. Mol Med Today 5: 250–256, 1999. [PubMed]
76. Wright D, Nakamichi R, Krause J, Butlin RK. QTL analysis of behavioral and morphological differentiation between wild and laboratory zebrafish (Danio rerio). Behav Genet 36: 271–284, 2006. [PubMed]
77. Yamada S, Maruyama I. HMGB1, a novel inflammatory cytokine. Clin Chim Acta 375: 36–42, 2007. [PubMed]
78. Yuan JS, Reed A, Chen F, Stewart CN. Statistical analysis of real-time PCR data. BMC Bioinformatics 7: 85, 2006. [PMC free article] [PubMed]
79. Zhang J, Underwood LE, D'Ercole AJ. Hepatic mRNAs up-regulated by starvation: an expression profile determined by suppression subtractive hybridization. FASEB J 15: 1261–1263, 2001. [PubMed]
80. Zhang K, Shen X, Wu J, Sakaki K, Saunders T, Rutkowski T, Back SH, Kaufman RJ. Endoplasmic reticulum stress activates cleavage of CREBH to induce a systemic inflammatory response. Cell 124: 587–599, 2006. [PubMed]

Articles from Physiological Genomics are provided here courtesy of American Physiological Society
PubReader format: click here to try


Related citations in PubMed

See reviews...See all...

Cited by other articles in PMC

See all...