Abstract

   In aquaculture, recurrence rates of amoebic gill disease (AGD) caused
   by the ectoparasite Paramoeba perurans are high and no prophylactic
   strategies exist for disease prevention. In this study, Atlantic salmon
   (Salmo salar) were initially inoculated with P. perurans and following
   the development of amoebic gill disease were treated with freshwater
   immersion on day 21 and day 35 post inoculation. Fish were
   re-inoculated following a negative qPCR analysis for the presence of P.
   perurans. The gill host immune response was investigated at 7, 14, and
   18 days post re-inoculation. Differential proteome expression of immune
   related proteins was assessed by comparison of each time point against
   naïve controls. In the gill, some proteins of the innate immune system
   were expressed in response to gill re-colonization by P. perurans,
   while no features of adaptive immunity were found to be differentially
   expressed. Many of the proteins identified are novel in the context of
   AGD and their expression profiles suggest that their roles in the
   response to disease development and progression in single or multiple
   infections warrant further investigation.

   Keywords: amoebic gill disease, immune response, gill proteome, repeat
   inoculation

1. Introduction

   First reported in 1986 [[32]1], the incidence of amoebic gill disease
   (AGD), an ectoparasitic condition caused by the protozoan parasite
   Paramoeba perurans, has escalated to become one of the most significant
   health threats in Atlantic salmon (Salmo salar) farming globally
   [[33]2]. Typical features of AGD include multifocal gill hyperplasia
   along with lamellar fusion, increased mucus production and macroscopic
   lesions on gill filaments [[34]3,[35]4]. Affected fish experience
   respiratory distress, and mortality rates of up to 80% in untreated
   cases have been reported [[36]5]. No prophylactic measures currently
   exist, and treatments are limited to freshwater or hydrogen peroxide
   baths, which remove the amoeba from affected gills and reduce gill
   mucus viscosity [[37]3,[38]6]. While the rate of re-infection has been
   reported to be delayed by hydrogen peroxide treatment when compared to
   fresh-water treatment [[39]7,[40]8], neither treatment prevents
   re-infection and repeated treatments may be required within one life
   cycle. Therefore the economic impact of AGD is not insignificant
   [[41]9].

   The high rate of AGD re-infection reported in aquaculture has prompted
   investigation of the host adaptive response in the case of recurrent
   infection. Many studies investigating gene expression after single
   infection have highlighted a transcriptional downregulation of key
   immune factors, such as interferon induced MHC I and MHC II pathways
   and anti-oxidant enzymes [[42]10,[43]11], while others reported
   upregulation of pro-inflammatory cytokine and immune-regulatory gene
   transcription [[44]12,[45]13,[46]14,[47]15] as a typical response to
   this infection. Those studies which investigated AGD re-infection
   yielded similarly disparate findings. Valdenegro-Vega et al. [[48]16]
   reported no significant differences in antibody gene expression in
   salmon plasma or skin following four successive infections, when
   compared to naïve controls, and only lesion specific upregulation of
   IgM was detected in the gill 31 days after infection exposure.
   Conversely, Pennacchi et al. [[49]17] detected a slight downregulation
   of IgM, IgT, and TCR mRNA in gills. Contradictory reports on the host
   immune response may be explained, at least for gill tissue, by
   differences in cellular composition and by the discrepancies commonly
   observed when comparing data reported from lesioned versus non-lesioned
   tissue [[50]14,[51]15,[52]18,[53]19]. Furthermore, the stage of disease
   progression has also been shown to influence immune marker expression,
   as evidenced by IgM elevation observed in moderately severe lesions,
   being absent in severe lesions [[54]18].

   In an attempt to elucidate the mechanisms underlying the host response
   to repeat infection with P. perurans, this study aimed to analyze the
   gill proteome of salmon experiencing re-infection in a highly
   controlled environment, at multiple timepoints, after freshwater
   treatment following initial AGD diagnosis.

2. Materials and Methods

2.1. Fish Husbandry

   Atlantic salmon reared on a land-based freshwater hatchery in the west
   of Ireland were transferred to a land-based indoor marine recirculating
   facility at the Marine and Freshwater Research Centre (MFRC) at the
   Galway-Mayo Institute of Technology (GMIT) in Galway, Ireland following
   smoltification.

   Salmon smolts (n = 35), average weight 80 g, were maintained in
   circular black 1000 L tank with the following conditions: water
   temperature 12 °C, artificial salinity of 30 parts per thousand (ppt)
   (Coral Pro salt, Red Sea), light cycle of 12 h light, 12 h dark. Fish
   were fed a commercial salmon diet at 1% body weight per day. This
   project was authorized by the Health Products Regulatory Authority
   (HPRA) in Ireland under project authorization number AE19137/P003, in
   compliance with the requirements under Directive 2010/63/EU transposed
   into Irish law by the European Union (Protection of Animals Used for
   Scientific Purposes) Regulations 2012 (S.I. No 543 of 2012 as amended).

2.2. Paramoeba Perurans Culture

   P. perurans were harvested from naturally infected salmon, sourced from
   a marine site in the west of Ireland in September 2017, and were
   cultured on 0.1 malt yeast agar plates in 30 ppt salinity seawater as
   previously described [[55]20]. Cultures underwent 16 passages before
   use in inoculation. The identity of the amoebae was confirmed by
   real-time polymerase chain reaction (PCR), as described in [[56]21], at
   the Marine Institute, Galway, Ireland.

2.3. Paramoeba Perurans Challenge and Sample Collection

   Following an acclimatization period of 7 days, naïve control gill
   samples (T0) were taken for proteomic analyses. Thirty-five fish were
   challenged with P. perurans (500 amoebae/L) in a volume of 300 L for 4
   h with oxygen saturation, and fish behavior and welfare were closely
   monitored. A further 35 control fish were also held at 300 L for 4 h.
   The control and treatment fish were moved into new tanks with
   containing 1000 L seawater. The identity of the amoebae was confirmed
   by real-time polymerase chain reaction (PCR), as before. Three weeks
   after inoculation, freshwater treatment was administered by immersion
   in freshwater tanks for 4 h. Fish were allowed to re-acclimatize and 14
   days after freshwater treatment the presence of P. perurans was
   assessed by diagnostic qPCR. The detection of P. perurans DNA at this
   time deemed a second freshwater treatment necessary. A further 14 days
   after the second freshwater treatment, fish were sampled (T1) and all
   qPCR samples were negative for the presence of P. perurans DNA at this
   time. The remaining fish were re-inoculated with P. perurans (2000
   amoebae/L), and fish welfare was monitored throughout as previously
   described. Further post re-infection samples were taken 7 (T2), 14
   (T3), and 18 (T4) days post re-infection (dpi) ([57]Figure 1). For
   sampling, fish were euthanized by overdose of anesthetic (400 mg L^−1
   tricaine methane sulfonate (MS-222) and when necessary, gill score (GS)
   was assessed in accordance with the standard commercial protocol
   adapted from Taylor et al. [[58]4]. This GS system categorizes gill
   infection across a range of levels from 0, which signifies no visible
   sign of infection, to 5, which indicates extensive areas of visible
   lesions across the gill surface. For gill proteomic analysis, the 2nd
   right gill arch from each fish was excised, the arch cartilage removed,
   and the remaining gill filaments snap frozen in liquid nitrogen and
   kept at −80 °C until required.

Figure 1.

   [59]Figure 1
   [60]Open in a new tab

   Experimental design of trial investigating the host response to repeat
   infection with P. perurans. Fish were acclimatized for 7 days.
   Immediately prior to the initial P. perurans inoculation, naïve control
   fish were sampled (T0). AGD was confirmed by gill score analysis 21
   days after inoculation and fish were treated by freshwater immersion.
   Some features of AGD persisted 14 days after freshwater treatment, and
   the freshwater immersion was repeated. Fish health was confirmed by
   gross gill analysis 14 days after the second freshwater treatment, and
   fish were sampled prior to re-infection with P. perurans (T1). Fish
   were sampled at 7 (T2), 14 (T3) and 18 (T4) days post re-infection.

2.4. Proteomics FASP

   Gills were weighed, and tissue lysis was performed, on ice, by
   sonication (4 × 30 s cycles at 25% amplitude; Fisherbrand™ Q500
   Sonicator, Waltham, MA, USA), in a 50 mM Tris, 1% SDS lysis buffer.
   Protein quantification was carried out using the Pierce BCA assay
   (Thermo Scientific, Waltham, MA, USA) and 300 µg of protein were
   incubated with 0.1 M DTT at 95 °C for 10 mins. Sample clean up and
   digestion steps were carried out using the FASP method [[61]22].
   Briefly, samples were mixed with UA buffer (8 M urea, 100 mM Tris-HCl
   pH 8.9) and contaminants and salts were removed by repeated
   ultrafiltration through 10 kDa Vivacon centrifugal concentrators
   (Sartorius, Göttingen, Germany).). Samples were incubated for 20 min in
   darkness with 100 µL 50 mM iodoacetamide in UA buffer, to facilitate
   blocking of reduced cysteine residues. After further washes with UA and
   50 mM ammonium bicarbonate (ABC), trypsin (1 µg per 100 µg protein;
   Sigma-Aldrich) in 40 µL ABC was added to the samples in the
   concentrator units. Digestion was performed at 37 °C for 4 h. The
   resulting peptide fragments were washed with ABC and digestion was
   halted by the addition of 5 µL 50% acetic acid. The eluates were
   desalted on C18-Stage pipette tips (Sigma-Aldrich, St. Louis, MO, USA)
   [[62]23], dried in a vacuum concentrator (Eppendorf Vacufuge
   concentrator 5301, Sigma-Aldrich) and stored at −20 °C until mass
   spectrometry analysis.

2.5. Mass Spectrometry

   The samples were analyzed by the Mass Spectrometry Resource (MSR) at
   University College Dublin on a Thermo Scientific Q Exactive mass
   spectrometer connected to a Dionex Ultimate 3000 (RSLCnano, Waltham,
   MA, USA) chromatography system. Peptides were separated on C18
   home-made column (C18RP Reposil-Pur, Ammerbuch, Germany, 100 × 0.075 mm
   × 3 μm) over 60 min at a flow rate of 250 nL/min with a linear gradient
   of increasing acetonitrile (ACN) from 1% to 27%. The mass spectrometer
   was operated in data dependent mode; a high resolution (70,000) MS scan
   (300–1600 m/z) was performed to select the twelve most intense ions and
   fragmented using high energy C-trap dissociation for MS/MS analysis.

2.6. Data Processing and Bioinformatics

   Raw data from the Q-Exactive were processed using the MaxQuant
   [[63]24,[64]25] (version 1.6.4.0) incorporating the Andromeda search
   engine [[65]26]. To identify peptides and proteins, MS/MS spectra were
   matched against Uniprot Salmo Salar database (2019_04) containing
   82,390 entries. All searches were performed using the default setting
   of MaxQuant, with trypsin as specified enzyme allowing two missed
   cleavages and a false discovery rate of 1% on the peptide and protein
   level. The database searches were performed with carbamidomethyl (C) as
   fixed modification and acetylation (protein N terminus) and oxidation
   (M) as variable modifications. For the generation of label free
   quantitative (LFQ) ion intensities for protein profiles, signals of
   corresponding peptides in different nano-HPLC MS/MS runs were matched
   by MaxQuant in a maximum time window of 1 min [[66]27]. The mass
   spectrometry proteomics data have been deposited to the ProteomeXchange
   Consortium via the PRIDE [[67]28] partner repository with the dataset
   identifier PXD024096.

2.7. Data Analysis

   Data were processed using the Perseus (v. 1.6.7.0) data analysis suite
   ([68]https://maxquant.net/perseus/ (accessed on 22 April 2020)). Label
   free quantitative (LFQ) ion current intensities were transformed
   (log[2]). Imputation was carried out to replace missing values with
   numbers drawn from a normal distribution. A Student’s t-test was
   performed to determined statistical significance (p < 0.05). For
   visualization using heat maps, data were z-score normalized. Fold
   change values were obtained by calculating the difference between the
   two compared conditions of the mean log[2] transformed values, and the
   Student’s t-test p-value. Pathway enrichment analysis was performed
   using the ClueGo (v2.5.4) [[69]29] and Cluepedia (v1.5.7) [[70]30]
   plugins in Cytoscape (v3.7.1) [[71]31] with the Salmo salar (8030)
   marker set. The gene ontology biological process, consisting of 22,097
   genes, was used [[72]32,[73]33]. GO tree levels (min = 3; max = 8) and
   GO term restriction (min#genes = 3, min% = 4%) were set and terms were
   grouped using a Kappa Score Threshold of 0.4. The classification was
   performed by the two-side hypergeometric statistic test, and the
   probability value was corrected by the Bonferroni method (Adjusted %
   Term p-value < 0.05) [[74]29].

3. Results

3.1. Diagnostic qPCR and Gill Score Analysis

   Gross assessment of gill health two weeks after the initial freshwater
   treatment demonstrated persistent mild to moderate gill pathology in 5
   of the 6 fish and diagnostic qPCR confirmed the presence of amoeba in
   gill samples of 6 fish at this timepoint. The freshwater treatment was
   repeated and after a further 2-week period, qPCR results were negative
   for P. perurans. Seven days post re-infection, only 1 fish of the 6
   sampled exhibited GS1. At 14 dpi 2 of the 6 fish sampled had reached
   GS1 and 3 fish were GS2. By 18 dpi 4 fish were GS2, and the remaining 2
   fish were GS1 ([75]Figure 2).

Figure 2.

   [76]Figure 2
   [77]Open in a new tab

   Distribution of gill scores from fish sampled 7 days (T2), 14 days
   (T3), and 18 days (T4) post re-infection.

3.2. Proteomic Analysis

   In excess of 4000 proteins were initially identified by mass
   spectrometry. Upon filtering based on matching more than one peptide in
   each case and performing Student’s t-test statistics on normalized and
   imputed data, 209 proteins were found to be differentially expressed at
   T1, 225 proteins at T2, 329 proteins at T3, and 288 proteins at T4 when
   compared to naïve control expression (T0). Between 14% and 28% of these
   proteins exhibited a t-test difference of at least +/− 1.5, across the
   4 timepoints (T1, T2, T3, and T4). Heatmap plots of the proteins
   exhibiting t-test differences of at least +/− 1.5 between control (T0)
   and post freshwater treatment samples (T1) and three post re-infection
   timepoints (T2, T3, and T4) are displayed in [78]Figure 3A–D. The
   ClueGO analysis of the key biological processes linked to those same
   proteins is displayed in Supplemental [79]Figure S1A,B. t-test
   difference values in excess of +/− 1.5 across all timepoints for
   proteins involved in immunity are displayed in [80]Table 1.

Figure 3.

   [81]Figure 3

   [82]Figure 3

   [83]Figure 3
   [84]Open in a new tab

   Heatmap plot of the proteins exhibiting t-test differences of at least
   +/− 1.5 in (A) control (T0) and post freshwater treatment samples (T1),
   (B) control (T0) and 7 days post re-infection samples (T2), (C) control
   (T0) and 14 days post re-infection samples (T3) and (D) control (T0)
   and 18 days post re-infection samples (T4) after LFQ values were
   z-score normalized. Hierarchical clustering was performed using
   Euclidian distance and average linkage using the Perseus software.

Table 1.

   Immune-related proteins differentially expressed in the gills of P.
   perurans inoculated salmon after freshwater treatment for initial P.
   perurans infection (T1) and at three timepoints post re-infection (T2,
   T3, and T4). Proteins were identified by mass spectrometry and analysis
   was performed by Max Quant and Perseus computational platforms. For
   each time point protein expressions were compared to naïve controls
   (T0). Reported fold change values were obtained by calculating the
   difference between the mean log[2] transformed values for the two
   compared conditions (p value < 0.05).
   Name Uniprot Id T0 v T1 T0 v T2 T0 v T3 T0 v T4 No. of Peptides
   Alpha-1-antitrypsin homolog A0A1S3SAE4 1.34 1.56 18
   Alpha-1-antitrypsin-like A0A1S3PJ46 1.90 2.35 2.40 2.56 5
   Alpha-2-macroglobulin-like A0A1S3SY27 1.67 1.59 2.08 37
   Alpha-2-macroglobulin-like isoform X2 A0A1S3SY28 2.09 1.94 2.22 19
   Apolipoprotein A-I precursor B5XBH3 1.32 1.66 2.28 2.28 29
   Apolipoprotein A-IV B5X8U6 2.57 3.47 4.09 4.58 15
   Apolipoprotein B-100 isoform X4 A0A1S3M896 3.11 2.62 3.60 61
   Apolipoprotein C-I isoform X1 A0A1S3RYY8 3.23 2.63 4
   Apolipoprotein C-I-like A0A1S3N6L4 1.53 7
   Coagulation factor VII-like A0A1S3MXK4 1.80 1.53 15
   Coagulation factor XIII B chain-like isoform X1 A0A1S3QZV0 1.41 1.59
   1.53 15
   Complement C3 A0A1S3QRL5 1.30 1.58 96
   Complement C3-like A0A1S3L7F1 1.80 1.66 1.99 2.35 35
   Complement C3-like A0A1S3PMF1 1.40 2.01 23
   Complement C4-B-like A0A1S3NRS7 2.85 1.77 9
   Complement C5 A0A1S3PRG9 1.96 1.73 1.93 2.43 41
   Complement factor H-like A0A1S3QR20 1.51 1.89 1.81 2.02 47
   Complement factor I-like A0A1S3QGL8 2.11 8
   E3 ubiquitin/ISG15 ligase TRIM25-like isoform X6 A0A1S3L7B4 2.80 2.52 3
   Fish-egg lectin precursor B5X7U2 1.96 6
   Gtpase IMAP family member 7-like B9ENC9 −2.33 −1.93 4
   Hemagglutinin/amebocyte aggregation factor-like isoform X1 A0A1S3MZH7
   2.38 1.71 4
   Hemopexin A0A1S3MVQ2 1.53 1.43 22
   Hemopexin A0A1S3PQV6 1.73 28
   Histidine-rich glycoprotein-like A0A1S3KK24 1.54 20
   Hyaluronan-binding protein 2 isoform X2 A0A1S3N344 2.12 1.69 1.83 2.18
   13
   Plasma protease C1 inhibitor-like A0A1S3SJI8 1.37 1.55 13
   Plasminogen A0A1S3MAB2 1.32 1.84 19
   Proteasome activator complex subunit 2 B5X6E1 −1.58 10
   Protein AMBP isoform X1 A0A1S3LTS1 1.50 10
   Type-4 ice-structuring protein B5X6Y1 2.08 4
   Type-4 ice-structuring protein LS-12 B5XDA4 3.19 4.76 4.79 5.88 9
   [85]Open in a new tab

4. Discussion

   The gravity of AGD in terms of fish mortality alongside the lack of
   prophylactic measures and the high rate of re-infection means that AGD
   is one of the most significant health and economic threats in salmon
   aquaculture. The frequency of infection recurrence, which may
   necessitate multiple freshwater immersion treatments within one life
   cycle in the industry setting, will contribute to an elevated stress
   response in the fish. The resultant increase in plasma cortisol has
   been shown to suppress immunity, which can contribute to increased fish
   susceptibility and further exacerbate the issue of re-infection
   [[86]34,[87]35]. This work attempts to try and elucidate the response
   of fish re-inoculated P. perurans and to determine whether an adaptive
   response is initiated. While some studies have investigated the gill
   transcriptomic profile after multiple inoculations [[88]16], and the
   gill and skin mucus proteomic profiles after repeated P. perurans
   infection [[89]36], here the gill, as the site of amoebic attachment,
   was chosen as an appropriate tissue for whole proteome analysis. As
   demonstrated, several proteins with roles in the immune response were
   found to demonstrate differential expression, when compared to T0 naïve
   control fish. Some of these proteins were previously described in the
   proteomic analysis of gill mucus, while others have not been discussed
   previously in the context of AGD.

   The range of complement factors observed to be differentially expressed
   in the gill upon re-infection reflects the significance of these
   fundamental components in the host immune response. Upregulation of
   complement 3, 4, and 5 persisted across all sampling points of
   re-infection. Complement factor I (CFI) negatively regulates the
   complement system by cleavage of the C3b and C4b complement factors
   [[90]37] while plasma protease C1 inhibitor inhibits the alternative
   complement pathway [[91]38]. Though both were upregulated at some
   timepoints in this study, they were elevated either prior to
   re-inoculation, or at a much later stage of the infection. However,
   complement factor H (CFH), the only complement inhibitory factor to
   remain elevated throughout this study, is particularly interesting in
   the context of AGD. CFH binds to complement C3b and blocks the
   downstream alternative pathway. Crucially, in pathogenic microbe
   studies in humans, pathogen expression of host-like glycans facilitates
   pathogen binding to complement factor H, which in turn stimulates host
   immune suppression as a strategy for microbe survival
   [[92]39,[93]40,[94]41]. In a previous study (data not published), our
   group detected a similar upregulation of CFH expression in gill tissue
   and serum of salmon which had experienced a single infection with P.
   perurans. In that case the upregulation of CFH was apparent in the
   context of a subdued immune response to infection. It is therefore
   interesting to note that in both the previous study, which demonstrated
   a limited immune response to a single P. perurans infection, and this
   study, which demonstrated a more robust immune response to P. perurans
   re-infection, the elevation of CFH remains consistent. This protein and
   its potential role in facilitating microbial survival in AGD must
   therefore be given further consideration.

   In addition to the role of apolipoproteins in lipoprotein structure and
   metabolism, their diverse and context dependent roles in innate
   immunity are becoming increasingly apparent. Apolipoprotein A-I
   (ApoA-I) and Apolipoprotein C-I (ApoC-I) have been implicated in the
   binding and neutralization of bacterial lipopolysaccharide (LPS)
   [[95]42] and the facilitation of LPS induced inflammation in human lung
   [[96]43] respectively. The upregulation of ApoA-I and two isoforms of
   ApoC-I upon P. perurans re-infection in this study may indicate an
   immuno-stimulatory response to gill colonization. However
   apolipoprotein A-IV (ApoA-IV) can negatively impact on leukocyte
   recruitment to the site of infection or injury [[97]44]. Further,
   ApoA-IV and ApoA-I have been identified as immune-inhibitory proteins
   which can impact eosinophil induced inflammation and T-cell signaling
   respectively [[98]45]. The elevation of these proteins suggest that the
   expression of the apolipoproteins much be considered carefully in this
   context. Whether ultimately positively or negatively impacting upon the
   host immune response, the upregulation of ApoA-I was modest prior to
   re-infection and showed a more substantial increase in expression with
   disease progression. The elevation of apoA-IV, however, was detected in
   fish in the T1 cohort, post freshwater treatment, and prior to
   re-infection, which implies that the differential expression of this
   protein must not only be considered in the context of gill
   colonization. Apolipoprotein B-100 (Apo B-100) predominantly functions
   as a lipid binding structural protein. It has however been implicated
   in the limiting of bacterial invasion in Staphylococcus aureus
   infection. Apo B-100 is thought to impact the accessory gene regulator
   quorum sensing system, which facilitates bacterial adaptation to
   environmental changes due to population density alterations [[99]46].
   Staphylococcus aureus colonizes mucosal and epithelial tissue and has
   the capacity to infiltrate further. Several members of the
   Staphylococcaceae family have been identified on the gills of
   AGD-affected fish [[100]47]. It is feasible that the significant and
   persistent upregulation of the apolipoprotein B-100 isoform X4 protein
   upon re-infection may point to a similar defense mechanism operating in
   the case of P. perurans colonization of gill tissue.

   Hyaluronan (HA) is a glycosaminoglycan which is found predominantly in
   the lung and has been shown to have stimulatory or inhibitory effects
   on lung pathology, in a manner which is dependent on molecular weight
   of the hyaluronan and binding of associated proteins. Proinflammatory
   cytokines, such as TNFα, IL-1β, and LPS induce HA production [[101]48]
   and HA has been associated with an increase in Muc5 expression in
   response to oxidative stress in the lung [[102]49]. However,
   Hyaluronan-binding protein 2 isoform X2, which is persistently
   upregulated in the current study, hinders HA mediated endothelial cell
   barrier protection and facilitates LPS induced endothelial barrier
   disruption in murine acute lung injury models [[103]50].

   Alpha-1-antitrypsin (A1AT), a member of the serine protease inhibitor
   family, is a potent neutrophil elastase inhibitor, which acts to
   protect lung tissue from excessive proteolytic damage in human models
   [[104]51]. More recently A1AT has been shown to exert effects on myriad
   other proteases, and is fundamental for maintaining a fine balance
   between sufficient protease activity for host immune response and the
   prevention of uncontrolled tissue injury [[105]52]. Long identified as
   part of the acute phase immune response, A1AT has since been recognized
   as having an important role in identifying appropriate responses to
   legitimate dangers, and avoiding aberrant tissue damage due to
   over-excitation of inflammatory responses [[106]53]. Protein AMBP
   isoform X1 is a precursor of the inter-alpha-trypsin inhibitor
   proteins, α1-microglobulin and bikunin, which, among a diverse range of
   other functions, exhibit similar serine protease inhibitory functions
   to A1AT [[107]54]. However, the elevated expression of one of the
   isoforms of A1AT, and protein AMBP occurs in advance of re-inoculation,
   and cannot therefore only be considered in the context of P. perurans
   infection. Many isoforms of the AMBP gene products have been localized
   to lung tissue [[108]55] and one of the major roles of the
   inter-alpha-trypsin inhibitor proteins is the production of HA
   complexes. As discussed, HA has been implicated in lung pathology, but
   also in wound healing in lung tissue [[109]56]. Given that in this
   study, fish were initially inoculated with P. perurans, and
   subsequently submitted to two cycles of freshwater treatment before
   post-freshwater sampling, the elevation of the protein AMBP isoform X1
   could be interpreted as an indicator of wound healing which would be
   expected in fish experiencing AGD recovery.

   Alpha-2-macroglobulin is also a broad spectrum protease inhibitor
   which, in addition to its effect on host proteases, confers protection
   by deactivating pathogen proteases [[110]57]. Both
   alpha-2-macroglobulin isoforms detected in the gill at various
   timepoint post-re-infection may be indicative of a host immune response
   to local colonization.

   The proteasome activator complex subunit 2, previously found to be
   upregulated in transcriptomic analyses of early stage AGD after single
   P. perurans inoculation [[111]58] and other salmon infections [[112]59]
   has a role in antigen processing by upregulating MHC class 1
   presentation. Here protein expression is undetected prior to 18 dpi, at
   which point proteasome activator complex downregulation is evident.
   Similarly the downregulation of GTPase IMAP family member 7-like, which
   has a role in lymphocyte survival [[113]60], is indicative of a
   suppression of some elements of immunity, however in this case the down
   regulation was evident before re-infection. Conversely, in this study,
   the hemagglutinin/amebocyte aggregation factor which has an established
   role in the mechanism of self/non- self-recognition [[114]61] and E3
   ubiquitin/ISG15 ligase TRIM25-like isoform X6, which has a role in
   innate anti-viral immunity [[115]62] were both found to be upregulated
   at the early and later timepoints respectively.

   Plasminogen is the precursor of plasmin, a broad range serine protease,
   but has numerous other functions which are highly context dependent.
   The minor pre-re-infection elevation of plasminogen expression and the
   more substantial increase at 18 days post re-infection may be
   indicative of inflammation stimulation, or suppression or indeed
   complement inhibition, extracellular matrix degradation to facilitate
   cell migration or leukocyte recruitment to the site of injury. However
   the concurrent upregulation of histidine-rich glycoprotein-like,
   coagulation factor VII-like and coagulation factor XIII B chain-like
   isoform X1, all of which are pro-thrombotic [[116]63], may support the
   role of plasminogen in coagulation in this case. The diversity of
   functions attributed to plasminogen suggest it should be considered
   further in the context of AGD.

   Heme release from ruptured red blood cells has been shown to elicit
   inflammation, through various mechanisms including presentation of LPS
   to immune cells [[117]64], a process which is suppressed by hemopexin
   [[118]65]. The context of this suppression is interpreted as protective
   against tissue damage in response to excessive or prolonged
   inflammation, or toxicity due to heme accumulation [[119]66]. As an
   additional defense mechanism, hemopexin scavenging of free heme
   restricts pathogenic nutrition [[120]67]. In previous salmo salar
   microarray studies hemopexin gene expression was found to be
   downregulated in Piscirickettsia salmonis [[121]68] and Aeromonas
   salmonicida inoculations [[122]69] in earlier stage single infections.
   The upregulation of two isoforms of hemopexin observed in this study
   are at the later stages of re-infection and are likely to be in
   response to prolonged inflammation.

   Type-4 ice-structuring protein LS-12 (LS-12), an antifreeze protein,
   the primary role of which is to lower the freezing point of blood and
   inhibit the formation of ice crystals, is one of the most profoundly
   elevated proteins in freshwater treated gills in this study.
   Interestingly, LS-12 possesses a 20% sequence homology to various
   members of the apolipoprotein family, which have been implicated in
   organogenesis and embryonic morphogenesis in Zebrafish [[123]70].
   Limited information exists on the expression of LS-12 in the context of
   immune response or infection. However, a tick-derived antifreeze
   protein was shown to exert anti-virulence properties in mouse models
   challenged with bacteria [[124]71]. In this study the initial elevation
   in expression detected in the freshwater treated gills at T1 is further
   stimulated by re-infection, with levels continuing to increase with
   disease progression. Whether the activity of the LS-12 protein mirrors
   the functions of those apolipoproteins which share sequence homology,
   or whether its function may be linked to an adaptive host immune
   response remains to be determined. To date this protein has not been
   reported in the context of AGD and warrants further investigation as a
   potential protein of interest in the adaptive host response.

   Lectins have an important role in innate immunity in teleosts. Members
   of the lectin family act as pathogen recognition molecules, which bind
   to the surfaces of pathogens, thus facilitating their recognition and
   clearance [[125]72]. Fish-egg lectin (FEL), a member of the
   rhamnose-binding lectin family is traditionally associated with immune
   defense in early stages of fish development, and is predominantly found
   in oocytes, although some family members have been isolated from fish
   skin mucus and various organs and cells of the immune system [[126]73].
   While in some fish species, FEL are shown to bind to a wide variety of
   both gram positive and gram negative bacteria [[127]74], pertinently
   for this study, FEL are thought to exhibit highly specific binding to
   rhamnose [[128]73]. Rhamnose is a monosaccharide and is one of the
   major constituents of the carbohydrate portion of the high molecular
   weight antigen in virulent P. perurans [[129]75]. The elevation of the
   fish egg lectin precursor protein at the latest timepoint in this study
   is suggestive of a response to the presence of infectious P. perurans,
   and this is the first time FEL has been implicated in the host response
   in AGD.

   Taken together, the data here demonstrate a host immune response to
   repeat inoculation with P. perurans. However, many of the proteins
   identified as exhibiting differential expression in response to
   re-inoculation are part of the innate immune response, and there are
   few indicators of an adaptive immune response in the gill in this
   study. Nonetheless, many of the proteins identified are novel in the
   context of AGD and their expression profiles suggest that their roles
   in the response to disease development and progression in single or
   multiple infections warrant further investigation.

Acknowledgments