Developmental and Comparative Immunology,Vol. 16, pp. 123-138, 1992 0145-305X/92$5.00 + .00 Printed in the USA. All rights reserved. Copyright © 1992 PergamonPress Ltd.
AGGLUTININ ACTIVITY IN PACIFIC OYSTER (Crassostrea gigas) HEMOLYMPH FOLLOWING IN VIVO Vibrio anguillarum CHALLENGE J a n A. O l a f s e n , * T h e l m a C. Fletcher,'l:l: a n d P a t r i c k T. G r a n t ' l *Department of Marine Biochemistry, The Norwegian College of Fishery Science, University of Tromse, Norway and I"NERC institute of Marine Biochemistry, St. Fittick's Road, Aberdeen AB1 3RA, UK
(Submitted January 1991; Accepted August 1991) OAbstract--Hemolymph from the Pacific oyster (Crassostrea gigas) contains iectins that agglutinate horse (Gigalin E) and human (Gigalin H) erythrocytes. The gigalins also agglutinate bacteria, including Vibrio anguillarum, and were adsorbed from oyster hemolymph at different temperatures by living, heat-killed, and freeze-dried V. anguillarum cells. Baseline activities of the two gigaiins were established by measuring their activities in oyster hemolymph over a period of 4 years. A normal distribution of Gigalin H activity (mean titer 139) w a s found, whereas the distribution of Gigalin E activity in the same samples was skew (mean titer 512). No covariance was observed between the two agglutinin activities. Increased lectin activity above this baseline was found in oysters exposed for varying time intervals to V. anguil/arum at different seasons and temperatures over a period of 2 years. Such exposure resulted in an increase in activity (titer) of four- to ninefold for Gigalin E and three- to seven-fold for Gigaiin H when compared with controls, and in augmentation in the hemolymph of a protein with the same electrophoretic mobility as affinity-purified oyster lectins (gigalins). Challenge with either living or heat-killed bacteria resulted in a significant increase of Gigaiin E activity, whereas results for Gigalin H were variable. Oysters challenged with bacteria were ob-
Address correspondence to Jan A. Olafsen, Department of Marine Biochemistry, The Norwegian College of Fishery Science, University of Troms¢, Dramsveien 201, N-9000 Troms¢, Norway. Present address: Department of Zoology, University of Aberdeen, UK.
served to filter normally with open shells during the experiments. Also, no increase w a s found in hemolymph calcium that could indicate anoxia following bacterial challenge (0.49 - 0.004 mg mL -t) compared to unexposed oysters (0.50 -e 0.001 mg mL-1). Increase in the concentration of free amino acids in oyster hemolymph was observed following exposure to bacteria (15.05 mM) and anaerobiosis (13.51 raM) compared to controls (9.06 mM), and changes (in tool %) of individual amino acids differed considerably between hemolymph from animals challenged with bacteria and animals kept anaerobic. The augmented lectin activity in oyster hemolymph, following in vivo exposure to increased bacteria in the seawater, suggests their involvement in enhancing bacterial clearance and defense in the oyster. OKeywords--Oyster; Agglutinin; Lectin; Induction; Challenge; Clearance; Defense.
Introduction
Bottom-dwelling marine bivalves, such as oysters, live in an environment with an abundant microflora where they feed by filtering and thus accumulate large n u m b e r s o f m i c r o o r g a n i s m s from the seawater. They harbour an exceptionally rich microflora in which Vibrio species predominate (2,3). Healthy bivalves may have bacteria in their soft tissues (4) and h e m o l y m p h (5), and can act as specific carriers of some vibrios (5). In contradiction to this, Vibrio and P s e u d o m o n a s species have been responsible for most
123
124
bacterial diseases affecting marine bivalves (6,7). Adult bivalves are thus normally populated by opportunistic pathogens without contracting disease, but the molecular basis for this interaction is not known. Molluscs appear to have effective systems for clearing invading bacteria from their tissues and hemolymph, following both intracardial injection or natural ingestion of the bacteria (5). The role of hemocytes in such clearance is well documented in bivalves, but the role of humoral factors like lectins remains unresolved. Lectins have been isolated from the hemolymph of most invertebrates, and their possible function in invertebrate defense has been extensively discussed (8-12). Even though most of these lectins were identified by their ability to agglutinate erythrocytes, natural agglutinins for bacteria have also been described in sea hares (13,14), hard clams (15), and oysters (16). It is generally inferred that invertebrate lectins may take part in the recognition and clearance of bacteria by the hemocytes. Such functions may be promoted by humoral lectins or by lectins on the hemocyte surface (17), or perhaps a combination of both. However, we still lack conclusive evidence to demonstrate the role of the different humoral and hemocytebound agglutinins in interaction with bacteria or defense of invertebrates. Agglutinins may increase phagocytosis by acting as opsonins, but the interactions b e t w e e n purified hemolymph lectins and bacterial pathogens have only rarely been investigated. We reported (18) that affinity-purified hemolymph lectins from oyster Crassostrea gigas hemolymph were opsonic, increasing the uptake of bacteria (Vibrio anguillarurn NCMB 6 and Escherichia coil K 235) by oyster hemocytes in vitro, i.e., bacteria preincubated with lectins were ingested three to five times more rapidly than untreated bacteria. Purified Mytilus edulis
J.A. Olafsen et al.
lectins have also been found to stimulate in vitro phagocytosis of yeast cells and erythrocytes by Mytilus hemocytes (19), Humoral factors (like lectins) in marine invertebrates are generally believed to be innate and noninducible. There have been reports that in vitro exposure to bacteria may induce activities of serum and hemocyte-bound enzymes, reviewed in (20), or increase humoral antibacterial activity (21). Increased secondary clearance of bacteria has been reported in sea hares (22), but agglutinin titers were not increased. More specifically, the bacterial cell wall molecules 2-keto-3-deoxy octonate and [3-glycerophosphate resulted in lectin induction in the Indian horseshoe crab, Carcinoscorpius rotundacauda (23). Hemolymph of the Pacific oyster C. gigas contains two erythrocyte lectin activities with the ability to agglutinate horse (Gigalin E) and human (Gigalin H) erythrocytes, respectively (1). Gigalin H was specific for sialic acid, but with much higher affinity for bovine submaxillary mucin (BSM), a glycoprotein with terminal sialic acid. Our demonstration (18) that in vivo exposure of C. gigas to bacteria led to an increase in gigalin activity in the hemolymph suggested its possible involvement in a defense reaction. The present investigation was undertaken to determine if increased levels of the opportunistic marine pathogen V. anguillarurn in the seawater would affect the activity of these erythrocyte agglutinins in oyster hemolymph. Since it has also been reported that V. anguillarurn infections in oysters may result in anoxia (24) and thus increase of calcium in hemolymph (25), the effect of anoxia on agglutinin activity was tested. Also, the effect of ambient t e m p e r a t u r e on hemolymph agglutinin activity was tested. This article de.scribes changes in lectin titers and hemolymph protein patterns resulting from such e n v i r o n m e n t a l changes and from exposure of oysters to
Oyster lectins and bacterial challenge
V. anguillarum, compared to baseline activities of lectins normally found in oyster hemolymph.
125
a 1-mL syringe after carefully cutting the adductor muscle and removing the upper shell. Hemolymph samples of about 0.25-0.5 mL could usually be obtained from one oyster. The individual samples were centrifuged for 2 min in an EppenMaterials a n d Methods doff centrifuge set to 12,000 g to remove Materials and Buffers hemocytes and debris. Larger samples of hemolymph, for some assays, were obBuffers used were modified Mytilus tained by separating the soft tissues from saline (26) (1.01 mM NaH2PO4; 2.35 mM the shells and collecting the supernatant Na2HPO4; 450 mM NaCI; 11.7 mM following centrifugation at 3,000 g. This CaCI2; 7.69 mM KCI; 25 mM MgCI2; 25 supernatant was then centrifuged at mM MgSO4), saline (0.19 M NaCI, pH 22,000 g for 30 min to remove cell debris 7.0), and phosphate buffered saline (0.19 and filtered through a Whatman no. I filM NaCI in 0.02 M phosphate buffer, pH ter paper at 4°C to remove lipid. 7.0). Unless otherwise stated, sodium Affinity chromatography was carried azide to a final concentration of 0.02% out using BSM (Sigma, St Louis, MO) was added to all buffers and tissue fluid conjugated with CNBr-activated Sephafractions. rose-4B (Pharmacia) as previously described (1,12). Centrifuged and filtered hemolymph was applied to a PharmaciaOysters and Hemolymph column containing the BSM-Sepharose and nonretarded material (effluent) colOysters (C. gigas), supplied by the lected. The column was washed thorFisheries Experiment Station of the Min- oughly with 50% Mytilus saline until no istry of Agriculture, Fisheries and Food protein or agglutinin activity could be de(Conwy, North Wales), were transported tected in the effluent. Bound lectins were from their natural beds in Wales and displaced by a linear gradient of increasmaintained in a circulating seawater ing ionic strength (0.5-5 M NaCI, pH aquarium at approximately 12°C. Oys- 7.0) in the eluant buffer. Fractions conters were adapted to aquarium condi- taining the active material of Gigalin E tions for at least 10 d, and not used if and Gigalin H were concentrated using held longer than 4 weeks. During expo- an Amicon U8 ultrafiltration cell with sure, oysters were kept in tanks contain- XMI00A membrane. For lectin subunits, ing 20 L seawater at 15-20°C in a dark an Amicon UM2 membrane was used. room. Oysters were adapted for not less than 72 h to experimental conditions before the bacterial suspensions were added to the seawater in the tanks. Care Experimental Conditions was taken to ensure that the animals were not otherwise disturbed during exThe effects of the ambient seawater periments. temperature and the incubation temperature of the agglutination assay on lectin activity were measured. Oysters were Collection of Hemolymph and kept in seawater at 4 or 20°C and the Purification of Oyster Lectins agglutinin activity of hemolymph from these animals was measured at 4 and Hemolymph was collected directly 20°C as described below. Bacteria used in the experiment were from the heart or pericardial cavity with
126
V. anguillarum NCMB 6 (National Collection of Marine Bacteria No. 6) obtained from Torry Research Station (Aberdeen, UK). Bacteria were grown in a seawater medium with yeast extract (0.3%) and peptone (0.5%). Cells were harvested by centrifugation at 5,000 g for 30 rain, washed twice, and suspended in Mytilus saline. Bacterial counts were made with a Neubauer hemocytometer. Resting cells of V. anguillarum, prepared as described above, were diluted in Mytilus saline to give an optical density of 1.0 (2.0 for freeze dried cells) at 620 nm. Nonviable bacterial cells were prepared by heating a resting cell suspension at 65°C for 1 h and viability tested by plating. Freeze-dried cells were suspended in Mytilus saline and washed by centrifugation before use. The agglutination of bacteria following incubation with hemolymph or affinitypurified oyster lectins was observed microscopically after 15 min and 1 h. To test the adsorption of lectins to bacterial cells, two-fold dilutions of bacteria were made in test tubes and hemolymph added at 1 part: 1 part bacterial dilution. Following incubation, bacteria were removed by centrifugation for 2 rain at 12,000 g before the supernatant was tested for remaining agglutinin activity. Oysters were maintained anoxic by tying the shell tightly closed with metal bands and keeping them in stainless steel anaerobic jars flushed with oxygen-free N2 for 30 min prior to sealing. The jars were kept closed and immersed in seawater tanks at 17°C to regulate the temperature. Calcium was determined in diluted hemolymph samples by atomic absorption s p e c t r o p h o t o m e t r y using a Varian T e c h t r o n M o d e l AA5. Hemolymph samples were prepared for analysis of free amino acids on a JEOL JLG-6AH analyzer as described (27). UItrafiltration of h e m o l y m p h was performed using an Amicon UM 8 (Mr 8,000 cut-off) membrane.
J.A. Olafsen et al.
Hemagglutination Assays Agglutination assays were performed in round-bottom, 80-well plastic trays (Baird and Tatlock, Cheshire, UK). Serial doubling dilutions of the test material (0.1 mL) were made in Mytilus saline and 0.1 mL 2% (v/v) horse or human group-O erythrocytes in 0.19 M NaCI added to each well. Titers (the highest dilution giving positive agglutination) were read after incubation for 1 h at 4°C. Agglutinin activities were read independently by two people without knowledge of the origin of the samples. This method was used for all oysters challenged with bacteria throughout this study. In some experiments, where stated, Mytilus saline was replaced by 0.19 M NaCI for the assay of Gigalin H activity and by 10 mM CaC12 in 0.19 M NaCI for the assay of Gigalin E activity. However, the inherent errors of the tray agglutination assay used to measure lectin activity should be considered. Difficulties in the visual determination of the end-point results in a two-fold effect on recorded activity, and thus deviations from the mean at high and low activity would not be comparable. All statistics quoted in this article were therefore calculated with log2 of titer values and no detectable agglutination (titer < 2) recorded as 0. For comparison of agglutinin activities between groups, results are reported as antilogs (i.e., Table 2).
General Analytical Methods Polyacrylamide gel electrophoresis in homogeneous rod gels was performed according to the method of Ornstein and Davis (28,29), using 8% separating gel with Tris-glycine at pH 8.3, and samples were dialysed against Tris-glycine prior to electrophoresis. Bromophenol blue (Merck) was used as tracking dye, and gels were stained with Coomassie Bril-
Oyster lectins and bacterial challenge
liant blue R250. Samples dialysed against 8 M urea were dialysed against Trisglycine sample buffer, immediately followed by electrophoresis. Statistical analyses and graphical presentations were made using an Apple M a c i n t o s h c o m p u t e r with s o f t w a r e Statview II (Abacus Concepts), Data D e s k P r o f e s s i o n a l ( O d e s t a ) , and MacDraw Pro (Claris), using t-test to compare groups of samples.
Results
Baseline Lectin Activity in C. gigas Oysters (C. gigas) used in this study varied in body weight from 4-25 g. The wet weight of soft tissues removed after centrifugation was 11.8 --+ 5.2 g (mean -+ SD for 40 oysters), and tissue fluid (hemolymph) was 12.9 -+ 8.5 g with 2.4 0.9 mg protein mL-~. Agglutinin activities in oyster hemolymph was monitored over a period of 4 years, and the variation in activities for agglutination of horse and human erythrocytes is demonstrated in Figure 1. Gigalin H activities appeared normally distributed with mean log2 titer (86 samples) of 7.1 --- 1.9. The mean titer (antilog of the mean log2 titer) was 139 with 95% of activities in the interval 104-186. Mean log2 titer of Gigalin E in the same samples was 9.0 +-- 2.2. The mean activity as titer (antilog of mean log2) was 512 with 95% of activities in the interval 368-711. In contrast to Gigalin H activities, the distribution of Gigalin E activities among the sampled animals appeared skewed (Fig. 1) with a negative kurtosis (fiat distribution) of - 1 . 2 as compared to - 0 . 2 for Gigalin H. This suggested that Gigalin E activity between the sampled oysters could be separated into populations with high and low activity. Each hemolymph sample was pooled from at least 30 animals, thus counteracting the effect of extreme indi-
127
viduai biological variation. In Figure 2, hemolymph titers (log2 titer -+ SD) are categorized by month. There was a slight tendency to increased activity of Gigalin E during the second half of the year, but almost overruled by high individual variation. This would be in agreement with the contention that generally serum components increase with food intake and are subsequently mobilized during gonad maturation (30,31). However, a specific relationship of increased activity with gonad development is unlikely since the seasonal variation was not predominant (Fig. 2). C. gigas hemolymph agglutinates a variety of cell types, including vertebrate red blood cells (human, horse, dog, rat, cow, sheep, rabbit, and plaice), some algae, and b a c t e r i a (V. anguillarum NCMB6 and E. coli K235). Affinitypurified gigalins also agglutinated these bacteria. No difference in agglutination of bacteria by hemolymph or affinitypurified oyster lectins was observed by microscopical examination. Agglutination of resting cells of V. anguillarum was microscopically visible within 15 rain, but was more marked, with large clumps containing many cells, after incubation for 1 h. Since V. anguillarum also autoagglutinates, the test for agglutination was only confirmative and the agglutinating activity or titer was not measured or scored. The hemolymph components responsible for agglutination of erythrocytes also appeared to be involved in agglutination of the bacteria, as judged by the adsorption of hemagglutinating activity of oyster hemolymph both by resting, heat-killed, and also freeze-dried cells of V. anguillarum (Table 1). The agglutinins reacting with horse and human erythrocytes (Gigalin E and H) were removed with the bacterial cells by centrifugation, and the remaining activity was inversely proportional to the number of bacteria added. It could appear that the adsorp-
128
J.A. Olafsen et aL
181 i 16i
A
14" 12 E
10 8
6 4 2
F/Z,
0 2
4
6 8 10 Activity GigalinE (log2 titer)
2
4
6 8 10 Activity GigatinH (log2 titer)
12
14
18 16 14 12 E
g
U.
10
6 4 2 0
12
14
Figure 1. Variation of lectin activities in different samples of C. gigas hemolymph. Graphs demonstrate frequency counts (occurrences) at discrete activities (x-axis) for 86 hemolymph samples through a period of 4 years. All oysters were from the same origin, but were kept in different tanks and at different temperatures. (A), Gigalin E; (B), Gigalin H.
tion of Gigalin E (at least for heat-killed bacteria) was more pronounced, also since this activity in the hemolymph was initially higher than that of Gigalin H. Only marginal differences were observed in activities of oyster agglutinins assayed at different temperatures. After keeping oysters for 1 week at either 4 or
20°C, no differences in agglutinin activity could be observed when assayed at the respective temperatures. However, agglutinin activities in hemolymph from oysters kept at 20°C when assayed at 20°C (Gigalin E, 984; Gigalin H, 360: mean of 12 animals) were higher than when assayed at 4°C (Gigalin E, 144; Gigalin H, 36). For oysters kept at 4°C, no
Oyster lectins and bacterial challenge
129
14
12 10
Feb(5)
Mar(7)
Alx (17)
May(5)
Jun (13) Jul (20) Aug Sept(23) Oct (8) Nov (8) Dec(20) Month Figure 2. Lectin activity in oyster hemolymph sampled at different times of the year over 4 years. Values represent mean log2 titer --- SD of Gigalin E and Gigalin H. Numbers in brackets are number of samples, each sample representing at least 30 animals.
difference in activity assayed at either 4 or 20°C could be observed. It thus seemed that incubation temperatures had little effect on agglutinin activities, but agglutinins in oysters adapted to higher temperatures were possibly sensitive to temperature decrease.
In Vivo Exposure of C. gigas to V. anguillarum Since h e m o l y m p h and affinitypurified gigalins were adsorbed to bacterial cells we decided to assay the activities of these agglutinins in oysters that had been exposed to V. anguillarum in vivo. In an initial experiment involving two groups of oysters (8 controls and 4 experimental animals) kept for 7 d at 22°C, Gigalin E activities were 3,042 (control) and 10,318 and 11,585 for oysters exposed to bacteria for 2 and 7 d, respectively. The results for Gigalin H were 116 for controls and 203 after 2 d and 430 after 7 d. When oysters were
moved from the aquarium to the experimental tanks, a transient increase in agglutinin titer could be observed. This effect probably resulted from disturbance due to handling of the animals since similar effects have been reported following disturbance or mechanical injury (23). In all experiments reported below, oysters were adapted for not less than 72 h to experimental conditions before the bacterial suspensions were added to the seawater in the tank and care was taken to ensure that the animals were not further disturbed during the experiment. Oysters were exposed in vivo to bacteria at different times of the year during a period of 2 years (Table 2). The ratio of activities (expressed as antilogs) in exposed vs. control oysters demonstrated a four to nine times increase in activity for Gigalin E. Similarly, the increase for Gigalin H was three to seven times. The activity (logz titer) of the challenged oysters were significantly higher (p ~< 0.05) than control oysters in all experiments except one, although sample sizes were
130
J.A. Olafsen et al.
Table 1. Remaining agglutlnln activity (tlter) In oyster (C. g/gas) hemolymph following adsorption with V. anguillarum cells.
A* Hemolymph + saline, control Hemolymph + resting cells of V. anguillarum
Activity Remaining in Hemolymph (titer)
Dilution of Bacteria
Gigalin E
Gigalin H
4,096 1,024 2,048 4,096
64
1"1 1:2 1:4
512
32
8
8 8
8 16 32
Bt Hemolymph + saline, control Hemolymph + heatkilled V. anguillarum
1:1 1:2 1:4 1:8 1:16
16 64 128 512
16 32 64
c$ Hemolymph + saline, control Hemolymph + freezedried V. anguillarum
64 16 32 64 64
1:1 1:2 1:4 1:8
32
8 16 16 32
All activities were measured in duplicate. * Adsorption with resting cells of V. anguillarum for 12 h at 4°C. t Adsorption with heat-killed cells of V. anguillarum for 2 h at 20°C. ~: Adsorption with freeze-dried cells of V. anguillarum for 2 h at 20°C.
small. This observed increase was unlikely to result from activity of a live pathogen since oysters responded in a similar way to live and heat-killed cells of V. anguillarum (Fig. 3). Following challenge with both live and heat-killed bacteria, Gigalin E was significantly higher (p ~< 0.05) than controls. For Gigalin H, exposure to heat-killed bacteria resulted in a more marked increase in titer than did live bacteria. The concentration of calcium in oyster hemolymph was measured following exposure to V. anguillarum and also in oysters kept anoxic. No increase of calcium in hemolymph was detected following exposure of oysters to bacteria (Fig. 4), with values of 0.49 _ 0.004 mg m L compared with 0.50 +- 0.001 mg m L - ~in controls. This would indicate that the increase in agglutinin titres for both Gigalin E and H (p
AGGLUTININ ACTIVITY IN PACIFIC OYSTER (Crassostrea gigas) HEMOLYMPH FOLLOWING IN VIVO Vibrio anguillarum CHALLENGE J a n A. O l a f s e n , * T h e l m a C. Fletcher,'l:l: a n d P a t r i c k T. G r a n t ' l *Department of Marine Biochemistry, The Norwegian College of Fishery Science, University of Tromse, Norway and I"NERC institute of Marine Biochemistry, St. Fittick's Road, Aberdeen AB1 3RA, UK
(Submitted January 1991; Accepted August 1991) OAbstract--Hemolymph from the Pacific oyster (Crassostrea gigas) contains iectins that agglutinate horse (Gigalin E) and human (Gigalin H) erythrocytes. The gigalins also agglutinate bacteria, including Vibrio anguillarum, and were adsorbed from oyster hemolymph at different temperatures by living, heat-killed, and freeze-dried V. anguillarum cells. Baseline activities of the two gigaiins were established by measuring their activities in oyster hemolymph over a period of 4 years. A normal distribution of Gigalin H activity (mean titer 139) w a s found, whereas the distribution of Gigalin E activity in the same samples was skew (mean titer 512). No covariance was observed between the two agglutinin activities. Increased lectin activity above this baseline was found in oysters exposed for varying time intervals to V. anguil/arum at different seasons and temperatures over a period of 2 years. Such exposure resulted in an increase in activity (titer) of four- to ninefold for Gigalin E and three- to seven-fold for Gigaiin H when compared with controls, and in augmentation in the hemolymph of a protein with the same electrophoretic mobility as affinity-purified oyster lectins (gigalins). Challenge with either living or heat-killed bacteria resulted in a significant increase of Gigaiin E activity, whereas results for Gigalin H were variable. Oysters challenged with bacteria were ob-
Address correspondence to Jan A. Olafsen, Department of Marine Biochemistry, The Norwegian College of Fishery Science, University of Troms¢, Dramsveien 201, N-9000 Troms¢, Norway. Present address: Department of Zoology, University of Aberdeen, UK.
served to filter normally with open shells during the experiments. Also, no increase w a s found in hemolymph calcium that could indicate anoxia following bacterial challenge (0.49 - 0.004 mg mL -t) compared to unexposed oysters (0.50 -e 0.001 mg mL-1). Increase in the concentration of free amino acids in oyster hemolymph was observed following exposure to bacteria (15.05 mM) and anaerobiosis (13.51 raM) compared to controls (9.06 mM), and changes (in tool %) of individual amino acids differed considerably between hemolymph from animals challenged with bacteria and animals kept anaerobic. The augmented lectin activity in oyster hemolymph, following in vivo exposure to increased bacteria in the seawater, suggests their involvement in enhancing bacterial clearance and defense in the oyster. OKeywords--Oyster; Agglutinin; Lectin; Induction; Challenge; Clearance; Defense.
Introduction
Bottom-dwelling marine bivalves, such as oysters, live in an environment with an abundant microflora where they feed by filtering and thus accumulate large n u m b e r s o f m i c r o o r g a n i s m s from the seawater. They harbour an exceptionally rich microflora in which Vibrio species predominate (2,3). Healthy bivalves may have bacteria in their soft tissues (4) and h e m o l y m p h (5), and can act as specific carriers of some vibrios (5). In contradiction to this, Vibrio and P s e u d o m o n a s species have been responsible for most
123
124
bacterial diseases affecting marine bivalves (6,7). Adult bivalves are thus normally populated by opportunistic pathogens without contracting disease, but the molecular basis for this interaction is not known. Molluscs appear to have effective systems for clearing invading bacteria from their tissues and hemolymph, following both intracardial injection or natural ingestion of the bacteria (5). The role of hemocytes in such clearance is well documented in bivalves, but the role of humoral factors like lectins remains unresolved. Lectins have been isolated from the hemolymph of most invertebrates, and their possible function in invertebrate defense has been extensively discussed (8-12). Even though most of these lectins were identified by their ability to agglutinate erythrocytes, natural agglutinins for bacteria have also been described in sea hares (13,14), hard clams (15), and oysters (16). It is generally inferred that invertebrate lectins may take part in the recognition and clearance of bacteria by the hemocytes. Such functions may be promoted by humoral lectins or by lectins on the hemocyte surface (17), or perhaps a combination of both. However, we still lack conclusive evidence to demonstrate the role of the different humoral and hemocytebound agglutinins in interaction with bacteria or defense of invertebrates. Agglutinins may increase phagocytosis by acting as opsonins, but the interactions b e t w e e n purified hemolymph lectins and bacterial pathogens have only rarely been investigated. We reported (18) that affinity-purified hemolymph lectins from oyster Crassostrea gigas hemolymph were opsonic, increasing the uptake of bacteria (Vibrio anguillarurn NCMB 6 and Escherichia coil K 235) by oyster hemocytes in vitro, i.e., bacteria preincubated with lectins were ingested three to five times more rapidly than untreated bacteria. Purified Mytilus edulis
J.A. Olafsen et al.
lectins have also been found to stimulate in vitro phagocytosis of yeast cells and erythrocytes by Mytilus hemocytes (19), Humoral factors (like lectins) in marine invertebrates are generally believed to be innate and noninducible. There have been reports that in vitro exposure to bacteria may induce activities of serum and hemocyte-bound enzymes, reviewed in (20), or increase humoral antibacterial activity (21). Increased secondary clearance of bacteria has been reported in sea hares (22), but agglutinin titers were not increased. More specifically, the bacterial cell wall molecules 2-keto-3-deoxy octonate and [3-glycerophosphate resulted in lectin induction in the Indian horseshoe crab, Carcinoscorpius rotundacauda (23). Hemolymph of the Pacific oyster C. gigas contains two erythrocyte lectin activities with the ability to agglutinate horse (Gigalin E) and human (Gigalin H) erythrocytes, respectively (1). Gigalin H was specific for sialic acid, but with much higher affinity for bovine submaxillary mucin (BSM), a glycoprotein with terminal sialic acid. Our demonstration (18) that in vivo exposure of C. gigas to bacteria led to an increase in gigalin activity in the hemolymph suggested its possible involvement in a defense reaction. The present investigation was undertaken to determine if increased levels of the opportunistic marine pathogen V. anguillarurn in the seawater would affect the activity of these erythrocyte agglutinins in oyster hemolymph. Since it has also been reported that V. anguillarurn infections in oysters may result in anoxia (24) and thus increase of calcium in hemolymph (25), the effect of anoxia on agglutinin activity was tested. Also, the effect of ambient t e m p e r a t u r e on hemolymph agglutinin activity was tested. This article de.scribes changes in lectin titers and hemolymph protein patterns resulting from such e n v i r o n m e n t a l changes and from exposure of oysters to
Oyster lectins and bacterial challenge
V. anguillarum, compared to baseline activities of lectins normally found in oyster hemolymph.
125
a 1-mL syringe after carefully cutting the adductor muscle and removing the upper shell. Hemolymph samples of about 0.25-0.5 mL could usually be obtained from one oyster. The individual samples were centrifuged for 2 min in an EppenMaterials a n d Methods doff centrifuge set to 12,000 g to remove Materials and Buffers hemocytes and debris. Larger samples of hemolymph, for some assays, were obBuffers used were modified Mytilus tained by separating the soft tissues from saline (26) (1.01 mM NaH2PO4; 2.35 mM the shells and collecting the supernatant Na2HPO4; 450 mM NaCI; 11.7 mM following centrifugation at 3,000 g. This CaCI2; 7.69 mM KCI; 25 mM MgCI2; 25 supernatant was then centrifuged at mM MgSO4), saline (0.19 M NaCI, pH 22,000 g for 30 min to remove cell debris 7.0), and phosphate buffered saline (0.19 and filtered through a Whatman no. I filM NaCI in 0.02 M phosphate buffer, pH ter paper at 4°C to remove lipid. 7.0). Unless otherwise stated, sodium Affinity chromatography was carried azide to a final concentration of 0.02% out using BSM (Sigma, St Louis, MO) was added to all buffers and tissue fluid conjugated with CNBr-activated Sephafractions. rose-4B (Pharmacia) as previously described (1,12). Centrifuged and filtered hemolymph was applied to a PharmaciaOysters and Hemolymph column containing the BSM-Sepharose and nonretarded material (effluent) colOysters (C. gigas), supplied by the lected. The column was washed thorFisheries Experiment Station of the Min- oughly with 50% Mytilus saline until no istry of Agriculture, Fisheries and Food protein or agglutinin activity could be de(Conwy, North Wales), were transported tected in the effluent. Bound lectins were from their natural beds in Wales and displaced by a linear gradient of increasmaintained in a circulating seawater ing ionic strength (0.5-5 M NaCI, pH aquarium at approximately 12°C. Oys- 7.0) in the eluant buffer. Fractions conters were adapted to aquarium condi- taining the active material of Gigalin E tions for at least 10 d, and not used if and Gigalin H were concentrated using held longer than 4 weeks. During expo- an Amicon U8 ultrafiltration cell with sure, oysters were kept in tanks contain- XMI00A membrane. For lectin subunits, ing 20 L seawater at 15-20°C in a dark an Amicon UM2 membrane was used. room. Oysters were adapted for not less than 72 h to experimental conditions before the bacterial suspensions were added to the seawater in the tanks. Care Experimental Conditions was taken to ensure that the animals were not otherwise disturbed during exThe effects of the ambient seawater periments. temperature and the incubation temperature of the agglutination assay on lectin activity were measured. Oysters were Collection of Hemolymph and kept in seawater at 4 or 20°C and the Purification of Oyster Lectins agglutinin activity of hemolymph from these animals was measured at 4 and Hemolymph was collected directly 20°C as described below. Bacteria used in the experiment were from the heart or pericardial cavity with
126
V. anguillarum NCMB 6 (National Collection of Marine Bacteria No. 6) obtained from Torry Research Station (Aberdeen, UK). Bacteria were grown in a seawater medium with yeast extract (0.3%) and peptone (0.5%). Cells were harvested by centrifugation at 5,000 g for 30 rain, washed twice, and suspended in Mytilus saline. Bacterial counts were made with a Neubauer hemocytometer. Resting cells of V. anguillarum, prepared as described above, were diluted in Mytilus saline to give an optical density of 1.0 (2.0 for freeze dried cells) at 620 nm. Nonviable bacterial cells were prepared by heating a resting cell suspension at 65°C for 1 h and viability tested by plating. Freeze-dried cells were suspended in Mytilus saline and washed by centrifugation before use. The agglutination of bacteria following incubation with hemolymph or affinitypurified oyster lectins was observed microscopically after 15 min and 1 h. To test the adsorption of lectins to bacterial cells, two-fold dilutions of bacteria were made in test tubes and hemolymph added at 1 part: 1 part bacterial dilution. Following incubation, bacteria were removed by centrifugation for 2 rain at 12,000 g before the supernatant was tested for remaining agglutinin activity. Oysters were maintained anoxic by tying the shell tightly closed with metal bands and keeping them in stainless steel anaerobic jars flushed with oxygen-free N2 for 30 min prior to sealing. The jars were kept closed and immersed in seawater tanks at 17°C to regulate the temperature. Calcium was determined in diluted hemolymph samples by atomic absorption s p e c t r o p h o t o m e t r y using a Varian T e c h t r o n M o d e l AA5. Hemolymph samples were prepared for analysis of free amino acids on a JEOL JLG-6AH analyzer as described (27). UItrafiltration of h e m o l y m p h was performed using an Amicon UM 8 (Mr 8,000 cut-off) membrane.
J.A. Olafsen et al.
Hemagglutination Assays Agglutination assays were performed in round-bottom, 80-well plastic trays (Baird and Tatlock, Cheshire, UK). Serial doubling dilutions of the test material (0.1 mL) were made in Mytilus saline and 0.1 mL 2% (v/v) horse or human group-O erythrocytes in 0.19 M NaCI added to each well. Titers (the highest dilution giving positive agglutination) were read after incubation for 1 h at 4°C. Agglutinin activities were read independently by two people without knowledge of the origin of the samples. This method was used for all oysters challenged with bacteria throughout this study. In some experiments, where stated, Mytilus saline was replaced by 0.19 M NaCI for the assay of Gigalin H activity and by 10 mM CaC12 in 0.19 M NaCI for the assay of Gigalin E activity. However, the inherent errors of the tray agglutination assay used to measure lectin activity should be considered. Difficulties in the visual determination of the end-point results in a two-fold effect on recorded activity, and thus deviations from the mean at high and low activity would not be comparable. All statistics quoted in this article were therefore calculated with log2 of titer values and no detectable agglutination (titer < 2) recorded as 0. For comparison of agglutinin activities between groups, results are reported as antilogs (i.e., Table 2).
General Analytical Methods Polyacrylamide gel electrophoresis in homogeneous rod gels was performed according to the method of Ornstein and Davis (28,29), using 8% separating gel with Tris-glycine at pH 8.3, and samples were dialysed against Tris-glycine prior to electrophoresis. Bromophenol blue (Merck) was used as tracking dye, and gels were stained with Coomassie Bril-
Oyster lectins and bacterial challenge
liant blue R250. Samples dialysed against 8 M urea were dialysed against Trisglycine sample buffer, immediately followed by electrophoresis. Statistical analyses and graphical presentations were made using an Apple M a c i n t o s h c o m p u t e r with s o f t w a r e Statview II (Abacus Concepts), Data D e s k P r o f e s s i o n a l ( O d e s t a ) , and MacDraw Pro (Claris), using t-test to compare groups of samples.
Results
Baseline Lectin Activity in C. gigas Oysters (C. gigas) used in this study varied in body weight from 4-25 g. The wet weight of soft tissues removed after centrifugation was 11.8 --+ 5.2 g (mean -+ SD for 40 oysters), and tissue fluid (hemolymph) was 12.9 -+ 8.5 g with 2.4 0.9 mg protein mL-~. Agglutinin activities in oyster hemolymph was monitored over a period of 4 years, and the variation in activities for agglutination of horse and human erythrocytes is demonstrated in Figure 1. Gigalin H activities appeared normally distributed with mean log2 titer (86 samples) of 7.1 --- 1.9. The mean titer (antilog of the mean log2 titer) was 139 with 95% of activities in the interval 104-186. Mean log2 titer of Gigalin E in the same samples was 9.0 +-- 2.2. The mean activity as titer (antilog of mean log2) was 512 with 95% of activities in the interval 368-711. In contrast to Gigalin H activities, the distribution of Gigalin E activities among the sampled animals appeared skewed (Fig. 1) with a negative kurtosis (fiat distribution) of - 1 . 2 as compared to - 0 . 2 for Gigalin H. This suggested that Gigalin E activity between the sampled oysters could be separated into populations with high and low activity. Each hemolymph sample was pooled from at least 30 animals, thus counteracting the effect of extreme indi-
127
viduai biological variation. In Figure 2, hemolymph titers (log2 titer -+ SD) are categorized by month. There was a slight tendency to increased activity of Gigalin E during the second half of the year, but almost overruled by high individual variation. This would be in agreement with the contention that generally serum components increase with food intake and are subsequently mobilized during gonad maturation (30,31). However, a specific relationship of increased activity with gonad development is unlikely since the seasonal variation was not predominant (Fig. 2). C. gigas hemolymph agglutinates a variety of cell types, including vertebrate red blood cells (human, horse, dog, rat, cow, sheep, rabbit, and plaice), some algae, and b a c t e r i a (V. anguillarum NCMB6 and E. coli K235). Affinitypurified gigalins also agglutinated these bacteria. No difference in agglutination of bacteria by hemolymph or affinitypurified oyster lectins was observed by microscopical examination. Agglutination of resting cells of V. anguillarum was microscopically visible within 15 rain, but was more marked, with large clumps containing many cells, after incubation for 1 h. Since V. anguillarum also autoagglutinates, the test for agglutination was only confirmative and the agglutinating activity or titer was not measured or scored. The hemolymph components responsible for agglutination of erythrocytes also appeared to be involved in agglutination of the bacteria, as judged by the adsorption of hemagglutinating activity of oyster hemolymph both by resting, heat-killed, and also freeze-dried cells of V. anguillarum (Table 1). The agglutinins reacting with horse and human erythrocytes (Gigalin E and H) were removed with the bacterial cells by centrifugation, and the remaining activity was inversely proportional to the number of bacteria added. It could appear that the adsorp-
128
J.A. Olafsen et aL
181 i 16i
A
14" 12 E
10 8
6 4 2
F/Z,
0 2
4
6 8 10 Activity GigalinE (log2 titer)
2
4
6 8 10 Activity GigatinH (log2 titer)
12
14
18 16 14 12 E
g
U.
10
6 4 2 0
12
14
Figure 1. Variation of lectin activities in different samples of C. gigas hemolymph. Graphs demonstrate frequency counts (occurrences) at discrete activities (x-axis) for 86 hemolymph samples through a period of 4 years. All oysters were from the same origin, but were kept in different tanks and at different temperatures. (A), Gigalin E; (B), Gigalin H.
tion of Gigalin E (at least for heat-killed bacteria) was more pronounced, also since this activity in the hemolymph was initially higher than that of Gigalin H. Only marginal differences were observed in activities of oyster agglutinins assayed at different temperatures. After keeping oysters for 1 week at either 4 or
20°C, no differences in agglutinin activity could be observed when assayed at the respective temperatures. However, agglutinin activities in hemolymph from oysters kept at 20°C when assayed at 20°C (Gigalin E, 984; Gigalin H, 360: mean of 12 animals) were higher than when assayed at 4°C (Gigalin E, 144; Gigalin H, 36). For oysters kept at 4°C, no
Oyster lectins and bacterial challenge
129
14
12 10
Feb(5)
Mar(7)
Alx (17)
May(5)
Jun (13) Jul (20) Aug Sept(23) Oct (8) Nov (8) Dec(20) Month Figure 2. Lectin activity in oyster hemolymph sampled at different times of the year over 4 years. Values represent mean log2 titer --- SD of Gigalin E and Gigalin H. Numbers in brackets are number of samples, each sample representing at least 30 animals.
difference in activity assayed at either 4 or 20°C could be observed. It thus seemed that incubation temperatures had little effect on agglutinin activities, but agglutinins in oysters adapted to higher temperatures were possibly sensitive to temperature decrease.
In Vivo Exposure of C. gigas to V. anguillarum Since h e m o l y m p h and affinitypurified gigalins were adsorbed to bacterial cells we decided to assay the activities of these agglutinins in oysters that had been exposed to V. anguillarum in vivo. In an initial experiment involving two groups of oysters (8 controls and 4 experimental animals) kept for 7 d at 22°C, Gigalin E activities were 3,042 (control) and 10,318 and 11,585 for oysters exposed to bacteria for 2 and 7 d, respectively. The results for Gigalin H were 116 for controls and 203 after 2 d and 430 after 7 d. When oysters were
moved from the aquarium to the experimental tanks, a transient increase in agglutinin titer could be observed. This effect probably resulted from disturbance due to handling of the animals since similar effects have been reported following disturbance or mechanical injury (23). In all experiments reported below, oysters were adapted for not less than 72 h to experimental conditions before the bacterial suspensions were added to the seawater in the tank and care was taken to ensure that the animals were not further disturbed during the experiment. Oysters were exposed in vivo to bacteria at different times of the year during a period of 2 years (Table 2). The ratio of activities (expressed as antilogs) in exposed vs. control oysters demonstrated a four to nine times increase in activity for Gigalin E. Similarly, the increase for Gigalin H was three to seven times. The activity (logz titer) of the challenged oysters were significantly higher (p ~< 0.05) than control oysters in all experiments except one, although sample sizes were
130
J.A. Olafsen et al.
Table 1. Remaining agglutlnln activity (tlter) In oyster (C. g/gas) hemolymph following adsorption with V. anguillarum cells.
A* Hemolymph + saline, control Hemolymph + resting cells of V. anguillarum
Activity Remaining in Hemolymph (titer)
Dilution of Bacteria
Gigalin E
Gigalin H
4,096 1,024 2,048 4,096
64
1"1 1:2 1:4
512
32
8
8 8
8 16 32
Bt Hemolymph + saline, control Hemolymph + heatkilled V. anguillarum
1:1 1:2 1:4 1:8 1:16
16 64 128 512
16 32 64
c$ Hemolymph + saline, control Hemolymph + freezedried V. anguillarum
64 16 32 64 64
1:1 1:2 1:4 1:8
32
8 16 16 32
All activities were measured in duplicate. * Adsorption with resting cells of V. anguillarum for 12 h at 4°C. t Adsorption with heat-killed cells of V. anguillarum for 2 h at 20°C. ~: Adsorption with freeze-dried cells of V. anguillarum for 2 h at 20°C.
small. This observed increase was unlikely to result from activity of a live pathogen since oysters responded in a similar way to live and heat-killed cells of V. anguillarum (Fig. 3). Following challenge with both live and heat-killed bacteria, Gigalin E was significantly higher (p ~< 0.05) than controls. For Gigalin H, exposure to heat-killed bacteria resulted in a more marked increase in titer than did live bacteria. The concentration of calcium in oyster hemolymph was measured following exposure to V. anguillarum and also in oysters kept anoxic. No increase of calcium in hemolymph was detected following exposure of oysters to bacteria (Fig. 4), with values of 0.49 _ 0.004 mg m L compared with 0.50 +- 0.001 mg m L - ~in controls. This would indicate that the increase in agglutinin titres for both Gigalin E and H (p
