In VitroActivities in Mussel Hemocytes as Biomarkers of Environmental Quality: A Case Study in the Abra Estuary (Biscay Bay)

ECOTOXICOLOGY AND ENVIRONMENTAL SAFETY ARTICLE NO. 0108

35, 253–260 (1996)

In Vitro Activities in Mussel Hemocytes as Biomarkers of Environmental Quality: A Case Study in the Abra Estuary (Biscay Bay) M. P. CAJARAVILLE,1 I. OLABARRIETA,

AND

I. MARIGOMEZ

Zitologi eta Histologi Laborategia, Biologia Zelularra eta Zientzia Morfologikoen Saila, Euskal Herriko Unibertsitatea, 644 PK, E-48080 Bilbo, Basque Country, Spain Received January 9, 1996

Mussels, Mytilus galloprovincialis, were collected from six coastal sites of dissimilar water quality (Zierbena, Santurtzi, Arrigunaga, Galea, Men˜akoz, and Plentzia) at Biscay Bay in September 1991, January 1992, June 1992, and September 1992. The extent of hemocyte infiltration in connective tissue of the digestive gland was quantified by stereology on histological sections in terms of volume density of hemocytes (HVD). HVD was elevated in mussels collected from Plentzia (the less polluted site) in January 1992 and September 1992, while such increases occurred in January 1992 in Santurtzi and Arrigunaga and in September 1991 and September 1992 in Galea. Conversely, HVD was reduced in Arrigunaga in September 1991 and in Galea in January 1992. Moreover, HVD was kept unchanged through the year in mussels collected from Men˜akoz and Zierbena. On the basis of this preliminary in vivo study, hemocytic activities of mussels collected in September 1994 from Arrigunaga and Plentzia were further investigated by means of four in vitro immunotoxicity assays: (a) the trypan blue exclusion assay, indicative of cell viability; (b) the zymosan phagocytosis assay, indicative of phagocytic activity; (c) the diaminobenzidine–manganese (DAB–Mn2+) oxidation assay for estimating reactive oxygen intermediate (ROI) production; and (d) the neutral red (NR) uptake assay, indicative of endocytic ability. These in vitro tests indicated some significant differences between Plentzia and Arrigunaga. Hemocytes from mussels collected in Plentzia exhibited a higher capability to phagocytose zymosan while, conversely, hemocytes from mussels collected in Arrigunaga endocytosed more NR and produced more ROI under nonstimulated conditions. These differences in the in vitro hemocytic activities of mussels from Plentzia (nonpolluted) and Arrigunaga (moderately polluted) suggest that in vitro assays may be used as biomarkers of environmental quality in coastal and estuarine areas. © 1996 Academic Press

INTRODUCTION

In vitro toxicology has bloomed during very recent years and the examination of cell cultures has partially replaced the routine sacrifice of experimental animals (Fry, 1993). Such a trend

1

To whom correspondence should be addressed.

is also noted in the field of environmental toxicology and toxicity tests with fishes or invertebrates are currently being substituted by in vitro testing of the toxicity of chemicals (Alvarez and Friedl, 1992; Zahn and Braunbeck, 1993; Seibert et al., 1994). Moreover, the in vitro testing approach could be also valuable for environmental monitoring programs based on the use of sentinel organisms, which could provide primary cultures of target cells whose functional responses could be used as measures of environmental quality (Anderson, 1988; Dunier and Siwicki, 1993). Bivalve molluscs and specially mussels are widely used as sentinels in monitoring programs of marine environmental quality (Goldberg, 1986). Chemical analysis of their tissues, measurement of biomarkers of exposure to specific chemicals, and measurement of pollution effects are increasingly performed in environmental monitoring programs (Widdows and Donkin, 1989; Rainbow, 1993). In most cases, sentinel mussels are killed to conduct these analyses and relatively large samples are required to attain feasible conclusions on water quality. Alternatively, in vitro assays might provide useful biomarkers of either exposure to pollutants or their biological effect (Anderson, 1993). Particularly, some functional responses of hemocytes and brown cells of molluscs have been proposed to develop in vitro toxicity assays indicative of the condition of sentinel mussels (Cheng, 1988a,b; Zaroogian et al., 1992; Zaroogian and Yevich, 1993) and, therefore, of the environmental quality of their habitat. Although the field of immunotoxicology of bivalve molluscs was in its infancy one decade ago (Anderson, 1988), it has grown up during the past years due to its interest for fisheries, aquaculture, and ecotoxicology (Anderson, 1993). Cell viability, phagocytic and endocytic activities, lysosomal enzyme activities, production of oxygen free radicals, and relative proportion of hemocyte types have been quantified in vitro after in vivo exposure to chemical pollutants (McCormick-Ray, 1987; Cheng, 1988a,b; 1990; Sami et al., 1992; Coles et al., 1994). Nevertheless, different chemicals or even different doses of the same chemical may provoke different hemocytic responses which, additionally, might be modified by either environmental factors or concurrence of other chemicals (Anderson, 1993; Auffret and Oubella, 1994).

253 0147-6513/96 $18.00 Copyright © 1996 by Academic Press All rights of reproduction in any form reserved.

254

CAJARAVILLE, OLABARRIETA, AND MARIGOMEZ

Therefore, the objective of the present investigation was to determine whether in vitro hemocytic activities in mussels living in environments of different water quality were significantly dissimilar and therefore could provide cost-effective biomarkers of environmental quality based on nondestructive analytical procedures. A main problem to face was the natural plasticity of the immune system of mussels. The numbers of circulating hemocytes in bivalves may change during the annual life cycle as well as a result of changes in environmental salinity, temperature, or food availability (McCormick-Ray, 1987; Santarem et al., 1992; Auffret and Oubella, 1994). It is not clear whether these changes are due to hemopoiesis or to the fact that not all hemocytes are found in the systemic circulation, since bivalves have open circulatory systems and cells may migrate from tissues to the circulatory system or vice versa (Auffret and Oubella, 1994). In fact, enhanced infiltration of hemocytes in connective tissues of various organs has been described in several marine molluscs under stress conditions (Seiler and Morse, 1988; Cajaraville et al., 1990a,b; Marigo´mez et al., 1990). The reproductive cycle seems to be a major factor governing the abundance of circulating hemocytes in bivalves (McCormick-Ray, 1987). Therefore, characterization of the reproductive cycle of the sentinel populations is an unavoidable requisite before immunotoxicity assays can be performed. Thus, in a preliminary study, mussels were collected through 1 year in various localities with different levels of pollution at Biscay Bay (North Iberian Peninsula; Soto et al., 1995) and the reproductive cycle of each population was characterized (Ogueta et al., 1995). The present study reports the results on the hemocyte infiltration of connective tissues as a measure of hemocyte abundance in relation to seasonal changes and environmental quality. For this, the volume density of hemocytes was measured by stereology on histological sections of the digestive gland tissue. Afterward, two localities exhibiting a similar pattern of seasonal variation in hemocyte abundance and a distinct degree of environmental contamination were selected for a comparative study of in vitro hemocyte activities. Various in vitro immunotoxicity assays were developed and applied to measure cell viability, phagocytic activity, production of oxyradicals, and endocytic activity as likely biomarkers of the immunological condition of mussels and hence of the quality of their surrounding environment. MATERIALS AND METHODS

Sentinel Species and Study Area Mussels, Mytilus galloprovincialis, of 3.5–4.5 cm shell length were used as experimental animals. Mussels were sampled from five different sites (Zierbena, Santurtzi, Arrigunaga, Galea, and Men˜akoz) at Abra estuary and from Plentzia (Biscay Bay, Iberian Peninsula, Fig. 1). Mussels from these six localities exhibit significantly dissimilar reproductive cycles (Ogueta et al., 1995).

FIG. 1. Location of the sampling sites within and outside the Abra estuary in the vicinity of Bilbo (Biscay Bay). Curved arrows show the main water currents within the estuary. Arrowed angle illustrates the main wind component in the study area.

The six localities are characterized by similar seawater salinities (around 32–33‰) and temperatures (average winter minimum around 10°C and summer maximum around 20°C). However, the degree of exposure to wave beating differs between localities. Men˜akoz is an open stony beach which receives strong wave beating. Plentzia is a sandy protected area where wave beating is weak. Intermediate levels of exposure to wave beating occur in the remaining four localities. Different levels of pollution from moderate to low have been reported in the study area (unpublished data and Soto et al., 1995). The highest average values of metal concentration (mg/g flesh wt) in the tissues of mussels occur in Galea (Cd: 2.2; Cr: 2.6) and Arrigunaga (Cd: 1.4; Cr: 2.2) and the lowest ones in Plentzia (Cd: 0.7; Cr: 1.2; Soto et al., 1995). In March 1992, the highest tissue concentration of polycyclic aromatic hydrocarbons (PAH) was recorded in Galea and Arrigunaga (total PAHs: 1.3 mg/g), while lower PAH tissue concentrations were found in mussels from Plentzia and Santurtzi (<0.5 mg/g; unpublished data). In addition, other previous reports discussed by Marigo´mez et al. (1996) indicate the presence of conspicuous levels of PCBs in the Abra estuary. In Vivo Studies Ten mussels were collected in September 1991, January 1992, June 1992, and September 1992 from Zierbena, Santurtzi, Arrigunaga, Galea, Men˜akoz, and Plentzia (Fig. 1). Digestive glands were excised in situ, placed in Bouin’s fixative at 4°C, and transferred immediately to the laboratory. Fixation was continued at 4°C for 36 hr. Then, tissue samples were dehydrated in alcohols, cleared in methylbenzoate, rinsed in benzene, and embedded in paraffin. Histological sections (7 mm) were cut in a Leitz 1512 microtome and stained with hematoxylin–eosin. The volume density of hemocytes (HVD) was estimated by point counting with the aid of a Weibel graticule (Multipurpose Test System M-42) (Weibel, 1979) by using a drawing-tube attachment on a Nikon Optiphot microscope (Marigo´mez et al., 1990). Eight randomly selected fields

255

HEMOCYTIC RESPONSES IN MUSSELS AS BIOMARKERS

(150,000 mm2) were measured per mussel at a final magnification of 1100×. The number of graticule points lying on either hemocytes (Xh) or connective tissue plus hemocytes (m) was recorded. According to Weibel (1979), HVD (volume of hemocytes per interstitial connective tissue unit) was calculated as HVD 4 Xh/m. Statistical differences of the changes in HVD were tested using a two-way analysis of variance to detect the effects of the sampling month (M), the site (S), and the interaction M × S. Significant differences were established at P < 0.05. Duncan’s test for multiple range comparison was used to detect significant differences between pairs of means (Sokal and Rohlf, 1979). The analyses were carried out with a 486 personal computer with the SPSS/PC+ statistical package (SPSS Inc., Microsoft Co.). In Vitro Studies Animals and hemolymph collection. Five mussels, M. galloprovincialis, of 3.5–4.5 cm shell length were collected at low tide from Plentzia (clean site) and Arrigunaga (moderately polluted site) in Biscay Bay (Fig. 1) each at the 12th (first sampling) and the 22nd (second sampling) of September 1994. Animals were transferred to the laboratory and used immediately for in vitro tests. The hemolymph of mussels was withdrawn immediately before use from the posterior adductor muscle with disposable plastic syringes. Viability test. A hemolymph aliquot was put on a hemocytometer to determine total cell counts and cell viability by the trypan blue exclusion assay (Ford and Haskin, 1988). Preparation of hemocyte monolayers. Aliquots of 1–2 × 106 living cells per milliliter were put on round glass coverslips (10 mm in diameter) located in 24-well microtiter plates. Cells were allowed to adhere and spread for 45 min at room temperature. Relative abundance of hemocyte types. After hemocyte spreading the hemolymph was discarded and adhered cells were stained with Giemsa. Cells were then rinsed in saline (2.5% NaCl, 7.5% sucrose), fixed in 2.5% glutaraldehyde, washed again in saline, and mounted in glycerine. Hemocytes were classified into two categories, granulocytes and hyalinocytes, according to previous studies (Cajaraville and Pal, 1995; Cajaraville et al., 1995) by scoring at least 200 cells randomly from each mussel preparation. Phagocytosis assays. Other sets of hemocyte monolayers were incubated with zymosan (1:10 cell/zymosan ratio) for 90 min. Suspensions of zymosan A (Sigma) were prepared according to Dikkeboom et al. (1987). After incubation, cells were rinsed in saline, fixed in 2.5% glutaraldehyde, washed again in saline, and mounted as above.

Detection of oxyradicals by the DAB–Mn2+ oxidation procedure. The generation of superoxide [a reactive oxygen intermediate (ROI); Anderson, 1993] was measured by the procedure developed by Briggs et al. (1986) based on manganesedependent diaminobenzidine (DAB) oxidation. This procedure was tested with nonstimulated and zymosan-stimulated hemocytes. In the latter case, after incubating cells with zymosan, they were washed in saline and incubated in the cytochemical medium containing 2.5 mM DAB, 0.5 mM Mn2+, and 1 mM NaN3 in saline at pH 7 at room temperature for 45 min. The incubation medium was prepared just before use. After incubation, cells were fixed in 2.5% glutaraldehyde, washed, and mounted as above for examination under light microscope. The extent of oxidized DAB was evaluated by counting the percentage of reactive cells (brown-colored cells) under light microscope (Olympus BX50, 400×). Every experiment was made in duplicate. At least 200 cells were examined at random from each preparation (one per mussel) and the number of hemocytes containing phagocytosed zymosan particles and/or brown oxidized DAB was recorded (Cajaraville et al., 1994). Neutral red uptake. This test was carried out according to Babich and Borenfreund (1992). A sample of 200 ml hemolymph per mussel was placed in each well of a 96-well culture microtiter plate. After 45 min, hemolymph was discarded and adhered hemocytes were incubated with 0.004% neutral red (NR) solution in saline. After 3 hr of incubation, the remaining NR solution was discarded. Then, cells were washed with saline and the NR incorporated into viable cells was released into the supernatant with 200 ml of an acetic acid (1%)–ethanol (50%) solution. Absorbance of NR was recorded at 540 nm in a Biotek EL-312 microplate reader. Differences between in vitro activities of hemocytes of mussels from the two sites were tested using Student’s t statistical test (a 4 0.05). RESULTS

In Vivo Studies Figure 2 presents the values of HVD in digestive gland tissues of mussels collected from six localities in Biscay Bay at four sampling periods. According to the two-way analysis of variance, significant seasonal changes occurred in HVD (Table 1). On the other hand, Duncan’s test revealed significant differences between sampling sites in September 1991, January 1992, and September 1992, but, however, differences between sites were not significant in June 1992. HVD values were high in mussels collected from Plentzia (the less polluted site) in January 1992 and September 1992, while such high HVD values were found in January 1992 in Santurtzi and Arrigunaga and in September 1991 and September 1992 in Galea. Conversely, HVD was low in Arrigunaga in September 1991 and in Galea in January 1992. Finally, HVD was kept unchanged through the year in mussels collected from Men˜akoz and Zierbena.

256

CAJARAVILLE, OLABARRIETA, AND MARIGOMEZ

FIG. 2. Hemocyte volume density (VD) in the interstitial connective tissue of the digestive gland, calculated on histological sections of mussels collected at six sampling sites (Zierbena, Santurtzi, Arrigunaga, Galea, Men˜akoz, and Plentzia) at four sampling periods (September 1991, January 1992, June 1992, and September 1992). Results are given as means of 10 animals and vertical segments represent confidence intervals (95%).

In Vitro Studies According to the trypan blue assay, no significant difference (Student’s t test, P > 0.05) was found in hemocyte viability between mussels collected from Plentzia and Arrigunaga (Fig. 3a). Similarly, no significant difference between the first and the second sampling was found in hemocyte viability (Fig. 3a). However, although the variation coefficient of viability (standard deviation × 100/mean) did not change between samplings, significantly lower values were recorded in Arrigunaga than in Plentzia. This indicates that mussels from Plentzia comprise a more heterogeneous population than mussels from Arrigunaga in terms of hemocyte viability (Fig. 3b). No significant difference was noted in any case in granulocyte/hyalinocyte ratios and in the relative occurrence of both hemocyte types (Fig. 4). Significant differences were found in the phagocytic index (percentage of phagocytic hemocytes), hemocytes of mussels TABLE 1 Summary of the Two-Way Analysis of Variance Performed to Detect Effects of the Month (M), Site (S), and M × S Interaction on Hemocyte Volume Density in the Interstitial Connective Tissue of Mussel Digestive Gland Source

Degrees of freedom

Month Site M × S interaction Remainder Total

3 5 15 176 199

* Significant effect: P(F) < 0.05.

F ratio

P(F ratio)

5.782 1.925 4.223

0.001* 0.092 <0.001*

collected at Plentzia exhibiting a higher capability to phagocytose zymosan than those of mussels collected at Arrigunaga on both sampling days (Fig. 5a). However, differences were only significant at the first sampling. According to the DAB– Mn2+ oxidation assay, significant differences were recorded in the capability to produce ROI by nonstimulated hemocytes (Fig. 5b). This capability was higher in mussels from Arrigunaga than in mussels from Plentzia, although only statistically significant (Student’s t test, P < 0.05) on the first sampling day (Fig. 5b). When hemocytes were incubated with zymosan, the release of ROI was increased in both localities at both sampling days and no differences between sites were detected (Fig. 5c). Neutral red uptake was higher in hemocytes from Arrigunaga mussels but the difference between the two sites was statistically significant only in the second sampling, possibly due to the large variability demonstrated by hemocytes from Arrigunaga mussels in the first sampling (Fig. 5d). DISCUSSION

It has been reported that hemocyte numbers vary on a seasonal basis, which could be due to changes in food availability and metabolite mobilization toward the reproductive tissue (Auffret and Oubella, 1994). Hemocytic lysosomal enzyme activities also experience seasonal changes, being higher in summer than in winter, in relation with the activation of the metabolism in summer (Huffman and Tripp, 1982; Santarem et al., 1992). Generally, in molluscs the number of circulating hemocytes decreases with starvation (Auffret and Oubella, 1994), exposure to pollutants (McCormick-Ray, 1987; Suresh and Mohandas, 1990), and in winter season (Auffret and Oubella, 1994) and increases with age (Dikkeboom et al.,

HEMOCYTIC RESPONSES IN MUSSELS AS BIOMARKERS

257

FIG. 3. (a) Viability (%) of hemocytes (trypan blue assay) in mussels collected from Arrigunaga (A) and Plentzia (P) on the 12th (first sampling) and 22nd (second sampling) of September 1994. Results are given as means of five animals ± standard deviations. (b) Estimated variation coefficient (%) for hemocyte viability (standard deviation × 100/mean).

1984) and as gonad development proceeds (Auffret and Oubella, 1994). On the other hand, migration of cells involved in immune defense (hemocytes and brown cells) to connective tissues has been described in molluscs on exposure to pollutants (Seiler and Morse, 1988; Marigo´mez et al., 1990; Cajaraville et al., 1992; Zaroogian and Yevich, 1993) and at the end of the reproductive cycle in relation to either gonad resorption or reserve mobilization (McCormick-Ray, 1987; Cajaraville et al., 1990a; Ogueta et al., 1995). In the present work, a seasonal pattern was found in the extent of hemocyte infiltration of digestive gland connective tissue. In mussels from all localities studied except Galea, maximum peaks were evidenced in January 1992 and Septem-

FIG. 4. Relative abundance (%) of hemocyte types in mussels collected from Arrigunaga (A) and Plentzia (P) on the 12th (first sampling) and 22nd (second sampling) of September 1994. G, granulocytes; H, hyalinocytes. Results are given as means of five animals ± standard deviations.

FIG. 5. In vitro hemocyte activities of mussels collected from Arrigunaga (A) and Plentzia (P) on the 12th (first sampling) and 22nd (second sampling) of September 1994. (a) Phagocytic index based on the zymosan uptake assay. (b) ROI production test based on the DAB–Mn2+ assay in nonstimulated cells. (c) ROI production test based on the DAB–Mn2+ assay in zymosan-stimulated cells. (d) Endocytic index based on the neutral red assay. Results are given as means of five animals ± standard deviations. Significantly dissimilar mean values according to the Student’s t test are indicated by asterisks (P < 0.05).

ber 1992 and minimum peak values in June 1992 and September 1991. This pattern is exactly the contrary of that described by Auffret and Oubella (1994) for the circulating hemocytes of Crassostrea gigas and Ruditapes philippinarum from Northern Biscay Bay, consisting of minimum peaks in winter and maximum peaks in June and September. Thus, it could be hypothesized that seasonal changes in hemocyte numbers are a result of the differential partition of hemocytes between systemic circulation and connective tissues, rather than a result of changes in hemopoietic activities. Nevertheless, this interesting hypothesis needs future confirmation by specifically targeted experimental works. The seasonal pattern of connective tissue infiltrating hemocytes described above was attenuated in Zierbena and Men˜akoz. Mussels from these localities are not contaminated with toxic metals (Soto et al., 1995) since they do not receive the direct impact of the industrial area around Bilbo (Abra estuary, Biscay Bay). According to various data including the biometric

258

CAJARAVILLE, OLABARRIETA, AND MARIGOMEZ

characteristics of mussels, the pattern of their reproductive cycle, the extent of reserve storage tissue, the prevalence of parasites, and various biomarkers of biological effect (Ogueta et al., 1995; Soto et al., 1995; Marigo´mez et al., 1996), water pollution in these two localities can be classified as moderately low and would unlikely affect infiltrating hemocyte numbers. On the other hand, both sites are rocky shores exposed to strong wave beating and therefore contribution of hemocytes to continuous shell repair (Bubel et al., 1977) could account for attenuated variations in cell numbers in digestive gland connective tissues. In contrast, Galea exhibited a completely different pattern of seasonal changes in HVD in interstitial connective tissue of the digestive gland. The immune system of mussels from Galea appears to work in a very different manner, which could be related to moderately high levels of organic chemical pollution (Marigo´mez et al., 1996). Accordingly, the numbers of circulating and infiltrating hemocytes vary significantly in mussels on exposure to organic pollutants (McCormick-Ray, 1987; Cajaraville et al., 1990a,b, 1992). In addition, other factors like parasites (Ford, 1988; Mourton et al., 1992; Ford et al., 1993) increase infiltrating hemocyte numbers and, interestingly, high parasite prevalences were found in Galea in a preliminary screening (unpublished observations). Therefore, further investigations will be especially conducted to elucidate the particular immunological state of mussels from Galea. Mussels from Santurtzi, Arrigunaga, and Plentzia presented a similar seasonal pattern for HVD, as described before. The low HVD values recorded in Santurtzi in September 1992 and in Arrigunaga in September 1991 might be related to the secondary peaks of spawning activity described by Ogueta et al. (1995) in both months. Among these three localities, Plentzia could be considered the cleanest and Arrigunaga the most polluted (Soto et al., 1995; Marigo´mez et al., 1996) and thus these two localities were selected for further in vitro studies of hemocytic activities. These in vitro investigations were carried out in September 1994 in order to use mussels not suffering from reproductive stress. It has been reported that hemocyte viability in vitro is depressed in polluted or stressed bivalve molluscs (Cheng, 1990; Alvarez and Friedl, 1992). Currently no difference in cell viability was found between Plentzia and Arrigunaga mussels in any of the replicate samplings. However, the variability in cell viability was very different, indicating that the population of mussels from Arrigunaga is less heterogeneous than that of mussels from Plentzia. Nevertheless, up to now consideration was made of the different types of hemocytes altogether, but at least two different types of hemocytes have been described in M. galloprovincialis (Cajaraville and Pal, 1995), as in other bivalve molluscs (Foley and Cheng, 1975; Rodrick and Ulrich, 1984; Auffret, 1989): granulocytes and hyalinocytes. Mussel granulocytes are phagocytic cells, while no definite function has been attributed to hyalinocytes (Auffret, 1989; Anderson, 1993; Cajaraville et

al., 1994; Cajaraville and Pal, 1995). Some reports indicate that the relative proportion of both cell types may change with either reproduction or pollution-induced stress (McCormickRay, 1987; Seiler and Morse, 1988; Cheng, 1990; Sami et al., 1992) and this could contribute to the different range of variation currently found in cell viability. However, mussels from Plentzia and Arrigunaga exhibited similar relative proportions of granulocytes and hyalinocytes. In addition, it has been reported that the granulocyte/hyalinocyte ratio increases on exposure to some metallic pollutants but decreases on exposure to others (Cheng, 1990), and thus, at present, this index would hardly indicate differences in water quality between both sites. One could think that no differences in hemocytic functions such as phagocytic or endocytic activities can be recorded if the proportion between granulocytes and hyalinocytes is kept constant, but, however, increased phagocytic activity has also been described when relative numbers of granulocytes decreased and vice versa (Cheng, 1990). These results indicate that the overall phagocytic capacity of hemocytes may be independent of the cell type composition of the immune system. Thus, in the present study the percentage of phagocytic hemocytes was higher in mussels sampled at Plentzia. On the other hand, mussels collected at Arrigunaga showed a higher capability to endocytose neutral red as well as to produce oxygen radicals under nonstimulated conditions. It has been demonstrated that in vivo exposure to metals and organic pollutants (McCormick-Ray, 1987; Cheng, 1990) and changes in environmental factors like salinity and temperature have influence on the mobility and phagocytic activity of molluscan hemocytes (Fisher and Tramplin, 1988; Ford and Haskin, 1988; Fisher et al., 1989). Nevertheless, since contradictory results have been obtained after in vitro exposure (Cheng and Sullivan, 1984) or after short-term in vivo exposures to either metals or organic compounds (Cheng, 1988a,b; Alvarez and Friedl, 1992), more research is needed to understand the effects of pollutants and environmental changes on the phagocytic responses of hemocytes in bivalves. At present, it can be assumed that in vivo exposure to pollutants depresses phagocytic activity in hemocytes (Anderson, 1993) and, therefore, the lower phagocytic index recorded in Arrigunaga than in Plentzia would be consistent with a higher level of environmental pollution in the former locality, in agreement with previous studies (Soto et al., 1995; Marigo´mez et al., 1996). On the other hand, NR uptake was higher in Arrigunaga than in Plentzia, which is apparently a contradictory result. NR assay is based on the incorporation of the cationic dye into the lysosomes of viable cells by pinocytosis or passive transport across the plasma membrane (Borenfreund and Babich, 1993). NR uptake is known to decrease in brown cells of the clam Mercenaria mercenaria exposed to pollutants (Zaroogian et al., 1992). Generally, the reduction in NR uptake after in vivo exposure to pollutants might result from cell damage or death (Borenfreund and Babich, 1993). In the present study, hemocytes from Arrigunaga and Plentzia did not exhibit dissimilar

HEMOCYTIC RESPONSES IN MUSSELS AS BIOMARKERS

abundance and viability rates and accordingly NR uptake was not lower in the polluted site than in the clean one. Conversely, phagocytosis was impaired to some extent in Arrigunaga as shown by zymosan uptake assays. The impairment of phagocytosis in hemocytes of mussels from this locality might result in an enhancement of alternative defense mechanisms such as endocytosis. Thus, increases in NR uptake could be related to a situation of moderate pollution not causing serious damage in hemocytic plasma membrane or cell death. In fact, hemocytes seemed to be even more active in defense and detoxification processes in the polluted than in the clean site, since production of ROIs was higher in Arrigunaga than in Plentzia. The effect of environmental pollutants on ROI production by bivalve hemocytes depends greatly on the nature and concentration of the pollutant (Anderson, 1993), but recent data indicate that exposure to environmentally relevant concentrations of the model polycyclic aromatic hydrocarbon fluoranthene stimulates the release of ROIs in mussels (Coles et al., 1994). CONCLUSIONS

In view of the present results, it is concluded that in vitro activities of the hemocytes of sentinel mussels can be useful biomarkers for marine pollution-monitoring programs. For this purpose, trypan blue exclusion, zymosan phagocytosis, DAB– Mn2+ oxidation, and NR uptake assays are valuable tests of low cost and rapid performance. The NR assay is also particularly well suited to in vitro toxicity testing since it is conducted in 96-well microtiter plates that can be read in only 5 sec, therefore allowing a large number of samples to be tested in a single day. Nevertheless, the present conclusions are not definitive since the immune defense mechanisms in marine molluscs are still not fully understood. One important factor seems to be seasonality since significant changes in the numbers of hemocytes infiltrating interstitial connective tissues occur through the annual cycle. In addition, the seasonal pattern is significantly altered in localities exhibiting high pollution levels (i.e., Galea) and therefore this effect provides itself a clear indication of low environmental quality. On the other hand, differences in seasonal patterns between sites must be carefully considered when performing intersite comparisons of in vitro hemocytic responses. ACKNOWLEDGMENTS This study has been supported by CICYT (AMB 93–0432). I. Olabarrieta was the recipient of grants from ‘‘Fundacio´n Caja de Madrid’’ and Education, Universities and Research Department of the Basque Government.

REFERENCES Alvarez, M. R., and Friedl, F. E. (1992). Effect of a fungicide on in vitro hemocyte viability, phagocytosis and attachment in the American oyster, Crassostrea virginica. Aquaculture 107, 135–140.

259

tal stressors. In Pathology of Marine and Estuarine Organisms (J. A. Couch and J. W. Fournie, Eds.), pp. 483–510. CRC Press, Boca Raton, FL. Auffret, M. (1989). Comparative study of the hemocytes of two oysters species: the European flat oyster, Ostrea edulis Linnaeus, 1750, and the Pacific oyster, Crassostrea gigas (Thunberg, 1973). J. Shellfish Res. 8, 367–373. Auffret, M., and Oubella, R. (1994). Cytometric parameters of bivalve molluscs: Effect of environmental factors. In Modulators of Fish Immune Responses. SOS Publ. USA 1, 23–32. Babich, H., and Borenfreund, E. (1992). Neutral red assay for toxicology in vitro. In In Vitro Methods of Toxicology (R. R. Watson, Ed.), pp. 39–66. CRC Press, Boca Raton, FL. Borenfreund, E., and Babich, H. (1993). Neutral red (NR) assay. In Cell and Tissue Culture: Laboratory Procedures (J. B. Griffiths, A. Doyle, and J. W. Newell, Eds.), pp. 4B: 7.1–7. 7. Wiley, Sussex, England. Briggs, R. T., Robinson, J. M., Karnovsky, M. L., and Karnovsky, M. J. (1986). Superoxide production by polymorphonuclear leukocytes. A cytochemical approach. Histochemistry 84, 371–378. Bubel, A., Moore, M. N., and Lowe, D. M. (1977). Cellular responses to shell damage in Mytilus edulis L. J. Exp. Mar. Biol. Ecol. 30, 1–27. Cajaraville, M. P., Marigo´mez, I., and Angulo, E. (1990a). Short-term toxic effects of 1-naphthol on the digestive gland–gonad complex of the marine prosobranch Littorina littorea (L.): A light microscopic study. Arch. Environ. Contam. Toxicol. 19, 17–24. Cajaraville, M. P., Marigo´mez, I., and Angulo, E. (1990b). Ultrastructural study of the short-term toxic effects of naphthalene on the kidney of the marine prosobranch Littorina littorea. J. Invert. Pathol. 55, 215–224. Cajaraville, M. P., Marigo´mez, I., and Angulo, E. (1992). Comparative effects of the water accommodated fraction of three oils on mussels. 1. Survival, growth and gonad development. Comp. Biochem. Physiol. 102C, 103–112. Cajaraville, M. P., Marigo´mez, I., and Olabarrieta, I. (1994). Phagocytosis and production of oxyradicals in molluscan blood cells or haemocytes. 36th Symposium of the Society of Histochemistry, Heidelberg, Germany. Cajaraville, M. P., and Pal, S. G. (1995). Morphofunctional study of the haemocytes of the bivalve mollusc Mytilus galloprovincialis with emphasis on the endolysosomal compartment. Cell Struct. Funct. 20, 355–367. Cajaraville, M. P., Pal, S. G., and Robledo, Y. (1995). Light and electron microscopical localization of lysosomal acid hydrolases in bivalve haemocytes by enzyme cytochemistry. Acta Histochem. Cytochem. 28, 409–416. Cheng, T. C. (1988a). In vivo effects of heavy metals on cellular defense mechanisms of Crassostrea virginica: Total and differential cell counts. J. Invert. Pathol. 51, 207–214. Cheng, T. C. (1988b). In vivo effects of heavy metals on cellular defense mechanisms of Crassostrea virginica: Phagocytic and endocytotic indices. J. Invert. Pathol. 51, 215–220. Cheng, T. C. (1990). Effects of in vivo exposure of Crassostrea virginica to heavy metals on hemocyte viability and activity levels of lysosomal enzymes. Pathol. Mar. Sci., 513–524. Cheng, T. C., and Sullivan, J. T. (1984). Effects of heavy metals on phagocytosis by molluscan hemocytes. Mar. Environ. Res. 14, 305–315. Coles, J. A., Farley, S. R., and Pipe, R. K. (1994). Effects of fluoranthene on the immunocompetence of the common marine mussel, Mytilus edulis. Aquat. Toxicol. 30, 367–379. Dikkeboom, R., van der Knaap, W. P. W., Meuleman, E. A., and Sminia, T. (1984). Differences between blood cells of juvenile and adult specimens of the pond snail Lymnaea stagnalis. Cell Tissue Res. 238, 43–47.

Anderson, R. S. (1988). Effects of anthropogenic agents on bivalve cellular and humoral defense mechanisms. Am. Fish. Soc. 18, 238–242.

Dikkeboom, R., Tijnagel, J. M. G. H., Mulder, E. C., and van der Knaap, W. P. W. (1987). Hemocytes of the pond snail Lymnaea stagnalis generate reactive forms of oxygen. J. Invert. Pathol. 49, 321–331.

Anderson, R. S. (1993). Modulation of nonspecific immunity by environmen-

Dunier, M., and Siwicki, A. K. (1993). Effects of pesticides and other organic

260

CAJARAVILLE, OLABARRIETA, AND MARIGOMEZ

pollutants in the aquatic environment on immunity of fish: A review. Fish Shellfish Immunol. 3, 423–438. Fisher, W. S., Chintala, M. M., and Moline, M. A. (1989). Annual variation of estuarine and oceanic oyster Crassostrea virginica Gm‘lin hemocyte capacity. J. Exp. Mar. Biol. Ecol. 127, 105–120. Fisher, W. S., and Tramplin, M. (1988). Environmental influences on activities and foreign particle binding by hemocytes of American oysters, Ostrea edulis. Can. J. Fish. Aquat. Sci. 45, 1309–1315. Foley, D. A., and Cheng, T. C. (1975). A quantitative study of phagocytosis by hemolymph cells of the pelecypods Crassostrea virginica and Mercenaria mercenaria. J. Invert. Pathol. 25, 189–197. Ford, S. E. (1988). Host–parasite interactions in eastern oysters selected for resistance to Haplosporidium nelsoni (MSX) disease: Survival mechanism against a natural pathogen. Am. Fish. Soc. Sp. Publ. 18, 206–224. Ford, S. E., Ashton-Alcox, K. A., and Kanaley, A. (1993). In vitro interactions between bivalve hemocytes and the oyster pathogen Haplosporidium nelsoni (MSX). J Parasitol. 79, 255–265. Ford, S. E., and Haskin, H. H. (1988). Comparison of in vitro salinity tolerance of the oyster parasite, Haplosporidium nelsoni (MSX) and hemocytes from the host, Crassostrea virginica. Comp. Biochem. Physiol. 90A, 183–187. Fry, J. R. (1993). Development in in vitro toxicology. Comp. Haematol. Int. 3, 4–7. Goldberg, E. D. (1986). The mussel watch concept. Environ. Monit. Assess. 7, 91–103. Huffman, J. E., and Tripp, M. R. (1982). Cell types and hydrolytic enzymes of soft shell clam (Mya arenaria) hemocytes. J. Invert. Pathol. 40, 68–74. Marigo´mez, I., Cajaraville, M. P., and Angulo, E. (1990). Histopathology of the digestive gland/gonad complex of the marine prosobranch Littorina littorea exposed to cadmium. Dis. Aquat. Org. 9, 229–238. Marigo´mez, I., Orbea, A., Olabarrieta, I., Etxeberria, M., and Cajaraville, M. P. (1996). Structural changes in the digestive lysosomal system of sentinel mussels as biomarkers of environmental stress in ‘‘Mussel-Watch’’ programmes. Comp. Biochem. Physiol. 113C, 291–297. McCormick-Ray, M. G. (1987). Hemocytes of Mytilus edulis affected by Prudhoe Bay crude oil emulsion. Mar. Environ. Res. 22, 107–122. Mourton, C., Boulo, V., Chagot, D., Hervio, D., Bachere, E., Mialhe, E., and Grizel, H. (1992). Interactions between Bonamia ostreae (Protozoa: Ascetospora) and hemocytes of Ostrea edulis and Crassostrea gigas (Mollusca: Bivalvia): In vitro system establishment. J. Invert. Pathol. 59, 235–240. Ogueta, S., Go´mez, B. J., Serrano, T., and Soto, M. (1995). Evaluacio´n del desarrollo gonadal y de las reservas energn˜ticas del manto del mejillo´n Mytilus galloprovincialis en relacio´n al nivel de metales pesados biodis-

ponibles presentes en áreas industrializadas. V Congr. Nac. Acuicul., Sant Carles de la Ra`pita, Catalonia. Rainbow, P. S. (1993). The significance of trace metals concentrations in marine invertebrates. In Ecotoxicology of Metals in Invertebrates (R. Dallinger and P. S. Rainbow, Eds.), pp. 3–23. Lewis, Boca Raton, FL. Rodrick, G. E., and Ulrich, S. A. (1984). Microscopical studies on the hemocytes of bivalves and their phagocytic interaction with selected bacteria. Helgol. Meeresunters. 37, 167–176. Sami, S., Faisal, M., and Huggett, R. J. (1992). Alterations in cytometric characteristics of hemocytes from the American oyster Crassostrea virginica exposed to a polycyclic aromatic hydrocarbon (PAH) contaminated environment. Mar. Biol. 113, 247–252. Santarem, M. M., Figueras, A. J., Robledo, J. A. F., and Caldas, J. R. (1992). Variation of the defence mechanisms in two groups of mussels, Mytilus galloprovincialis Lmk. Seasonal and environmental effects—Preliminary results. Aquaculture 107, 185–188. Seibert, H., Gu¨lden, M., and Voss, J.-U. (1994). An in vitro toxicity testing strategy for the classification and labelling of chemicals according to their potential acute lethal potency. Toxicol. in Vitro 8, 847–850. Seiler, G. R., and Morse, M. P. (1988). Kidney and hemocytes of Mya arenaria (Bivalvia): Normal and pollution-related ultrastructural morphologies. J. Invert. Pathol. 52, 201–214. Sokal, R. R., and Rohlf, F. J. (1979). Biometría. Blume, Madrid. Soto, M., Kortabitarte, M., and Marigo´mez, I. (1995). Bioavailable heavy metals in estuarine waters as assessed by metal/shell-weight indices in sentinel mussels, Mytilus galloprovincialis. Mar. Ecol. Prog. Ser. 125, 127– 136. Suresh, K., and Mohandas, A. (1990). Effect of sublethal concentrations of copper on hemocyte number in bivalves. J. Invert. Pathol. 55, 325–331. Weibel, E. R. (1979). Stereological Methods, Vol. 1. Academic Press, London. Widdows, J., and Donkin, P. (1989). The application of combined tissue residue chemistry and physiological measurements of mussels (Mytilus edulis) for the assessment of environmental pollution. Hydrobiologia 188/189, 455– 461. Zahn, T., and Braunbeck, T. (1993). Isolated fish hepatocytes as a tool in aquatic toxicology: Sublethal effects of dinitro-o-cresol and 2,4dichlorophenol. Sci. Tot. Environ. Suppl., 721–734. Zaroogian, G., Yevich, P., and Anderson, S. (1992). Individual and combined effects of cadmium, copper, and nickel on brown cells of Mercenaria mercenaria. Ecotoxicol. Environ. Saf. 35, 41–45. Zaroogian, G., and Yevich, P. (1993). Cytology and biochemistry of brown cells in Crassostrea virginica collected at clean and contaminated stations. Environ. Pollut. 79, 191–197.