Lack of a contact requirement for direct antibacterial activity of lymphocyte subpopulations in ginbuna crucian carp

Fish & Shellfish Immunology 39 (2014) 178e184

Contents lists available at ScienceDirect

Fish & Shellfish Immunology journal homepage: www.elsevier.com/locate/fsi

Full length article

Lack of a contact requirement for direct antibacterial activity of lymphocyte subpopulations in ginbuna crucian carp Haitham M. Tartor a, Yuta Matsuura b, Gamal El-Nobi a, Teruyuki Nakanishi b, * a b

Faculty of Veterinary Medicine, Zagazig University, Zagazig, Egypt Department of Veterinary Medicine, Nihon University, Fujisawa, Kanagawa 252-0880, Japan

a r t i c l e i n f o

a b s t r a c t

Article history: Received 16 November 2013 Received in revised form 3 May 2014 Accepted 3 May 2014 Available online 22 May 2014

Cytotoxic T lymphocytes (CTL) recognize and kill cells infected with viruses, intracellular bacteria and tumors with MHC restriction and antigen specificity. In addition to these activities, recent studies in mammals have suggested that CTL can exhibit direct microbicidal activity. In our previous study we documented direct antibacterial activity of CD4þ T cells and sIgMþ cells as well as CD8aþ T cells from immunized fish. However, we also found weak non-specific killing activity of lymphocytes against bacteria. In the present study we further analyzed the weak killing activity of lymphocytes, increasing the effector cell to target bacteria ratio from 10:1 to 103:1. Sensitized and non-sensitized effector lymphocytes (CD8aþ, CD4þ and sIgMþ) separated by MACS were incubated with target bacteria. CD8aþ T cells from Edwardsiella tarda-immunized ginbuna crucian carp killed 98%, 100% and 70% of E. tarda, Streptococcus iniae and Escherichia coli, respectively. CD8aþ T cells from non-immunized fish showed similar but slightly lower killing activity than sensitized cells. CD4þ and sIgMþ lymphocytes also showed high killing activity against E. tarda and S. iniae as found for CD8aþ T cells, although the activity was lower against E. coli. Supernatants from all three types of lymphocytes showed microbicidal activity, although the activity was lower than that evoked by effector lymphocytes. Furthermore, the presence of a membrane between effectors and targets did not affect the killing activity. The present results suggest that both sensitized and non-sensitized lymphocytes non-specifically killed target bacteria without the need of contact. The major difference between the present and previous experiments is the E:T ratio. We suspect that there are two different mechanisms in the direct bacterial killing by lymphocytes in ginbuna. © 2014 Elsevier Ltd. All rights reserved.

Keywords: Antibacterial activity CD8aþ cells CD4þ cells Surface IgMþ cells Ginbuna crucian carp

1. Introduction All over the world one of the fastest rising sectors of food production is aquaculture. However, this growth is hindered by severe economic losses due to many infectious diseases [1]. Antibiotics are not the best choice in countering infectious diseases because of accumulation in the environment and the emergence of antibiotic resistant strains [2]. The best alternative to antibiotics for protection against infectious diseases is vaccination and usage of effective vaccines to control persistent and emerging diseases have significant positive impact on the reduced usage of antibiotics [3]. However, evaluation of protection for most of vaccines was based on

* Corresponding author. Department of Veterinary Medicine, College of Bioresource Sciences, Nihon University, 1866 Kameino, Fujisawa, Kanagawa 252-0880, Japan. Tel.: þ81 466 84 3383; fax: þ81 466 84 3380. E-mail address: [email protected] (T. Nakanishi). http://dx.doi.org/10.1016/j.fsi.2014.05.006 1050-4648/© 2014 Elsevier Ltd. All rights reserved.

parameters relevant to humoral immunity and not cell-mediated immunity. In spite of accumulating evidences that cell-mediated immunity play important roles in the protection against viruses and intracellular bacteria, little is known about cell-mediated immunity in fish [4]. Immunity against extracellular bacteria depends on antibodies, complement factors and phagocytic cells while protection against intracellular pathogens is highly dependent on cytotoxic T lymphocytes (CTLs) [5] and activated monocytes/macrophages stimulated with IFN-g [6]. Effective immunity to intracellular bacterial infection often requires lysis of infected cells as well as killing of the invading pathogen [7]. CTLs are required for clearance of intracellular bacteria and recognize bacterial peptides presented by MHC molecules on the surface of infected cells. However, in mammals MHC-independent direct antimicrobial activities of T cells against different pathogens have been reported [8]. Recently, we reported the direct antibacterial activity of CD8a, CD4þ T cells and sIgMþ cells against both intracellular and

H.M. Tartor et al. / Fish & Shellfish Immunology 39 (2014) 178e184

extracellular bacteria in ginbuna crucian carp [9]. Although we demonstrated higher killing activity of immunized effector cells when compared to that of non-immunized cells at an E:T ratio of 10:1, we detected low (15e20%) but significant killing activity even in non-immunized effector cells. Furthermore, in mammals, results from in vitro studies indicate that contact is required between T cells and pathogens such as fungi and parasites to achieve direct killing activity [8], while the contact is not required for bacteria [10]. At present, a contact requirement between fish effector cells and target bacteria remains unknown. In the present study we further analyzed the weak killing activities of non-immunized effector cells, increasing the E:T ratio from 10:1 to 1000:1. We also examined the requirement for contact between effector cells and target bacteria together with the specificity of the killing. We found high antibacterial activity in T cell subpopulations (CD8aþ and CD4þ) and sIgMþ cells from both immunized and non-immunized ginbuna crucian carp at an E:T ratio of 1000:1, without contact between bacterial targets and effector cells. Our findings suggest that there are two different mechanisms in the direct bactericidal activity of lymphocytes, and that lymphocytes are involved in innate cellular immunity in fish. 2. Material and methods 2.1. Fish Ginbuna crucian carp, Carassius auratus langsdorfii (OB1 clone, collected from Okushiri Island) weighing 25e40 g were used in this study. Fish were maintained at 25 ± 1  C in 60 l tanks with running water and were fed once daily with commercial pellets. 2.2. Bacterial strains Two gram-negative bacteria, Edwardsiella tarda FPC347 as an intracellular bacterium and Escherichia coli IAM1239 as an extracellular bacterium as well as Streptococcus iniae no. 20 as a grampositive extracellular bacterium were used in this study. These were kindly provided by Dr. Mano in the Department of Marine Science and Resources of our university. 2.3. Monoclonal antibodies Monoclonal antibodies (MAbs) against ginbuna CD8a and CD4 were produced in rat according to Akashi et al. [11] and the characteristics of the MAbs have been described in our previous papers [12,13]. A MAb against ginbuna IgM was produced in mice using a standard protocol and has been used to separate sIgMþ and sIgM cells [14,15]. 2.4. Preparation of bacterial antigen E. tarda and S. iniae were grown separately in Tryptic Soy (TS) broth (Eiken chemical Co. Ltd, Japan) at 26  C for 24 h. E. coli was grown in TS broth at 37 C for 24 h. Bacterial cells were harvested by centrifuging the broth at 1000 g for 10 min at room temperature. Bacterial antigen was prepared by inactivating the live cells in PBS (pH 7.3) with 1% formalin overnight at 4  C. The bacterial cells were then collected as described above and finally suspended in PBS after three washes.

179

booster injection, fish were killed and head and trunk kidney from three immunized ginbuna were aseptically removed to prepare effector lymphocytes. For the studies with neutrophils, tissues were sampled three days after the final immunization because of their short life span. 2.6. Preparation of effector cells Head and trunk kidney tissues from both immunized and nonimmunized fish were aseptically disaggregated through a sterilized 150-gauge mesh stainless steel sieve in HBSS (Nissui Pharmaceutical Co. Ltd, Japan) supplemented with 10% heat-inactivated fetal bovine serum (FBS), 0.5% bovine serum albumen (BSA) (Wako Pure Chemical Industries, Ltd, Japan), 100 U/ml of penicillin and 100 mg/ml of streptomycin (Wako Pure Chemical Industries, Ltd, Japan). 2.6.1. Preparation of lymphocyte subpopulations The leucocyte suspension was layered over a Percoll (GH Healthcare, USA) density gradient (r ¼ 1.08 g/ml) and centrifuged at 450 g for 30 min at 4  C. The lymphocyte-rich fraction at the interface was collected and washed three times with HBSS. Magnetic activated cell sorting (MACS; Mini Macs, Miltenyi Biotec) was used to obtain different cell fractions (CD8aþ/CD4þ/sIgMþ cells) from the lymphocyte-rich fraction after incubation with specific mAb against the respective cell types according to Toda el al [12]. Briefly, 1.0  107 cells/ml of kidney lymphocytes in HBSS were incubated with 1:104 diluted rat anti-ginbuna CD8a MAb (mouse ascites) for 45 min on ice. Cells were then washed three times with HBSS and adjusted to 1  108 cells/ml, incubated for 30 min at 4  C with 1 ml of a 1:5 dilution of magnetic bead-conjugated goat antirat Ig antibody (Miltenyi Biotec GmbH, Germany), and washed three additional times. Each cell suspension was then applied to a plastic column equipped with an external magnet to separate the CD8aþ and CD8a fractions. Similarly, CD4þ and sIgMþ cells were separated using rat anti-ginbuna CD4 MAb and mouse antiginbuna IgM MAb, respectively. 2.6.2. Preparation of neutrophil cells The kidney cell suspension was layered over two gradients (r ¼ 1.075 and 1.09 g/ml) of Percoll and centrifuged at 400 g for 30 min at 4  C. Cells above the 1.09 Percoll layer were then collected followed by three washes with HBSS and the cells were then counted. 2.7. Viability and purity of individual cell fractions The viability of lymphocytes and neutrophils was assessed by Trypan blue dye exclusion while the purity of the MACS-sorted lymphocyte cell fractions was confirmed by flow cytometry. A portion of MACS-sorted CD8aþ and/or CD4þ cell fractions was incubated with FITC-conjugated goat anti-rat IgG þ M þ A antibody (Rockland) following incubation with anti-ginbuna CD8a and/or CD4 MAbs. Similarly, a portion of the sIgMþ cells was incubated with FITC-conjugated goat anti-mouse Ig G þ M antibody with antiginbuna IgM MAb. In order to remove any antibiotics, effector cells were washed three times and re-suspended in antibiotic-free RPMI 1640 (Nissui pharmaceutical co. ltd, Japan) supplemented with 10% fetal bovine serum (FBS) and 2 mM L-glutamine (Wako Pure Chemical Industries, Ltd, Japan).

2.5. Immunization of effector donors 2.8. Antibacterial activity of different effector cells Ginbuna were intra-peritoneally immunized with inactivated bacterial antigen (E. tarda at 108 CFU/fish) followed by one booster dose seven days after the first immunization. Seven days after the

The direct killing activity of effector cells (CD8aþ, CD4þ, sIgMþ cells and neutrophils) was evaluated using the colony-forming unit

180

H.M. Tartor et al. / Fish & Shellfish Immunology 39 (2014) 178e184

Table 1 The composition of MACS and Percoll-isolated cells. Fraction

Viability (%)

Purity (%)

Composition of isolated cells Lymphocytes

þ

CD8a T cells CD4þ T cells sIgMþ cells Cell fraction above 1.09 Percoll layer

92.3 ± 1.5 93.8 ± 3.5 92.5 ± 4.9 97.6% ± 0.8

90.8 92.0 93.2 93.6

± ± ± ±

2.0 1.5 1.0 1.4

82.9 83.8 77.9 1.5

± ± ± ±

5.1 2.7 3.6 0.5

Neutrophils

Macrophage and blast cells

2 0.1 ± 0.4 1.5 ± 0.5 1.6 ± 1.3 93.6% ± 1.4

14.3 12.7 15.3 2.8

± ± ± ±

2.6 3.1 0.9 1.3

Kidney cells from OB1 ginbuna were magnetically separated into CD8aþ, CD4þ and sIgMþ subpopulations, or were enriched for neutrophils by Percoll separation. Numbers indicated the percentage of cell types in each fraction as the mean ± SEM (n ¼ 3).

(CFU) assay according to Markham et al. [16]. Briefly, 100 ml of 1  106 cells/ml (105 cells/well) of effector cells in antibiotic-free RPMI 1640 was placed in a well of a flat-bottomed 96-well plate. For each type of target bacteria, 100 ml containing 1  102 bacteria (suspended in RPMI 1640) was added to the well at an effector (E):target (T) ratio of 1000:1 and then incubated for 4 h at 26  C. After incubation, 100 ml of the culture was withdrawn from each well and the CFU assay was conducted to determine the number of live bacteria. All experiments were performed in triplicate and the percentages of bacterial killing activity (% killing) were determined according to the following equation described by Nayak and Nakanishi [9].

% killing ¼1  ½ðviable cells in test cultureÞ =ðviable cells in control culture  ½bacteria þ RPMI mediumÞ  100

2.10. Statistical analysis Statistical analysis was performed by using student ‘t’ test to determine significant differences (p  0.05) in the mean percentage of bacterial cells killed at a given E/T ratio. 3. Results 3.1. Purity and viability of isolated cells Flow cytometric analysis of MACS-sorted CD8aþ, CD4þ, sIgMþ and Percoll-isolated neutrophils showed that purities were more than 90%, 92%, 93% and 93%, respectively. Percentages of lymphocyte in MACS-sorted CD8aþ, CD4þ, sIgMþ cells and Percoll-isolated neutrophils were 80%, 81%, 70% and 1.5% respectively. Viabilities of above cells were more than 91%, 90%, 90% and 97%, respectively (Table 1, Figs. 4 and 5).

2.9. Contact requirement assay

3.2. Effect of immunization

2.9.1. Effect of tissue culture insert The killing activity of effector cells separated from target bacteria was tested using 24-well tissue culture inserts (Thin Cert, Greiner Bio-One) according to Farouk et al. [17]. Briefly, 300 ml of 1  103 cells/ml (3  102 cells/well) of E. tarda was added to the lower compartment of 24-well plate wells and 300 ml of 1  106 cells/ml (3  105 cells/well) of effector cells were added to the upper compartment of the tissue culture inserts. The two compartments were separated by a membrane of 1 mm pore size. Plates were then incubated at 26  C for 4 h. After incubation, 100 ml of the culture was withdrawn from the lower compartment and the CFU assay was conducted to determine the percentage of killing activity as described above. As a negative control, 300 ml of 1  103 cells/ml (3  102 cells/ well) of E. tarda was added to the lower compartment of a 24-well plate well and 300 ml of cell-free RPMI was added to the upper compartment of the tissue culture insert.

The direct antibacterial activity of different effector cells (CD8aþ, CD4þ, sIgMþ and neutrophils) collected from immunized and nonimmunized ginbuna against three different bacteria (E. tarda, S. iniae and E. coli) was examined. All cell types from both immunized and non-immunized ginbuna exhibited antibacterial activities to a similar extent against the different bacterial targets. However, killing activity of immunized effectors was slightly higher than that of non-immunized effectors (Fig. 1).

2.9.2. Killing activity of supernatants from effector cells The killing activity of supernatants from effector cells incubated with or without target bacteria was evaluated. Briefly, 100 ml of supernatant from effector cells incubated with E. tarda at an E:T ratio of 1000:1 was collected by centrifugation at 1000 g for 10 min following vigorous pipetting of the culture. One hundred ml of supernatant was sterilized by filtration through a 0.45-mm Millex filter (Millipore S.A.S., Molsheim, France) to remove any bacteria or cells and then incubated with 100 ml of 1  103 cells/ml of E. tarda for 4 h. The CFU assay was conducted to determine the percentage of killing activity. As a negative control, 100 ml of RPMI was added to 100 ml of 1  103 cells/ml of E. tarda before conducting CFU assay. We used the same equation described in co-culture experiment, although supernatant was used instead of test culture.

3.3. Bacterial killing activity of different effector cells 3.3.1. CD8aþ T cells CD8aþ T cells collected from ginbuna immunized with inactivated E. tarda showed high killing activity against both E. tarda and S. iniae at 98.8 ± 1.2% and 99.6%, respectively. Similar but lower killing activity against E. coli (70.3 ± 4.2%) by CD8aþ T cells from immunized ginbuna was observed (Fig. 1A). There was a significant difference in the killing activity of CD8aþ T cells between nonimmunized and immunized ginbuna against E. tarda and S. iniae. 3.3.2. CD4þ T cells CD4þ cells from immunized fish were found to kill 100% and 99.6% of E. tarda and S. iniae, respectively. CD4þ T cells also showed higher direct killing activity against E. tarda and S. iniae than against E. coli (21.3 ± 2.4%). Immunized CD4þ T cells showed significantly higher killing activities against E. tarda, S. iniae and E. coli than nonimmunized CD4þ T cells (Fig. 1B). 3.3.3. sIgMþ cells Similar to the results with CD8aþ and CD4þ T cells, sIgMþ cells collected from immunized ginbuna killed 77.5 ± 2.0% of E. tarda and 77.8 ± 1.7% of S. iniae, while the killing activity was lower against E. coli (45.7 ± 1.2%, Fig. 1C). Immunized sIgMþ cells showed

H.M. Tartor et al. / Fish & Shellfish Immunology 39 (2014) 178e184

significantly higher killing activities against E. tarda, S. iniae and E. coli than non-immunized sIgMþ cells. 3.3.4. Neutrophils Neutrophils showed lower antibacterial activities than those of lymphocytes against the bacteria tested, especially against E. tarda. We found significant differences in direct killing activity between

181

immunized and non-immunized neutrophils. Neutrophils from E. tarda-immunized ginbuna killed 11.3 ± 0.5% of E. tarda, 49.6 ± 3.5% of S. iniae and 47.4 ± 2.5% of E. coli, while nonimmunized neutrophils killed only 5.6 ± 2% of E. tarda, 30.3 ± 3% of S. iniae and 26.6 ± 4.5% of E. coli (Fig. 1D). E. tarda

E. tarda

* *

S. iniae E. coli

80 60 40 `

20 0

Immunized with E. tarda

40

20 Non-immunized

Immunized with E. tarda

Bacteria killed (%)

B. 100

*

80 60 40

60 40

20

*

20

C. 100 Non-immunized

Immunized with E. tarda

CD4+ T cells

*

80

*

60

*

*

80

0

Bacteria killed (%)

Bacteria killed (%)

Bacteria killed (%) Bacteria killed (%)

*

100

* Non-immunized

Immunized with E. tarda

CD4+ T cells

80

*

60

40 20

40 0

20

Non-immunized

Immunized with E. tarda +

sIgM cells

0 D.

60

CD8α T cells Non-immunized

B.

C. 100

E. coli

+

CD8α+ T cells

0

S. iniae

80

0

100

Non-immunized

Immunized with E. tarda

sIgM+ cells

80 60

*

*

40 20

0

* Non-immunized

Immunized with E. tarda

D. 100

Bacteria killed (%)

Bacteria killed (%)

A. 100

Bacteria killed (%)

A. 100

80 60 40 20 0

Non-immunized

Immunized with E. tarda

Neutrophils

Neutrophils Fig. 1. Comparison of the direct killing activity of lymphocyte subpopulations against different bacterial targets (E. tarda, S. iniae and E. coli) from immunized and nonimmunized ginbuna crucian carp (n ¼ 3). A) CD8aþ B) CD4þ C) sIgMþ and D) neutrophils. Effector lymphocytes (E, 1  105 cells/well) and target bacteria (1  102 CFU/ well) were co-cultured in 96-well microtiter plate wells at an E:T ratio of 1000:1 for 4 h at 26  C. *Significantly higher killing activity of cells from immunized fish when compared to those from non-immunized fish (p  0.05). The columns with error bars represent the mean of triplicate assays ± standard deviation (SD).

Fig. 2. Comparison of the direct killing activity of lymphocyte subpopulations against different bacterial targets (E. tarda, S. iniae and E. coli) from immunized and nonimmunized ginbuna crucian carp (n ¼ 3) using tissue culture inserts. A) CD8aþ B) CD4þ and C) sIgMþ. Effector lymphocytes (E, 3  105 cells/upper compartment) and target bacteria (3  102 CFU/well lower compartment) were cultured in 24-well microtiter plate wells at an E:T ratio of 1000:1 for 4 h at 26  C. *Significantly higher killing activity of cells from immunized fish compared to those from non-immunized fish (p  0.05). The columns with error bars represent the mean of triplicate assays ± standard deviation (SD).

182

H.M. Tartor et al. / Fish & Shellfish Immunology 39 (2014) 178e184

3.4. Requirement for contact between effector cells and target bacteria for the induction of antibacterial activities We examined the requirement for contact between effector cells and target bacteria using tissue culture inserts and supernatants of effector cells.

Fig. 4. Purity and viability of whole kidney granulocytes separated with Percoll (two density gradients). After separation, 1  106 granulocytes were stained with Propidium iodide (PI), 2.5 mg/ml. Granulocytes were gated on FS & SS dot plot and granulocytes excluding debris were analyzed for staining with PI. Percentages of purity and viability of granulocytes are shown as the mean ± SEM; (n ¼ 3). Bars indicate gating. A) FS-SS patterns of freshly-isolated granulocytes showed purity (93.6% ± 1.4). B) Distribution of fluorescence intensity of PI showed high viability of the isolated granulocytes (97.6% ± 0.8).

3.4.1. Effect of tissue culture inserts In the presence of a membrane separating effector cells from target bacteria, non-immunized and immunized CD8aþ, CD4þ and sIgMþ effector cells showed considerably high killing activities against E. tarda and S. iniae, while quite low killing activity was observed in CD4þ and sIgMþ cells against E. coli. Immunized CD8aþ T cells killed 66.4 ± 3.5%, 72 ± 1.1% and 36.7 ± 0.8% of E. tarda, S. iniae and E. coli, respectively (Fig. 2A), while immunized CD4þ T cells killed 63.4 ± 1.6%, 73.3 ± 1.4% and 8.6 ± 1.4% of E. tarda, S. iniae and E. coli, respectively (Fig. 2B). Immunized sIgMþ cells killed 64.3 ± 3.0%, 47.5 ± 1.8% and 14.0 ± 2.0% of E. tarda, S. iniae and E. coli, respectively (Fig. 2C). Neutrophils collected from immunized fish killed 2.6 ± 2.8%, 9 ± 4.5% and 5.6 ± 3.0% of E. tarda, S. iniae and E. coli, respectively (Fig. 2D).

Fig. 3. Comparison of the killing activity of supernatants from cultured lymphocyte subpopulations from immunized and non-immunized ginbuna crucian carp (n ¼ 3) against different bacterial targets (E. tarda, S. iniae and E. coli). A) CD8aþ B) CD4þ C) sIgMþ and D) Neutrophils. One hundred micro liter of supernatant collected from immunized or non-immunized effector cells (E, 1  106 cells) were incubated in vitro with target bacteria (T, 1  102 CFU) in 96-well microtiter plate wells for 4 h at 26  C. *Significantly higher killing activity of culture supernatants from cells of immunized fish compared to those of non-immunized fish (p  0.05). The columns with error bars represent the mean of triplicate assays ± standard deviation (SD).

3.4.2. Effect of supernatants Supernatants collected from cultures of immunized CD8aþ T cells killed 71.6 ± 2.0% of E. tarda, 78.6 ± 1.5% of S. iniae and 57.33 ± 2.5% of E. coli (Fig. 3A). Supernatants from immunized CD4þ T cells killed 67.1 ± 1.5% of E. tarda, 81.6 ± 1.5% of S. iniae and 50.6 ± 1.1% of E. coli, and sIgMþ cell supernatants exhibited killing activities of 26.3 ± 4%, 71.6 ± 1.5% and 39.2 ± 1.0% against E. tarda, S. iniae and E. coli, respectively (Fig. 3B and C).

H.M. Tartor et al. / Fish & Shellfish Immunology 39 (2014) 178e184

Supernatants from immunized neutrophils killed 35 ± 2.0% of S. iniae and 10.3 ± 1.5% of E. coli, while the killing activity of supernatants from non-immunized neutrophils against E. tarda was very low, 1.6 ± 0.5% (Fig. 3D). 4. Discussion In the present study we found that T cells subpopulations (CD8aþ, CD4þ cells) and sIgMþ cells have direct antibacterial activity at the high E:T ratio of 1000:1, similar to the results reported in our

Fig. 5. Purity of MACS-sorted cell types of kidney leucocytes using monoclonal antibodies (MAbs) against ginbuna CD8a, CD4 and IgM. A) CD8aþ fraction B) CD4þ fraction C) sIgMþ fraction. Percentages of fluorescent positive cells stained with the above MAbs are shown as the mean ± SEM; (n ¼ 3). Bars indicate gating.

183

previous paper with an E:T ratio of 10:1 [9]. Here we demonstrated that contact between effector cells and target bacteria is not required by using cell culture inserts to separate effector cells from target bacteria, and by examining the bactericidal activity of cell culture supernatants. We also found that effector cells from both immunized and non-immunized fish have killing activity against different pathogens. Our results agree with the previous report by Petkovic et al. [18] who described high antifungal activity of non-sensitized CTL and NK cells against Cryptococcus neoformans. In our previous study, however, bactericidal activity was observed only when effector donor fish were immunized more than twice [9]. The major differences between our previous and present experiments are the E:T ratio and incubation period. We conducted killing assays at an E:T ratio of 10:1 with a 2h incubation period in the previous experiment, while we carried out the assay at the much higher E:T ratio of 1000:1 and incubated for four hours in the present experiment. In our previous study, effector cells from non-immunized fish showed low percentages of killing activity (15e20%) against tested pathogens. Furthermore, immunized effector cells showed slightly higher killing activity than the non-immunized cells in the present study. These results suggest that we are looking at magnified activity of nonimmunized effector cells by increasing the number from 10 to 1000 against each target bacterium. In the present study we found that direct bactericidal activity mediated by T cell subsets was non-specific. Our results are consistent with the study by Markham et al. [16] who found that T cells from immune mice produced a lymphokine which has bactericidal activity against a broad range of both gram-negative and gram-positive bacteria after in vitro re-exposure to the immunogen. However, Powderly et al. [19] found that the final mediator of T cells protection in mice in vivo is non-specific in its effect, but this non-specific protection can only be elicited by reexposure to the original immunizing antigen. There is a slight difference in the specificity of killing between the present and our previous study. In the previous study effector cells exhibited around 40% of non-specific killing activity against third party bacteria, although effector cells killed more than 90% of target bacteria used as immunogen. These results suggest that “strong” killing activity induced by sensitized effector cells is specific killing, while “weak” killing activity of non-sensitized effectors is non-specific killing. Li et al. [20] found that rainbow trout B cells have phagocytic and antibacterial activity. In the present study we also found that killing activity was higher in sIgMþ cells and neutrophils coincubated with bacteria than those separated from target bacteria by insert. Lower killing activity was also observed in the supernatants. These findings suggest that phagocytosis is partly involved in the killing by B cells expressing sIgM and neutrophils. It is also possible that NK cells are also included in sIgMþ cells, since NK cells are known to have Fc receptors in channel catfish [21], and antiginbuna IgM MAb is considered to bind sIgM that is bound to NK cells via their Fc receptors. Antibacterial activity of NK cells mediated by granulysin/NK-lyin have been reported in mammals [22,23]. Accordingly, NK cells may be partly responsible for the observed killing in the sIgMþ fraction. Direct microbicidal activity of macrophages is considered to be low, although we have no data about the cytotoxicity of macrophages. Macrophages and neutrophils are considered to take bacteria and kill by phagocytosis, while T lymphocytes such as cytotoxic T cells kill target cells or even bacteria by secreting killing molecules. Therefore, the contact is required for the killing by phagocytic cells such as macrophages and neutrophils. In the present study, however, microbicidal activity of neutrophils was quite low in the supernatant and their killing activity was greatly

184

H.M. Tartor et al. / Fish & Shellfish Immunology 39 (2014) 178e184

suppressed by the presence of membrane, while killing activity of lymphocytes was considerably high. These results suggest that killing mechanism is different between lymphocytes and neutrophils/macrophages, and the contribution of contaminated macrophages to the microbicidal activity of lymphocyte subsets is small in the present experiments. Previous in vitro studies have shown that contact between effector cells and target microorganisms is required to exert antimicrobial activity by human and mice T cells against parasites [24] and fungi [25], although contact was not needed for bacteria [10]. Our result showed that both immunized and non-immunized effectors killed bacterial targets even in the presence of a membrane between effector cells and bacterial targets. These results suggest that contact requirement varies based on the type of microorganism used as target. Using supernatants of effector cell cultures we confirmed that contact between effector cells and target bacteria is not required for direct bactericidal activity of lymphocytes. Our results are consistent with the previous report by Markham et al. [16] who found that supernatants from cultures of murine splenic CD8þ T cells previously exposed in vitro to both macrophages and Pseudomonas aeruginosa non-specifically killed the bacteria in a cell-free environment. It is clear that killing activity is only effective against targets which are close to effector cells, and the concentration of the killing substance must be quite low, as previously reported by Markham et al. [16]. This is evident by our observations that supernatant diluted more than 10 times did not show any killing activity against the pathogens tested, and supernatants from effector cell cultures gave no inhibition zone to tested pathogens in agar well diffusion assays, although spreading supernatant over bacterial lawn cultures inhibited their growth (Data not shown). Furthermore, we found that supernatant containing soluble mediator/s lost its antibacterial activity after heat treatment at 56  C for 15 min. Neutrophils from immunized ginbuna showed higher antibacterial activity than those from non-immunized fish. Ainsworth [26] reported that neutrophils from blood of channel catfish were metabolically activated when challenged with bacterial and nonbacterial stimulants. Nagamura and Wakabayashi [27] reported that the glycogen content of eel neutrophils stimulated with formalin-killed Vibrio anguillarum is higher than those exposed to 2% casein or saline and stimulated neutrophils display enhanced phagocytosis and bactericidal activity. Therefore, it is possible that the higher antibacterial activity of neutrophils from immunized fish can be attributed to their increase in metabolic rate and/or glycogen content following stimulation with formalin-killed bacteria. In conclusion, ginbuna CD8þ T cells, CD4þ T cells and sIgMþ cells showed direct bactericidal activity which is apparently not executed via MHC. The present study conducted at an E:T ratio of 1000:1 showed that direct contact between effectors and targets was not required, and even culture supernatants showed killing activity, suggesting that contact is not needed for this killing. Sensitization of effector donors is not needed and the killing is non-specific. However, in our previous study sensitization was essential to induce killing activity and the killing was partially specific when the assay was conducted at the E:T ratio of 10:1. These findings suggest that at least two mechanisms are involved in the direct bactericidal activity of lymphocytes. One is partially specific and sensitization is required. The other is non-specific and sensitization is not needed. The activity of former is stronger than that of latter since 10 effectors killed one target bacterium in the former, while 1000 effectors were required in the latter. Acknowledgment We gratefully acknowledge the financial support from the Ministry of Higher Education, Egypt. We also thank Dr. Namba

(Department of Marine Science and Resources of our university) for providing expertise in bacterial culture and immunization. References ~ os B, Romalde JL. A review of the main bacterial fish [1] Toranzo AE, Magarin diseases in mariculture systems. Aquaculture 2005;246:37e61. [2] Harikrishnan R, Jin CN, Kim MC, Kim JS, Balasundaram C, Heo MS. Effectiveness and immunomodulation of chemotherapeutants against scuticociliate (Philasterides dicentrarchi) in olive flounder. Exp Parasitol 2010;124(3): 306e14. [3] Brudeseth BE, Wiulsrød R, Fredriksen BN, Lindmo K, Løkling KE, Bordevik M, et al. Status and future perspectives of vaccines for industrialised fin-fish farming. Fish Shellfish Immunol 2013;35(6):1759e68. [4] Nakanishi T, Aoyagi1 K, Xia C, Dijkstra JM, Ototake M. Specific cell-mediated immunity in fish. Vet Immunol Immunopathol 1999;72:101e9. [5] Hahn H, Kaufman SHE. The role of cell-mediated immunity in bacterial infections. Rev Infect Dis 1981;13:1221e50. [6] Zhao Z, Fux B, Goodwin M, Dunay IR, Strong D, Miller BC, et al. Autophagosome-independent essential function for the autophagy protein Atg5 in cellular immunity to intracellular pathogens. Cell Host Microbe 2008, Nov. 13;4(5):458e69. [7] Stenger S, Mazzaccaro RJ, Uyemura K. Differential effects of cytolytic T cell subsets on intracellular infection. Science 1997;276(5319):1684e7. [8] Oykhman P, Mody CH. Direct microbicidal activity of cytotoxic T-lymphocytes. J Biomed Biotechnol; 2010:249482. [9] Nayak SK, Nakanishi T. Direct antibacterial activity of CD8/CD4 T-cells in ginbuna crucian carp, Carassius auratus langsdorfii. Fish Shellfish Immunol 2013;1(34):136e41. [10] Levitz SM, Mathews HL, Murphy JW. Direct antimicrobial activity of T cells. Immunol Today 1995;16(8):387e91. [11] Akashi S, Saitoh S, Wakabayashi Y, Kikuchi T, Takamura N, Nagai Y. Lipopolysaccharide interaction with cell surface toll-like receptor 4-MD-2: higher affinity than that with MD-2 or CD14. J Exp Med 2003;198:1035e42. [12] Toda H, Shibasaki Y, Koike T, Ohtani M, Takizawa F, Ototake M, et al. Alloantigen-specific killing is mediated by CD8-positive T-cells in fish. Dev Comp Immunol 2009;33:646e52. [13] Toda H, Saito Y, Koike T, Takizawa F, Araki K, Yabu T, et al. Conservation of characteristics and functions of CD4 positive lymphocytes in a teleost fish. Dev Comp Immunol 2011;35(6):650e60. [14] Somamoto T, Yoshiura Y, Sato A, Nakao M, Nakanishi T, Okamoto N, et al. Expression profiles of TCRb and CD8a mRNA correlate with virus-specific cell-mediated cytotoxic activity in ginbuna crucian carp. Virology 2006;7: 348e70. [15] Takizawa F, Mizunaga Y, Araki K, Moritomo T, Ototake M, Nakanishi T. GATA3 mRNA in ginbuna crucian carp (Carassius auratus langsdorfii): cDNA cloning, splice variants and expression analysis. Dev Comp Immunol 2008;32(8): 898e907. [16] Markham RB, Goellner J, Pier GB. In vitro T cell mediated killing of Pseudomonas aeruginosa I. Evidence that a lymphokine mediates killing. J Immunol 1984;133(2):962e8. [17] Farouk SE, Mincheva-Nilsson L, Krensky AM, Dieli F, Troye-Blomberg M. gd T cells inhibit in vitro growth of the asexual blood stages of Plasmodium falciparum by a granule exocytosis-dependent cytotoxic pathway that requires granulysin. Eur J Immunol 2004;34:2248e56. [18] Petkovi c L, Arsenijevi c VA, Skvarc M, Kopitar A, Kotnik V, Sim ci c S, et al. Anticryptococcal cytotoxicity of murine nonadherent cells is perforin and nonperforin mediated. Mycologia 2010;102(3):552e9. [19] Powderly WG, Pier GB, Markham RB. Lymphocyte-mediated protection against Pseudomonas aeruginosa infection in granulocytopenic mice. J Clin Invest 1986;78:375e80. [20] Li J, Barreda DR, Zhang YA, Boshra H, Gelman AE, LaPatra S, et al. B lymphocytes from early vertebrates have potent phagocytic and microbicidal abilities. Nat Immunol 2006;7:1116e24. [21] Shen L, Stuge TB, Evenhuis JP, Bengten E, Wilson M, Chinchar VG, et al. Channel catfish NK-like cells are armed with IgM via a putative Fc micro R. Dev Comp Immunol 2003;27:699e714. ~ arrubia P, Koster FT, Kelley RO, McDowell TD, Bankhurst AD. [22] Garcia-Pen Antibacterial activity of human natural killer cells. J Exp Med 1989;169(1): 99e113. [23] Endsley JJ, Endsley MA, Estes DM. Bovine natural killer cells acquire cytotoxic/ effector activity following activation with IL-12/15 and reduce Mycobacterium bovis BCG in infected macrophages. J Leukoc Biol 2006;79(1):71e9. [24] Ellner JJ, Olds GR, Lee CW, Kleinhenz ME, Edmonds KL. Destruction of the multicellular parasite Schistosoma mansoni by T lymphocytes. J Clin Invest 1982;70(2):369e78. [25] Beno DWA, Mathews HL. Growth inhibition of Candida albicans by interleukin-2-activated splenocytes. Infect Immun 1992;60(3):853e63. [26] Ainsworth AJ. Fish granulocytes: morphology, distribution and function. Annu Rev Fish Dis 1992;2:123e48. [27] Nagamura Y, Wakabayashi H. Changes in glycogen content of neutrophils in eel, Anguilla japonica by bacterial infection. Fish Pathol 1985;20(389): 94.