Immune- and enzyme histochemical characterisation of leukocyte populations within lymphoid and mucosal tissues of Atlantic halibut (Hippoglossus hippoglossus)

Fish & Shellfish Immunology 20 (2006) 693e708 www.elsevier.com/locate/fsi

Immune- and enzyme histochemical characterisation of leukocyte populations within lymphoid and mucosal tissues of Atlantic halibut (Hippoglossus hippoglossus) Søren Grove a,*, Renate Johansen a, Liv Jorun Reitan a, Charles McL. Press b a

b

Section of Immunoprophylaxis, National Veterinary Institute, P.O. Box 8156 Dep. 0033 Oslo, Norway Department of Basic Sciences and Aquatic Medicine, Norwegian School of Veterinary Science, P.O. Box 8146 Dep. 0033 Oslo, Norway Received 6 June 2005; revised 21 July 2005; accepted 26 August 2005 Available online 18 October 2005

Abstract Leukocyte populations within the kidney, spleen, posterior intestine and gills of Atlantic halibut were investigated using a panel of histological, enzyme- and immunohistochemical methods. In the kidney and spleen, a diverse population of leukocytes was associated with the extensive network of sinusoids and larger blood vessels present in these tissues. IgMC cells (B-cells, plasma cells and IgM-bearing macrophages) and large mononuclear cells showing reactivity for non-specific esterase (NSE) and acid phosphatase (ACP), representing macrophage populations, were often associated with vessel walls that were also the site of trapping of fluorescent microspheres. In the kidney, trapping of 0.1 and 0.5 mm diameter microspheres occurred at these sites but in the spleen, the 0.1 mm diameter microspheres were retained in ellipsoids. The lymphoid tissues of the kidney and spleen possessed a spread population of 5#-nucleotidaseC (5#NC) cells but compartmentalisation of the splenic white pulp was suggested by an absence of these 5#NC reticular cells in areas associated with melanomacrophage accumulations and in areas rich in IgMC cells. A striking feature of the mucosal tissues was the diversity of leukocyte populations within the epithelium particularly of the posterior intestine, including IgMC cells and NSEC, ACPC and 5#NC mononuclear cells. Although limited in numbers in the posterior intestine, IgMC cells were more common in the epithelium than in the lamina propria. In the gills, leukocytes as detected by enzymatic reactivity were scarce, but IgMC cells were very abundant in the stratified epithelium of the gill arch and filaments. The difference in distribution of these leukocyte populations between the intestines and gills suggested a compartmentalisation of the mucosal immune system and the need to assess the immunological competence of mucosal tissues in Atlantic halibut. Ó 2005 Elsevier Ltd. All rights reserved. Keywords: Atlantic halibut; Leukocytes; Lymphoid; Mucosal; Immunoglobulin-positive cells; Immunohistochemistry; Enzyme histochemistry

1. Introduction The introduction of Atlantic halibut (Hippoglossus hippoglossus) to aquaculture in temperate climates has required the formulation of strategies to control bacterial and viral disease that are likely to be encountered during intensive cultivation. A number of studies have focussed on the investigation of immunological parameters and particularly * Corresponding author. Tel.: C47 23 21 63 36; fax: C47 23 21 63 02. E-mail address: [email protected] (S. Grove). 1050-4648/$ - see front matter Ó 2005 Elsevier Ltd. All rights reserved. doi:10.1016/j.fsi.2005.08.009

694

S. Grove et al. / Fish & Shellfish Immunology 20 (2006) 693e708

their behaviour following vaccination or challenge with an infectious agent [1,2]. However, the characterisation of leukocyte populations within lymphoid and mucosal tissues of halibut has not received similar attention. The structural organization of the immune system within tissues provides information directly relevant to the uptake of antigens and to the likely sites of initiation and elaboration of an immune response. The lymphoid tissues of the kidney and spleen make important contributions to the generation of systemic immunity in teleosts and perform the vital function of removal of foreign or antigenic substances from the blood [3,4]. The mucosal surfaces of the gills and the gut are of equal significance as these are sites of entry of pathogens in teleost species [5,6] and the contribution of these mucosal tissues to immunity is being increasingly recognised [7]. Thus, the tissues of the kidney, spleen, gut and gills are central for the uptake of pathogens and/or antigens and the subsequent progression of immunity or disease. The organization of the systemic lymphoid organs and the mucosa associated lymphoid tissues has been investigated in a number of fish species [8]. These tissues contain many cell populations known to occupy the lymphoid tissues in mammals but they do not show the same well defined morphological and functional compartments. The fish kidney is divided into the aglomerular anterior part and the glomerular posterior part. The lymphoid and haematopoietic parenchyma are supported by a reticulo-endothelial stroma and are interspersed between an extensive network of vascular sinusoids and renal tissue [9,10]. The lymphoid parenchyma comprises a rich population of leukocytes, including lymphocytes and macrophages. Several of the stromal cell types have phagocytic capacity, and sinusoidal cells are able to trap material from the bloodstream. Accumulations of melanomacrophages are prominent in the haematopoietic tissue and these cells have been shown to retain antigen for long periods [11e13]. The spleen can be divided into red and white pulp. Blood sinuses with their attendant populations of erythrocytes and macrophages dominate the red pulp while the white pulp consists of stretches of lymphoid tissue scattered throughout the splenic parenchyma [14,15]. Specialised terminations of arterioles, the ellipsoids, are present in the teleost spleen and these structures trap blood-borne substances [3,4,16]. Accumulations of melanomacrophages are also present in spleen. The gut-associated lymphoid tissues (GALT) represent a dispersed population of lymphoid cells, granulocytes and macrophages present in and under the intestinal epithelium [17]. The distribution of both B- and T-cells in the intestine has been investigated specifically [18,19]. Absorption of antigen has been reported in posterior intestine epithelium of Atlantic halibut [20]. The intestinal epithelium in teleosts is known to absorb antigen and subsequently transport to intestinal macrophages or to the circulation [17,21]. The gills possess an intricate vasculature [22] and an immune competent leukocyte population [23,24]. Antibody producing cells have been isolated from gills [25,26] and IgM has been detected in gill mucus [27] along with induced levels of specific antibodies [28]. Recently both MHC I and MHC II expressing cells have been demonstrated in the mucosal tissues of the gills [29,30]. The aim of the present study was to characterise the leukocyte populations and their cellular microenvironments in the lymphoid tissues of the kidney and spleen and the mucosal tissues of the hindgut and gills of the Atlantic halibut (H. hippoglossus). We employed a panel of enzyme histochemical methods to characterise lymphoid and nonlymphoid cell populations and two immunohistochemical methods to demonstrate IgM and cytokeratin positive cells. To investigate the ability of tissues to retain particles from the bloodstream, fish were given intravenous injections of two sizes of fluorescent microspheres.

2. Materials and methods 2.1. Fish and tissue samples Ten Atlantic halibut (H. hippoglossus) (mean weight 40 g) were obtained from a commercial fish farm. Fish were anaesthetised in 0.05& Benzocaine, killed and bled from the caudal vein. Spleen, head kidney, body kidney, posterior intestine and gill samples were dissected from each fish. The samples were split in two and one part was immediately frozen in MeforexÒ M55 (Ausimont, Bollate (MI), Italy) chilled in liquid nitrogen and stored at ÿ80  C until use (frozen tissue). The other part was fixed in 4% buffered formaldehyde (pH 7.0) (fixed tissue). The head kidney was divided by a sagital cut into two identical halves. The body kidney was cut either as the head kidney or by a transverse cut into an anterior (which was frozen) portion and a posterior (which was fixed) portion. The most distal part (5e8 mm) of the hindgut was fixed. The adjoining approximately 5e8 mm of the hindgut (in an oral direction) was cut open and placed with the luminal surface resting on a liver slice and then frozen.

S. Grove et al. / Fish & Shellfish Immunology 20 (2006) 693e708

695

Another 16 Atlantic halibut (mean weight 20 g) were obtained from a different commercial fish farm. From 10 of these, tissues were sampled as above, but were fixed in 4% formaldehyde buffered in isotonic PBS (7.5 mM Na2HPO4$2H2O, 2.5 mM NaH2PO4$2H2O, 195 mM NaCl, pH 7.2). After 48 h, samples were transferred to isotonic PBS and kept at 4  C for 48 h, and then stored in 70% ethanol until use. Samples were embedded in paraffin wax. The remaining 6 fish were anaesthetised and received an intravenous injection (vena caudalis) of either green fluorescent Flouresbrite YG Microspheres 0.1 mm (3 fish) or these spheres in combination with yellow/red fluorescent Flouresbrite Polychromatic Red Microspheres 0.5 mm (Polysciences Inc, Warrington, Pennsylvania USA) (3 fish). The total injection volume was 50e70 ml. After 1 h, the fish were killed and spleen, head kidney, body kidney, distal hindgut, and gill samples were collected and frozen in liquid nitrogen. The samples were stored at ÿ80  C until use. The purpose of the injections of microspheres was to detect ellipsoids and other blood clearance structures in tissues. 2.2. Immuno-affinity chromatography (IAC) purification of polyclonal rabbit anti-halibut IgM antiserum Rabbit anti-halibut IgM antiserum (K629) [31] was purified by affinity chromatography modified from Buchmann et al. [32]. Briefly, purified halibut IgM was coupled to divinyl sulfone activated agarose (Mini-LeakÔ, Kem-En-Tec, Copenhagen Denmark) according to the manufacturer. IgM specific rabbit antibodies bound to the immobilised halibut IgM were eluted at low pH. Whereas K629 showed some reactivity for renal tubules, the purified antibodies (IACK629) showed no such reactivity (data not shown). 2.3. Histochemistry Sections from both formalin-fixed and frozen tissues were stained for argyrophilic fibres according to Gomori’s method for silver impregnation [33]. Sections of frozen material were fixed in a solution of 4% formaldehyde and 7.8% ethanol in water. 2.4. Enzyme histochemistry 2.4.1. 5#-Nucleotidase (5#N) The modified method of Mu¨ller-Hermelink et al. [34] as described by Press et al. [35] was used, with the exception that sections were heated in 0.05 M Trisemaleate (pH 7.4) (80  C for 5 min) to obtain control sections. 2.4.2. Non-specific esterase (NSE) The modified method of Pearse [36] as described by Press et al. [35] was used. 2.4.3. Acidic phosphatase (ACP) The modified method of Lodja et al. [37] as described by Press et al. [35] was used, with the exception that sections were heated in water (80  C for 10 min) to obtain control sections. 2.4.4. Alkaline phosphatase (ALP) The method of Press et al. [35] was used with slight modifications. Sections from fish that had received an intravenous injection of fluorescent microspheres were fixed for 5 min at 4  C in formolecalcium (4% formaldehyde, 67 mM CaCl, pH 7.2). For control, sections were heated in water (80  C for 10 min). 2.5. Immunohistochemistry Sections of fixed tissue (5e6 mm) were dewaxed and rehydrated and then demasked by a combination of microwave oven treatment and trypsin digestion. Briefly, sections were heated to 90  C in 0.01 M citric acid (pH 6.0) for 2 ! 5 min and then allowed to cool for 15 min, before transfer to PBS. Sections were then incubated in a trypsin (T8642, Sigma) solution (0.1 mg mlÿ1 in 0.1 M TriseHCl buffer, pH 8.0, 0.1% CaCl2) for 60 s at 37  C. Sections of frozen tissue were fixed for 5 min at 4  C in formolecalcium.

S. Grove et al. / Fish & Shellfish Immunology 20 (2006) 693e708

696

2.5.1. Detection of immunoglobulin Immunohistochemical detection of IgM was performed with a streptavidinebiotin-peroxidase technique. Endogenous peroxidases were inhibited by incubating sections in 3% H2O2 in methanol for 2 min. Then sections were incubated for 20 min with goat normal serum (1:50) (X 0907, DAKO) and avidin (1:6) (SP-200, Vector Laboratories, CA, USA) in PBS with 5% bovine serum albumin (5%BSA/PBS). Next, the sections were incubated overnight at 4  C with IAC-K629 diluted 1:200 in 2.5%BSA/PBS. This incubation solution also contained biotin, in a dilution of 1:6 (SP-2001, Vector Laboratories). Sections were then incubated for 30 min with biotinylated goat anti-rabbit IgG (E 0432, DAKO) 1:500 in 2.5%BSA/PBS and finally for 30 min with streptavidin coupled peroxidase (Roche Diagnostics, Mannheim Germany, 1 089 153) diluted 1:500 in 2.5%BSA/PBS. Freshly made AEC (3-amino-9-ethyl carbazole, A-5754, Sigma) solution was used as the chromogenic substrate (1.3 mM AEC, 0.03% H2O2, 0.9 M N,N-dimethyl formamide, 0.09 M Na-acetate, pH 5.2). The sections were washed three times in PBS between each incubation step. If not specifically stated, incubations were carried out in a humidity chamber at room temperature. The sections were counter-stained with haematoxylin and examined by light microscopy. As a negative control for the method, sections were incubated with a pre-immune rabbit serum instead of the primary antibody. 2.5.2. Detection of cytokeratin To detect epithelial cytokeratin, a monoclonal mouse anti-cytokeratin (ZS18-0132; Zymed Laboratories, San Francisco, USA) was employed as primary antibody. Immunostaining was performed according to Ref. [38], except that counterstaining was performed with Mayer’s haematoxylin. 3. Results All control incubations used in the immune and enzyme histochemical methods gave the expected result. 3.1. Kidney The kidney formed a dorsal-caudal, half-moon shaped trunk, which in the cranial direction gradually narrowed into a Y-shaped anterior part. The trunk or body kidney was rich in renal tissue with glomeruli, tubuli and collecting ducts. Stretches of haematopoietic tissue were present between the renal tissue. Only the most anterior 1/10e1/5 of the kidney (head kidney) was effectively devoid of renal tissue, i.e. aglomerular. The head kidney contained cords of hormonal cells of the suprarenal tissue. 3.1.1. Argyrophilic fibres Silver impregnation revealed that the parenchyma of the kidney was traversed by an extensive net of relative coarse argyrophilic reticular fibres (Fig. 1A). The reticular fibre network did not extend to involve the melanomacrophage accumulations (MMas) and evidence of a capsule delimiting the accumulations was not present. In the body kidney, the renal tubules and collecting ducts were ensheathed in a comprehensive network of argyrophilic fibres and the basement membrane of the tubuli was distinctly silver stained (not shown). 3.1.2. Enzyme histochemistry 5#N: In the body kidney, a small to medium sized 5#NC cell population was seen, often in close proximity to the renal tubules (Fig. 1B). This cell population showed a dark, granular 5#NC staining. A network of weak to moderate 5#NC cells was present in haematopoietic tissue in the body kidney but could not be conclusively identified in the head kidney. A population of large mononuclear cells in the epithelium of renal tubules showed reactivity for 5#NC. The walls of large blood vessels possessed only weak reactivity for 5#N and no 5#N reactivity was detected in melanomacrophages (Fig. 1B). NSE: In the kidney parenchyma, endothelial cells of sinusoidal capillaries showed moderate reactivity for NSE. These capillaries formed extensive networks, which traversed the haematopoietic parenchyma (Fig. 1C). Strong reactivity for NSE was present in single cells that were closely associated with the NSEC capillaries. In the remaining parenchyma, NSE reactivity was mainly present in melanomacrophages, which were often distinctly positive. Strongly NSE positive macrophages were associated with the peritubular sinuses.

S. Grove et al. / Fish & Shellfish Immunology 20 (2006) 693e708

697

Fig. 1. Silver impregnation and enzyme- and immunohistochemical stains of Atlantic halibut kidney. (A) Silver impregnation. Relative coarse argyrophilic fibres are traversing the parenchyma. The argyrophilic fibre network appears not to extend to involve the melanomacrophage accumulations (MMa, arrow). (B) 5#N. Reactivity for 5#N (grey brown colour) is seen in tubuli (T). Solitary 5#N positive cells (arrowheads) are observed between epithelial cells of the tubules epithelium and also in the haematopoietic parenchyma (arrows). (C) NSE. Moderate to strong reactivity for NSE (red brown colour) is present in cells lining small sinusoids. An accumulation of melanomacrophages (arrow) is positive, whereas the endothelium of a larger blood vessel (V) is negative. (D) ACP. Distinct reactivity for ACP (red colour) is present in MMa (large arrows) and solitary melanomacrophages (small arrow). Scattered ACPC single cells (arrowheads) are present in parenchyma. (E) ALP, light microscopy. Strong reactivity (red staining) is seen in tubules (T), in Bowman’s capsule of a glomerulus (G) and in the walls of blood vessels (V). Reactivity is also seen scattered in the parenchyma. (F) ALP, fluorescent microscopy of the same section as seen in (E). The ALP stain emits orangeered light after UV excitation. Tubuli (T), blood vessel (V), Glomerulus (G). Note the green fluorescent 0.1 mm microspheres retained in peritubular sinuses and other narrow blood vessels. (G) ALP, light microscopy. Single ALPC reticular-like cell with cytoplasmic projections (arrowheads). (H) IHC for IgM. IgM positive staining (red colour) is seen in scattered solitary cells in the parenchyma, some IgMC cells are located adjacent to capillaries (C) and larger blood vessels (V).

ACP: In both the head kidney and body kidney, the haematopoietic parenchyma showed a widespread, weak to moderate ACP reactivity. Numerous distinct ACPC cells were scattered within the parenchyma and the melanomacrophages were often strongly ACPC (Fig. 1D). Strong ACP reactivity was present in a population of large mononuclear cells within the epithelium of the renal tubuli and collecting ducts and in scattered cells associated with the peritubular sinus (not shown). ALP: Moderate ALP reactivity was present in walls of larger blood vessels, whereas sinusoids showed weaker reactivity. There was widespread ALP reactivity in the parenchyma within different cell populations (Fig. 1E). The most abundant ALPC cell population was a stellate cell with long cytoplasmic extensions (Fig. 1G). A less abundant ALPC cell population consisted of relatively large cells that possessed globular cytoplasmic reactivity. The melanomacrophages were in many cases ALPÿ, but occasionally a weak reactivity in these cells was observed.

698

S. Grove et al. / Fish & Shellfish Immunology 20 (2006) 693e708

3.1.3. Immunohistochemistry The immunoglobulin-positive (IgMC) cells in fixed tissue sections were predominantly strongly labelled throughout the cytoplasm. The size of positive cells varied from small to very large. IgMC cells were found in both the head and body kidney. Numerous IgMC cells were present in the haematopoietic parenchyma, scattered as single cells or less frequently as small relatively compact clusters (Fig. 1H). IgMC single cells and clusters were often observed in association with vascular vessels and occasionally appeared to form sheaths around smaller vessels. An association between IgMC cells and melanomacrophage accumulations was observed but was not pronounced. IgMC cells were observed in association with the vascular sinuses adjacent to renal tubules and glomeruli. 3.1.4. Fluorescent microspheres In the kidney, the simultaneous administration of the two sizes of microspheres resulted in co-localisation of the microspheres, with the yellow fluorescence of the 0.5 mm microspheres tending to mask the green fluorescence of the 0.1 mm microspheres (data not shown). In the head kidney, the microspheres were mainly associated with the vascular sinusoids. In the body kidney, both sizes of microspheres were also retained in large amounts in association with peritubular and periglomerular sinuses and large blood vessels (Fig. 1F). Microspheres were not commonly found in the epithelium of the renal tubules and were very seldom observed within the lumen of tubules. 3.1.5. Summary e kidney The kidney contained a vascular network of sinusoids. The haematopoietic parenchyma was distributed between the sinusoids, in addition to being associated with larger blood vessels. The kidney parenchyma was traversed by an extensive network of relatively coarse argyrophilic reticular fibres. The presence of a reticular cell network was further indicated by reactivity patterns for ALP and in the body kidney 5#N in a scattered stellate cell population. In the head kidney, a scattered macrophage population within the parenchyma showed moderate NSE and strong ACP reactivity. Another macrophage population with strong NSE and ACP reactivity was associated with the vascular sinusoids. IgMC cells were scattered in the parenchyma and tended to associate with larger blood vessels. Accumulations of melanomacrophages were also scattered in the parenchyma. The argyrophilic fibres of the reticular network did not extend to involve the melanomacrophage accumulations, which did not possess a capsule. Within the accumulations, the melanomacrophages showed strong reactivity for NSE and ACP and were PASC but demonstrated little or no reactivity for ALP and 5#N. In the body or glomerular kidney, the tissue was dominated by renal tubules interspersed with lymphoid parenchyma that was similar to the head kidney. There was a large population of sinusoidal macrophages showing strong NSE and ACP reactivity in association with the peritubular sinuses. A mononuclear cell, presumably macrophage, population was present within the epithelium of the renal tubules. This basally located population showed ACP and 5#N but not NSE reactivity. The morphology of cells showing ALP reactivity differed, consistent with the presence of reticular cell and granulocyte populations. 3.2. Spleen 3.2.1. Argyrophilic fibres Argyrophilic reticular fibres were sparse in most parts of the spleen (Fig. 2A). Some arterioles were supported by relatively coarse argyrophilic fibres. The ellipsoids, as identified by the retention of i.v. injected 0.1 mm microspheres, were surrounded by a relatively dense sheath of fine reticular fibres. 3.2.2. Enzyme histochemistry 5#N: A web-like pattern of 5#N reactivity was present in the spleen (Fig. 2B, I). This dense meshwork of reactivity was most prominent in the red pulp. The white pulp showed less reactivity but a scattered solitary 5#NC cell population was present. However, in certain expanded areas of white pulp, this 5#NC cell population appeared to be absent (Fig. 2I). The ellipsoids contained a small population of relatively large 5#NC cells. Melanomacrophages did not show reactivity for 5#N. NSE: NSEC cells were present in both the red pulp, associated with the sinusoids, and to some degree in the parenchyma of white pulp. A distinctly NSEC cell population was located along the outer border of and occasionally within the wall of the ellipsoids (Fig. 2D, E). Reactivity for NSE in cells of the melanomacrophage accumulations was variable. Both ellipsoid and sinusoid endothelium showed a faint reactivity for NSE.

S. Grove et al. / Fish & Shellfish Immunology 20 (2006) 693e708

699

700

S. Grove et al. / Fish & Shellfish Immunology 20 (2006) 693e708

ACP: Cells reacting for ACP were abundant in the spleen. ACPC cells tended to be associated with blood vessels including both sinusoids and ellipsoids. The melanomacrophage accumulations were distinctly ACPC (data not shown). ALP: Reactivity for ALP showed considerable variation between individual fish. In some individuals, only weak ALP reactivity was observed in the parenchyma but in other individuals, a stronger reactivity was present in the white pulp around large blood vessels and associated with melanomacrophage accumulations (Fig. 2F, G). A weak to moderate reactivity for ALP was observed in the walls of arteries and large arterioles of all fish.

3.2.3. Immunohistochemistry The spleen contained many IgMC cells that were present as scattered single cells and as accumulations, predominantly in the white pulp (Fig. 2H, J). The accumulations of IgMC cells were generally larger but less dense than the clusters of the kidney. The accumulations were usually associated with melanomacrophage accumulations and/or larger blood vessels (Fig. 2J). Some solitary IgMC cells tended to associate with the outer boundary of the ellipsoids.

3.2.4. Fluorescent microspheres The two sizes of microspheres were largely retained in different compartments of the spleen (Fig. 2C). The green fluorescent 0.1 mm microspheres were mainly retained in the ellipsoids forming characteristic circular or tubular deposits (Fig. 2C, E). The yellow fluorescent 0.5 mm microspheres were mainly retained in the red pulp showing more scattered deposits. The parenchyma of the white pulp including the melanomacrophage accumulations were essentially devoid of the microspheres.

3.2.5. Summary e spleen The red pulp of the spleen contained an extensive network of sinusoids that was rich in 5#N reactivity but showed little ALP reactivity. The sinusoids of the red pulp contained many ACPC NSEC macrophages. The white pulp could show considerable ALP reactivity and contained populations of ACPC macrophages and 5#NC reticular cells. Large populations of IgMC cells were present in the white pulp in association with melanomacrophage accumulations and blood vessels. Certain expanded areas of white pulp lacked 5#N reactivity but contained numerous IgMC cells. The ellipsoids located at the margin between the white and red pulp contained a large leukocyte population including an outer rim of NSEC cells and solitary IgMC cells.

Fig. 2. Silver impregnation and enzyme- and immunohistochemical stains of Atlantic halibut spleen. (A) Silver impregnation. Relatively coarse argyrophilic fibres (arrowheads) are lining an arteriole (A) that bifurcates before entering ellipsoids. The ellipsoids (E) are surrounded by a more comprehensive net of finer argyrophilic fibres. Melanomacrophage accumulations (MMa) (arrow). There are few argyrophilic fibres observable in the parenchyma. (B) 5#N, light microscopy. Strong staining for 5#N (red brown) is seen as comprehensive stretches of cells (arrowheads). Scattered solitary cells also show reactivity for 5#N (blue arrows) whereas MMa (black arrows) are negative. (C) Fluorescent microscopy. Green (0.1 mm) and yellow (0.5 mm) fluorescent microspheres are retained in different tissue compartments. Note the circular and tubular outline of the green deposits. (D) NSE, light microscopy. Moderate reactivity for NSE (red brown colour) is present in MMa (arrows) whereas stronger positive solitary macrophages (arrowheads) line the outer rim of ellipsoids (E). (E) Fluorescent microscopy of the same section seen in Fig. 2D. Green fluorescent 0.1 mm microspheres are retained in ellipsoids. (F) ALP, light microscopy. Strong reactivity (red staining) is present in and adjacent to MMa (arrow) and as smaller scattered foci (arrowheads). A rim around the lumen of blood vessels (V) is also ALP positive, whereas ellipsoids (E) are only faintly stained (G) ALP, fluorescent microscopy of the same section seen in Fig. 2F. The ALP stain emits an orangeered light after UV excitation. MMa (arrow), blood vessels (V), ellipsoids (E), ALP positive foci (arrowheads). Note the green fluorescent 0.1 mm microspheres retained in ellipsoids. (H) IHC for IgM, light microscopy. Staining for IgM (red colour) is mainly seen around blood vessels (V) and MMa (arrow), but also around the outer rim of ellipsoids (E). (I, J) Two serial sections, stained for 5#N and IgM, respectively. (I) 5#N, light microscopy. Strong staining for 5#N (red brown) is seen in stretches of cells. Scattered solitary cells also show reactivity for 5#N whereas MMa (large arrows) are negative. Expanded areas of white pulp around arterioles (arrowheads) lack 5#N reactivity. Ellipsoids (E1eE4), artefact (Art). (J) IHC for IgM, light microscopy, cryosection. Expanded areas of white pulp around arterioles (arrowheads) contain accumulations of IgMC cells. IgMC cells also line the outer rim of ellipsoids (some shown, E1eE4). MMa (large arrows). Light blue stained stretches (red pulp) alternate with ellipsoids and white pulp (darker blue areas). Careful comparison of (I) and (J) reveals that the strongly 5#N stained stretches in (I) correspond to the light blue stained stretches of red pulp in (J).

S. Grove et al. / Fish & Shellfish Immunology 20 (2006) 693e708

701

3.3. Posterior intestine 3.3.1. Argyrophilic fibres Both the inner circular and the outer longitudinal muscle layers contained an extensive, relatively coarse network of argyrophilic reticular fibres (data not shown). The lamina propria and submucosa were enmeshed by finer and more brownish fibres. 3.3.2. Enzyme histochemistry 5#N: Intense staining for 5#N was seen in the brushborder of the epithelial cells (Fig. 3A). The lamina propria and submucosa were moderately 5#NC, and distinct reactivity in single cells was occasionally detected. Solitary

Fig. 3. Enzyme- and immunohistochemical stains of Atlantic halibut posterior intestine. (A) 5#N. A strong reactivity for 5#N (dark brown) is present in the brushborder of epithelial cells (BB), while the lamina propria (LP) shows only weak to moderate reactivity. Cells stained for 5#N are seen in the epithelium (arrows). (B) NSE. Brushborder and lamina propria of the intestine villi (V) show distinct reactivity, while submucosa (SM) and the muscularis (M) lack reactivity. Pancreatic tissue (P) is strongly NSEC. (C) NSE. Strong reactivity for NSE (red brown) is present in the brushborder (BB) of epithelial cells. NSEC foci are seen basally in the epithelium (arrows) and both epithelium and lamina propria (LP) show some reactivity for NSE. (D) ACP. Distinct reactivity for ACP (red colour) is seen in the supranuclear compartments of epithelial cells (small arrows). ACPC foci are present in the lamina propria (arrowheads) and in the basal part of epithelium (large arrows). Note that the latter are often relatively larger. (E) ALP. The lamina propria (LP) is distinctly ALP positive (red colour), while the brushborder of epithelial cells (arrows) is stained very strongly. (F) Immunohistochemistry for IgM. A superficial, longitudinal section of a villous reveals IgMC cells (red colour) (mainly) in the epithelium (arrows). A single IgMC cell (arrowhead) is seen surrounded by goblet cells (G). (G) Immunohistochemistry for IgM. IgMC cells are seen in the basal part of the epithelium (arrowheads) and in the lamina propria (arrows). A distinct basal membrane (red arrow) is present. Note the weak interstitial IgMC staining in the lamina propria (LP). Goblet cells (G).

702

S. Grove et al. / Fish & Shellfish Immunology 20 (2006) 693e708

mononuclear 5#NC cells were observed in the epithelium. Both muscular layers showed distinct reactivity for 5#NC (not shown). NSE: In the intestinal epithelium, the brushborder of epithelial cells showed strong reactivity for NSE, while the remainder of the epithelial cell showed a weaker reactivity (Fig. 3B, C). Occasional solitary cells located basally within the epithelium showed strong reactivity for NSE. The lamina propria exhibited an uneven but distinct reactivity, whereas the reactivity in the submucosa was weaker and more variable. Moderate to strong foci of NSE reactivity were observed in the lamina propria. The outer longitudinal muscle layer showed weak NSE reactivity, as did large nerve cells in the inner circular muscle layer and between the two muscle layers. ACP: The supranuclear compartment in the epithelial cells was moderately to strongly ACPC (Fig. 3D). Large cells with strong ACP reactivity were present basally in the epithelium. Single cells with distinct ACP reactivity were frequent to abundant in the lamina propria and submucosa. ALP: The brushborder of epithelial cells was strongly ALPC while a usually more diffuse ALP reactivity was observed in the lamina propria and the submucosa (Fig. 3E). Single cells with distinctly stronger ALP reactivity were sometimes detectable within the diffuse reactivity of the lamina propria and submucosa. 3.3.3. Immunohistochemistry IgMC cells were predominantly located in the epithelium and in the lamina propria (Fig. 3F, G), with few labelled cells in the submucosa and muscular layers. In the epithelium, the IgMC cells were most commonly located basally within the epithelium, but occasionally labelled cells were observed in more apical locations (Fig. 3F). IgMC cells in the lamina propria often had a flattened elongated appearance. IgMC cells were more frequent in the epithelium than in the lamina propria. The presence of IgMC cells varied between individuals. Interstitial IgMC labelling could be distinct in the lamina propria and submucosa. 3.3.4. Fluorescent microspheres Both sizes of microspheres were only found in a very limited number in the posterior intestine. The microspheres were confined to the lumen of small blood vessels in submucosa and lamina propria (data not shown). In fish given both sizes of microspheres, the microspheres were co-localised. 3.3.5. Summary e posterior intestine In the posterior intestine, the lamina propria contained a diverse population of NSEC and ACPC macrophages and IgMC cells. The epithelium contained a prominent population of leukocytes showing reactivity for NSE, ACP and 5#N, presumably representing macrophages. A relatively limited population of IgMC cells was present, mainly in the epithelium of the posterior intestine. 3.4. Gills 3.4.1. Argyrophilic fibres Argyrophilic fibres were only identified in association with supportive elements of the gill arch and filaments, i.e. cartilage and muscle (data not shown). 3.4.2. Enzyme histochemistry 5#N: Distinct to strong reactivity for 5#N was present in the cell membrane of pillar cells (Fig. 4A). In arteries and veins, 5#N reactivity was observed in the layers of connective tissue, but not in the endothelium. A very narrow rim of strong 5#N reactivity was present around the filament rod cartilage and filament muscularis. Sub-epithelial connective tissue in the gill arch and rakers was variably 5#NC. Some 5#N reactivity was also observed in filament muscle. NSE: Epithelial cells covering the filaments, gill arch and rakers were distinctly to strongly NSEC (Fig. 4C). Of these, the most apical cells always showed reactivity, whereas epithelial cells in basal and especially mid-layers of the epithelium could be weaker or do not show reactivity. Interlamellar and lamellar epithelial cells were NSEÿ. Reactivity for NSE was observed in the lamella arterioles and in the outer periphery of the lamellae, which was most likely associated to the endothelium of the outer marginal channel (Fig. 4B). A moderate NSE reactivity of the vessel walls was seen in the efferent, but not in the afferent filament artery. This reactivity was sometimes located to endothelial

S. Grove et al. / Fish & Shellfish Immunology 20 (2006) 693e708

703

Fig. 4. Enzyme- and immunohistochemical stains of Atlantic halibut gills. (A) 5#N staining of a longitudinal section of filaments. Strong reactivity for 5#N (dark brown colour) is present in the walls of filament arteries (FA) and in the pillar cells in the lamellae (arrows). Note the ‘transition’ from filament artery to lamella capillaries (arrowheads). (B) NSE reactivity in a transverse section of lamella. NSEC staining of cells is seen along the outer margin of the lamellae (arrows). (C) NSE reactivity in a section of stratified epithelium on a gill arch. NSE reactivity (red brown colour) is seen in apical epithelial cells, whereas cells basally in epithelium are NSEÿ. Goblet cells (G) and basal membrane (arrowhead). (D) ALP (light microscopy) reactivity in a longitudinal section of a filament, where some lamellae have been sectioned obliquely. Strong ALP reactivity (red colour) is present in a filament artery (arrowheads) and adjoining capillaries in lamellae (arrows). In obliquely sectioned lamellae, the cubical pillar cells are ALPC. (E) ALP (fluorescent microscopy) reactivity in a longitudinal section of a filament, where one lamella has been sectioned transversely. Note the orangeered auto-fluorescence of ALP reactivity. Strong ALP reactivity (red colour) is present in a filament artery (arrowheads) and adjoining capillaries in lamellae (arrows). Yellow fluorescent 0.5 mm microspheres are retained in lamella capillaries, mainly in the apical parts (presumably the outer marginal channel). (F) Immunohistochemistry for cytokeratin in a transverse section of a gill arch. Labelling for cytokeratin (red colour) is seen in the epithelium covering all parts of the gill. The labelling is sharply delineated by the basement membrane (arrows). Goblet cells in apical epithelium (arrowhead). (G) Immunohistochemistry for IgM of a longitudinal section of filaments. IgMC cells are present in the epithelium, whereas more outstretched IgMC cells (small arrows) are seen in very narrow vessels running parallel to a filament artery (FA). A pigment cell (large arrow) in a neighbouring filament is seen in a narrow vessel. Goblet cells (G). (H) Immunohistochemistry for IgM in section of stratified epithelium on a gill arch. Cells labelled for IgM (red colour) are scattered mainly in the basal and middle part of the epithelium. Extracellular labelling for IgM is seen in the apical layers of the epithelium (arrows). Goblet cells (G) and basal membrane (arrowhead).

704

S. Grove et al. / Fish & Shellfish Immunology 20 (2006) 693e708

cells. Muscle tissue in the filament was strongly NSEC, as were gill cartilage and tissues supporting nerve fibres. Pigmented cells were NSEÿ. ALP: The pillar cells in the lamella showed strong ALP reactivity (Fig. 4D, E). The walls of larger blood vessels, especially the arteries, were also ALPC. A very narrow rim of strong ALP reactivity was observed around the filament rod cartilage. This reactivity possibly related to vessels of the interlamellar or the nutrient vascular system. In the gill arch and rakers, large stretches of connective tissue were ALPC, and occasionally ALPC single cells could be observed in the stratified epithelium of gill arch and rakers. Muscle tissue in the filament was also observed to show reactivity for ALP. 3.4.3. Immunohistochemistry IgMC cells were abundant in the gills, and the great majority were present in the stratified gill epithelium. In the filament, the stratified epithelium covering the apex and afferent and efferent edges were rich in scattered IgMC cells. Many IgMC cells were also observed in epithelium covering the gill arch and rakers (Fig. 4H). In the simple epithelium covering the interlamellar spaces, IgMC cells were also observed, but more variably and in considerably lower numbers. IgMC cells were only rarely seen in the lamellar epithelium. Extracellular IgMC staining was regularly present in the sub-epithelial sinuses of the lamellas, and in the apical parts of the stratified epithelium. IgMC cells were frequently observed in narrow and thin-walled blood vessels, adjacent and parallel to the filament arteries or filament cartilage (Fig. 4G). These vessels may belong to the interlamellar and/or nutrient circulatory systems [22]. In the gill arch, IgMC cells were occasionally observed in the lamina propria immediately beneath the basement membrane. To distinguish the epithelium from subjacent connective tissue particularly in regions with stratified epithelium, the gill tissue was labelled with the epithelial marker cytokeratin (Fig. 4F). A clear delineation was present between the broad cytokeratin labelled epithelium and an unlabelled basement membrane. 3.4.4. Fluorescent microspheres Dense deposits of 0.1 mm microspheres were observed in the lamella capillaries, most often distally, suggesting a deposition in the outer marginal channel. The 0.5 mm microspheres showed a similar though not so clear distribution (Fig. 4E). These larger microspheres were also occasionally seen obstructing the lumen of filament arteries. 3.4.5. Summary e gills IgMC cells were prominent in the regions of stratified epithelium in the gills and less frequent in regions of simple epithelium. Other leukocyte populations were scarce although ACPC cells were detected and some NSE reactivity associated with blood vessels could have been related to the presence of a macrophage population. 4. Discussion The characterisation of leukocyte populations within kidney, spleen, posterior intestine and gills of Atlantic halibut showed that the general organization of lymphoid and mucosal tissues was similar to that described for other teleost species such as Atlantic salmon [8,35,39], rainbow trout [40], carp [18,41], sea bass [15,42] and channel catfish [43,44]. However, some novel features were observed. Intra-epithelial leukocyte populations were prominent in the mucosal tissues of Atlantic halibut. The presence of IgMC cells and macrophage populations in the epithelium and lamina propria of the posterior intestine has been reported in teleost species [17,18,45,42]. In teleost species such as carp, many IgC cells (B-cells and plasma cells) are usually present in the lamina propria throughout the posterior intestine, while most Igÿ cells (T-cells) and large Ig-binding macrophages are the populations present in the intestinal epithelium [18]. In halibut, the relatively limited numbers of IgMC cells that were present in the posterior intestine were predominantly in the epithelium rather than in the lamina propria. The presence of intra-epithelial IgMC cells was even more prominent in the gills. The stratified epithelium of halibut gills harboured numerous IgMC cells, although IgMC cells were also found associated with blood vessels and in the lamina propria. Antibody producing cells have been isolated from the gills of sea bass (Dicentrarchus labrax) [26] and dab (Limanda limanda) [25]. Transcription of immunoglobulin heavy m-chains have been reported for gills from turbot (Scophthalmus maximus) [46] and fugu (Takifugu rubripes) [47,48] and in Atlantic halibut (Grove et al., unpublished observations). In Atlantic cod (Gadus morhua), plasma cells were detected by in situ hybridisation in

S. Grove et al. / Fish & Shellfish Immunology 20 (2006) 693e708

705

vessels and associated connective tissue but seemingly not in the epithelium [19]. In the gills of Atlantic salmon (Salmo salar), MHC II was expressed in epithelial cells, and sparsely in other cell populations [30]. While the nature of the IgMC cells was not determined in the present study, the function of these cells in relation to epithelial cells and intra-epithelial leukocyte populations is worthy of further investigation. Interestingly, the majority of IgMC cells in intestinal and gill epithelium of halibut showed strong immunohistochemical labelling throughout the cell cytoplasm, although some large cells with membrane labelling were observed, possibly representing Ig-binding macrophages rather than B-cells. These observations would suggest that the majority of IgMC intra-epithelial leukocytes in halibut were plasma cells. It is possible that the present immunohistochemical technique was not sensitive enough to demonstrate the membrane labelling of intra-epithelial B-cells. Even so, the further definition of intra-epithelial lymphocyte populations is needed particularly the identification of T-cells (Igÿ lymphocytes) and the study of the co-expression of MHC II. In mammals, B-cells but not plasma cells express MHC II [49]. However, it should be noted that Koppang et al. [30] were unable to report evidence for MHC II expression in B-lymphocytes of Atlantic salmon. A diverse population of non-lymphoid leukocytes was also present in mucosal epithelia, as demonstrated using enzyme histochemical techniques. Studies have shown that NSE reactivity could be detected in cells isolated from the gills of dab and Atlantic salmon and that some of these cells had a macrophage-like morphology [23]. However, most enzyme reactivity detected in these mucosal tissues was localised to epithelial cells. In the present study, the intravenous injection of microspheres provided important information on the organization of non-lymphoid leukocytes in the kidney and spleen of Atlantic halibut. Two distinct patterns of localisation of microspheres were identified. Microspheres were retained in association with the extensive network of sinusoids in the aglomerular and glomerular kidney and the red pulp of the spleen. Sinusoidal macrophages that showed reactivity for NSE and ACP were presumably involved in the trapping of microspheres. In the spleen, sinusoidal macrophages of the red pulp appeared to be responsible for the trapping of the larger microspheres (0.5 mm diameter) while the ellipsoids retained the smaller (0.1 mm diameter) microspheres. A role in the selective filtration and retention of blood-borne particles has been described for ellipsoids in rainbow trout [50]. The further involvement of macrophage populations associated with sinusoids and ellipsoids in the processing and presentation of trapped antigen requires further investigation. It has been speculated that trapped antigens after removal from blood are transported to melanomacrophage accumulations where immune responses are initiated and long-term storage of antigens undertaken [16,51]. However, the relationship between the NSEC and ACPC macrophages of sinusoids and ellipsoids, the moderate NSEC cells of the lymphoid parenchyma and the NSEC ACPC melanomacrophages of the kidney and spleen remains to be determined, as does the interaction of these macrophage populations with the clusters of IgMC cells identified in the spleen and kidney. Further observations in the present study that are worthy of pursuit include the differences in enzyme histochemical profile of melanomacrophage accumulations in the kidney and spleen and the nature of the intraepithelial cell population showing 5#N and ACP reactivity in renal tubuli of the kidney. In spleen, 5#N reactive reticular-like cells were observed in stretches of white pulp, but not in certain expanded areas of white pulp. This interesting difference may reflect a differential presence of reticular cells or that the phenotype of reticular cells may vary between tissue compartments. It is increasingly recognised that mammalian stromal cell populations, by differential expression of receptors and ligands, participate in the regulation of site-specific leukocyte trafficking [52]. In the posterior intestine and gills, microspheres were confined to the lumen of small blood vessels suggesting embolic obstruction. However, the concurrence of microsphere retention in the gills and NSE reactivity in lamella arteriole endothelium could suggest an endocytic capacity of a specialised endothelium [53]. Similarly, the observation of IgMC cells and pigment containing cells in narrow blood vessels adjacent to filament arteries and supportive structures in the filament is interesting. In addition to the respiratory arterio-arterial vascular system, the gills contain two separate post-lamellar vascular networks, i.e. the interlamellar system and the nutrient circulation [22]. Both systems are fed by efferent arteries and form extensive, often parallel networks in the central filament body, making the distinction between them difficult. The interlamellar circulation has strong anatomical similarities to mammalian lymphatic capillaries [22]. A possible localisation of IgMC cells and large pigment containing cells in this circulatory system may therefore have immunological significance. Labelling of interstitial IgM was often seen in apical epithelium and in the sub-epithelial sinuses in lamellae. The presence of interstitial IgM may be an important defence mechanism in the lamellae, where access of immune cells may be physically restricted.

706

S. Grove et al. / Fish & Shellfish Immunology 20 (2006) 693e708

The observation that mucosal tissues of Atlantic halibut contained a rich and diverse population of leukocytes supports the postulated separation of mucosal and systemic immune systems in teleosts [7,18,28,54] and emphasises the need to consider route of vaccination in the formulation of disease control strategies for this teleost species. It has been shown in various situations that immersion vaccination of teleosts invokes a stronger local immune response than does intraperitoneal vaccination in the absence of powerful adjuvants [7,55,56]. Furthermore, the differences in cell populations observed between the gills and posterior intestine would argue for a compartmentalisation of the mucosal immune system in halibut. Studies in sea bass have shown that immersion vaccination induced extremely high numbers of antibody-secreting cells in the gills but did not provoke a response of comparable magnitude in the gut or in systemic lymphoid organs [26]. The present morphological study would suggest that the detailed immunological investigation of mucosal tissues in Atlantic halibut is warranted. Acknowledgements The authors thank Laila Aune, Inger Bo¨ckermann, Marisel Sa´nchez Rodriguez and Randi Faller for their technical advice and skill with enzyme- and immunohistochemical techniques. Further, we thank Dr Agnar Kvellestad for fruitful discussions on gill morphology. This work was financed by a grant (124210/110) from The Norwegian Research Council. References [1] Bricknell IR, Bowden TJ, Verner-Jeffreys DW, Bruno DW, Shields RJ, Ellis AE. Susceptibility of juvenile and sub-adult Atlantic halibut (Hippoglossus hippoglossus L.) to infection by Vibrio anguillarum and efficacy of protection induced by vaccination. Fish Shellfish Immunol 2000;10:319e27. [2] Grove S, Johansen R, Dannevig BH, Reitan LJ, Ranheim T. Experimental infection of Atlantic halibut Hippoglossus hippoglossus with nodavirus: tissue distribution and immune response. Dis Aquat Organ 2003;53:211e21. [3] Espenes A, Press CM, Reitan LJ, Landsverk T. The trapping of intravenously injected extracellular products from Aeromonas salmonicida in head kidney and spleen of vaccinated and nonvaccinated Atlantic salmon, Salmo salar L. Fish Shellfish Immunol 1996;6:413e26. [4] Ferguson HW, Claxton MJ, Moccia RD, Wilkie EJ. The quantitative clearance of bacteria from the bloodstream of rainbow trout (Salmo gairdneri). Vet Pathol 1982;19:687e99. [5] Drolet BS, Rohovec JS, Leong JC. The route of entry and progression of infectious hematopoietic necrosis virus in Oncorhynchus mykiss (Walbaum) e a sequential immunohistochemical study. J Fish Dis 1994;17:337e47. [6] Holzer AS, Sommerville C, Wootten R. Tracing the route of Sphaerospora truttae from the entry locus to the target organ of the host, Salmo salar L., using an optimized and specific in situ hybridization technique. J Fish Dis 2003;26:647e55. [7] Huising MO, Guichelaar T, Hoek C, Verburg-van Kemenade BML, Flik G, Savelkoul HFJ, et al. Increased efficacy of immersion vaccination in fish with hyperosmotic pretreatment. Vaccine 2003;21:4178e93. [8] Press CM, Evensen O. The morphology of the immune system in teleost fishes. Fish Shellfish Immunol 1999;9:309e18. [9] Zapata AG. Ultrastructural study of the teleost fish kidney. Dev Comp Immunol 1979;3:55e65. [10] Meseguer J, Lopezruiz A, Garciaayala A. Reticuloendothelial stroma of the head-kidney from the seawater teleost gilthead seabream (Sparus aurata L) e an ultrastructural and cytochemical study. Anat Rec 1995;241:303e9. [11] Grove S, Hoie S, Evensen O. Distribution and retention of antigens of Aeromonas salmonicida in Atlantic salmon (Salmo salar L.) vaccinated with a Delta aroA mutant or formalin-inactivated bacteria in oil-adjuvant. Fish Shellfish Immunol 2003;15:349e58. [12] Herraez MP, Zapata AG. Structure and function of the melano-macrophage centres of the goldfish Carassius auratus. Vet Immunol Immunopathol 1986;12:117e26. [13] Lamers CHJ, De Haas MJ. Antigen localization in the lymphoid organs of carp (Cyprinus carpio). Cell Tissue Res 1985;242:491e8. [14] Zapata A. Lymphoid organs of teleost fish. III. Splenic lymphoid tissue of Rutilus rutilus and Gobio gobio. Dev Comp Immunol 1982;6: 87e94. [15] Quesada J, Villena MI, Agulleiro B. Structure of the spleen of the sea bass (Dicentrarchus labrax) e a light and electron-microscopic study. J Morphol 1990;206:273e81. [16] Espenes A, Press CM, Dannevig BH, Landsverk T. Immune-complex trapping in the splenic ellipsoids of rainbow trout (Oncorhynchus mykiss). Cell Tissue Res 1995;282:41e8. [17] Rombout JHWM, Bot HE, TaverneThiele JJ. Immunological importance of the second gut segment of carp. II. Characterization of mucosal leukocytes. J Fish Biol 1989;35:167e78. [18] Rombout JHWM, TaverneThiele AJ, Villena MI. The gut-associated lymphoid-tissue (GALT) of carp (Cyprinus carpio L.): an immunocytochemical analysis. Dev Comp Immunol 1993;17:55e66. [19] Schrøder MB, Flan˜o E, Pilstro¨m L, TØ Jørgensen. Localisation of Ig heavy chain mRNA positive cells in Atlantic cod (Gadus morhua L.) tissues identified by in situ hybridisation. Fish Shellfish Immunol 1998;8:565e76. [20] Strand HK, Dalmo RA. Absorption of immunomodulating beta(1,3)-glucan in yolk sac larvae of Atlantic halibut, Hippoglossus hippoglossus (L.). J Fish Dis 1997;20:41e9.

S. Grove et al. / Fish & Shellfish Immunology 20 (2006) 693e708

707

[21] O’Donnell GB, Reilly P, Davidson GA, Ellis AE. The uptake of human gamma globulin incorporated into poly (D,L-lactide-co-glycolide) microparticles following oral intubation in Atlantic salmon, Salmo salar L. Fish Shellfish Immunol 1996;6:507e20. [22] Olson KR. Vascular anatomy of the fish gill. J Exp Zool 2002;293:214e31. [23] Lin SH, Davidson GA, Secombes CJ, Ellis AE. A morphological study of cells isolated from the perfused gill of dab and Atlantic salmon. J Fish Biol 1998;53:560e8. [24] Lin SH, Ellis AE, Davidson GA, Secombes CJ. Migratory, respiratory burst and mitogenic responses of leucocytes isolated from the gills of rainbow trout (Oncorhynchus mykiss). Fish Shellfish Immunol 1999;9:211e26. [25] Davidson GA, Lin SH, Secombes CJ, Ellis AE. Detection of specific and ‘constitutive’ antibody secreting cells in the gills, head kidney and peripheral blood leucocytes of dab (Limanda limanda). Vet Immunol Immunopathol 1997;58:363e74. [26] dos Santos NMS, Taverne-Thiele JJ, Barnes AC, van Muiswinkel WB, Ellis AE, Rombout JHWM. The gill is a major organ for antibody secreting cell production following direct immersion of sea bass (Dicentrarchus labrax, L.) in a Photobacterium damselae ssp piscicida bacterin: an ontogenetic study. Fish Shellfish Immunol 2001;11:65e74. [27] Zilberg D, Klesius PH. Quantification of immunoglobulin in the serum and mucus of channel catfish at different ages and following infection with Edwardsiella ictaluri. Vet Immunol Immunopathol 1997;58:171e80. [28] Lumsden JS, Ostland VE, MacPhee DD, Ferguson HW. Production of gill-associated and serum antibody by rainbow-trout (Oncorhynchus mykiss) following immersion immunization with acetone-killed Flavobacterium branchiophilum and the relationship to protection from experimental challenge. Fish Shellfish Immunol 1995;5:151e65. [29] Dijkstra JM, Kollner B, Aoyagi K, Sawamoto Y, Kuroda A, Ototake M, et al. The rainbow trout classical MHC class I molecule OnmyUBA*501 is expressed in similar cell types as mammalian classical MHC class I molecules. Fish Shellfish Immunol 2003;14:1e23. [30] Koppang EO, Hordvik I, Bjerkas I, Torvund J, Aune L, Thevarajan J, et al. Production of rabbit antisera against recombinant MHC class II beta chain and identification of immunoreactive cells in Atlantic salmon (Salmo salar). Fish Shellfish Immunol 2003;14:115e32. [31] Grove S, Tryland M, Press CM, Reitan LJ. Serum immunoglobulin M in Atlantic halibut (Hippoglossus hippoglossus): characterisation of the molecule and its immunoreactivity. Fish Shellfish Immunol 2006;20:97e112. [32] Buchmann K, Østergaard L, Glamann J. Affinity purification of antigen-specific serum immunoglobulin from the European eel (Anguilla anguilla). Scand J Immunol 1992;36:89e97. [33] Romeis B. Mikrosopiche Technik. Mu¨nich: Leibniz Verlag; 1948. [34] Mu¨ller-Hermelink HK, Heusermann U, Stutte HJ. Enzyme histochemical observations on the localisation and structure of the T cell and B cell regions in the human spleen. Cell Tissue Res 1974;154:167e79. [35] Press CM, Dannevig BH, Landsverk T. Immune and enzyme-histochemical phenotypes of lymphoid and nonlymphoid cells within the spleen and head kidney of Atlantic salmon (Salmo salar L.). Fish Shellfish Immunol 1994;4:79e93. [36] Pearse AGE. Histochemistry: theoretical and applied. Edinburgh: Churchill-Livingstone; 1968. [37] Lodja Z, Gossrau R, Schiebler TH. Enzymhistochemische Metoden. Berlin, Heidelberg: Springer-Verlag; 1976. [38] Falk K, Press CM, Landsverk T, Dannevig BH. Spleen and kidney of Atlantic salmon (Salmo salar L.) show histochemical changes early in the course of experimentally induced infectious salmon anaemia (ISA). Vet Immunol Immunopathol 1995;49:115e26. [39] BakkeMcKellep AM, Press CM, Baeverfjord G, Krogdahl A, Landsverk T. Changes in immune and enzyme histochemical phenotypes of cells in the intestinal mucosa of Atlantic salmon, Salmo salar L., with soybean meal-induced enteritis. J Fish Dis 2000;23:115e27. [40] Grace MF, Manning MJ. Histogenesis of the lymphoid organs in rainbow-trout, Salmo gairdneri Rich-1836. Dev Comp Immunol 1980;4:255e64. [41] Lamers CHJ. Histophysiology of a primary immune response against Aeromonas hydrophila in carp (Cyprinus carpio L.). J Exp Zool 1986;238:71e80. [42] Abelli L, Picchietti S, Romano N, Mastrolia L, Scapigliati G. Immunohistochemistry of gut-associated lymphoid tissue of the sea bass Dicentrarchus labrax (L). Fish Shellfish Immunol 1997;7:235e45. [43] Hebert P, Ainsworth AJ, Boyd B. Histological enzyme and flow cytometric analysis of channel catfish intestinal tract immune cells. Dev Comp Immunol 2002;26:53e62. [44] PetrieHanson L, Ainsworth AJ. Differential cytochemical staining characteristics of channel catfish leukocytes identify cell populations in lymphoid organs. Vet Immunol Immunopathol 2000;73:129e44. [45] Georgopoulou U, Vernier JM. Local immunological response in the posterior intestinal segment of the rainbow trout after oral-administration of macromolecules. Dev Comp Immunol 1986;10:529e37. [46] Low C, Wadsworth S, Burrells C, Secombes CJ. Expression of immune genes in turbot (Scophthalmus maximus) fed a nucleotidesupplemented diet. Aquaculture 2003;221:23e40. [47] Saha NR, Suetake H, Suzuki Y. Characterization and expression of the immunoglobulin light chain in the fugu: evidence of a solitaire type. Immunogenetics 2004;56:47e55. [48] Saha NR, Suetake H, Suzuki Y. Analysis and characterization of the expression of the secretory and membrane forms of IgM heavy chains in the pufferfish, Takifugu rubripes. Mol Immunol 2005;42:113e24. [49] Piskurich JF, Lin KI, Lin Y, Wang Y, Ting JPY, Calame K. BLIMP-I mediates extinction of major histocompatibility class II transactivator expression in plasma cells. Nat Immunol 2000;1:526e32. [50] Espenes A, Press CM, Dannevig BH, Landsverk T. Investigation of the structural and functional features of splenic ellipsoids in rainbowtrout (Oncorhynchus mykiss). Cell Tissue Res 1995;279:469e74. [51] Agius C, Roberts RJ. Melano-macrophage centres and their role in fish pathology. J Fish Dis 2003;26:499e509. [52] Parsonage G, Filer AD, Haworth O, Nash GB, Rainger GE, Salmon M, et al. A stromal address code defined by fibroblasts. Trends Immunol 2005;26:150e6.

708

S. Grove et al. / Fish & Shellfish Immunology 20 (2006) 693e708

[53] Dannevig BH, Lauve A, Press CM, Landsverk T. Receptor-mediated endocytosis and phagocytosis by rainbow-trout head kidney sinusoidal cells. Fish Shellfish Immunol 1994;4:3e18. [54] Davidson GA, Ellis AE, Secombes CJ. Route of immunization influences the generation of antibody secreting cells in the gut of rainbow trout (Oncorhynchus mykiss). Dev Comp Immunol 1993;17:373e6. [55] dos Santos NM, Taverne-Thiele JJ, Barnes AC, Ellis AE, Rombout JH. Kinetics of juvenile sea bass (Dicentrarchus labrax, L.) systemic and mucosal antibody secreting cell response to different antigens (Photobacterium damselae spp. piscicida, Vibrio anguillarum and DNP). Fish Shellfish Immunol 2001;11:317e31. [56] Vervarcke S, Ollevier F, Kinget R, Michoel A. Mucosal response in African catfish after administration of Vibrio anguillarum O2 antigens via different routes. Fish Shellfish Immunol 2005;18:125e33.