Physiol. Genomics Track the topics, authors and articles important to you
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

Physiol. Genomics 26: 15-34, 2006. First published February 21, 2006; doi:10.1152/physiolgenomics.00320.2005
1094-8341/06 $8.00
This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Supplemental Table
Right arrow All Versions of this Article:
26/1/15    most recent
00320.2005v1
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via ISI Web of Science (14)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Morrison, R. N.
Right arrow Articles by Nowak, B. F.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Morrison, R. N.
Right arrow Articles by Nowak, B. F.
Received 22 December 2005; accepted in final form 20 February 2006.
Physiological Genomics 26:15-34 (2006)
1094-8341/06 $8.00 © 2006 American Physiological Society

Transcriptome profiling the gills of amoebic gill disease (AGD)-affected Atlantic salmon (Salmo salar L.): a role for tumor suppressor p53 in AGD pathogenesis?

Richard N. Morrison1, Glenn A. Cooper2, Ben F. Koop2, Matthew L. Rise3, Andrew R. Bridle1, Mark B. Adams1 and Barbara F. Nowak1

1 Aquafin Cooperative Research Centre, School of Aquaculture, Tasmanian Aquaculture and Fisheries Institute, University of Tasmania, Tasmania, Australia
2 Centre for Biomedical Research, University of Victoria, British Columbia, Canada
3 Great Lakes Wisconsin Aquatic Technology and Environmental Research Institute, University of Wisconsin-Milwaukee, Milwaukee, Wisconsin


   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 GRANTS
 REFERENCES
 
Neoparamoeba spp. are amphizoic amoebae with the capacity to colonize the gills of some marine fish, causing AGD. Here, the gill tissue transcriptome response of Atlantic salmon (Salmo salar L.) to AGD is described. Tanks housing Atlantic salmon were inoculated with Neoparamoeba spp. and fish sampled at time points up to 8 days postinoculation (pi.). Gill tissues were taken from AGD-affected fish, and a DNA microarray was used to compare global gene expression against tissues from AGD-unaffected fish. A total of 206 genes, representing 190 unique transcripts, were reproducibly identified as up- or downregulated in response to Neoparamoeba spp. infection. Informative transcripts having GO biological process identifiers were grouped according to function. Although a number of genes were placed into each category, no distinct patterns were observed. One Atlantic salmon cDNA that was upregulated in infected gill relative to noninfected gill at 114 and 189 h pi. showed significant identity with the Xenopus, mouse, and human anterior gradient-2 (AG-2) homologs. Two Atlantic salmon AG-2 mRNA transcripts, designated asAG-2/1 and asAG-2/2, were cloned, sequenced, and shown to be predominantly expressed in the gill, intestine, and brain of a healthy fish. In AGD-affected fish, differential asAG-2 expression was confirmed in samples used for microarray analyses as well as in AGD-affected gill tissue taken from fish in an independent experiment. The asAG-2 upregulation was restricted to AGD lesions relative to unaffected tissue from the same gill arch, while p53 tumor suppressor protein mRNA was concurrently downregulated in AGD lesions. Differential expression of p53-regulated transcripts, proliferating cell nuclear antigen and growth arrest and DNA damage-inducible gene-45ß (GADD45ß) in AGD lesions, suggests a role for p53 in AGD pathogenesis. Thus AGD may represent a novel model for comparative analysis of p53 and p53-regulated pathways.

anterior gradient-2; microarray; salmonid


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 GRANTS
 REFERENCES
 
AMOEBIC GILL DISEASE (AGD) is an ectoparasitic condition of some marine fish (reviewed in Ref. 54) including Atlantic salmon (Salmo salar) (55), rainbow trout (Oncorhynchus mykiss) (55, 56), Chinook salmon (O. tshawytscha), coho salmon (O. kisutch Walbaum) (37), turbot (Scophthalmus maximus L.) (19, 21), and seabass (Dicentrarchus labrax) (24). The infectious agent is Neoparamoeba pemaquidensis (20, 37); however, N. branchiphila have also been isolated from fishes affected by AGD, raising the possibility that AGD is a disease of mixed etiology (22). Therefore, in the interim until the issue of etiology is addressed, AGD is considered to be caused by Neoparamoeba spp. infection.

AGD is amoebic branchialitis with clinical signs that include reduced appetite, lethargy, respiratory distress, loss of equilibrium, and, if untreated, mortality (55, 65). The characteristic gross sign of AGD is focal, white, raised mucoid patches, the result of a pronounced cellular inflammatory response (55). Histological examination of these patches reveals hyperplasia of so-called undifferentiated epithelial cells (3, 37) that can fuse secondary lamellae. Occasionally, fusion of the secondary lamellae entraps amoebae in interlamellar vesicles. As a consequence of lamellar fusion, there is a reduction in the surface area of the respiratory epithelium. Mitochondrion-rich cells (MRCs), also known as chloride cells, are displaced from their typical interlamellar residence and are absent from larger lesions (3). Mucous cell hyperplasia on the gill surface (2, 91) concurrent with excessive mucus production is also observed, as is an infiltration of leucocytes within the central venous sinus adjacent to AGD lesions and lesions themselves (3, 4). Leucocytes also migrate into interlamellar vesicles containing amoebae and presumably destroy the pathogen (4). Some of these infiltrating leucocytes express major histocompatibility complex class IIß (MHC IIß) chain and therefore have peptide antigen presentation capacity (53) but equally may play a role in the ensuing inflammatory response. Inflammation associated with AGD may be mediated by the proinflammatory cytokine IL-1ß, as mRNA transcripts of this molecule have been shown to be upregulated in the gills in response to Neoparamoeba spp. infection of rainbow trout (12) and Atlantic salmon (Bridle AR, Morrison RN, Cupit-Cunningham P, and Nowak BF, unpublished data). Beyond these studies, little is known about the host response to Neoparamoeba spp. infection.

A number of teleost fish-specific microarray chips have been developed for research in nutrition (34), toxicology (38, 40), immunology (13, 14, 41, 46, 47, 62), pathology (27, 48, 63), physiology (30, 35, 39, 67, 78), reproduction (80, 81), and development (45, 67, 76). We have utilized genomics resources developed by the Genomic Research on Atlantic Salmon Project (GRASP) (64) including a cDNA microarray chip (79) to examine the gill transcriptome response to Neoparamoeba spp. infection over time with a view of elucidating the molecular mechanisms of AGD pathogenesis. This study identified 190 different host transcripts that were reproducibly dysregulated in AGD-affected gills relative to AGD-unaffected gills. The functional annotations of these candidate informative genes lead to new hypotheses regarding mechanisms involved in AGD. For example, differential expression of Atlantic salmon orthologs of the Xenopus anterior gradient-2, p53 tumor suppressor antigen, and p53-regulated mRNA transcripts in AGD lesions suggests that the perpetual hyperproliferative response to Neoparamoeba spp. infection may be mediated by downregulation of p53.


   MATERIALS AND METHODS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS
 DISCUSSION
 GRANTS
 REFERENCES
 
Experimental inoculation of Atlantic salmon with Neoparamoeba spp. and sampling of fish.
All procedures involving the use of vertebrate animals during this study were approved by the University of Tasmania Animal Ethics Committee. The committee approves procedures that meet the requirements of the Tasmanian Animal Welfare Act 1993 and the Australian Code of Practice for the Care and Use of Animals for Scientific Purposes 7th edition 2004 (enacted under the Act). Amoebae isolation and inoculation of tanks housing Atlantic salmon with Neoparamoeba spp. were performed as described previously (52). Briefly, seawater-adapted Atlantic salmon (mean 166.2 ± 8.7 g, n = 48 fish/tank) were placed into two autonomous 4,000-liter recirculation systems. Gill-derived amoebae were appraised by light microscopy, immunocytochemistry (11), PCR (85), and fluorescence microscopy (23) to confirm the presence of Neoparamoeba spp. and placed in one of two recirculation systems at 500 cells/l (naive and Neoparamoeba spp. inoculated). Immediately after inoculation of the tank with amoebae, a total of four fish from each system were euthanized (5 ml/l Aqui-S NZ, Lower Hutt, New Zealand), and the first right gill arch was removed. A piece of tissue (~100 mg) was taken from an area adjacent to the gill arch in the medial region, flash frozen in liquid nitrogen, ground, placed into Trizol (Invitrogen, Mount Waverly, Australia), and stored at –80°C until required. The second left gill arch was removed, placed in Seawater Davidson’s fixative, processed for routine histology, sectioned (5 µm), hematoxylin and eosin stained, and mounted. This sampling procedure was repeated at 44, 114, and 189 h postinoculation (pi.).

For verification of differential mRNA expression by quantitative real-time RT-PCR (qRT-PCR), two independent Neoparamoeba spp. inoculations were performed. Briefly, for the first independent Neoparamoeba spp. inoculation, one of two 4,000-liter recirculation systems housing Atlantic salmon (~100 g) was inoculated with Neoparamoeba spp. (450 cells·l–1·day–1 over 3 consecutive days) as described (52). At 0, 7, and 14 days pi., 18 fish from the control (AGD-unaffected) and 18 fish from the AGD-affected tanks were euthanized as described (52), and gill tissue was placed in RNAlater RNA Stabilization Reagent (Qiagen, Clifton Hill, Victoria, Australia) at 4°C overnight and then at –20°C until required. For the second independent Neoparamoeba spp. inoculation, one of two 4,000-liter recirculation systems housing Atlantic salmon (~100 g) was inoculated with Neoparamoeba spp. (500 cells/l) as described previously (52). An independent recirculation system housed AGD-naive control fish. At 38 days pi., fish were euthanized as described above, and gill tissue from four AGD-affected and AGD-naive fish was placed into RNAlater, also as described above.

Reverse transcription and hybridization of cDNA to microarray slides.
Microarray experiments were designed to comply with minimum information about a microarray experiment (MIAME) guidelines (10). Experiments were conducted using the GRASP 16K array version 1 microarray chip (79). To minimize technical variability, hybridizations were carried out using the same print batch of the microarray. Microarray construction and fabrication have been described previously (79). Gill tissue mRNA obtained from three replicate infected and control fish at each of the four sampling points (0, 44, 114, and 189 h) was hybridized. Samples were labeled with different fluorophores, and one of the three replicates at each time point was reversed (dye flip) to compensate for cyanine fluorophore bias.

Gill tissue stored in Trizol was thawed, total RNA was purified according to the manufacturer’s instructions (Invitrogen), 40 U of RNaseOUT recombinant ribonuclease inhibitor (Invitrogen) were added, and samples were frozen at –80°C until required. Before preparation for microarray analysis, total RNA concentration was determined using a spectrophotometer, and RNA integrity was checked by agarose gel electrophoresis. Labeled cDNA was generated using a SuperScript Indirect cDNA Labeling System kit (Invitrogen) according to the manufacturer’s instructions. For each sample, 5.0 µg of total RNA were used to generate modified cDNAs which were then labeled with Cy5 or Cy3 cyanine fluorophores (Amersham Biosciences, Castle Hill, NSW, Australia). All microarrays (60) [Gene Expression Omnibus (GEO) platform accession no. GPL2716] were prepared for hybridization by washing 2 x 5 min with 0.1% SDS, 5 x 1 min with MilliQ water, and 1 x 3 min with MilliQ water at 95°C followed by drying by centrifugation (5 min, 514 g in 50-ml tube). Arrays were prehybridized in 5x SSC, 0.1% SDS, and 0.2% BSA for 1.5 h at 49°C, briefly washed 2 x 20 s in MilliQ water, and dried by centrifugation. Labeled cDNAs were hybridized to prewarmed microarrays in a formamide-based buffer (25% formamide, 4x SSC, 0.5% SDS, 2x Denhardt’s solution, and 4 µl of Genisphere LNA dT blocker) for 16 h at 49°C. The arrays were washed 1 x 10 min at 49°C (2x SSC, 0.1% SDS). All subsequent washes were at room temperature (2 x 5 min in 2x SSC, 0.1% SDS, 2 x 5 min in 1x SSC, and 2 x 5 min in 0.1x SSC) and arrays dried by centrifugation. Images of hybridized arrays were acquired immediately at 10-µm resolution, using ScanArray Express (PerkinElmer, Fremont, CA). The Cy3 and Cy5 cyanine fluorophores were excited at 543 and 633 nm, respectively, at the same laser power (90%) with the photomultiplier tube settings adjusted between slides to balance the Cy5 and Cy3 channels. Fluorescent intensity data were extracted from TIFF images using ImaGene 5.6.1 software (Biodiscovery, El Segundo, CA).

For analysis, elements were deemed present if they were equal to or greater than one standard deviation above the background signal as detected by ImaGene software. Control spots (6 Arabidopsis cDNAs, each in quadruplicate) (79) were used to calculate fluorescent threshold for considering genes present. In addition, data with control values less then the average base/proportional value were excluded. Data transformation (background correction), Lowess normalization (87), and the generation of gene lists based on statistical significance (Student’s t-test, {alpha} = 0.05) were performed using GeneSpring 6.1 software (Silicon Genetics, Agilent Technologies, Palo Alto, CA). Only genes significantly affected by AGD (P < 0.05) on all slides tested at each time point were incorporated into the data set. The raw data set has been deposited into the GEO (http://www.ncbi.nlm.nih.gov/geo, series GSE3857).

Cluster software (http://rana.lbl.gov/EisenSoftware.htm) was used to analyze differentially expressed genes. Before analysis, gene expression data were natural log transformed to create a normal distribution about zero so that inductions were positive and repressions were negative. Clustering was performed using a noncentered hierarchical clustering approach (25), and data were visualized with TreeView software (http://rana.lbl.gov/EisenSoftware.htm). cDNAs represented on the microarray chip were identified by the GRASP consortium as described previously (79). For experiments described here, the identities of differentially expressed genes (P < 0.05) were verified by basic local alignment search tool (tBLASTX) query (6), and the GRASP-designated Gene Ontologies (GOs) were also verified manually by interrogation of the UniProtKB/Swiss-Prot database (http://au.expasy.org/sprot/). Genes were assigned to higher-order GOs (GO Slim) by batch analysis using GO TermMapper software (http://go.princeton.edu/cgi-bin/GOTermMapper). Lists were also annotated manually.

RNA extraction and cDNA synthesis for qRT-PCR.
For microarray data confirmation, two pools of cDNA were created at each sampling point, one from AGD-naive fish (n = 3 fish) and the other from AGD-affected fish (n = 3 fish). Total RNA from the same samples used for microarray analysis was purified as described and then repurified using an RNeasy RNA extraction kit (Qiagen) including an on-column DNase I DNA digestion step. Five hundred nanograms of total RNA from each of the three fish at each time point were pooled and reverse transcribed according to the manufacturer’s instructions (Bioline, Alexandria, NSW, Australia). First-strand cDNA synthesis was performed using 5 µM oligo(dT)18, 2 mM dNTP mix, reaction buffer, 10 U RNAseOUT (Invitrogen), and 50 U of Moloney murine leukemia virus (MMLV) RNase H– RT. Samples were incubated for 60 min at 37°C followed by 10 min at 70°C. For the first of two independent Neoparamoeba spp. inoculations, RNAlater-stabilized gill tissue samples were thawed and total RNA extracted using an RNeasy RNA extraction kit (Qiagen), a Dounce homogenizer (Wheaton Scientific, Millville, NJ), and QIAshredders (Qiagen), following the manufacturer’s instructions. Gill tissue taken from fish in the second of two independent Neoparamoeba spp. inoculations was dissected under a stereomicroscope to isolate either normal tissue or AGD-affected tissue. Normal tissue did not contain AGD lesions according to observations using the stereomicroscope. Total RNA was then isolated, as described above, and 500 ng reverse transcribed according to the manufacturer’s instructions (Invitrogen). Briefly, first-strand cDNA synthesis was performed using 2.5 µM oligo(dT)20, 0.5 mM dNTP mix, reaction buffer, 10 U RNAseOUT, 10 mM DTT, 5 mM MgCl2, and 200 U SuperScript III RT (Invitrogen). Samples were incubated for 50 min at 50°C and 5 min at 85°C, and 2 U of RNase H– were added before incubation at 37°C for 20 min.

qRT-PCR and data analysis.
The relative abundance of mRNA from a select number of genes identified as differentially expressed during AGD by microarray analysis was confirmed by qRT-PCR (Table 1). To circumvent the amplification of genomic DNA during qRT-PCR, nucleotide sequences were obtained from genomic DNA, corresponding to the expressed sequence tags (ESTs) of interest. Briefly, oligonucleotides were designed to amplify ESTs of interest using Atlantic salmon liver genomic DNA (DNeasy kit, Qiagen) as template. PCR products were either directly sequenced or ligated into pGEM-T easy plasmid vector (Promega, Annandale, Australia). After transformation into Escherichia coli (DH10ß), positive clones were identified first by blue-white color selection (BlueTech, Mirador, Montreal, QC, Canada) followed by PCR. Clones were inoculated into Luria broth, and plasmids were purified (QIAprep Spin miniprep kit, Qiagen). PCR amplification of each product for nucleotide sequencing was performed using a DTCS Quick Start Dye terminator kit (Beckman Coulter, Gladesville, Australia), M13 forward or reverse oligonucleotides (plasmids), or gene-specific oligonucleotides (PCR products). Samples were analyzed using a CEQ 8000 sequencer (Beckman Coulter).


View this table:
[in this window]
[in a new window]
  		Table 1. Various genes and corresponding oligonucleotides used for estimation of relative expression by qRT-PCR

 
Apart from one EST of interest (Table 1, accession no. CA056958) that represented a single exon (data not shown), nucleotide sequences obtained from genomic DNA were used to design oligonucleotides that anneal adjacent exons. Initially, the annealing temperature for each oligonucleotide pair was determined empirically using pooled gill cDNA, 2x PCR mastermix (Quantitect SYBR Green PCR kit, Qiagen), gene-specific oligonucleotides (0.3 µM each), and an Eppendorf Mastercycler Gradient Thermal cycler (Eppendorf, North Ryde, NSW, Australia). Subsequently, the amplification efficiency of each gene of interest was assessed by qRT-PCR using five threefold dilutions of cDNA template, gene-specific oligonucleotides, PCR mastermix, and a Real-Time PCR Detection System (MyiQ, Bio-Rad). All samples amplified by qRT-PCR and those described herein were assessed in triplicate. qRT-PCR efficiencies were calculated according to the equation E = 10(–1/S), where S is the slope of a standard curve generated using dilutions of template described above. cDNAs were amplified using the following thermal cycling parameters: 1 cycle of 95°C for 15 min followed by 40 cycles of 94°C for 20 s, 56°C for 20 s, and 72°C for 20 s. PCR reactions were performed in triplicate in 25-µl volumes consisting of 2x PCR mastermix (Quantitect SYBR Green PCR kit, Qiagen), gene-specific oligonucleotides (0.3 µM each), 1 µl of cDNA, and nuclease-free water. All qRT-PCR reactions were subjected to postamplification melt-curve analysis, and initially PCR products from each target were analyzed by gel electrophoresis and nucleotide sequencing as described above to ensure the appropriate target cDNA was amplified. In addition, no-template controls and a subsample of "RT-free cDNA" samples were analyzed. Amplification data were analyzed by a relative method ({Delta}{Delta}CT, where CT is threshold cycle; Ref. 59) after correction for minor discrepancies in PCR amplification efficiency between genes of interest and the endogenous control (ß-actin) using the Relative Expression Software Tool (REST version 2) (60). Relative expression ratios (control vs. treatment) X > 1 were upregulated, while ratios 0 < X < 1 were downregulated. Downregulated ratios were inversely transformed to create "fold" change values consistent with upregulated data.

Nucleotide sequencing of Atlantic salmon anterior gradient-2 ortholog cDNAs by rapid amplification of cDNA ends.
For rapid amplification of cDNA ends (RACE), total RNA was purified from the gill tissue of a healthy individual Atlantic salmon. cDNA was prepared from 2.5 µg of total RNA using a GeneRacer kit (Invitrogen) according to the manufacturer’s instructions. RACE was performed using PCR or nested PCR with proprietary 5'- or 3'-oligonucleotides and gene-specific oligonucleotides (5'-RACE reverse oligonucleotide, 5'-acctctgaagtcgcaggacaatgg-3'; 5'-RACE reverse nested oligonucleotide, 5'-catgttgctcagtaagagtttg-3'; 3'-RACE forward oligonucleotide 5'-gctgtcaacgatacgctacgtaacg-3') designed using an Atlantic salmon anterior gradient-2 (AG-2) mRNA transcript, asAG-2 EST (accession no. CB504403). PCR products were ligated into a plasmid vector and cloned, and nucleotide was sequenced as described above.

Western blotting and immunohistochemical detection of proliferating cell nuclear antigen.
Gill tissue (~50 mg) was harvested from a normal healthy Atlantic salmon, placed in 1 ml of lysis buffer (50 mM Tris·HCl, pH 7.4, 150 mM NaCl, 1% Triton-X 100, and protease inhibitor cocktail; Sigma), and homogenized at 4°C. Lysates were centrifuged, and supernatants were removed, diluted with SDS-PAGE reducing buffer, and boiled for 5 min. Reduced proteins were applied to precast 10–20% Tris-glycine gradient polyacrylamide gels (SoftGel, Mirador, CA) and electrophoresed. Proteins were semi-dry transferred to nitrocellulose (Hybond-C extra, Amersham Biosciences, Castle Hill, NSW, Australia), incubated with casein (Vector laboratories, Burlingame, CA), and then probed with monoclonal antibody (MAb) anti-proliferating cell nuclear antigen (anti-PCNA) (0.4 µg/ml, Clone PC10, IgG2a; Zymed Laboratories, San Francisco, CA). MAb IgG2a against an irrelevant antigen (Vector Laboratories) at an equivalent protein concentration served as a negative control. Alkaline phosphatase-conjugated goat anti-mouse IgG was added (1:10,000, Sigma), and then blots were developed using DuoLux (Vector Laboratories) and exposed to film (BioMax Light film, Kodak). Membranes were washed three times with Tris-buffered saline (50 mM Tris·HCl, 150 mM NaCl, pH 7.4) in between each step. Immunohistochemical detection of PCNA was performed as described previously (2) using the monoclonal anti-PCNA and isotype control described above.


   RESULTS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
DISCUSSION
 GRANTS
 REFERENCES
 
Partial purification of amoebae and confirmation of AGD.
Because no virulent cultured Neoparamoeba spp. are currently available (51), Atlantic salmon affected by AGD were used as a source of virulent amoebae. Therefore, a thorough process was undertaken for each infection of Atlantic salmon to confirm the presence of Neoparamoeba spp. in the inoculum. Light microscopic, fluorescence microscopic, immunocytochemical, and 18S rDNA PCR results suggested that the inoculums contained Neoparamoeba spp. consistent with the causative agent of AGD (Fig. 1). After inoculation of tanks with Neoparamoeba spp., four fish were sampled daily to monitor for clinical signs of AGD. At 6 days pi., the typical gross sign of AGD-like lesions, white raised mucoid patches on the gills, was evident (see Fig. 2B). The number and size of these AGD-like lesions increased thereafter until sampling was terminated at 9 days pi. No histological evidence of AGD lesions was observed from negative control fish, whereas gills from fish in the system inoculated with Neoparamoeba spp. clearly displayed AGD lesions together with amoebae (Fig. 2). Histopathology of AGD-affected gill was consistent with that described previously (3–5, 91). No gross or histological signs of any other disease were observed.


Figure 1
View larger version (65K):
[in this window]
[in a new window]
  		Fig. 1. Amoebae inoculum used to initiate amoebic gill disease (AGD) for microarray analyses contained Neoparamoeba sp. Amoebae trophozoites (T) were assessed by light microscopy (A), where the nucleus (N) and parasome (P) were clearly evident. This observation was confirmed using DAPI staining and fluorescence microscopy (B). Anti-Neoparamoeba spp. antiserum bound to trophozoites [normal serum control (C), rabbit anti-Neoparamoeba spp. (D)], and genomic DNA isolated from amoebae in the inoculum was used to amplify Neoparamoeba spp. 18S rDNA (E). +ve, immune serum; –ve, normal serum control.

 

Figure 2
View larger version (127K):
[in this window]
[in a new window]
  		Fig. 2. Gross pathology and histopathology of gills from AGD-affected Atlantic salmon used in microarray analyses. A: Seawater Davidson’s-fixed gill arch from an AGD-naive fish. B: AGD-like lesions in a Seawater Davidson’s-fixed gill arch at 189 h postinoculation (pi.) with Neoparamoeba spp. C: at 19 h pi., amoebae had colonized the gill, but no hyperplastic epithelial cells were evident. Bar = 50 µm. D: an AGD lesion 44 h pi., showing hyperplastic epithelial cells leading to fusion of secondary lamellae. Edema had caused epithelial cells to lift (L). Bar = 50 µm. E: AGD lesion at 114 h pi., showing fusion of several secondary lamellae. Bar = 100 µm. F: AGD lesion at 114 h pi., showing extensive fusion of secondary lamellae. Bar = 200 µm. Arrows indicate amoebae.

 
Summary of microarray analyses.
At 19 h pi., there were signs of initial attachment and colonization of the gills by Neoparamoeba spp. with only subtle pathology that included hypertrophy of squamous epithelial cells, edema causing epithelial lifting and lamellar synechiae (Fig. 2C). Therefore, apart from fish sampled immediately after inoculation of tanks with Neoparamoeba spp., only tissues harvested beyond 19 h pi. were used in microarray analyses. Three sampling times (44, 114, and 189 h pi.) were chosen that represented distinctly different stages of infection. These stages of infection were characterized by the following: 1) 44 h pi., histological signs of initial epithelial cell hyperplasia and lamellar fusion affecting approximately two lamellae per AGD lesion (Fig. 2D); 2) 114 h pi., histological signs of epithelial cell hyperplasia increasing the number of fused lamellae to approximately four per AGD lesion (Fig. 2E); and 3) 189 h pi., detection of gross signs of AGD commensurate with a significant increase in the number of fused secondary lamellae (~18/lesion, Fig. 2F).

Genes incorporated into the data set were those greater than twofold up- or downregulated in all replicate hybridizations including the reverse fluorophore hybridization within a sampling time point. A total of 206 genes representing 190 unique transcripts were differentially regulated in response to Neoparamoeba spp. infection at all the time points assessed [Table 2, Supplemental Table S1 (the online version of this article contains supplemental data); GEO sample series GSE3857]. Only nine genes were greater than twofold up- or downregulated at more than one time point (Table 3). Two different microarray features with ESTs identified as differentially regulated trout protein-1 (DRTP-1) were found to be greater than twofold induced in AGD-affected samples relative to AGD-unaffected samples at both the 114-h (EST accession no. CB502879) and 189-h (CA046376) time points (Table 2).


View this table:
[in this window]
[in a new window]
  		Table 2. List of annotated transcripts >2-fold up or downregulated in the gills of AGD-affected Atlantic salmon

 

View this table:
[in this window]
[in a new window]
  		Table 3. Genes differentially expressed at more than one time point

 
During microarray construction, genes were selected and identified using BLASTX or BLASTN (6) alignments against the GenBank nonredundant protein or nucleotide databases, respectively, as described by von Schalburg et al. (79). Of 190 unique transcripts described here, 87 had no significant (E >1e–15) hit and were designated unknown. Only 77 of the remaining 103 identified genes had been assigned to at least one of three fundamental GO identifiers (molecular function, biological process, or cell component) (Fig. 3A).


Figure 3
View larger version (31K):
[in this window]
[in a new window]
  		Fig. 3. Summary of differentially expressed genes shows that a minority of sequences that have significant identity with Atlantic salmon transcripts have been designated a molecular function, biological process, or cell component Gene Ontology (GO) category. A: ~40% of genes identified as differentially expressed in microarray analyses have been designated to a molecular function, biological process, and/or cell component GO category. B: differentially expressed genes described here that have significant tBLASTX identity (E <1e–15) with genes assigned to the GO biological process category. Genes were grouped within higher-order GO biological process categories. GO identifications (IDs) are provided with each term.

 
In an attempt to identify biological processes that may be affected by AGD, genes that were differentially regulated were assigned to high-order GO biological process categories (Fig. 3B). The five most populated categories were protein metabolism (GO: 0019538), transport (GO: 0006810), cell growth and maintenance (GO: 0008151), response to biotic stimulus (GO: 0009607), and response to external stimulus (GO: 0009605). Categories hypothesized to be important in the pathogenesis of AGD were further analyzed by cluster analysis; however, disparate expression profiles within each of these categories were observed (Fig. 4).


Figure 4
View larger version (54K):
[in this window]
[in a new window]
  		Fig. 4. Cluster analysis of differentially expressed genes that have significant tBLASTX identity (E < 1e–15) with genes assigned to the GO biological process category. Genes grouped within higher-order GO biological process categories exhibit variable levels of suppression and/or induction.

 
Given that information generated by cluster analysis on transcripts with GO annotation was limited, further analysis was performed on the entire data set, the purpose being to identify trends in expression regardless of the level of gene annotation available. Within this expression cluster profile was a group of upregulated cDNAs containing an Atlantic salmon anterior gradient-2 (asAG-2) ortholog of Xenopus anterior gradient-2 (xAG-2) (Fig. 5). This gene had a clearly defined expression pattern based on multiple spots on the microarray, providing at the very least some degree of internal validation of the microarray results. In fact, on closer inspection of all the Atlantic salmon ESTs similar to the zebrafish (Danio rerio) AG-2 ortholog, it was discovered that two transcripts of asAG-2 were present. These transcripts (CB498161, CB504403, CB990918, and CB503191 and CK990462, CB492865, and CB496677) were hypothesized to represent two asAG-2 isoforms, possibly the retention of two functional genes from a genome duplication event that occurred in the evolution of Atlantic salmon and indeed salmonids in general (7).


Figure 5
View larger version (25K):
[in this window]
[in a new window]
  		Fig. 5. Cluster analysis of differentially expressed genes in the gills of AGD-affected Atlantic salmon. A: all differentially expressed transcripts identified during microarray hybridization experiments were subjected to cluster analysis. Each gene is represented as a single row of boxes. Redundant genes were not removed from the analysis. B: the Atlantic salmon anterior gradient-2 ortholog (asAG-2) is represented multiple times on the GRASP 16K v1.0 chip, and each spot consistently indicated that asAG-2 is upregulated at the 114- and 189-h pi. sampling points.

 
qRT-PCR confirmation of microarray analyses using a subset of transcripts.
For the purposes of confirming microarray results, RNA from each of three fish used for hybridizations was pooled at each time point before reverse transcription to reduce the number of PCR reactions required. Seven genes of interest were selected on the basis of their putative functional role in the immune response but also because they represented genes that were either up- or downregulated in response to AGD. The housekeeping gene ß-actin was used to normalize qRT-PCR data, as ß-actin mRNA is not affected by AGD (Bridle AR, unpublished observations). Of the seven genes selected, six were found to be in agreement with the microarray data in terms of up- or downregulation, although the magnitude of fold difference between AGD-affected and -naive control samples was generally lower (Table 1). Illustrating this point were two transcripts (immunoglobulin L-chain and T cell receptor {alpha}-chain) that were upregulated at the same time point as per the microarray data, and yet in qRT-PCR analyses, they were less than twofold upregulated. The tissue inhibitor of matrix metalloproteinase transcript was found to be significantly downregulated in microarray analyses; however, on closer inspection by qRT-PCR, this transcript was in fact 2.3-fold upregulated in response to AGD. No nonspecific amplification (e.g., primer-dimers) was detected in any of the qRT-PCR assays. In addition, no RT free control templates were amplified within 15 PCR amplification cycles of the reactions using cDNA as template.

Identification of two Atlantic salmon orthologs of xAG-2.
Using one of two putative asAG-2 nucleotide sequences (accession no. CB504403), oligonucleotides were designed for 5'- and 3'-RACE. Serendipitously, both asAG-2 transcripts were cloned and sequenced using the same oligonucleotides. These transcripts were designated asAG-2/1 and asAG-2/2 (GenBank accession nos. DQ288664 and DQ288665, respectively). asAG2/1 is a transcript of 860 bp in length comprising a 77-bp 5'-untranslated region (UTR), an open reading frame (ORF) of 513 bp encoding a 171-amino acid protein, and a 270-bp 3'-UTR (Fig. 6A). Similarly, the asAG-2/2 transcript is 851 bp in length with a 76-bp 5'-UTR, an ORF of 513 bp encoding a 171-amino acid protein, and a 262-bp 3'-UTR (Fig. 6B). The asAG-2/1 and asAG-2/2 ORFs are 98% identical and 99% similar at the amino acid level, suggesting that they may have retained the same function. Both translated asAG-2 sequences have predicted cleavage sites [SignalP 3.0 (9)] between the 22nd and 23rd residues producing mature peptides of 149 amino acids in length, respectively. Both asAG-2 sequences were included in a multiple amino acid sequence alignment of AG-2 sequences from a divergent group of animals, and this shows that there is a high degree of conservation across divergent vertebrates (Fig. 6C).


Figure 6
View larger version (40K):
[in this window]
[in a new window]
  		Fig. 6. Nucleotide and predicted amino acid sequences of asAG-2/1 (A) and asAG-2/2 (B) orthologs. Arrowheads illustrate the predicted signal peptide cleavage sites. Endoplasmic reticulum (ER) retention sequences are underlined, and the putative polyadenylation signal is in bold. C: multiple alignment of anterior gradient-2 (AG-2) amino acid sequences. Atlantic salmon (asAG-2/1 and asAG-2/2), zebrafish (zfAG-2, GenBank accession no. BC093250), Xenopus (xAG-2, GenBank accession no. AF025474), mouse (mAG-2, GenBank accession no. NM_011783), and human (hAG-2, GenBank accession no. NM_006408) nucleotide sequences were translated and then aligned using Clustal X software (74a). Differences between the two asAG-2 sequences are highlighted by boxes. Identical (*) and similar (: or .) amino acid residues are shown beneath the sequences.

 
xAG-2 mRNA expression is restricted to the anterior region of the dorsal ectoderm in developing Xenopus (1). Therefore, the distribution of expression in a normal healthy Atlantic salmon was determined by RT-PCR (Fig. 7A). Like xAG-2 expression, asAG-2 transcripts were detected predominantly in epithelial-rich tissue of the gill and intestine but also in the brain, a pattern of expression consistent with the role of xAG-2 in Xenopus (1). PCR oligonucleotides used for this RT-PCR and microarray data validation were designed to amplify both asAG-2 transcripts, as both were upregulated at the same time points pi. Therefore, it was important that dual PCR amplification of asAG-2 was assessed empirically. Restriction mapping showed that HAEII was unique to the asAG-2/1 while FOKI and NspI were unique to the asAG-2/2 PCR amplicon. When the PCR product(s) was purified of oligonucleotides and then digested with HAEII and/or FOKI, it appeared that two different products were present, although sAG-2/1 was the most abundant transcript (data not shown). These data were consistent with restriction digestion of purified PCR amplicon with HAEII and/or NspI (data not shown). Therefore, it is likely that qPCR amplification is indicative of asAG-2/1 and asAG-2/2 mRNA expression combined.


Figure 7
View larger version (6K):
[in this window]
[in a new window]
  		Fig. 7. asAG-2 transcripts have a restricted distribution of expression. RT-PCR analysis of the distribution of asAG-2 transcripts in a healthy Atlantic salmon. asAG-2 mRNA is predominantly expressed in the intestine, brain, and gill. No. of PCR amplification cycles is shown at right. L, liver; HK, head kidney; Sp, spleen; Int, intestine; Br, brain; G, gill; H, heart; PBL, peripheral blood leucocytes; NTC, no template control.

 
Differential expression of asAG-2 orthologs in AGD-affected tissue: overrepresentation on the microarray chip or meaningful response to infection?
Because there was asAG-2 redundancy on the microarray chip, we sought to clarify whether upregulation of these transcripts was specifically relevant to AGD and not an overrepresentation of a nugatory element. cDNA samples were taken from independent experiments, within the same time pi. as the microarray analyses and beyond, and then assessed by qRT-PCR. Initially, samples were taken from fish at 7 and 14 days pi., and cDNA was prepared from pooled gill tissue total RNA. As shown in Fig. 8A, asAG-2 was upregulated relative to samples taken from AGD-naive fish at both day 7 (2.2-fold, P = 0.414) and day 14 pi. (2.2-fold, P = 0.223); however, these data were not statistically significant. AGD is a (multi)focal condition affecting, at least at a cellular level, discrete areas of the gill in close juxtaposition to parasites. We speculated that the magnitude of asAG-2 upregulation was accurate, but variability in transcript abundance due to multiple lesion loci precluded detection of statistically significant data. In other words the low proportion of AGD-affected tissue relative to "normal tissue" in samples affected the analysis. This was addressed by excising gill tissue from fish at 38 days pi. with Neoparamoeba spp. and preserving in proprietary RNA stabilization reagent. AGD-like lesion integrity was maintained in RNA-preserved tissue (Fig. 8, B and C), and lesions were easily dissected out from unaffected tissue with the aid of a stereomicroscope. From AGD-affected fish, three sections of lesion ~5 mm in length were taken from each arch, as were three equivalent pieces of "lesion-free" tissue. Tissue was also taken from AGD-naive control fish, and samples were subjected to qRT-PCR. asAG-2 was significantly upregulated (2.9-fold, P = 0.001) in AGD lesions relative to tissue from normal healthy fish (Fig. 8D) but not in AGD lesions relative to normal tissue taken from the same gill arch (1.9-fold, P = 0.055) and in normal tissue from AGD-affected fish compared with tissue from healthy fish (1.5-fold, P = 0.251).


Figure 8
View larger version (36K):
[in this window]
[in a new window]
  		Fig. 8. asAG-2 mRNA is upregulated in gill tissue taken from independent Neoparamoeba spp. infection experiments. A: asAG-2 mRNA is upregulated in gill tissue sampled at days 7 and 14 pi. from an independent Neoparamoeba spp. infection experiment. Pooled total RNA from gill tissue taken from 6 fish (n = 3 pools from 6 fish) was reverse transcribed, and asAG-2 expression was analyzed by quantitative PCR (qPCR). asAG-2 was 2.2-fold upregulated at days 7 and 14, relative to samples taken from AGD-naive controls (P > 0.05). B: AGD-affected and -unaffected tissue is clearly visible in RNAlater-preserved gill arches. AGD-affected and -unaffected tissue was dissected from AGD+ or AGD– gill tissue, mRNA was reverse transcribed, and the cDNA was analyzed by qPCR. C: higher magnification of boxed section in B. D: asAG-2 mRNA is upregulated whereas p53 tumor suppressor antigen is significantly downregulated in AGD lesions relative to AGD-unaffected tissue. Further analysis showed that genes encoding p53-regulated proteins, proliferating cell nuclear antigen (PCNA), and growth arrest and DNA damage-inducible ß-protein (GADD45ß) were also differentially regulated; however, murine double minute-2 (MDM2) mRNA was not significantly different from mRNA isolated from unaffected tissue. *Significant difference (P < 0.05).

 
Human AG-2 has been implicated in inhibiting tumor suppressor protein p53-induced transactivation function upon ultraviolet light damage to cells (61). Therefore, p53 expression was assessed in samples dissected as described above. Interestingly, the Atlantic salmon p53 mRNA was significantly downregulated in AGD lesions relative to either normal tissue form the same arch (5.9-fold, P = 0.001) or normal tissue from healthy fish (4.3-fold, P = 0.001), and yet there was no difference in expression between normal tissue from AGD-affected and AGD-unaffected fish (P = 0.146) (Fig. 8D). These data support the notion that variability in transcript abundance may be due to the proportion of lesions relative to normal tissue. p53 is a transcription factor important in tumor suppression (83). We proposed that proliferation of epithelial cells in response to Neoparamoeba spp. infection may be mediated by suppression of p53. Therefore, it was important to confirm p53 suppression by assessing relative expression of transcripts involved in p53 signaling pathways. These transcripts were growth arrest and DNA damage-inducible gene-45ß (GADD45ß), PCNA, and murine double minute-2 (MDM2).

qRT-PCR analysis of GADD45ß, PCNA, and MDM2 expression; members of p53 regulatory pathways; and detection of PCNA in AGD lesions.
By use of the same cDNA that was produced from the excised tissue described above, relative expression of GADD45ß, PCNA, and MDM2 was analyzed. These transcripts were chosen on the basis that Atlantic salmon sequences were available and have been documented to be relevant to either p53 itself or to the effects of p53 in higher vertebrates. For example, human GADD45ß activation can be p53 dependent, while suppression of p53 in turn reduces responsiveness of GADD45ß (89). Similarly, p53 binds the PCNA promoter, inducing expression of PCNA at low concentrations while repressing expression at high concentrations (50). MDM2, on the other hand, is a protein that binds p53, leading to the ubiquitination and eventual degradation of p53 by the 26S proteasome (44). Data presented here show that Atlantic salmon GADD45ß is significantly downregulated in AGD lesions relative to tissue taken from healthy fish (2.1-fold, P = 0.049) (Fig. 8D). Conversely, PCNA was significantly upregulated in AGD lesions (2.9-fold, P = 0.001) relative to tissue from healthy fish; however, there was no difference in expression between samples from AGD lesions and normal tissue from the same gill arch (1.3-fold upregulated, P = 0.408) (Fig. 8D). Relative expression of MDM2 was not significantly different between any of the tissue types tested (P > 0.05). Interestingly, these three transcripts are represented on the microarray but were not identified as differentially expressed in earlier experiments. This provides further support that the proportion of AGD lesions relative to normal tissue may have influenced the detection of transcripts significantly affected by Neoparamoeba spp.

Given that the PCNA gene product is well conserved across the phylogenetic spectrum, a MAb anti-PCNA antibody that binds PCNA from all vertebrate species and insects was used to confirm that PCNA was actively expressed in proliferating cells within AGD lesions. Initially, it was shown that this antibody binds a protein in reduced gill tissue lysate that is consistent with the molecular weight of PCNA in humans (36 kDa) (Fig. 9A). No binding of the isotype control antibody to gill lysates was observed. PCNA-positive cells were identified on and within the basal epithelium and occasionally on the secondary lamellar epithelium (Fig. 9C). In AGD-affected tissue, anti-PCNA antibody bound the majority if not all of the cells associated with AGD lesions (Fig. 9, B, D, and E), whereas no binding was evident using the biotinylated isotype control (Fig. 9F).


Figure 9
View larger version (119K):
[in this window]
[in a new window]
  		Fig. 9. Detection of PCNA in tissue lysates and in AGD-affected tissue. A: monoclonal anti-PCNA binds to a protein in gill tissue lysate consistent in size to PCNA (36 kDa). Gill tissue was homogenized, boiled in reducing buffer containing DTT, and electrophoresed through a polyacrylamide gel. Proteins were transferred to nitrocellulose and probed with monoclonal anti-PCNA antibody. Bound antigen was detected using chemiluminescence. B: cells associated with AGD lesions express PCNA. Bar = 200 µm. C: AGD-unaffected tissue boxed in B. Bar = 50 µm. D: AGD-affected tissue boxed in B, showing lesion-associated cells expressing PCNA. Bar = 50 µm. E: higher magnification of AGD-affected tissue with cells expressing PCNA. Bar = 50 µm. F: AGD-affected tissue probed with an isotype control antibody. Note lack of nuclear staining. Bar = 50 µm. Gills were processed for histology, sectioned, and placed on glass slides. Epitope retrieval was performed before incubation with monoclonal anti-PCNA antibody. Horseradish peroxidase-conjugated streptavidin and diaminobenzidine were used in the detection of bound antibody. Arrowheads indicate amoebae.

 

   DISCUSSION
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
GRANTS
 REFERENCES
 
Microarray analyses revealed that a total of 190 unique transcripts were dysregulated in AGD-affected fish at the four sampling points combined. Transcriptome profiling provides a global perspective on transcriptional variations in response to an event such as disease. However, given the lack of GO annotations for a significant proportion of the 190 differentially expressed transcripts, the data do not provide such understanding. Some transcriptional variants may have provided, at least on an individual basis, an insight into AGD pathogenesis. For example, it is noteworthy that a mannose-binding protein C (MBP-C) ortholog was significantly downregulated at 189 h pi. Parasitic protozoa use complex mechanisms to evade the host innate immune system (66), and downregulation of a lectin such as MBP-C that binds carbohydrate moieties on pathogen membranes activating the classical complement cascade is interpreted as a potential immune evasion strategy on behalf of Neoparamoeba spp. Other transcripts that are important in the innate immune response were also downregulated at 189 h pi. These included LPS-binding protein (LBP), important in transducing cellular signals from LPS (32), and NADPH oxidase cytosolic protein p40phox, a protein involved in the generation of superoxide anions during respiratory burst (26). In contrast, NADPH oxidase cytosolic protein p40phox is upregulated in rainbow trout administered a DNA vaccine against infectious hematopoietic necrosis virus (IHNV) G protein (62). In fact, analysis of the growing number of transcriptome profiling studies in fish during infection (27, 41, 48, 63) or immunostimulation (46) or after vaccination (47, 62) reveals that there are a number of modulated genes in common with those described here. DRTP-1 for example is upregulated in Atlantic salmon head kidney during infection with Piscirickettsia salmonis (63) and in the liver of Atlantic salmon intraperitoneally injected with an Aeromonas salmonicida vaccine (47). DRTP-1 is hypothesized to be an ortholog of CD59 (8) which prevents the formation of membrane attack complexes on host cells after complement activation. Other differentially expressed genes described here in common with those described in similar studies include guanine nucleotide-binding protein, FK506-binding protein-3, peroxiredoxin-1 (46), immunoglobulin-{tau} (62), selenoprotein P, T cell receptor {alpha}-chain, C-type lectin 2-1, retinol dehydrogenase, immunoglobulin M light chain, hemopexin, collagen type-1 {alpha}1, complement C3 (63), collagen type-1 {alpha}1, immunoglobulin M light chain, metalloproteinase inhibitor, selenoprotein P, guanine nucleotide-binding protein, mannose-binding lectin (48), C-type lectin 2-1, HSPC300 (47) immunoglobulin M light chain, C-type lectin 2-1, hemopexin (27), and collagenase (41). Further research using microarray technology may not only build on individual gene commonalities but also processes that are involved in the development of infection and immunity in teleost fishes.

Modulation of AG-2 mRNA, like those genes described above, is not restricted to Neoparamoeba spp. infection of Atlantic salmon. Indeed, AG-2 mRNA is downregulated during Mycobacteriium marinum infection of zebrafish (48) and in the gills of Atlantic salmon administered an Aeromonas salmonicida vaccine (47). AG-2 is a protein that was first described in Xenopus (xAG-2), and in this species it influences ectodermal fate, specifically in the formation of the cement gland, a mucus-secreting organ (1). The human homolog of xAG-2 (hAG-2) is overexpressed in prostate (90), colorectal (75), breast (29, 43, 74, 90), and pancreatic (49) tumors and is also upregulated in the Barrett’s epithelium, a bile acid reflux damage-mediated hyperproliferative response of the human esophageal epithelium (61). Most importantly, hAG-2 inhibits posttranslational modifications of p53, giving hAG-2 biological relevance in a hyperproliferative environment (61). p53 is a tumor suppressor protein with its importance in regulating cell fate highlighted by the fact that it is the most frequently mutated gene found in human cancers (88). Typically, p53 activity is increased in compromised cells, resulting in either cell cycle arrest or apoptosis (44, 71). However, in cells transfected with hAG-2, phosphorylation of serine residues at both the NH2- and COOH-termini of p53 (Ser15 and Ser392) is inhibited after ultraviolet light damage to cells (61). This p53 inhibitory effect by hAG-2 produced a cellular response similar to p53 mutants that lose function and was abrogated after the truncation at the COOH-terminus of hAG-2, which effectively removed an endoplasmic reticulum (ER) retention sequence. Interestingly, the ER retention sequence (KTEL) is conserved in both asAG-2 sequences. Phosphorylation of p53 has a variety of effects that include protein stabilization, activation of p53-mediated downstream responses (86), and induction of an anti-proliferative effect (77). Thus, during AGD, upregulation of asAG-2 in hyperproliferative tissue was seen as a potential link to p53 and p53-regulated pathways. However, before qRT-PCR was attempted, designation of the Atlantic salmon EST as p53 was verified. One of six frame translations of the Atlantic salmon p53 sequence showed 92% identity and 95% similarity with a rainbow trout p53 amino acid sequence over 140 residues (GenBank accession no. AAA49605) (17). Previously, this rainbow trout sequence has been shown to contain the p53 signature (M-C-N-S-S-C-[MV]-G-G-M-N-R-R), and therefore it is likely that the Atlantic salmon EST is indeed p53.

Downregulation of tumor suppressor protein p53 mRNA, like the upregulation of asAG-2, was restricted to AGD-affected tissue. However, to place these observations in context, it is necessary to know whether p53 function is conserved across the phylogenetic spectrum. Not only is p53 found in both vertebrates and invertebrates, but it shows remarkably similar function. p53 has been characterized in species as diverse as rainbow trout (17), Xenopus (82), chicken (Gallus gallus) (73), European flounder (Platichthys flesus) (15), Drosophila melanogaster (33), and Caenorhabditis elegans (70). In Drosophila, for example, the p53 homolog (dp53) shows significant homology to the human p53 protein, particularly in the DNA-binding region, and has the ability to bind the human consensus p53-binding site, activating transcription in vitro (33). Furthermore, C. elegans p53 is required for DNA damage-induced apoptosis, suggesting that p53-mediated transcriptional regulation is part of an ancestral pathway (70). Data described here may be interpreted based on the existing p53 paradigm in higher vertebrates. Therefore, in AGD lesions it might be expected that p53 would be affected at transcriptional and potentially posttranslational levels if asAG-2 function is conserved among metazoans. However, functional analysis of asAG-2 requires further study.

In actively proliferating cells of mammals, p53 accumulates in the nucleus, transactivating a cascade of factors that eventually induces cell cycle arrest or apoptosis (44). Human p53 has the ability to transactivate ~100 genes (44), and yet equally, there are pathways to silence the effects of p53, enabling cells to proliferate as required. If p53 suppression is maintained via loss of function by mutation or at transcriptional or posttranscriptional levels, cells may remain in cycle indefinitely. Therefore, we searched for Atlantic salmon ortholog members of p53 regulatory pathways to confirm our earlier observations. In AGD-affected tissue, PCNA, a p53-induced protein (58), was not only upregulated in terms of mRNA but its gene product was also shown to be extensively expressed in hyperproliferative cells associated with AGD lesions. At low levels, p53 induces PCNA, but at higher concentrations, repression is observed (72). Similarly, p53 transactivates GADD45ß (89), and therefore its downregulation in AGD-affected tissue was predictable. Given that p53 has extensive transactivation functions, there must be mechanisms to control p53 activity. Indeed, p53 is regulated by proteasome-mediated degradation. Degradation is ubiquitin dependent with MDM2 commonly associated with the final ubiquitylation of p53 by virtue of its E3 ligase activity (31). Before final ubiquitylation of p53, a cascade of enzymatic reactions occurs, (44) and on inspection of differentially expressed genes described here, one enzyme in this cascade (ubiquitin-conjugating enzyme E2 C) was upregulated at 189 h pi. However, MDM2 mRNA expression was not significantly affected in AGD lesions. Therefore, in AGD lesions, downregulation of p53 transcription may initiate the hyperproliferative epithelial cell response. Alternatively, the p53 gene product may be degraded by MDM2 in the absence of transcriptional change, degraded by another factor such as Jun kinase (JNK) (44), inhibited posttranslationally by asAG-2 or by an as yet unidentified p53 regulatory product of Neoparamoeba spp. akin to that produced by a number of viral pathogens (16). For example, viral proteins produced by simian virus 40 (SV40) (42), human papillomavirus (HPV) (69, 84), hepatitis B virus (28), adenovirus (18, 36), human cytomegaly virus (68), and Kaposi’s sarcoma-associated herpesvirus (KSHV) (57) target p53 for inactivation. KSHV produces a viral interferon regulatory factor that not only inhibits transcriptional activation but also affects acetylation and phosphorylation of the p53 gene product (57).

In summary, the results described here provide evidence that the hyperproliferative response of Atlantic salmon gill epithelial cells may be mediated by the inhibition of p53. Supporting this notion are data that show the downstream effects of p53 downregulation on p53-regulated transcripts; however, it is not yet known how asAG-2 is upregulated nor how p53 is inhibited.


   GRANTS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 GRANTS
REFERENCES
 
This work formed part of a project of Aquafin Cooperative Research Centre (CRC) and received funds from the Australian Government’s CRCs Program, the Fisheries Research and Development, and other CRC Participants. This project was also proudly sponsored by the International Science Linkages program established under the Australian Government’s innovation statement, "Backing Australia’s Ability."


   ACKNOWLEDGMENTS
 
We thank E. MacDonald, M. Attard, H. Statham, P. Crosbie (all Univ. of Tasmania), and M. Cook (Commonwealth Scientific and Industrial Research Organisation) for laboratory support and the School of Human Life Sciences (Univ. of Tasmania) for access to the molecular laboratory for qPCR and, in particular, D. Kunde and A. Crawford.


   FOOTNOTES
 
Address for reprint requests and other correspondence: R. N. Morrison, Aquafin CRC, School of Aquaculture, Univ. of Tasmania, Tasmanian Aquaculture and Fisheries Institute, Locked Bag 1370, Launceston, Tasmania, Australia 7250 (e-mail: rmorriso{at}utas.edu.au)

Article published online before print. See web site for date of publication (http://physiolgenomics.physiology.org).


   REFERENCES
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 GRANTS
 REFERENCES
 

  1. Aberger F, Weidinger G, Grunz H, and Richter K. Anterior specification of embryonic ectoderm: the role of the Xenopus cement gland-specific gene XAG-2. Mech Dev 72: 115–130, 1998.[CrossRef][ISI][Medline]
  2. Adams MB and Nowak BF. AGD sequential pathology: field re-infection and experimental primary interactions/effect of physical trauma on attachment of Neoparamoeba sp. Third Scientific Conference of the Atlantic Salmon Aquaculture Subprogram. Hobart, Tasmania: CSIRO Marine Laboratories, 2003, p. 15–18.
  3. Adams MB and Nowak BF. Amoebic gill disease: sequential pathology in cultured Atlantic salmon, Salmo salar L. J Fish Dis 26: 601–614, 2003.[CrossRef][ISI][Medline]
  4. Adams MB and Nowak BF. Distribution and structure of lesions in the gills of Atlantic salmon, Salmo salar L., affected with amoebic gill disease. J Fish Dis 24: 535–542, 2001.[CrossRef][ISI]
  5. Adams MB and Nowak BF. Experimental amoebic gill disease of Atlantic salmon, Salmo salar L.: further evidence for the primary pathogenic role of Neoparamoeba sp (Page, 1987). J Fish Dis 27: 105–113, 2004.[CrossRef][ISI][Medline]
  6. Altschul SF, Gish W, Miller W, Myers EW, and Lipman DJ. Basic local alignment search tool. J Mol Biol 215: 403–410, 1990.[CrossRef][ISI][Medline]
  7. Bailey GS, Poulter RT, and Stockwell PA. Gene duplication in tetraploid fish: model for gene silencing at unlinked duplicated loci. Proc Natl Acad Sci USA 75: 5575–5579, 1978.[Abstract/Free Full Text]
  8. Bayne CJ, Gerwick L, Fujiki K, Nakao M, and Yano T. Immune-relevant (including acute phase) genes identified in the livers of rainbow trout, Oncorhynchus mykiss, by means of suppression subtractive hybridization. Dev Comp Immunol 25: 205–217, 2001.[CrossRef][ISI][Medline]
  9. Bendtsen JD, Nielsen H, von Heijne G, and Brunak S. Improved prediction of signal peptides: SignalP 3.0. J Mol Biol 340: 783–795, 2004.[CrossRef][ISI][Medline]
  10. Brazma A, Hingamp P, Quackenbush J, Sherlock G, Spellman P, Stoeckert C, Aach J, Ansorge W, Ball CA, Causton HC, Gaasterland T, Glenisson P, Holstege FCP, Kim IF, Markowitz V, Matese JC, Parkinson H, Robinson A, Sarkans U, Schulze-Kremer S, Stewart J, Taylor R, Vilo J, and Vingron M. Minimum information about a microarray experiment (MIAME)—toward standards for microarray data. Nat Genet 29: 365–371, 2001.[CrossRef][ISI][Medline]
  11. Bridle AR, Butler R, and Nowak BF. Immunostimulatory CpG oligodeoxynucleotides increase resistance against amoebic gill disease in Atlantic salmon, Salmo salar L. J Fish Dis 26: 367–371, 2003.[CrossRef][ISI][Medline]
  12. Bridle AR, Morrison RN, and Nowak BF. The expression of immune regulatory genes in rainbow trout, Oncorhynchus mykiss, during amoebic gill disease (AGD). Fish Shellfish Immunol 20: 346–364, 2006.[CrossRef][ISI][Medline]
  13. Byon J, Ohira T, Hirono I, and Aoki T. Comparative immune responses in Japanese flounder, Paralichthys olivaceus, after vaccination with viral hemorrhagic septicemia virus (VHSV) recombinant glycoprotein and DNA vaccine using a microarray analysis. Vaccine 24: 921–930, 2006.[CrossRef][ISI][Medline]
  14. Byon JY, Ohira T, Hirono I, and Aoki T. Use of a cDNA microarray to study immunity against viral hemorrhagic septicemia (VHS) in Japanese flounder (Paralichthys olivaceus) following DNA vaccination. Fish Shellfish Immun 18: 135–147, 2005.
  15. Cachot J, Flaman JM, Frebourg T, and Leboulenger F. The European flounder (Platichthys flesus) TP53 functions as a temperature-sensitive transcription factor which inhibits cell growth in yeast. Gene 324: 97–104, 2004.[CrossRef][ISI][Medline]
  16. Collot-Teixeira S, Bass J, Denis F, and Ranger-Rogez S. Human tumor suppressor p53 and DNA viruses. Rev Med Virol 14: 301–319, 2004.[CrossRef][ISI][Medline]
  17. Defromentel CC, Pakdel F, Chapus A, Baney C, May P, and Soussi T. Rainbow trout p53—cDNA cloning and biochemical characterization. Gene 112: 241–245, 1992.[CrossRef][ISI][Medline]
  18. Dobner T, Horikoshi N, Rubenwolf S, and Shenk T. Blockage by adenovirus E4orf6 of transcriptional activation by the p53 tumor suppressor. Science 272: 1470–1473, 1996.[Abstract]
  19. Dykova I, Figueras A, and Novoa B. Epizoic amoebae from the gills of turbot Scophthalmus maximus. Dis Aquat Organ 38: 33–38, 1999.[ISI][Medline]
  20. Dykova I, Figueras A, Novoa B, and Casal JF. Paramoeba sp., an agent of amoebic gill disease of turbot Scopthalmus maximus. Dis Aquat Organ 33: 137–141, 1998.[Medline]
  21. Dykova I, Figueras A, and Novoa B. Amoebic gill infection of turbot, Scopthalmus maximus. Folia Parasitol (Praha) 42: 91–96, 1995.[Medline]
  22. Dykova I, Nowak BF, Crosbie PB, Fiala I, Peckova H, Adams MB, Machackova B, and Dvorakova H. Neoparamoeba branchiphila n. sp., and related species of the genus Neoparamoeba Page, 1987: morphological and molecular characterization of selected strains. J Fish Dis 28: 49–64, 2005.[CrossRef][ISI][Medline]
  23. Dykova I, Fiala I, Lom J, and Lukes J. Perkinsiella amoebae-like endosymbionts of Neoparamoeba spp., relatives of the kinetoplastid Ichthyobodo. Eur J Protistol 39: 37–52, 2003.
  24. Dykova I, Figueras A, and Peric Z. Neoparamoeba Page, 1987: light and electron microscopic observations on six strains of different origin. Dis Aquat Organ 43: 217–223, 2000.[ISI][Medline]
  25. Eisen MB, Spellman PT, Brown PO, and Botstein D. Cluster analysis and display of genome-wide expression patterns. Proc Natl Acad Sci USA 95: 14863–14868, 1998.[Abstract/Free Full Text]
  26. El-Benna J, Dang PMC, Gougerot-Pocidalo MA, and Elbim C. Phagocyte NADPH oxidase: a multicomponent enzyme essential for host defenses. Arch Immunol Ther Exp (Warsz) 53: 199–206, 2005.[Medline]
  27. Ewart KV, Belange