Enumeration of Cultivatable and Non-cultivatable Bacteria in Egg Case Sheaths of Loligo opalescens Joe Tyburczy Advisors: Dr. Dave Epel and Dr. Chris Preston Hopkins Marine Station 175H, Spring 1999 Permission is granted to Stanford University to use the citation and abstract of this paper. Abstract Background Symbiotic bacteria inhabit the accessory nidamental gland of the squid, Loligo opalescens, and are transmitted vertically to its egg cases. Previous work has resulted in the cultivation, isolation, and characterization of two dominant bacterial strains (genera Shewanella and Roseobacter) from homogenates of egg case sheath and accessory nidamental gland. Questions Do these cultivated isolates accurately represent the diversity of the sheath bacterial community? Are there bacteria present in the sheaths that are not cultivatable and therefore have not been described or characterized? What are the numbers of bacteria in the sheaths and what percent of these can be cultured? Results In order to obtain release of the bacteria from the sheath for quantitative enuneration, a means of disrupting the the tough egg sheath was developed using sodium periodate and glass beads. Total bacterial counts ranged between 10° and 10' bacteria per sheath which is slightly higher than the 10° to 10° estimated by Biggs and Epel (1991) with electron microscopy. A live/dead staining assay with DAPI and propidium iodide indicated about 68% of bacteria in sheath homogenate were alive. Serial dilutions of the native sheath homogenate were plated on various culture media and colony forming units were counted. Using comparison of direct bacterial counts from the homogenate, the viability stain, and the number of colony forming units on plates, the percent of bacteria cultivatable was found to be between 30 and 50%. Conclusions Because a significant portion of the bacteria were not cultivatable, further investigation of egg sheath bacteria may be necessary to adequately describe its diversity and composition. Introduction The market squid, Loligo opalescens, mates and deposits finger-like egg cases on sandy bottoms. During the three to five weeks required for the eggs to mature and hatch, the seemingly vulnerable eggs suffer little mortality due to predation or infection (Fields, 1965). One obvious protection mechanism is the sheath that surrounds the egg cases. This sheath arises from secretions of the nidamental gland and the accessory nidamental (AN) gland (Fields, 1965). Previous work has found that at least two strains of bacteria, of genus Roseobacter and Shewanella, are transmitted vertically from the AN gland of the adult female to the egg cases (Thompson, 1997; Kaufman et al., 1998). One strain, Shewanella spp., has been found to have antifungal properties (Kaufman, unpublished results). This isolate, like its close relative Shewanella putrefaciens, was found to produce the noxious polyamide putrescine which was shown to deter potential predators (Hoerner, 1996). It was hypothesized that these symbiotic bacteria may be responsible for anti-fungal protection as well as other natural defenses of the eggs. It is accepted that many bacterial strains cannot be cultivated with conventional techniques and that the classical procedure of cultivation, isolation, and identification is therefore inadequate to describe bacterial communities. The description of two strains previously isolated from the AN gland and egg case may therefore be an underestimate of their bacterial diversity. While Biggs and Epel (1991) estimated that total bacterial population of an egg sheath to be between 106-108 using electron microscopy, there has been no work done on what portion of the bacteria present are cultivable or on how the number of bacteria changes during development. In this paper, DAPI staining with epifluorescent microscopy was used to address the question of the numbers of bacteria present in the egg sheath. Another assay using both DAPI and propidium iodide was employed to determine what portion of the bacteria were viable. These numbers were compared to the number of bacteria cultivatable as determined by plating on culture dishes. The results of these experiments indicate that a limited portion (30 to 50%) of the viable bacteria present in the egg sheath are cultivatable under conventional cultivation conditions. In addition, Biolog plates were used to measure the metabolic capacity of two cultivatable strains as well as the metabolic capacity of the entire community. Materials and Methods Egg cases were collected from Monterey Bay at depths of 16 to 20 meters by scuba diving. They were brought into the laboratory and kept in flow-through seawater tanks at 13°C. Bacterial Count Assay Development Egg cases were rinsed with filtered seawater (FSW) and the sheath dissected away from the eggs, and weighed. The middle third of the sheath (Fig. 1) was then cut, weighed and fixed overnight in ImL FSW with a final concentration of 3.7% formaldehyde at 4°C. In order for bacterial enumeration with DAPI staining, the material to be examined, the egg case sheath in this case, must be homogenized. This ensures that when the bacteria are then pulled down onto a filter, they will be evenly distributed so that areas on the filter viewed with epifluorescent microscopy will be representative of the entire filter area. Thus the development of a homogenization procedure was essential to this method of counting bacteria. The fixed sheath section was then subjected to experimental homogenizing treatments. Chemical treatments included: a control with no additional treatment; acid (pH 3); 50mM dithiothreitol at pH 9; 5OmM sodium periodate with 25mM sodium acetate at pH 9; collagenase at 276U/mL. Physical homogenization methods included use of a glass homogenizer, a high speed electric rotary "Tissue Tearor," and the addition of glass beads with vortex agitation. After homogenization, 25uL of homogenate was loaded onto a filtration column along with 3mL FSW, and stained for three minutes with the DNA-binding, fluorescent dye, DAPI (4’, 6- diamidino-2-phenylindole) at a final concentration of lOug/mL. The bacteria in the homogenate were then collected onto a 0.22 micron black polycarbonate filter using a vacuum pump at -4psi. The filter was then placed on a slide and viewed with a DAPI fluorescence filter set with oil at 1000x magnification. The homogenizing techniques were evaluated on the basis of the abundance and brightness of bacteria, evenness of bacterial distribution on the slide, and breakdown of the sheath material. To determine whether significant bacterial lysis resulted from sodium periodate treatment and bead-beating, samples of Roseobacter were suspended in FSW, centrifuged, and resuspended in a small volume of FSW. Equal amounts of this concentrated suspension were pipetted into different tubes. The control was resuspended in ImL FSW, while the other was incubated in the sodium periodate solution for an hour in the dark and then beaten with beads. Equal volumes of prepared suspension were stained, collected onto filters, placed on slides, and counted. Bacterial Counts Egg sheaths were prepared with the periodate technique as described above involving sequential dissection, fixation, homogenization, bead agitation, staining, and filtration. The filters were mounted on slides and then examined with a UV fluorescent DAPI filter set with immersion oil at 1000x magnification. Bacteria were counted in an ocular grid with known dimensions. Ten fields were counted on each slide and the counts averaged. The number of bacteria per gram of sheath material and bacteria per sheath were calculated using the total sheath weight, the weight of the section homogenized, the volume of homogenate prepared, the volume of homogenate loaded into the filtration column, the total filtration area, and the area of the ocular grid. A relative age for each egg sheath was found by examining the squid embryos and determining their developmental stage using stages and diagrams from Arnold (1965) and Segawa, Yang, Marthy, and Hanlon (1988). Number of Viable Bacteria The percent of viable bacteria was determined by homogenizing an egg sheath section using only the hand homogenizer and glass beads (no periodate treatment). A small volume was then stained with two different DNA-binding fluorescent stains, the membrane-permeant DAPI, and the relatively impermeant propidium iodide and incubated five minutes before being collected onto a filter. The filter was examined under 1000x magnification using the DAPI filter set to view DAPI¬ stained cells and a Rhodamine filter set to view cells stained with propidium iodide. The total number of cells was counted with the DAPI filter, while those stained with propidium iodide and seen with the Rhodamine filter were counted as dead. Viable cells were calculated by subtracting the number viable from the total. The percent viable was calculated by dividing the number viable by total bacteria. Bacterial Culture Egg sheaths were dissected, ground with the glass homogenizer and bead beaten without any chemical treatment. The homogenate was serially diluted to concentrations of 5x103, 5x10“, and 5x10° with FSW. A known volume of the dilutions was plated on plates with complete 2216 marine agar or a low nutrient agar containing 5% 2216. The plates were incubated for three days at room temperature. The number and type of bacterial colonies that arose were counted. A known volume of the initial homogenate was stained, filtered, and the number of bacteria counted with DAPI fluorescence. Using the same calculation methods as for the other bacterial counts, the number of bacteria present in the sample was determined. Based on this number, the dilution series, and volume plated, the number of bacteria on each plate was determined. This was then multiplied by the percent viable as determined above, to approximate the number of viable bacteria plated onto each culture dish. This number was then compared to the number of colonies found on each plate. Sole-Carbon-Source Utilization The egg sheath was dissected, homogenized and beaten with beads. It was digested with a large enough volume that the fluid supernatant alone could be used without the flocculent material on the bottom. This supernatant was then plated into the 96 wells of a Biolog GN2 Microplate. A total of four Biolog assays were performed on egg sheath material. The first plate was prepared with homogenate from a single egg sheath with stage 24 embryos and incubated at 15°C. Another, slightly older sheath (stage 28 embryos) was homogenized, plated and incubated at 15°C. Two egg sheaths from the same clutch (stage 27 embryos) were ground together (in order to provide sufficient bacterial density) and the homogenate divided into two plates; one was incubated at 15°C and the other at room temperature. Thus the four plates were prepared from three different homogenates: one younger sheath, an older sheath, and a pair of plates prepared from the same homogenate (composed of two sheaths of similar age), but incubated at different temperatures. Plates were incubated for up to three days, or until distinct readable patterns arose on the plates, and analyzed for the presence of purple color in each well. Two different bacterial isolates, PKI and WH2, from culture plating were suspended in FSW and plated into Biolog plates, which were incubated for three days at room temperature and read. These isolate plates were done in duplicate. Results Bacterial Count Assay Development Treatment with sodium periodate solution (SOmM with 25mM sodium acetate at pH 4.5) for 45 minutes followed by thorough hand homogenization of about 0.25 g of sheath material in one ml FSW in a glass homogenizer for at least two minutes resulted in the best dissolution of the sheath. To maximize the release of bacteria, 0.1 g of Imm diameter glass beads were added and the homogenate was subjected to one minute vortex agitation. This procedure resulted in effective breakdown of sheath material, high counts, even distribution, and bright clearly visible bacteria with low background under DAPI fluorescence. Qualitative observations on the results of the various homogenization methods are shown in Table 1. A control for potential cell lysis resulting from the periodate treatment and homogenization was performed using Shewanella samples. Equal volumes of concentrated bacterial suspension were used. One was subjected to periodate and homogenization identical to that used for the sheath while the other was resuspended in FSW. The samples were then counted using DAPI staining. The numbers of bacteria counted with the control Shewanella samples were slightly lower than the counts without the homogenization treatment (Table 2). Bacterial Counts The total bacterial population per egg sheath varied from 3.9x108 to 3.5x10° (Fig. 2). There appeared to be a trend of increasing bacterial number with embryo developmental stage. The variation of bacterial numbers among sheaths also increased with embryo stage. Number of Viable Bacteria Of 714 bacteria stained with DAPI, 226 were also stained with propidium iodide indicating that about 32% of the bacteria were nonviable, while the remaining 68% were viable (Fig. 3). Bacterial Culture At least three distinct colony morphotypes arose on the culture plates. The first formed pink colonies and was dubbed PKI, the second formed white colonies and was designated WH2. and the third which was also white, but formed smaller colonies was designated TN3. The PK1 and WH2 colonies were easily visible after only one day incubation on the 2216 agar. The TN3 colony type was only visible on the plates after three day incubation and were by far the most numerous. The 5x10° dilution was overgrown with bacterial colonies and could not be counted. Each colony type was about ten times as numerous on the 5x10" dilution plates than on the corresponding 5x10° dilution plates. The colonies on the full concentration 2216 marine agar plates grew faster than those on the 5% plates. The full concentration 2216 plates exhibited rapid colony growth and supported all three colony types. Growth of all colony types on the 5% 2216 plates was much slower. The PKl colony type did not grow at all on the lower nutrient plates, while much higher numbers of the WH2 grew on them. The TN3 type grew in nearly equal numbers on both agar concentrations. Sole-Carbon-Source Utilization The Biolog system is an assay used to classify bacteria by determining the different carbon sources that they can use. Biolog plates contain 96 wells, each of which contain a different carbon source as well as a tetrazolium violet reduction dye, which turns purple upon reduction, indicating that the bacteria were capable of metabolizing that source. For isolated bacterial strains, the pattern obtained from this assay can then be entered into a database and compared to those of known strains and used to identify the bacteria. For bacterial communities, this essay is not useful as a means of identifying its constituents, but it can still indicate differences in metabolic capabilities between communities and individual components. Though it is possible for strains to have different metabolic capacities when alone than when in association with each other, if a bacterial community can use carbon sources that none of its known isolates can, it suggests that there are strains within the community that have not been isolated. The Biolog metabolic fingerprint for all of the sheaths tested were very similar (Table 3). The patterns for the replicate for the PKI plates were nearly identical, as were the patterns of the WH2 isolate plates. The plates of the PKl and WH2 isolates showed patterns distinct from each other. There were at least four carbon sources that tested negative for metabolism by both isolates that tested positive in the sheath homogenates. These carbon sources in which utilization was unique to the sheath homogenate were a-cyclodextrin, D-arabitol, D-sorbitol, and sucrose. The Biolog database indicated match between the pattern of the PKl isolate and Shewanella putrefaciens A with a similarity value of 0.843. The pattern of the WH2 strain did not match any bacteria in the Biolog database to propose any identification. Discussion The finding that the number of bacteria ranges from 3.9x108 to 3.5x10° per sheath (Fig. 2), is significantly higher than the previous estimate of 10° to 108 made by Biggs and Epel (1991). The numbers of bacteria also appeared to increase with the age of the egg sheath. Propidium iodide and DAPI staining indicated that 68% of the bacteria in the sheaths were viable. Upon plating of the sheath homogenate on media, it was found that of the viable bacteria, a significant fraction, between 50 and 70%, cannot be cultivated with conventional techniques (Fig. 4). Biolog analysis of sheath homogenate showed that the metabolic characteristics of the bacterial community were fairly consistent. There were at least four carbon sources that could not be used by either the PKI or the WH2 isolates that could be used by the community. The isolate PKI was also identified as a close relative of Shewanella putrefaciens A. Development and refinement of a homogenization procedure is important for further study of bacterial communities in egg sheaths since manual homogenization alone leads to clumpy distribution of the bacteria. Incubation in an acidic solution of sodium periodate was an effective means of breaking down sheath material. The chemical mechanism for this breakdown can probably be ascribed, at least partially, to oxidative cleavage of polysaccharides in the sheath material. This procedure allows direct quantification of the number of bacteria present and should also prove useful in future research as a preparation for immunofluorescent identification of specific bacterial strains within the community. This procedure seems innocuous since there was no bacterial lysis due to the homogenizing procedure. While the homogenizing procedure was reasonably effective, there was invariably some unhomogenized, bacteria-laden, sheath material which means that the counts derived with this procedure are at least mild underestimates. The number of bacteria was surprisingly high especially considering this limitation. Correlation of bacterial numbers with relative age using squid embryo developmental stage provided a crude means of comparison of bacterial numbers with age of the egg sheaths. There appears to be a trend of increasing bacterial numbers with age of the egg cases. However, due to the large variation between individual egg cases and the extremely low level of replication in this experiment, further study is necessary before any firm assertions can be made. Ideally future experimentation will include the use of squid eggs laid in captivity so that their absolute age will be known and the course of bacterial proliferation can be tracked more directly. The number of viable bacteria was relatively high - over two thirds. This assay was only performed on a single egg sheath so further experimentation should include replication on many more sheaths. In future experimentation, all egg sheath preparations will include propidium iodide as well as DAPI so that the number of viable bacteria can be determined along with the total number. The main problem with the use of the propidium iodide/DAPI assay is that it is limited to determination of membrane integrity and is not a measure of actual biological function. This means that there may be some nonviable bacteria whose membranes are still intact that may be incorrectly considered viable. In order to address this problem, an experiment with a compound such as fluorescein diacetate which indicates cellular metabolic activity should be used in combination with propidium and DAPI staining. This would allow an assessment of the validity of propidium staining as an indicator of viability. The most interesting result of this experiment was the finding that a significantfraction of total viable bacteria present in the egg sheath that cannot be cultured. The fact that the PKI strain did not grow on the 5% medium is illustrative of the dependence of different bacteria on culturing conditions. Therefore it is likely that some of the non-cultivable, viable bacteria represent entire strains that have not been characterized in previous studies of the egg cases. Future investigation should include efforts to cultivate more strains by varying culture environments including even lower nutrient concentrations, and anaerobic and microaerophilic conditions. The AN gland of mature squid should also be examined in a similar fashion to determine what percentage of those bacteria are cultivable and whether any additional strains discovered in the egg cases are acquired vertically from their parent or horizontally from the environment. It would be ideal if the bacteria from freshly laid egg cases and the AN gland of their parent were plated and their isolates compared. Another possibility for research efforts to elucidate the role of bacteria in the egg cases would be to incubate a female squid in antibiotics for several days prior to spawning in order to eliminate any bacteria and then observe any differences between development of the resulting clutch and that of untreated egg cases. The results of the Biolog assay indicated that the PKI isolate is related to Shewanella putrefaceins A. It is likely that this is the same strain isolated previously by Kaufman (unpublished) in the AN gland and egg sheaths. The uncertainty in this identification and the lack of any identification for the WH2 isolate shows the need for more definitive and rigorous identification methods. Further research will employ the use of modern molecular and genetic techniques to identify isolated strains. The patterns of carbon usage for the different sheaths were nearly identical, indicating that the metabolic characteristics, though not necessarily the composition, of their bacterial communities remains fairly uniform despite differences in incubation temperature and egg case age. Furthermore the fact that at least four of the carbon sources in the plates could not be used by either the PKI or the WH2 isolates indicates that some component strains of the community may not have been cultivated. One possibility for further study is to attempt to culture any such strains by plating them from the Biolog wells that are unique to the sheath homogenates. A corollary is to plate out directly from new sheath homogenate to culture plates with those carbon sources as the only nutrients. This might allow the cultivation of any novel species capable of metabolizing these carbon sources. Acknowledgements 1 would like to thank Dr. Dave Epel for his advice, enthusiasm, and patience. I would also like to thank Dr. Chris Preston for her indispensable help, expertise, and generosity. Lastly, I would like to thank Dr. Chris Patton for a last-minute slide shoot, and Dr. Jim Watanabe for his timely encouragement and counsel. Literature Cited Arnold, J. M. 1965. Normal embryonic stages of the squid, Loligo pealii (Lesueur). Biol. Bull. 128:23-32. Biggs, J. and D. Epel. 1991. Egg capsule sheath of Loligo opalescens Berry: structure and association with bacteria. J. Exp. Zool. 259:263-267. Fields, G. 1965. The structure, development, food relations, reproduction, and life history of the squid Loligo opalescens Berry. Calif. Fish. Bull. 131:1-108. Hoerner, M. 1996. Analysis of putrescine in the egg case of Loligo opalescens : potential roles in predator deterrence. Hopkins Marine Station Final Papers Bio 175H. Kaufman, M., Y. Ikeda, C. Patton, G. V. Dykhuizen, and D. Epel. 1998. Bacterial symbionts colonize the accessory nidamental gland of the squid Loligo opalescens via horizontal transmission. Biol. Bull. 194:36-43. Kaufman, M., C. A. Reeb, J. Thompson, and D. Epel. in press. Phylogenic identification of Loligo opalescens reproductive gland endosymbionts as novel Shewanella and Roseobacter species. Segawa, S., W. T. Yang, H. J. Marthy, and R. T. Hanlon. 1988. Illustrated embryonic stages of the Eastern Atlantic squid Loligo forbesi. The Veliger. 30(3):230-243. Thompson, J. R. 1997. Verification of vertical transmission of reproductive gland symbionts to the egg sheaths of Loligo opalescens. Hopkins Marine Station Final Papers Bio 175H. Figure Legend Fig. 1 Intact egg case with incision line dotted (A). Egg sheath with middle third used for homogenates shaded (B). Fig. 2 Squid egg sheath bacterial counts from DAPI staining. Numbers are in numbers of bacteria per sheath. Fig. 3 Bacterial viability as determined by counts of sheath bacteria stained with DAPI and propidium iodide. All cells stained with DAPI, while those stained with propidium iodide were considered nonviable. Fig. 4 Bacterial numbers from cultivation and direct counts. Numbers of cultivatable bacteria are divided into identifiable colony types. Numbers of bacteria are expressed as the number expected from the entire homogenate (containing two egg sheaths). Fig. 1 A. B. reeug led eiseg 5 peujeis sijes jusblaa 9. H. (ojdules uj sjje5 jo lequinu pojejnojeo) ellsjoeg 8 . Table 1 Descriptions of the results of different egg sheath homogenization treatments when stained with DAPI and viewed under an epifluorescent microscope with an ocular grid. Treatment Result — Glass homogenizer Uneven fields, low bacterial counts Tissue Tearor High counts, poor staining Poor staining Collagenase Poor staining, debris in fields Dithiothreitol Acid (pH3) Higher counts Sodium Periodate Very high counts, even fields, bright staining Table 2 Total bacterial counts from ten 30x30 um fields. Experimental bacterial aliquots were subjected to periodate homogenization before filtration and counting. Control aliquots were merely resuspended in FSW. % Diff. Exp. Replicate Control Experimenta vs. Control 100 11.9% B 758 814 7.4% Table 3 Results of Biolog sole carbon source utilization assay for egg sheath homogenates and bacterial isolates. Asterisk (*) indicates positive result; dash (-) indicates marginal result. enbpo Sage 24 27 27 28 Incubation Temperature (°C) 15 15 22 15 solates Sole Carbon Source yctodextnn — ven 80 — acetyl-D-glactosamine etyl-D-glucosamine Dalad r zucrose Lomithir L-phenylal L-prolin L-pyroglutamic acid D-serine amno bulprie acd rocanic acid osine resene 2.3-butane dlycerol D.L-o-glycerol phosphate ucose-1-p Ucose-6-phosphate