| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
1 School of Biology, University of St. Andrews, Fife, United Kingdom
2 Moscow State University of Instrument Engineering and Computer Science, Moscow, Russia
3 Department of Biology, University of Southern Georgia, Statesboro, Georgia
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMETHODSRESULTSDISCUSSIONGRANTSREFERENCES |
|---|
microarray; genes; differential expression
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONGRANTSREFERENCES |
|---|
Although hypothesis-driven approaches have successfully identified many ion and water transporters as well as hormonal systems that are involved in osmoregulation in FW and/or SW environments, there are undoubtedly many other, as yet unknown or uncharacterized, genes involved in salinity adaptation in the eel. Both genomic and proteomic methods are currently available for the screening of differentially expressed genes within the osmoregulatory tissues of the eel. In this paper we describe the generation and use of an eel tissue-specific microarray to identify novel genes that exhibit differential expression within the brain, gill, kidney, and gut when fish are transferred from FW to SW.
Microarray technology has been employed with a variety of model organisms, including various mammalian (92), fish (32, 115), and insect (41) species to study differential gene expression in association with a wide range of biological phenomena, including aspects of both normal body function and the study of pathologies of various diseases. Array methodologies have been used successfully to screen for salinity-dependent gene expression in a number of plant (98), yeast (51, 89), and bacterial (18) systems, and recently a limited number of studies have also examined the effects of environmental salinity on gene expression in fish (14, 39, 63). To date, there have been no reports describing transcriptomic approaches for the study of differential gene expression in the euryhaline European eel during acclimation to SW. This communication reports the results of such a study, where microarray methodologies enabled the comprehensive screening of the expressed genome in selected tissues to identify novel genes with uncharacterized biological functions that were associated with ion and water transport within the major osmoregulatory tissues of the European eel.
Subtractive suppressive hybridization methods (34) were used to construct a 6,144-feature, tissue-specific microarray comprising DNA fragments amplified from genes expressed in the brain, gill, kidney, and intestine of the eel. These arrays were used to screen for differential gene expression within these tissues during the acclimation of FW silver eels to the SW environment. Although previous studies in our laboratory have indicated that changes in expression of a few osmoregulatory-relevant genes are affected by the silvering process, the changes in expression of the majority of genes studied so far were correlated with the transfer of fish to SW (27–31). The results presented in this article are therefore related only to changes in gene expression initiated by FW/SW transfer. In a previous article we presented preliminary details of the production and validation of the microarray technique for salinity acclimation studies using known genes previously implicated in osmoregulation including prolactin, growth hormone, the Na-K-2Cl-cotransporter (NKCC2) and the Na-K-ATPase (52). In this article we present the results and analyses from the full microarray data set, where in addition to the identification of well-known salinity-stress genes other salinity-sensitive genes are highlighted, the role of which in osmoregulation awaits future investigations.
|
METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONGRANTSREFERENCES |
|---|
|
5% then refilled back to initial levels over a 1-h period with either FW (controls) or SW (salinity stressed). Tanks thereafter received free-flowing FW or SW [940 mosM/kg, with ionic composition (in meq/l) of Na+ (407), K+ (8.8), Ca2+ (9.0), Mg2+ (45.7), Cl– (470), pH 8.1] at ambient temperatures (5–10°C) for the remainder of the acclimation periods. Tanks containing six fish were set up for each time point. Five tissues (brain, renal and head kidney, gill, and intestine) were removed and sampled from each of six FW control and six SW-transferred fish at 6 h, 2 and 7 days, and 5 mo. The long-term time point of 5 mo was included as it has been calculated that eels will take at least this length of time to swim back to the Sargasso Sea to breed (116).
RNA preparation.
Total RNA was prepared by a modification of the method described by Chomczynski and Sacchi (19) as detailed in Mahmmoud et al. (64), and concentrations were determined with spectrophotometric absorbance at 260 nm. Quality assessment of isolated RNA was performed with a 2100 Bioanalyzer (Agilent Technologies, Palo Alto, CA). Poly(A)+ mRNA was prepared with oligo (dT)-cellulose according to the classical procedure presented in Sambrook et al. (96) from pools of total RNA prepared for each tissue from individual fish sampled for each time point and condition.
Construction of suppression subtraction hybridization libraries and microarray production.
Four subtracted cDNA libraries based on the suppression subtraction hybridization (SSH) technique (34) were constructed for brain, gill, renal plus head kidney, and intestine using a PCR-Select cDNA Subtraction Kit as detailed by the manufacturers (Clontech, Basingstoke, UK). Poly(A)+ RNA (4 µg) was used from each sample to generate the ds cDNAs then digested by the restriction endonuclease RsaI. cDNAs produced from each tissue (tester) were subtracted against an 18-fold excess of cDNA prepared from the combination of the three other tissues taken in equimolar amounts (driver) to obtain the representative collection of tissue-specific expressed sequence tags. The subtracted cDNA was purified, cloned to the pCR4-TOPO vector (Invitrogen, Paisley, UK), and then transformed into DH5
Escherichia coli (Eurogentec). Following cloning, 1,536 colonies (4 plates of 384 wells) were randomly picked from each tissue-subtracted library, and 50 clones were randomly selected and sequenced. The inserts from all clones were amplified using SSH nested primers (Clontech, Basingstoke, UK), and their integrity was verified by agarose gel electrophoresis. Successful PCR products (6,144 cDNAs) were spotted onto GAPS II-coated slides (Corning, Corning, NY) (three technical replicates per slide) by the Liverpool Microarray Facility (School of Biological Sciences, The University of Liverpool).
Probes, labeling, and hybridization.
In the initial experiments the cDNAs used for hybridization were prepared from 20-µg samples of total RNA pooled from six salinity-acclimated (FW/SW) or six control (FW/FW) fish for each time point (6 h, 2 and 7 days, 5 mo) and for each tissue (gill, intestine, brain, and renal kidney). In subsequent experiments, cDNAs obtained from 20 µg of total RNA isolated from each of the six individual SW-acclimated fish (at the 6-h and 5-mo time points for gill and intestine) were hybridized against cDNAs prepared from 20 µg of total RNA pooled from the same tissues of six FW/FW fish at the equivalent time points. The cDNAs were labeled with Cy3 or Cy5 (SW-acclimated or control fish) in dye swap according to the manufacturer's instructions (CyScribe Post-Labelling Kit; Amersham Biosciences, Little Chalfont, UK). The microarrays were prehybridized for 1 h and then hybridized with labeled probes overnight at 44°C, washed, and dried as described in the manufacturer's protocol for GAPS II-coated slides (Corning). Of the 6,144 clones spotted, 96 contained control cDNAs comprising either known eel genes implicated in osmoregulation and previously studied in our laboratory or coding regions of genes sequenced from other species (i.e., Aspergillus nidulans, Arabidopsis thaliana, Platichthys flesus, Carcharhinus leucas, Scyliorhinus canicula, Squalus acanthias, and Salmo salar). The labeled cDNA probes derived from the eel RNAs hybridized only with the spotted cDNAs from the eel clones indicating the stringency and probe specificities of the hybridization conditions were appropriate.
Microarray design and reproducibility.
A simple and effective design for the direct comparison of two samples in dye-swap initial experiments was chosen (20). In this initial experiment, labeled cDNAs derived from pools of six FW samples for each tissue and time point were compared with the equivalent labeled, time- and tissue-matched cDNAs from RNA pooled from six SW fish. Maximum differences in gene expression were mainly found between the long-term acclimated (5 mo) FW and SW groups. The interfish variability was investigated in the long-term (5 mo) FW/FW- and FW/SW-acclimated groups by reference design methodology (20). Labeled cDNAs were prepared from the gill and intestine of six individual FW/SW-transferred fish and then hybridized along with a common (for each tissue) reference obtained from a pool of six FW/FW fish RNAs. There was high reproducibility between successive hybridizations using the same sample (52).
Data acquisition, normalization, and analysis.
Hybridized slides were scanned at 10 µm resolution using a ScanArray Lite, Microarray Analysis System (Perkin Elmer Life Sciences, Beaconsfield, UK) and saved as *TIF files. The fluorescent signal intensities of both dyes for each spot (at 532 nm for Cy3 and 635 nm for Cy5) were measured using ScanArray and Quantarray microarray analysis software (Perkin Elmer Life Sciences).
Madscan software, developed by INSERM U 533 in Nantes, France (59), was used for the stepwise normalization of gene intensities. Following background subtraction, results from features on the array were considered to be valid if the signal to noise fluorescence intensity ratio in both channels was >3. A Rank invariant method (114) was applied to select a sufficient number of nondifferentially expressed genes on each slide for the construction of normalization curves (Lowess fitness normalization) (54). We calculated an average value for each triplicate spot within the array before completing various methods of data analysis. Hierarchical clustering with Cluster and Tree View software developed by Eisen et al. (35) was performed to obtain a good visual data representation. The Cluster program calculates the difference between gene expression levels, which allows the clustering of genes with similar expression profiles.
A classical "fold change" method was used for the selection of differentially expressed genes. Using the fold change method, we considered genes overexpressed if the ratio between salinity stress and control normalized expression values was reproducibly (including dye swap) >1.5 (mRNA more abundant in fish encountering salinity stress) and underexpressed if the ratio was <0.5 (mRNA is less abundant in fish encountering salinity stress).
Identification of unique, nonredundant clones exhibiting differential expression and clustering of genes with similar functions and/or expression profiles was performed using Clcopy software (Zaguinaiko VA, unpublished data).
Sequencing analysis.
Differentially expressed clones were sequenced (Macrogen), edited from expression vector and adaptor sequences, quality assessed by electrophoretogram viability, and aligned with BLAST (basic local alignment search tool) against available databases using Trace2dBest (5) before being assembled into contigs using ClustalX software (111).
Northern blotting and analysis.
Total RNAs (20 µg) extracted from tissues isolated from individual fish were size fractionated by 1% agarose-formaldehyde gel electrophoresis and electroblotted onto Zeta Probe nylon membranes (Bio-Rad) as described previously (64). Complementary DNA fragments of 28 genes identified as being differentially expressed by microarray analyses were amplified and used as specific probes for Northern blotting. The cDNA fragments were radiolabeled using [32P]dCTP and hybridized to the blots in UltraHyb hybridization solution (Ambion) at 52°C for 16 h as detailed previously (64). After repetitive washing in 0.2x SSC, 0.1% SDS at 52°C, the blots were dried, and the RNA was cross-linked to the membrane using the optimal crosslink function on a Spectrolinker XL-1500 UV Crosslinker (Spectronics, Westbury, NY). Membranes were exposed to Kodak film for different times (according the level of gene expression) to visualize specific mRNA detection. The extent of radiolabeled probe hybridizing to the mRNA of interest was quantitatively assessed with an Instant Imager (Packard) and normalized against intensities for 28S and 18S RNA as detailed previously (27, 64) (1x SSC = 0.15 M NaCl, 15 mM sodium citrate, pH 7.0).
Statistical analysis.
Student's t-tests were used to determine the statistical significance in gene expression between FW- and SW-acclimated fish in both Northern blot and microarray data analysis conducted using the biological replicates at the 5-mo time point.
MIAME compliance.
All data sets, hybridization analyses, sequence, and metadata files were collected and processed in compliance with MIAME standards with all information submitted to the Array Express database, European Bioinformatics Institute (EBI), Oxford, UK [www.ebi.ac.uk/aerep/login(accession no. E-MAXD-24)].
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONGRANTSREFERENCES |
|---|
Selection and sequencing of differentially expressed genes.
The effect of SW transfer on gene expression within the brain, gill, intestine, and kidney was investigated using pooled RNA samples from six SW and six FW fish at up to four time points during acclimation, 6 h, 2 and 7 days, and 5 mo. Genes with consistent SW/FW expression ratios of >1.5 or <0.5 in both initial and dye-swap experiments were selected. In addition, at the 5-mo time point, samples from the intestine and gill of six individual SW-acclimated fish were taken for analysis against pooled RNA samples from six time-matched FW/FW controls. Genes showing SW/FW expression values with a two-standard deviation difference in Student's t-test were selected.
Sequencing and functional analysis of differentially expressed genes.
The majority of genes identified as exhibiting differential expression were detected in the 5-mo FW/SW transfer group. In total, 782 clones were selected as potential differentially expressed genes and sequenced: 61 from brain and 222, 190, and 309 from gill, intestine, and kidney, respectively. The percentages of unique sequences were 39 for brain, 61 for gill, 54 for intestine, and 33 for kidney, the remainder being either redundant clones or unknown sequences. The sequences were analyzed using BlastX and BlastN searches against the National Center for Biotechnology Information data base (www.ncbi.nlm.nih.gov/BLAST).
The significance level used for identification of a known functional protein was equivalent to a BlastX score of E < 10–5 and identity equal to at least 30% over a minimum range of 40 sequential amino acids. Table 2 represents the result of BlastX analysis of sequences for all time points.
|
|
Cluster analysis allows the combination and visualization of genes exhibiting similar expression profiles. Typical hierarchical clustering of genes with similar up- or downregulated expression profiles is shown for the intestine in Fig. 2. The expression profiles and possible functions of some of these genes along with those exhibiting differential expression in other tissues are described in more detail in DISCUSSION.
|
The majority of the genes presented in Table 2 exhibited approximately twofold changes in expression between FW- and SW-acclimated fish, but for a few genes this difference was much greater and up to-10 fold for some transcripts [e.g., zymogen granule membrane protein 16 (ZG-16)]. A few genes exhibited simultaneous differential expression in more than one tissue (e.g., myo-inositol monophosphatase 1, which was upregulated in both gill and in kidney of SW-acclimated fish), whereas the majority were expressed and differentially regulated in a tissue-specific manner [e.g., angiotensin-converting enzyme (ACE), upregulated in the kidney of 5-mo acclimated SW eels].
Validation of the microarray analysis by Northern blotting.
To validate the microarray results, 28 genes from three libraries, which were identified as showing differential expression in one or more tissues following FW/SW transfer were selected and used as probes to compare transcript abundance by Northern blotting (n = 6 for SW or FW samples). These clones (21 known genes implicated in different physiological functions and 7 unknown genes) were chosen to represent the wide range of genes listed in Table 2 that were either upregulated or downregulated and therefore functionally implicated in the process of fish adaptation to salinity transfer by the microarray study.
A comparison of the results obtained for these genes selected as differentially expressed by both microarray and the subsequent Northern blot analyses is presented in Fig. 3, A–C. There were high levels of consistency in the levels of expression for both up- and downregulated genes between the two methods. For some genes (see for example expression of nephrosin in gill) the variability between individual fish within the same group (SW or FW) is large, and this may account for the "differential expression" found after microarray analysis when pooled samples from FW fish were hybridized against pooled samples of SW fish. This observation of high variability of expression of the nephrosin gene agrees with the results reported by Boutet et al. (14) in the gill. Figure 3D presents examples of typical Northern blots for some up- or downregulated genes in samples isolated from intestine, kidney, and gill of FW- and SW-acclimated fish.
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONGRANTSREFERENCES |
|---|
Four tissues (kidney, intestine, gill, and brain) were investigated at four selected time points (6 h, 2 and 7 days, and 5 mo after transfer of fish to SW). The procedure included the use of the SSH method that had recently been successfully used for studies on osmoregulation in tilapia (Orechromis mossambicus) (39). The SSH technique was used to enrich for low abundance tissue-specific transcripts from each tissue. Clones selected from each of the tissue libraries were amplified and used to prepare microarrays to screen RNA samples for genes exhibiting salinity-selective expression. A variety of salinity-sensitive genes were identified from each of the tissues investigated, ranging from a few expected and well-characterized ion and water transporters to various other genes products not normally associated with osmoregulation, including proteins associated with cell signaling, cell structure, transcription, translation, and the immune response. Moreover, a number of other gene products exhibited differential expression but these were not identified due to limited sequence homologies with transcripts already deposited in the gene banks. Additional sequence information will be required before any further information can be gleaned about the possible functions of these gene products.
Salinity-sensitive genes were classified into eight different functional groups, and some of those identified and listed in these groups are discussed below.
Genes involved in cell protection and the immune system.
The largest of the functional groups exhibiting differential expression following SW transfer (18% of the identifiable genes) comprise genes implicated in some aspects of cell/tissue protection (immune system and detoxification). The large number of differentially expressed genes involved in cell protection may be explained by the possibility that the eel immune system is constantly responding to various external stressors related to habitat changes, which include perturbations in environmental salinity and the association with a variety of autochthonous microorganisms living in either freshwater or marine environments (13, 82).
The majority of the genes implicated in cell protection were downregulated 5 mo after fish were transferred to SW. For example, C-type lectin, which was downregulated by two- to fivefold in gill, is known to be implicated in pathogen recognition and immunoresponse. Similar results were found by Mistry et al. (74) who reported the downregulation of C-type lectin in the gills of Japanese eels transferred from FW to SW. The C-type lectins are involved in many immune-system functions, from inflammation to identification of tumors and virally infected cells (25). They can cause erythrocyte agglutination and may help to repress bacterial growth (73). The changes in expression of C-type lectin suggest that the functional role of this gene is more important in FW than SW.
The renal expression of myeloid protein-1 mRNA (the most redundant clone sequenced) was found to be doubled in long term SW-acclimated eels. The gene was represented by 53 different clones forming three different contigs. Northern blot analysis confirmed the significant increase in expression of this transcript in SW-acclimated fish (Fig. 3D). Myeloid protein-1 is stored within the secretory granules of promyelocytes (83) and expression is reported to be specifically induced during granulopoiesis (90). The biological function of this protein has not been fully elucidated although it is known to have acetyltransferase activity, suggesting a possible role in gene regulation (2). This protein is coded by the mim-1 gene, an endogenous marker of myeloid differentiation (84) that is directly regulated by the product of the v-myb oncogene (83). It is interesting to note that in adulthood the major site of hematopoiesis in the zebrafish is the kidney (12). The relationship between the hematopoietic myeloid protein functions and osmoregulation needs to be elucidated.
The expression of major histocompatibility complex (MHC) class II mRNA was suppressed by two- to threefold in the gill, kidney, and intestine 5 mo after eel transfer to SW. This protein belongs to a group of molecules known as the immunoglobulin supergene family, which presents the peptide products of lysosomal degradation for CD4+ T cell recognition and therefore has potential relevance to viral immunity, tumor immunity, and autoimmunity (77). Macroautophagy is an universal process that eukaryotic cells employ to reutilize the constituents of the cytoplasm and organelles and that has been proposed to play an essential role during development (77) and oncogenesis (22), as well as in response to physiological stresses associated with starvation (6). The MHC class II molecules have important roles in starvation-induced cellular autophagy, where peptides produced by incomplete lysosomal proteolysis exhibit increased presentation at the cell surface (24, 33). Under certain conditions, MHC class II proteins are also known to signal apoptosis in the antigen presenting cell (123). The suppression of MHC class II expression may therefore be related to cellular atrophy and/or apoptosis associated with regulated changes in epithelial functions in the gill, kidney, and intestine when nonfeeding migratory silver eels enter SW and start to drink.
Genes involved in energetic metabolism.
NADH dehydrogenase was downregulated in the kidney by twofold 2 days after fish transfer to SW. This enzyme is a component of Complex I of the mitochondrial respiratory chain, which catalyzes the transfer of electrons from NADH to coenzyme Q (CoQ). In this process, the complex translocates protons across the inner membrane and helps to build the electrochemical potential used to produce ATP. These results suggest SW acclimation is associated with decreased respiratory activity and energy production in kidney.
Genes involved in detoxification.
Pi-class glutathione S-transferase (GST), which was downregulated by 2.5-fold in the intestine 5 mo after eel transfer to SW, plays an important role in the detoxification of a variety of chemical toxins and mutagens (9). It was noticed that expression was significantly diminished in diet-restricted mice in all age groups (81), suggesting another possible link to the nutritional status of the nonfeeding SW-acclimated eels.
Genes involved in signal transduction.
Inositol (myo)-1 (or 4)-monophosphatase 1 was upregulated by 3.5- to 4-fold in the kidney and gill of SW-acclimating fish at 2 and 7 days and 5 mo after eel transfer. This enzyme has an important role in cell signaling, catalyzing the dephosphorylation of various myo-inositol monophosphates to free myo-inositol (99). Inositol monophosphatase is known to be a central enzyme in the phosphatidyl-inositol signaling system in all tissues and especially the brain. Its activity is inhibited by lithium, which is a mood-stabilizing medication used in patients suffering from various psychiatric disorders including manic depression (88). As well as a possible involvement in cellular signaling in the kidney and gill, this enzyme may be responsible for the formation of free inositol, a known osmolyte in mammals. When urine is concentrated as a result of the hypertonic conditions within mammalian renal medulla, cells in the thick ascending loop of Henle lose water to the hyperosmotic surroundings and therefore concentrate ions and other solutes trapped within the kidney tubule. Under these conditions myo-inositol enters the tubular epithelial cells along with sodium to offset the changes in osmolality induced by the absorption of water (44). In the mammalian kidney, renal medullary cells use inositol as well as other osmolytes to adjust their intracellular osmolality (and thereby their volume) to rapid and profound changes in the ionic composition of the immediate extracellular environment (56), indicating that this low molecular weight, cyclic alcohol is also acutely involved in cellular osmoregulation (17). Although the teleost kidney does not have a loop of Henle, tubular cells will still have to respond to any temporal changes in the osmolality of the tubular or serosal fluids induced by ion transport. It is possible that inositol may play a role as an intracellular osmolyte within the kidney, especially when fish enter SW.
Growth hormone (GH) mRNA was upregulated by twofold in brain 6 h and 5 mo after eel transfer to SW. This is in agreement with a number of studies in other teleost fish including the catfish (107). It is generally well accepted that GH is a hormone required for SW adaptation in several teleost species from the more primitive Euteleostei such as salmonids to the more evolutionarily advanced Acanthopterygii, including the cichlid fish such as the tilapia (93). The physiological actions of GH are mediated, at least in part, through the actions of insulin-like growth factor I, which, together with cortisol, can increase the size and number of SW-type chloride cells within the branchial epithelium (95). Exogenous treatment by GH/IGF-I has been reported to induce osmoregulatory adaptations in most salmonids, tilapia, and killifish but surprisingly has no apparent effect on several osmoregulatory parameters in the gilthead sea bream (Sparus auratus) (65), indicating that the mechanisms of regulation of ion and water balance may not be equivalent in all teleosts.
Messenger RNA transcripts for prolactin were increased by nearly threefold in the brain 6 h posttransfer, but then downregulated by >10-fold in the same tissue after 5 mo in SW. Prolactin is known to perform a variety of important functions in vertebrates and is known as a FW-adapting hormone in euryhaline fish (38, 106, 122). The decrease in mRNA abundance in the long-term SW-acclimated eel would be consistent with the proposed function of this extracellular signaling molecule, although the increase in mRNA abundance within the first few hours of SW acclimation was a surprising finding and needs further investigation.
Expression of somatolactin, a novel teleost-specific pituitary hormone structurally related to GH and prolactin (71), was decreased by 50% in brain, 5 mo after SW transfer. A number of different physiological functions have been attributed to this signaling molecule (71), including the regulation of skin coloration changes (15).
Genes involved in protein regulation.
Aminopeptidase N (also known as cluster differentiation antigen 13, CD13), is a single-pass transmembrane protein implicated in the hydrolysis and modulation of activity of a variety of peptides and proteins including kinins, neuropeptides, chemotactic mediators, and MHC class II molecules involved in cell adhesion and cell-cell interactions (53). This peptidase, which was upregulated by over twofold in the kidney 5 mo after SW transfer, also plays an important role in the regulation of local concentrations of various signaling peptides such as the growth factors, hormones, and cytokines (42). The protein, which also exists in a soluble, membrane-detached form, has been used as a clinical marker as it is known to be overexpressed in various tumors, having an important role in tumor angiogenesis (117) and in lymphocytes and neutrophils in several inflammatory diseases (10). Aminopeptidase N is likely to play an important role in the degradation of osmotically active peptides such as ANG II and vasotocin in fish, and high levels of enzyme activity have been detected in trout and sea bream plasma, gill, and kidney (1). The levels of aminopeptidase N activity in the SW-acclimated sea bream tissues were consistently higher than those recorded in the FW-acclimated trout, leading these workers to suggest that differences between species could be related to different mechanisms of osmoregulation in SW- and FW-adapted fish (1).
Ornithine decarboxylase antizyme is a polyamine-induced cellular protein that binds and inhibits ornithine decarboxylase (ODC) and signals its rapid ubiquitin-independent degradation by the 26S proteasome (43). ODC antizyme is one of a number of genes transiently upregulated in the intestine after SW transfer. Highest levels of mRNA abundance were found 2 days after SW transfer where transcript abundance nearly doubled. ODC is the rate-limiting enzyme in the synthesis of polyamines, which are ubiquitous aliphatic polycations that have regulatory roles in many fundamental cellular processes such as growth, differentiation, transformation, apoptosis, and cell osmoregulation (43). A role for polyamines in the regulation of several classes of cation channels has also been reported (55). When the antizyme binds to ODC, enzymatically inactive heterodimers are formed and the ODC is targeted for degradation. As well as suppressing the synthesis of polyamines through reduction in ODC activity, the antizyme has also been implicated in the suppression of polyamine transport into the cell by unknown mechanisms (75). Osmotic stress in mammalian cells is known to be associated with antizyme expression and function with increases in extracellular tonicity stimulating cellular abundance and activity (76). Mammalian cells contain another regulatory protein, antizyme inhibitor, a protein with strong homology to ODC but lacking the decarboxylase activity (80). It binds to antizyme with a higher affinity than ODC, therefore releasing active ODC from the inactive antizyme-ODC heterodimer (85). Recently, two ODC antizyme isoforms and an antizyme inhibitor protein have been cloned and characterized in the zebrafish (48), suggesting that similar feedback mechanisms for polyamine metabolism might operate in both mammals and fish.
Cathepsin B, which was downregulated by over twofold in the kidney 5 mo after transfer of eels to SW, is a hydrolytic cysteine protease that is a member of the lysosomal proteolytic apparatus (45). In combination with other cysteine (cathepsins C, H, K, L, and others), serine (cathepsin A) and aspartic (cathepsin D) proteases, and tripeptidyl-peptidase I, it is responsible for the efficient digestion of a diverse range of proteins transported into the lysosomes, releasing free amino acids and dipeptides that are transported back to the cytoplasm and reused according to the metabolic needs of the cell (45). Potential roles for this protease in osmoregulation are currently unknown; however, the abundance of cathepsin B, which has been implicated in the processing of prorenin to renin in the kidney, decreases by 50% in renal juxtaglomerular cells when rats are fed on a high-salt diet (113).
Chitotriosidase (chitinase) expression in the intestine (Fig. 2) was downregulated by twofold in the intestine of 2 and 7 days and 2.5-fold in long-term SW-transferred fish. Chitinase is characterized as an enzyme involved in the degradation of chitin-containing pathogens with unclear function in mammals and humans, although it has been associated with lipid metabolism, being proposed as a marker of lipid accumulation in macrophages in different lipid-storage diseases, including atherosclerosis (16). Three chitinase isozymes, HoChiA, HoChiB, and HoChiC, were recently purified by Matsumiya et al. (70) from the stomach of the greenling, Hexagrammos otakii. Their results demonstrated that fish and insect chitinases possess unique substrate specificities that are correlated with their proposed physiological roles in the digestion of food or cuticle.
Renal ACE (peptidyl dipeptidase A) mRNA was increased by over twofold, 5 mo after transfer of fish to SW. Although well known as a key enzyme involved in the renin-angiotensin system (RAS) catalyzing the conversion of angiotensin I (AI) to angiotensin II (AII) and therefore associated with the regulation of blood pressure and blood volume, this is not the only function of ACE. This enzyme is also involved in the regulation of a range of other physiological processes including control of cell proliferation (36) and drinking and sodium intake (120). In a wide range of species, circulating levels of the dipsogenic AII (105, 120) have been found to be much higher in SW- than FW-acclimated fish (91, 112), implicating ACE and the RAS as essential requirements for osmoregulation.
Secretogranin 5 (neuroendocrine 7B2 protein), was upregulated by over twofold in the kidney of long-term SW-acclimated eels. The gene product is a developmentally and hormonally regulated acidic protein residing in secretory granules of neuroendocrine cells, where it functions as a specific chaperone of prohormone convertase-2, allowing trafficking from the endoplasmic reticulum and activation of the zymogen in vesicles of the secretory pathway (72, 79). A role for this protein within the renal kidney awaits further study.
One of the genes that exhibited the highest FW/SW expression ratio on the arrays (10- to 11-fold change) and subsequently in the Northern blots (Fig. 3D) was found in the gill and had a putative 41% amino acid identity with rat ZG-16. ZG-16 was initially localized to the zymogen granule component of the fractionated rat pancreas where it is thought to have a functional role in the in the regulation of granule secretion within the acinar cells (23). The putative role(s) of this ZG-16-like protein in osmoregulation in the eel branchial epithelium is presently unknown.
Genes involved in transport functions.
Apolipoproteins, synthesized mainly in liver and intestine, play important roles in the transport and cellular uptake of a variety of lipids, phospholipids, and cholesterol throughout the circulatory system (125). The abundance of the teleost-specific 14-kDa apolipoprotein mRNA (which exhibits highest homology with mammalian apoA-II-like protein) was suppressed by >50% in the intestine of 2 and 7 days and 5 mo after transfer of eels to SW (Fig. 2). This novel plasma apolipoprotein appears to be specific to fish (58) and has been cloned from the eel (57). The related eel apolipoprotein-I (Apo-I) exhibited similar levels of downregulation in both kidney and brain of 5-mo SW-acclimated fish. Apo-I and Apo-II are the major protein components of high-density lipoproteins found in the higher vertebrates (62), and equivalent proteins are present in high concentrations in fish plasma (7). The lipid content of teleost plasma is 3–12 times higher than mammals, and as a consequence plasma apolipoprotein levels are generally three- to fourfold higher in fish (7). In addition recent studies have suggested that certain apolipoproteins have antiviral and antibacterial actions (21, 104), implicating these lipid transporting proteins as part of the intrinsic defense systems in fish.
The di- and tripeptide transporter PepT-1 was downregulated 1.5-fold in the intestine of 2 days and 7 days and by up to 70% in the intestine of long-term SW-acclimated eels (Fig. 2). The PepT-1 gene product has been immunolocalized to the villous brush border of enterocytes from all segments of the rat small intestine (100) and is the proton-coupled transporter for a variety of small peptides and drugs in both the intestine and the kidney (40, 108). In addition to functions within the intestine PepT-1 is important for the uptake of peptides in numerous tissues including the renal and lung epithelial and in glial cells within the central nervous system (40). The zebrafish PepT-1 has recently been cloned and characterized (118). In contrast to homologues cloned from other higher vertebrate groups, the maximal transport activities of the functional zebrafish PepT-1 transporter were found to be markedly increased, by as much as fourfold, at alkaline extracellular pH, conditions that would certainly be present within the intestinal lumen in the SW-drinking marine teleost (119). Therefore, the lowered mRNA abundance and any consequential reductions in translation of this protein may be compensated by higher activities of the expressed transporter in vivo.
Microarray data showed that the expression of the absorptive form of the NKCC2 increased (Fig. 2) twofold in the intestine of 2-day, threefold in the intestine of 7-day and three- to sixfold in the intestine of long-term (5 mo acclimated) SW-transferred fish. The microarray results were well correlated with Northern blot data (Fig. 3A) and our previous results with this isoform (29; Cutler CP, Cramb G, unpublished observations). The capacity for ion absorption (especially for Cl–) and solute-linked water flux within the intestine of the eel have previously been reported to increase three- to fourfold following SW acclimation (4, 50, 102, 103), which is also in agreement with the upregulation of cotransporter expression.
Expression of aquaporin 1 (AQP1) in the intestine of 2 and 7 day SW-acclimated eels (Fig. 2) was increased by over twofold, and expression remained elevated (by threefold) in the long-term SW-acclimated fish group. This is consistent with our previous studies, which determined that the expression of this aquaporin isoform is mainly localized to the endothelium of blood vessels and the apical brush border membrane of intestinal enterocytes and particularly in the posterior regions of the gut (69). These results suggest that AQP1 plays a major role in water absorption in the posterior regions of the SW-acclimated eel intestine.
Na-K-ATPase 1 subunit mRNA expression was elevated twofold in the intestine of 2 and 7 day and twofold in both the gill and intestine of 5-mo SW-acclimated fish (Fig. 2), which is consistent with our previous studies on yellow eels acclimating to SW (26). The basic function of the Na+-K+-ATPase (NKA), or sodium pump, is to maintain the high Na+ and K+ gradients across the plasma membrane of all animal cells (110), which in turn are used by other ion channels and transporters to produce the resting membrane potential and to allow absorption or secretion of other ions and solutes by secondary active transport. It has been reported that branchial NKA activity is generally higher in most SW-residing teleosts compared with FW teleosts [reviewed by Sakamoto et al. (94) and Marshall (67)] with most euryhaline species showing increases in activity when fish move from FW to SW environments (61).
Genes encoding structural proteins.
Claudin 27, a member of a large family of claudin genes that encode proteins important in tight junction formation and function, was found to be downregulated by >50% in the gills 5 mo after fish were transferred to SW. As claudins are implicated in the structure of tight junctions in various epithelia, the function of these proteins is associated with epithelial resistance and the paracellular movement of both monovalent and divalent cations (3, 8, 78). The downregulation of this claudin isoform correlates well with the many previous reports indicating there is a decrease in branchial epithelium resistance due to the appearance of "leaky" tight junctions when fish enter SW (68, 97). Interestingly, recent reports have indicated that claudin gene expression is frequently altered in several human cancers (49).
Barrier to autointegration factor 1 (BAF-1) mRNA was also found to be downregulated by >60% in the gill. BAF-1 is a human DNA- and nuclear filament-binding protein thought to be involved with the organization of chromatin structure and gene expression that is essential for cell proliferation and embryogenesis (11, 124). The protein was first described by Lee and Craigie (60) as a host cell protein that is required for the assembly of the preintegration complex that is essential for the integration of viral DNAs, such as human immunodeficiency virus type 1, into the host genome (66). The role of this protein within the teleost gill and the reasons for the decrease in expression in SW remain unknown.
Genes regulating transcription and translation.
Following transfer to SW, ribosomal proteins L28 and L18a were downregulated by two-fold in both the intestine and kidney and ribosomal proteins L27, L4, and L26 were similarly downregulated in gill. In contrast, the expression of the 5S ribosomal RNA gene was upregulated by threefold in the kidney of 5-mo, SW-acclimated eels. The changes in expression of ribosomal proteins may relate to a general suppression of cellular transcriptional/translational machinery as many genes appear to be downregulated in these tissues after eel transfer to SW.
The microarray data were consistent with a number of previous studies that implicated specific genes as being involved in cellular and/or whole animal osmoregulation such as myo-inositol monophosphatase 1, GH, prolactin, ACE, NKCC2, AQP1, and the Na-K-ATPase (alpha subunit). The microarray results for myo-inositol monophosphatase 1 and NKCC2 were confirmed by Northern blot analyses. The salinity-induced changes in expression of a number of other genes including C-type lectin, myeloid protein 1, and the ZG-16 homolog were unexpected, and the physiological roles of these gene products and their relationship to osmoregulation in the eel remain to be elucidated. Many of the genes identified as downregulated following SW acclimation were found in the intestine. It is well known that following metamorphosis to the silver migratory stage, eels stop feeding in FW and the role of the intestine is limited to an osmoregulatory organ during their spawning migration to the Sargasso Sea (87). Although changes in gastrointestinal (GI) tract structure and function are usually associated with the silvering process, our results suggest that the subsequent SW transfer further inhibits the expression of many genes known to be involved with digestion and food absorption within the GI tract, and this was particularly noticeable in eels acclimated to SW for 5 mo. Genes associated with digestion and also downregulated include PepT-1 (101), cathepsin B (24), chitotriosidase (chitinase) (47, 113), and Pi-class GST (81). The microarray analysis also revealed a list of genes with as yet unknown identities. Creation of full-length DNA collections of these genes using 5'- and/or 3'-rapid amplification of cDNA ends (RACE) will enable these genes to be identified and characterized. Such investigations will allow a better understanding and give us a much clearer picture of all physiological adaptations associated with the salinity stress response in the European eel.
In conclusion, the microarray approach allowed the identification of a large number of genes that exhibit differential expression at certain time points during the acclimation of silver FW eels to SW. The use of this methodological approach, in relation to salinity adaptation in the eel, was validated by simultaneously conducting a number of Northern blots with cDNA probes for a subset of the identified up- and downregulated genes. The consistency between the microarray and Northern blot analyses along with the confirmation of expected changes in gene expression of previously characterized osmoregulatory-sensitive genes indicates that this approach was a valid and sensitive method for screening changes in gene expression in the European eel. Using this approach we have identified a number of novel and as-yet uncharacterized genes that appear to be associated with salinity acclimation. Due to limited sequence information for some of the clones these genes and their role in osmoregulation remains to be defined. Expanding the use of DNA microarrays to nonmodel species that exhibit specialized adaptive properties is important when studying certain physiological functions and will be valuable for the identification of the genes associated with these processes (121). The discovery that a large number of unrelated genes, with a wide range of functions, alter expression following the movement of fish to SW greatly increases our current knowledge base on eel salinity adaptation and will permit the development of new experimental hypothesis regarding the molecular, cellular, and physiological bases of osmoregulation and migration in the European eel and other euryhaline teleosts.
|
GRANTS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONGRANTSREFERENCES |
|---|
|
ACKNOWLEDGMENTS |
|---|
|
FOOTNOTES |
|---|
Article published online before print. See web site for date of publication (http://physiolgenomics.physiology.org).
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONGRANTSREFERENCES |
|---|
1 subunit in the European eel (Anguilla anguilla). Comp Biochem Physiol 111B: 567–573, 1995.[CrossRef][Medline]. Am J Physiol Gastrointest Liver Physiol 291: G851–G856, 2006.This article has been cited by other articles:
|
|
 |
|
  C. K. Tipsmark, P. Kiilerich, T. O. Nilsen, L. O. E. Ebbesson, S. O. Stefansson, and S. S. Madsen Branchial expression patterns of claudin isoforms in Atlantic salmon during seawater acclimation and smoltification Am J Physiol Regulatory Integrative Comp Physiol, May 1, 2008; 294(5): R1563 - R1574. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C. K. Tipsmark, D. A. Baltzegar, O. Ozden, B. J. Grubb, and R. J. Borski Salinity regulates claudin mRNA and protein expression in the teleost gill Am J Physiol Regulatory Integrative Comp Physiol, March 1, 2008; 294(3): R1004 - R1014. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Visit Other APS Journals Online |
fold change 