Abstract Bacteria were observed on the embryos of Pachygrapsus crassipes, Pagurus samuelis, and Hopkinsia rosacea. Strains that grew in culture were isolated, and were identified as Pseudomonadaceae and Enterobacteriaceae. There were eighteen isolates: eight from Pachygrapsus crassipes, six from Pagurus samuelis, and four from Hopkinsia rosacea. None of the strains showed anti-fungal activity against Lagenidium callinectes. The role of the bacteria isolated from Pachygrapsus crassipes embryos were studied in greater detail. Embryos treated with penicillin G, which was effective against seven of the eight strains of bacteria, had a higher mortality rate, and a greater likelihood of being infested with other microbes. This finding suggests a defense role for the bacteria. Introduction Bacteria have been cited as a pathogen of eggs and embryos of aquatic organisms, especially in commercial hatcheries (Fisher, 83, and Harper and Talbot, 1984). They have been noted to restrict oxygen and foul crustacean eggs considerably, causing the mother to drop them. However, studies on shrimp and lobsters have shown that bacteria found on embryonic surfaces can also perform a useful service to the developing animal, by keeping off parasites that would otherwise block gas exchange of the embryo with its surroundings, and compete for oxygen. The bacteria keep the embryos clear of Lagenidium callinectes, for example, by producing anti-fungal agents (Gil-Turnes, 1989, 1992). The intent of this study was to describe the kinds of bacteria present on embryos of marine invertebrate species with long incubation periods, long enough that colonization by bacteria and fungi may become a potential problem. Bacteria on Pachygrapsus crassipes, Pagurus samuelis, and Hokinsia rosacea were characterized. Pachygrapsus crassipes embryos were studied in greater detail. The female Pachygrapsus crassipes produces large yolky eggs that are internally fertilized, and then extruded and attached to the setae on the abdominal pleopods. The female tends this brood mass, containing some 50,000 eggs, by cleaning and ventilating it for an average period of 29 days, whereupon the larvae hatch (Strathmann, 1987, and Hiatt, 1948). Despite her careful preening, numerous bacteria are present on the surface of the embryos. In some Brachyura species, gram-negative bacteria, including known potential pathogens, such as Vibrio, Pseudomonas and Aeromonas species, are seen in healthy specimens (Brock and Lightner, 1990). It is interesting to ask whether bacteria covering the surface of the Pachygrapsus crassipes embryo, too, are a normal, or even beneficial factor in the crab's development. This study will compare success of embryos with and without the bacterial factor. The effect of inhibiting bacterial growth on the embryos was found to be harmful to the embryo. The embryo suffered greater infestation by fungi, protozoa and diatoms when bacterial growth was inhibited. Materials and Methods Organisms Ovigerous crabs (Pachygrapsus crassipes and Pagurus samuelis), and nudibranchs (Hopkinsia rosacea) were collected from the rocky shore at Hopkins Marine Station, Pacific Grove, CA. The specimens were stored in running sea water aquaria at 15 degrees Celcius. Äfter several days, the nudibranchs produced an egg ribbon. Visualization of Bacteria Use of the DNA specific stain, DAPI (4', 6-diamidino-2- phenyl-indole), allowed visualization of bacteria on the invertebrate embryos. DAPI (1 mg/ml stock solution) was diluted 1 to 1,000 in ultraviolet filtered sea water (UVSW). The embryos were soaked for 10 minutes in the stain, were rinsed three times in UVSW, and observed under the fluorescence microscope under oil immersion. Isolation of Bacteria Embryos were collected from each species, and bacteria were isolated. To isolate surface strains from Pachygrapsus crassipes and Pagurus samuelis, embryos were held with tweezers and dragged along the surface of "ÖZR" agar (15g DIFCO 2216 marine agar, 1g tryptone, 1 g yeast extract, 5 mg ferric citrate, 1 L sea water). A second procedure was used to isolate other strains that might be within the egg sheath beneath the surface. Embryos were homogenized in sterile sea water and diluted ten-fold. This second procedure was used for the H. rosacea eggs, additionally. In the attempt to isolate strains of each morphological type, each colony of distinct morphology that appeared on the agar was chosen for isolation, in the attempt t. Eighteen strains were selected in total: eight strains were chosen from the Pachygrapsus crassipes embryos, strains 1, 2, 4, 5, 7, 8, 19, and 20, six from the Pagurus samuelis, strains 9, 10, 11, 12, 13, and 14, and four from H. rosacea, strains 15, 16, 17, and 18. Pure cultures were obtained after two to three streakings. Bacteria Identification Colony characteristics on ÖZR agar were noted with respect to color, edge continuity, and general texture. Gram staining was performed on each isolate. Live cells were examined for motility. The strains were defined as strict aerobes or facultative anaerobes based upon their growth being strictly on the agar surface or additionally, beneath the agar at a depth of greater than 2 cm. Fermentation of glucose was ascertained by a pH change. The strains were grown up in a medium containing a pH indicator; acid production caused a color change. Strains were tested for the presence of catalase, an enzyme that catalyzes the breakdown of peroxides formed in the cell during respiratory metabolism, and for the presence of cytochrome C (oxidase test). A result of the cell's breakdown of tryptophan is indole accumulation. This property was recorded, also. The methyl red test was done to detect the compound acetoin, indicative of a distinct fermentation pathway leading to the product 2,3-butanediol, in which acetoin is the immediate precursor to butanediol. Cells were examined for H»S production, and nitrate reduction, additionally. The results of all these tests are presented in Table 1. Testing for Anti-Fungal Properties Among Isolates A competition experiment was established to ascertain whether Lagenidium callinectes growth was inhibited by the various bacterial species, L. callinectes was grown on ÖZR agar, and a plug removed and transferred to a fresh plate of ÖZR agar. Hyphae grew out from the plug after several days, at which point, each of the eighteen strains were streaked 5 mm away from the hyphae in order to determine whether any of the isolated strains inhibited the growth of L. callinectes. It was concluded that if the hyphae growth continued to extend to the bacterial streak, then there was no inhibition (Figure Bacterial Strain Response to Antibiotics Each bacterial strain was used to inoculate 3 ml of agar. The liquefied agar was immediately poured over solidified agar. Discs containing antibiotics were placed on the overlay. The commercially manufactured ampicillin discs contained 10 mcg of ampicillin. The penicillin G discs were prepared by soaking filter paper discs in a 1 g/L solution of penicillin G for one hour. Growth was checked at 48 hours. In the case that penicillin G inhibited bacterial growth, a circular zone of inhibition was observed surrounding the disc (Figure 2). Bacterial growth that covered the plate, including the area of the plate immediately surrounding the discs, marked a lack of inhibition, a negative result in Table 3. Antibiotic and Fungal Treatments Eggs from five specimens of ovigerous Pachygrapsus crassipes were examined microscopically, and those from different mothers were found to be in various stages of development. The female with embryos at the earliest stage of development was chosen to be used experimentally. These embryos had completed gastrulation; the blastomeres were numerous, and there was a slight withdrawal of blastomeres from one section of the capsule, revealing perivitelline space. They lacked early appendages, heartbeats, or eye spots. It was determined that they were approximately 5 days into development (Hiatt, 1947) Sterile tweezers were used to remove a cluster of approximately 200-300 eggs from the female. Altogether, four clusters were removed. Each cluster was placed in a separate vial within separate beakers, filled with 900 ml of UVSW. The sides of the vials were made of Nytex mesh, which allowed water to flow through, but prevented the embryos from escaping. Each beaker was aerated by a tube connecting to an air filter system. The first cluster received no treatment. The second cluster received 90mg penicillin G (0.1g/L) for 6 hours on the first day of experimentation, and 6 hours the next. The third cluster was treated in the same way as the second, and in addition, was infected with L. callinectes. The fungus had been grown up on agar, and agar plugs approximately 0.5cm on a side, thick with fungal hyphae, were placed in the vial containing the embryos. The beaker holding the fourth cluster was infected with L. callinectes, only. Results Observation of DAPI stained embryos Colony forming rods of several shapes and sizes were observed by fluorescence microscopy, using oil immersion. The rods formed long strings or patches. They covered the surface of the egg itself, and had also colonized the fibers binding the eggs together. The bacterial cover was not uniform, but patchy. The rod-shaped bacteria formed small colonies of ten to fifteen cells, and larger colonies of approximately fifty cells. Rods also formed colonies in long strings. Identification of Bacterial Isolates The bacteria that grew on ÖZR agar were identified to family. The tests performed on the strains are presented in Table 1. Gram-negative, aerobic strains were tentatively placed in the family Pseudomonadaceae (Table 2). Furthermore, if the colonies were pigmented yellow or orange, the strain was classified as being in the flavobacterium or cytophaga genera. If the strain was aerobic and gram negative, but lacked pigment (white or tan) it was classified as being in the alteromonas or xanthomonas genera (Hayes, et.al, 1979). Gram negative facultative anaerobes were tentatively placed in the Enterobacteriacea family. Four strains in the Pseudomonadacea family, and one in the Enterobacteriacea family were found on the Pachygrapsus crassipes embryos. Six strains from the Pseudomonadacea family were isolated from the Pagurus samuelis embryos; three strains from the Pseudomonadacea family, and one from the Enterobacteriacea family were isolated from the Hopkinsia rosacea embryos. Anti-fungal Properties None of the 18 isolates inhibited hyphae growth. The hyphae growth proceeded right up until the edge of the bacterial streak in each case. A problem with the L. callinectes culture obtained from the ATCC was an association of bacteria with the culture. Growth of the bacterial contaminant was apparently stimulated with any perturbation of the fungal culture, i.e., when agar plugs were cut from it, and transferred to new plates. It is difficult to ascertain whether the contaminant may have affected the results. Response to Antibiotic Treatment Ten of thirteen of the strains showed sensitivity to ampicillin, and thirteen of eighteen strains to penicillin G (table 3). Some strains showed greater sensitivity than others, but any noticeable inhibition is recorded as positive. Effects of Removing Bacteria with Antibiotic While seven of the eight strains isolated from Pachygrapsus crassipes may have been sensitive to penicillin in the antibiotic disc assay, penicillin did not reduce the bacterial density on the Pachygrapsus crassipes egg surface. Using DAPI to stain the penicillin G-soaked embryos, and comparing them to stained control embryos, showed little difference in the density of bacteria on the surface. As on the control embryos, the bacteria grew in patchy and string-like colonies. This observation prompted a quantitative assessment. Äfter a ten-fold dilution, the density of colonies from the penicillin G-soaked embryos was 7.1x10° colony forming units (cfu)/ug, compared to 3.8x104 cfu/ug for the control (Table 3). The penicillin+L. callinectes treatment resulted in 4.6x10° cfu/ug, and the L. callinectes treatment showed 1.2x10° cfu/ug (Table 4). Embryos were examined for development, mortality, and number of epibionts 10, 13, and 18 days after the experimental setup (corresponding to days 15, 18, and 23 of embryo development, since embryos were taken from their mother approximately 5 days into development). In the assessments on days 10 and 13, only 10-20 embryos per treatment comprised the sample. On day 18, all remaining embryos were examined (100-300 embryos per treatment). Epibionts While fungal hyphae were not observed on the penicillin G or penicillin G+L.callinectes-treated embryos 18 days after the initial penicillin soaking, many other epibionts appeared heavily on the penicillin G-treated egg surfaces. By 10 and 13 days into the treatment, diatom and protozoan activity was high on the eggs receiving penicillin G. Eighteen days after the start of the treatments, the eggs were thickly covered with masses of fibers, made up of smaller, round bodies. This could be an egg mass or a fungus (apparently not Lagenidium callinectes). Several transparent worms crawled the surface of the embryos receiving the L.callinectes -treatment. Besides this change, the surfaces of eggs in the control treatment and the L.callinectes -treatment remained as clean at 18 days into the treatment as they had been at the start of the experiment. They hosted only a few protozoa. Mortality Discoloration indicated a dead embryo. In dead embryos, either the cells were orange, or the extracellular fluid was pink. In the first assessment, 10 days after the start of the experiment, 50% of the sample of embryos taken from the control group were dead. None of the penicillin G- treated were dead. 29% of the penicillin+L.callinectes-treated were dead. None of the L.callinectes- treated were dead. Development was not retarded in any case, in those that remained alive (figures 3 - 6). On day 13, 25% of the control embryos were dead, 42.5% of the penicillin-soaked were dead, 45% of the penicillin+L.callinectes¬ treated were dead, 17.5% of the L.callinectes-treated were dead. Development appeared retarded in the penicillin+L.callinectes treated embryos. By day 18, 30.5% of the controls, and 73.3% of the penicillin- treated were dead, and essentially all of the penicillin+L.callinectes, and L.callinectes -treated were dead (98.7% and 99.5%, respectively). Discussion The embryos of the three invertebrates under investigation in this study were shown to host a number of Gram negative, rod-shaped bacteria. The bacteria were seen in dense coverage over as much as 50% of the surface of Pachygrapsus crassipes embryos. The bacteria were identified as Pseudomonadacea and Enterobacteriacea. They did not inhibit growth of Lagenidium callinectes on agar. Seven of the eight Pachygrapsus crassipes strains were sensitive to penicillin-G. After treatment with penicillin G, Pachygrapsus crassipes embryos suffered from colonization by more protozoa, diatoms, and worms than control embryos receiving no treatment. Their mortality rate increased dramatically. The question under scrutiny in this investigation was whether the bacterial guests of the embryo protect it from unwanted parasites, or if these bacteria are only parasites themselves. Microbial protection of embryos has been reported for the shrimp, Palaemon macrodactylus, which, like Pachygrapsus crassipes, broods its eggs externally. In that study, however, it was shown that the bacteria produced an anti-fungal compound, 2,3-indolinedione (Gil-Turnes et. al, 1989). Similarly, the bacteria found on Homarus americanus embryos produced an anti-fungal compound, tyrosol (Gil-Turnes and Fencial, 1992). The bacteria isolated in this study showed no anti-fungal activity, yet the possibility exists that not all bacteria inhabiting the embryo surface were isolated. Some strains may not have been able to grow on ÖZR agar, and some strains may have been overlooked during the initial screening in which bacteria were chosen for further study from the first platings. The penicillin G treatments provide an insight into another possible role of the bacteria. It was found that treatment with penicillin G resulted in more bacteria, as noted. Because some strains were resistant to the antibiotic, it may be that after the sensitive strains were eliminated, the population of the remaining resistant strains may have exploded. In a study of the epibiotic flora associated with embryos of Homarus americanus, the embryos of laboratory spawned females hosted different types of epibionts from the wild spawned controls, notably fewer colonial rods. While the wild spawned embryos seemed unaffected by bacterial associations, a relationship was found between increase in bacteria and decline in egg clutch size for the lab spawned embryos. The authors suggest that colonial rods could be protecting eggs from an unchecked population explosion of other kinds of bacteria (Harper and Talbot, 1984). In a study of the bacterial strains on pelagic fish eggs, some antibiotic treatments stimulated bacterial growth. The author suggested that the antibiotics in these particular cases either provided a growth-promoting substance, or merely repressed one or more strains of bacteria, allowing the other strains better growth conditions (Oppenheimer, 1955). In another study, epibionts on the embryos of Cancer magister were compared before and after treatment with penicillin G and streptomycin sulfate. The antibiotics prevented non-filamentous fouling, but algal filaments were resistant to the antibiotic, and increased in number. In this case, mortalities were reduced. Antibiotics inhibited harmful epibionts (Fisher, 1976). For this study, the antibiotic-sensitive strains of bacteria may have been preventing epibiont infestation. Their removal prompted the explosion of various protozoa and diatoms. The fibrous masses covering the penicillin G-Lagenidium callinectes-treated embryos could be a type of fungus, or the egg masses of a predatory worm. Indeed, evidence of a parasitic nemertean, Carcinonemertes epialti, has been reported on Pachygrapsus crassipes embryos (Roe, 1979). However, Carcinonemertes epialti is pigmented, unlike the observed in this study. Its egg masses are as thick in diameter as the worm itself, also inconsistent with these findings. It can be hypothesized that symbiont strains were those sensitive to penicillin G, and their elimination meant an opportunity for pathogenic strains to exploit the embryo. The fact that penicillin treated embryos have higher mortalities and retarded development supports this hypothesis. In conclusion, while it may be true that the bacteria themselves stress the developing Pachygrapsus crassipes embryo by their own oxygen and nutrient requirements, it is likely that they protect the embryos from more demanding parasites, capable of harming the embryo to a greater extent. Acknowledgements My thanks go to David Epel, Robin Hayes, and Paul Sund for their help and expertise. Literature Cited Brock, J. A. and Lightner, D.V. (1990) "Diseases Caused by Microorganisms. Disease of Marine Animals, Vol.III, ed. O. Kinne. Hamburg, Germany: Biologische Anstalt Helgoland. Fisher, W.S. (1983) Eggs of Palaemon macrodactylus: II. Association with Aquatic Bacteria. Biol. Bull., 164: 201-213. Fisher, W.S., and D.E. Wickham. (1976) "Mortalities and Epibiotic Fouling of Eggs from Wild Populations of the Dungenous crab, Cancer magister. U.S. Fish Widl. Serv. Fish Bull.,74. 201-207. Fisher, W.S. (1976) Relationships of Epibiotic Fouling and Mortalities of Eggs of the Dungenous Crab. J. Fish. Res. Board. Can.,33: 2849-2853. Gil-Turnes, M. S. and W. Fencial. (1992) Embryos of Homarus americanus are Protected by Epibiotic Bacteria. Biol. Bull., 182: 105-108. Gil-Turnes, M.S. Mark E. Hay, and W. Fencial. (1989) Symbiotic Marine Bacteria Chemically Defend Crustacean Embryos from a Pathogenic Fungus. Science, 246: 116-118. Harper, R.E. and P. Talbot. (1984) Analysis of the Epibiotic Bacteria of Lobster (Homarus) Eggs and their Influence on the Loss of Eggs from the Pleopods. Aquaculture, 36: 9-26. Hayes, R.R., T.A. McMeekin, and J.M. Shewan. (1979) "The Identification of Gram Negative, Yellow Pigmented Rods.' Identification Methods for Microbiologists, second edition, ed. F.A. Skinner, and D.W. Lovelock. New York: Academic Press. 177-187. Hiatt, R.W. (1948) The Biology of the Lined Shore Crab, Pachygrapsus crassipes Randall. Pacific Science, Vol. II, No. 3: 135-213. Oppenheimer, Carl H. (1955) The Effect of Marine Bacteria on the Development and Hatching of Pelagic Fish Eggs, and the Control of Such Bacteria by Antibiotics. Copeia, No. 1. 43-49. Roe, P. (1979) Aspects of the Development and Occurrence of Carcinonemertes epialti (Nemertea) from Shore Crabs in Monterey Bay, California. Biol. Bull., 156: 130-140. Strathmann, M. (1987) Reproduction and Development of Marine Invertebrates of the Northern Pacific Coast. Seattle: University of Washington Press. 451-460. Table 1 02 Tex MR 8 8 gb 14 gf 20 P g 1- E Key: St = strain Co = color, y = yellow, w = white, t = tan, o = orange Ed = edge, c = continuous, s = rough, d = diffuse Tex = texture, g = glossy, b = blurry, f = flowery, c = clumpy, t = transparent G = Gram stain M = Motility O2 = oxygen requirement (+ indicates strict aerobe, - indicates facultative anaerobe) F = glucose fermentation C = catalase = oxidase (cytochrome C) H2s I = indole MR = methyl red VP = Voges Proskauer 28 = hydrogen sulfide production N = nitrate reduction Table 2 Strain f, organism 1 P. crassipes 2 P. crassipes 4 P. crassipes 5 P. crassipes 7 P. crassipes 8 P. crassipes 9 P. samuelis 10 P. samuelis 11 P. samuelis 12 P. samuelis 13 P. samuelis 14 P. samuelis 15 H. rosacea 16 H. rosacea 17 H. rosacea 18 H. rosacea 19 P. crassipes Family Pseudomonadacea unidentified Enterobacteriacea Pseudomonadacea unidentified Pseudomonadacea Pseudomonadacea Pseudomonadacea Pseudomonadacea Pseudomonadacea Pseudomonadacea Pseudomonadacea Pseudomonadacea Pseudomonadacea Pseudomonadacea Enterobacteriacea Pseudomonadacea Genera flavobacterium or cytophaga xanthomonas or alteromonas flavobacterium or cytophaga xanthomonas or alteromonas flavobacterium or cytophaga flavobacterium or cytophaga xanthomonas or alteromonas flavobacterium or cytophaga flavobacterium or cytophaga xanthomonas or alteromonas xanthomonas or alteromonas flavobacterium or cytophaga xanthomonas or alteromonas 20 P. cri sipes Pseudomonadacea xanthomonas alteromonas Table 3 Strain 13 17 18 19 20 Ampicillin sensitive + +/- +- Penicillin sensitive +- +/- Table 4 Treatment control penicillin penicillin + L.callinectes L. callinectes cfu/ug, day 13 of treatment 3.8X104 7.1x 105 46X 105 12x106 Figure Legends Figure 1: Bacterial strains do not inhibit L. callinectes growth from the transferred plug (pictured in center). Figure 2: Strain #20 is sensitive to the antibiotics penicillin G and ampicillin. Figure 3: Control group mortalities on day 10, 13 and 18 of artificial incubation. Figure 4: Penicillin G-treated embryo mortalities on day 10, 13, and 18 of artificial incubation. Figure 5: Penicillin G+-L.callinectes-treated embryo mortalities on day 10, 13, and 18 of artificial incubation. Figure 6: L.callinectes-inoculated embryo mortalities on day 10, 13, and 18 of artificial incubation. Figure 1 Figure 2 igure 20 - Figure 4 80 60 Da Figure 5 0 Day Figure 6 00 Day