Abstract
The resistance of Loligo opalescens embryos to predation may be due to symbiotic
bacteria which colonize the egg sheath. It has been hypothesized that these symbiotic
microbes are transferred to the egg sheath from the accessory nidamental (AN) gland of the
mature female. Cultures of bactena from freshly spawned egg sheath tissue have produced
two to three morphological types of bacteria. Sheath and gland isolates of the same
morphology were shown to express identical patterns of protease activity, antifungal
activity, and H,S production. Whole cell hybridizations of purified sheath bacteria with
l6s TDNA probes supports the hypothesis that one sheath isolate is the same species as a
previously charactenzed AN gland isolate. RAPD PCR analysis confirms the correlation
between sheath and gland isolates by showing both produce similar genomic fingerprints.
Comparative analysis of 16S FDNA sequences verifies this homology between sheath and
gland isolates, indicating vertical transmission of the AN gland symbionts to the egg sheath
during spawning.
Introduction
Monterey Bay is a spawning ground for the squid Loligo opalescens which
deposits numerous eggs cases on the ocean floor every spring. The squid embryos,
encased within a mucopolysaccharide egg sheath, have been observed to develop virtually
undisturbed during their three to five week incubation period. Studies have shown that the
sheaths of these egg cases are inhabited by bacteria (Biggs and Epel, 1991). The presence
of these bacteria in the egg sheath may protect the embryos from micro and macro
predation. Previous studies have indicated that these sheath bacteria secrete putrescine, a
polyamine compound found in rotting material, which has been shown to deter potential
predators from the egg cases (Hoerner, 1996).
The presence of bacteria within newly laid cases suggests that the bacteria and the
egg sheath may be secreted simultaneously, a process known as vertical transmission. A
newly identified species of bacteria in the genus Shewanella has been shown to colonize the
accessory nidamental gland (AN gland) of the female squid which participates in the
secretion of the egg case (Kaufman. personal communication). The AN gland is also
known as the reproductive gland. as its color signals the reproductive maturity of the squid.
This change in color occurs upon the colonization of the gland by pigmented bacteria. Due
to its association with bacteria and secretory structure, the AN gland has been indicated as a
likely source for bacteria found in the egg sheath (Biggs and Epel, 1991).
Verification that the bacteria in the freshly laid egg sheaths were of the same new
Shewanella species as the bacteria from the AN gland would be strong evidence to support
the theory of vertical transmission. Several techniques were employed to investigate
whether the bacteria in the AN gland were vertically transmitted to the egg sheath. In this
study we illustrated that the bacteria in the egg sheath express the same potentially
symbiotic activities as the bacteria in the AN gland, antifungal activity, protease activity and
the production of hydrogen sulfide. RAPD PCR analysis of the egg sheath isolates
genomic DNA demonstrated that egg sheath isolates and AN gland isolates produced
similar genomic fingerprints. Whole cell hybridization studies showed that 165 IDNA
probes specifically designed to label the AN gland symbiont would also label egg sheath
isolates. Additionally, l6s ribosomal DNA amplification, sequencing, and homology
searches revealed that egg sheath isolates had almost identical 16S rDNA sequences as
previous AN gland isolates. and indicated the presence of a second uncharacterized species
in both the AN gland and the egg sheath.
Pathogenic associations have long been the focus of bacterial-animal interactions.
However all animals maintain cooperative associations with a complement of bacteria
(Ruby, 1996) and would not be able to survive without microbial assistance. Models of
non pathogenic bacterial-animal associations can reveal much about the w ay prokarvotes
interact with eukaryotic cells. provide valuable information to the medical and
pharmace utical community. and contribute to an understanding of the dy namics of the
biosphere. Squid-bacteria symbiosis have been previously characterized in both loliginid
squid such as Loligo pealei (Bloodgood. 1977/ Barbieri.19%6) and sepiod squid such as
Euprymna scolopes (Ruby. 19%6). Loligo opalescens is the first of such squid-bacteria
models hy pothesi zed to vertically transmit its bacterial symbionts to the egg sheaths of its
embryos.
Materials and Methods
Plating experiments:
Isolation of colonies:
Freshly spawned egg cases from Loligo opalescens Berry were collected from the
squid culture facility at Hopkins Marine Station. The egg cases were immersed in 95%
ethanol for 2 minutes to remove any exogenous bacteria on their surfaces, then rinsed twice
in sterile seawater. A section of the sheath, approximately lem’, was cut from each of the
egg cases and its mass was recorded. Each section was subjected to one of the following
additional ethanol treatments: O. 2. 4,6, and 8 minutes to sterilize the outer lavers of the
egg sheath section (Table 1). After the second immersion in ethanol, the sections were
again rinsed twice in sterile seawater.
Next, each section was homogenized in ImL of sterile seawater with a sterile
dounce and the homogenate was diluted to concentrations of 10', 10, 103, 10 and 105.
100uL of the homogenate and each of the dilutions, was plated onto marine agar and
allowed to incubate for I to 2 days. After the bacterial colonies had developed on the
plates, the colonies formed per gram of sheath tissue (cfulg) was determined by counting
the number of morphologically identical colonies observed on the plate. multiplied by the
reciprocal of the original homogenate 's dilution, and then divided by the mass of the
section from which the homogenate had been made.
Colonies of different morphologies were purified by removing the desired colony
trom the sheath tissue cultures on a sterile inoculation loop and then re-streaking the colony
across a fresh. sterile. marine agar plate.
Protease Activity Assar:
Each of the morphologically different sheath isolates was patched onto replicate
marine agar plates containing ISce skim milk and allowed to grow at 1S Cor at room
temperature. It was obser ed whether the isolate expressed protease activity by
determining whether a zone of clearing formed around the isolate in the agar plate.
indicating the proteolytic degradation of the protein casein contained in the milk.
H.S Production Assay:
Each of the morphologically different sheath isolates was patched onto replicate
Marine triple-sugar-iron (TSI) plates and allowed to grow at 15°C or at room temperature.
Isolates were positive for the production of H,S if a black precipitate of iron sulfide formed
on the plate within the patch of bacteria.
Antifungal Assay:
Top agar plates containing the yeast against which the antifungal activity would be
tested were made by making overnight cultures the yeast (l colony of the yeast grown
overnight in SmL of marine broth) and adding lul, lOul, 5Oul or 10Oul of the overnight
culture to Aml of 1% molten top agar. This mixture was vortexed briefly and immediately
poured onto à manine agar plate, forming a top layer of agar containing the yeast. The tor
agar was allowed to cool, then each of the morphologically different sheath isolates was
patched onto the plate. After several days of growth it was determined whether the sheath
isolate exhibited antifungal activity by observing a zone of clearing in the lawn of veast
around the isolate.
Antibiotic Sensitivity Assay:
lop agar plates containing the sheath isolate whose antibiotic sensitivity would be
tested were made by making ovemnight cultures of the isolate (I colony of the isolate grown
overnight in Sml of marine broth) and adding 25ul of the culture to dml of 1% melted tot
agar. This mixture was vortexed and immediately poured onto a sterile. marine agar plate.
wo such plates were made for each isolate. Upon cooling. twenty disks containing the
following antibiotics were placed upon the two top agar plates made for each isolate (IO per
plate):
PLATE41
PLATES2
Ampicillin
lOug
Neomycin
3Omcg
30ug
Bacitracin
10 units
Nalidixic acid
30ug
Cefamanoole
Penicillin
10 IU
3Oug
Cefazolin
Polymyxin B
300 IU
Sug
Sug
Ciprofloxacin
Rifampin
Streptomycin
Chloramphenicol
3Oug
Soug
Clindamycin
Sulfisoxazole
2ug
25mg
93 -
Erythromycin
Sulfamethoxazole
D.15ug
ISug
Kanamycin
3Omc
and Trimethoprim 1.25ug
Moxalactam
30ug
Tetracycline
30ug
Tobramycin
lOug
After several days of growth. the plates were observed for zones of clearing in the bacterial
lawn around the antibiotic disks. A zone of clearing would imply that the bacteria had been
prevented from growing in that region due to its sensitivity to the antibiotic.
Microscopic Morphology and Metabolic fingerprinting
Electron microscopy:
Isolates from the egg sheath in log and in stationary phase were examined by
electron microscopy to observe possible differences in morphology. Cells were applied to
an EM grid by floating the grid for 2 minutes on a lOOul drop of log phase culture or
overnight culture. The cells were fixed onto the grid. and negatively stained with a 2 minute
treatment with 1% phosphotungstic acid.
Biolog Identification Kit:
Potential identities for the egg sheath isolates were examined using the BIOLOG
assay and computer program which compared the metabolic fingerprint of each isolate,
against the charactenstic metabolic fingerprints of other bacteria in the Biolog database
Molecular Studies
Isolation of genomic DNA:
The genomic DNA was extracted from several isolates of each morphology found
in the egg sheath, including purified colonies of bacteria previously isolated from the AN
gland, from purified colonies previously isolated from the egg sheath, and from several
control strains of bacteria. The genomic DNA (gDNA) was isolated using the Rapid Prep
Micro Genomic DNA Isolation Kit (Pharmacia Biotech). This kit employs the use of
individual premanufactured resin-packed columns to purify the DNA from 2x 10
microbial cells (approximately Iml of an overnight culture), and then the DNA is eluted in
400ul of buffer. The DNA is then precipitated from this suspension with the addition of
320ul of isopropanol and collected by centrifugation. After use of the RapidPrep kit the
gDNA pellet from each sample was re-suspended in 30ul of ddH.O and was stored at 4°C.
RAPD PCR Analvsis:
The genomic differences between the previously isolated AN gland symbionts.
previously isolated egg sheath bacteria, and the new egg sheath isolates were investigated
using Random Amplified Polymorphic DNA analysis (RAPD). This technique detects
genomic polymorphisms by using short oligonucleotide primers (7 to 15 bp) of random
sequence to initiate the polymerase chain reaction. generating a fingerprint of amplification
products specific to the species and/or strain level in question.
Ready to Go RAPD Analysis Beads (Pharmacia Biotech) were used to implement
the RAPD Analysis. These beads contained all the reagents. excluding the primer.
nécessary to carry out the reactions.
lo prepare the samples for RAPD Analysis one Ready to Go" bead was placed in
each stenle PCR tube with Sul of the primer to be used in the reaction. 2ul of sample DNA
was added to each tube, and brought up to a volume of 25ul with the addition of 18ul of
water. For each trial a no DNA control was included. The samples were placed on a Perkin
Elmer thermal cycler at a setting of:
95°-5 min
95° - 1 min
cycles
36°- 1 min
) X 45
1 min
4°- hold
When the reaction ended the products were visualized via agarose gel electrophoresis,
Banding patters were compared by calculating a ratio of similarity for pair-wise
comparisons of isolates the formula for which is:
ratio of - number of common bands
similarity
total number of bands
Whole cell hybridisation
Ribosomal DNA probes were used to determine whether probes designed
specifically for the AN gland symbiont would hybridize to the egg sheath isolates.
Bacterial cells were harvested from Iml of log phase cultures of three morphologically
different egg sheath isolates, and two control strains. The harvested cells were washed in
cold sterile seawater and then resuspended in Iml cold sterile seawater. This resuspension
was added to Iml of 8% paraformaldehyde fixative (Ag paraformaldehyde flakes, Smls
IOXPBS. 45mls ddH,O) and fixed for 1 hour. After 1 hour the cells were pelleted from the
fixative, washed in sterile seawater, and resuspended in cold sterile seawater. 3Oul of the
fixed cells were applied to slides coated with gelatin and chrome alum CrK(SO.), The
slides were allowed to air dry. and then were post-fixed for 20 minutes in a solution of
3Omls 37% formaldehyde and 270mls MeOH. After the post-fixation the slides were
rinsed briefly in water and dried in the dark.
20ul of pre-warmed hybridization buffer (lOmls 25xSET. S00ul 10%SDS.
39.5mss dH. O)was added to the prepared cells on each slide and lul of the probe
(SOng/ ul) was added to the buffer. Four different probe treatments were applied to each
group of cells. A eubacterial probe (EUB) as a positive control was used to label all
isolates at a conserved region in the 16S FDNA gene. A probe specific to Shewanella
putrefaciencs (SPN) was used as a negative control to show that the hybridization reaction
was specific on the species level and not across the genus Shewanella. A probe designed
to label the AN gland symbiont (SQ26) which would hybridize along the same variable
region of the l6s gene as the SPN probe. And a control hybridization, which contained no
probe, was also performed with each group of cells. The cells on the slide were covered
with à small square of parasilm and were placed in a SOml Falcon tube containing a
kimwipe saturated in hybridization buffer to incubate overnight at 50C. After the
incubation the cells were washed twice in. 2X SET then were observed for probe binding
under phase contrast and fluorescence microscopy
l6s Ribosomal DNA Amplification and Sequencing
los ribosomal DNA was amplified from the genomic DNA prepared from four of
the egg sheath isolates using Perkin Elmer Gene AMP PCR Core Reagents and two
primers which hybridized at opposite ends of the l6s gene. The amplification was
achieved via PCR with the following reaction profile:
CX 3 min
95°Cx30 sec
SSCX30 sec
28 cycles
CxI min¬
Verification that the l6s Ribosomal DNA gene had been amplified was achieved by
observing a 1500bp band after gel electrophoresis of the product.
The amplified 16S FDNA was purified with the Promega Wizard PCR Preps DNA
purification system. The purified product was sent to the Stanford University Protein and
Nucleic Acid facility for nucleotide sequencing. The sequences were analy zed for
homology to the previous AN gland isolate with the program BESTHT. Searches of the
Genbank database with the FASTA algorithm were then used to locate homologous
sequences.
Results
Plating experiments:
Isolation of colonies from egg sheath tissue
Bactena were isolated from 15 section of egg sheath in 4 trials as described in the
Materials and Methods (Table 1). Three distinet bacterial morphologies were isolated from
the 15 egg sheaths examined. Based upon their color they were labeled brown-orange
(BO), orange-green (OG), and white (W). The BO and Wisolates were present in cultures
from each of the 15 sheaths. The OG isolate was found in 7 of the 15 sheaths. The number
of colonies formed per gram of egg sheath tissue (cfu/g) for each of the 15 sheaths
examined are presented in Table 2. The cfu present on the marine agar decreased linearl
with the amount of time the sections of sheath were immersed in ethanol (Figure 1.) The
decrease in cfu per minute of EtOH treatment is present between the BO and Wisolates
from all the sheaths examined, indicating that the two morphologies' relative frequency is
maintained upon exposure to ethanol. A similar plot of cfulg formed vs. ethanol time
treatment was generated for sheaths (J3, J5-J10) which included the OG isolate (Figure
2).
Protease Activity Assay. H.S Production Assay, Antifungal Assay:
The BO and OG isolates were shown to express protease activity against the protein
casein as indicated by a zone of clearing around the isolate after several days of growth on a
marine agar plate containing skim milk (Figure 3a). The BO and OG isolates were both
found to produce hydrogen sulfide. as evidenced by a black precipitate of iron sulfide on
marine triple-sugar-iron-agar plates (Figure 3b).The Wisolates do not express protease
activity nor did they indicate production of hydrogen sulfide. The BO isolates have also
been shown to express antifungal activity by preventing the growth of the marine veast
Rhodotorula glutinis, and Candidatropicalis on marine agar plates (Table 3).
Antibiotic Sensitivity Assay.
The three isolates were tested for their sensitivity to twenty antibiotics which
targeted a wide range of cellular activities such as cell wall formation and protein synthesis
(Table 4). The BO, and OG isolates shared a90% identical pattern of antibiotic
resistance and sensitivity. The W bacteria showed sensitivity to all the antibiotics tested.
The three isolates shared sensitivities to thirteen out of the twenty antibiotics tested.
Morphological characterisation and metabolic fingerprinting
Electron microscopy,
Each of the isolates were examined via electron microscopy. The BO and QG
isolates appeared to have identical morphologies. Both isolates were rod shaped with
unsheathed polar flagella. The Wisolates were morphologically different than the BO and
CG isolates. The Wisolates were smaller, shorter rods with a single appendage (the E
pilus) emanating from the side of the cell. (Figure 4).
Biolog Assay:
The BIOLOG identification assay indicated that the BO and OG isolates were of the
genus Shewanella, and were closely related to the species Shewanella putrefaciens (Table
5). The system was unable to indicate any close relatives of the Wisolate,
Molecular studies:
RAPD PCR Analvsis:
RAPD PCR analysis consistently produced similar genomic fingerprints among the
BO egg sheath isolates. and the previous isolates from the AN gland and the egg sheath.
8010 and SQ26. RAPD PCR analysis also consistently produced similar genomic
fingerprints among the Wegg sheath isolates and previous WAN gland isolates (Figure
5). The fingerprints of the OG isolates were shown to consistently differ from the BO
isolates. indicating genomic differences between the two similar morphologies. The ratios
of similarity (described in materials and methods) among pairwise comparisons of isolates
show the AN gland symbionts and egg sheath isolates of the same morphologies produced
the highest ratios of similarity (Table 6). A chart of these values graphically illustrates the
amplification product similarity ratios (Figure 6).
Whole cell hybridisations
A l6s ribosomal DNA probe desi gned to specifically label the newly identified AN
gland symbiont (SQ26) also labeled the BO and OG egg sheath isolates, indicating that the
two egg sheath isolates were closely related to, if not the same species as the AN gland
symbiont (Figure 7). The W bacteria were not labeled by this probe. The control probe
EUB which hybridizes at a conserved region of the l6s gene and thus labels all eubacteria,
labeled all the isolates tested. The negative control probe, specific to Shewanella
putrefaciens, and differing from the newly identified Shewanella species probe (SO26) by
4 bases, did not hybridize to any of the isolates, only to the control strain of S. putrefaciens
(SPATCC). The no probe treatment indicated the level of autofluorescence emitted by the
bactena, and provided the standard against which the positive or negative results of the
other probe treatments were judged. The cumulative results of all the hybridizations and
controls are shown in Table 7.
l6s Ribosomal DNA Amplification and Sequencing.
The l6s ribosomal DNA was amplified via PCR as described in the Materials and
Methods section and the amplification product viewed on a.7% agarose gel. A band near
150Obp was observed indicating the amplification reaction had been successful.
Sequencing of the purified product was performed by the Stanford University Protein
Nucleic Acid facility (PAN). Sequences of the BO and OG isolates were found to be at
least 96% identical to that of the newly identified AN gland symbiont. The OG isolates
were found to be at least 93% identical to the sequence of the AN gland symbiont. The
BO and OG isolates were also 93%e identical. Due to an incomplete sequence (10 out of
1300 bp total), the given percentages of identity among the OG, BO and AN gland
symbiont are likely to be higher than indicated.
Using the computer algorithm FASTA, the OG and W sequences were compared to
the bacterial l6s FDNA sequences in the Genbank database to locate instances of sequence
homology (Table 8). The closest match to the OG bacteria that Genbank presented was a
barophilic Shewanella strain (DB5S0 1) with a similarity percentage of 92.1%, followed
closely by 9 Shewanella species and strains with similarity percentages from 91.9-893%.
The closest match presented by Genbank is over 3% lower than the match between the
newly identified species of Shewanella from the AN gland and the OG bacteria. The W
isolate 's sequence was also searched for in the Genbank database. The closet match
presented by Genbank was Roseobacter algicola a newly identified symbiont of the
dinoflagellate Prorocentrum lima. (Lafay, B et al, 1995). The percent of similarity between
the Wisolate and Roseobacter was 86.6%.
Discussion
In this study we verified that bacteria isolated from freshly spawned squid eggs
were the same as AN gland symbionts, indicating the vertical transmission of the bacterial
symbionts from the squid to their egg cases. Bacterial isolates from the egg sheath were
tested to verify that they expressed the same potentially symbiotic activities as isolates from
the AN gland. The BO and OG bacteria were shown to express protease activity. The
protease activity expressed by these isolates may serve to create an inhospitable
environment within the egg sheath to prevent additional microbial colonization. This
protease activity may also serve weaken the egg sheath when the embryos are ready to
emerge from their incubation.
The antifungal activity shown to be expressed by BO egg sheath isolates and AN
gland isolates may also be a symbiotic function as the egg sheaths are observed to be
resistant to fungal parasites during their incubation period. The bacterial production of an
antifungal compound compatible with the tissue of newly developing squid has widespread
pharmaceutical implications. Most antifungal compounds will indiscriminately damage all
eukaryotic cells, not just fungal ones, so a fungicidal compound specific only to fungal
cells could revolutionize the treatment of fungal infections.
Both the BO and the OG bacteria were shown to produce hydrogen sulfide.
Hydrogen sulfide is a toxic compound and its presence in the egg sheath suggests that
perhaps the squid have developed a defense to this compound. H,S, like putrescine, is part
of the chemical signal of rotting food, and may also serve to deter predators by chemically
disguising the egg cases as rotting material.
Molecular studies were used to determine whether the similarities observed in
morphology and activity profiles among the BO and OG egg sheath isolates and the AN
gland symbionts could be observed at the genomic level. A four primer RAPD PCR
analysis was used to assign ratios of similarity for pairwise comparisons of egg sheath
isolates and AN gland isolates. The highest ratios of similarity achieved were between the
white AN gland isolates and white egg sheath isolates, and between BO egg sheath
isolates, and the newly identified Shewanella species from the AN gland. The OG isolates
were shown to genomically differ from the BO ones, as they consistently produced a
fingerprint distinct from the AN gland isolate or BO egg sheath isolate. This fingerprint
analysis indicated some genomic difference between the BO and the OG bacteria. The ratios
of fingerprint similarity obtained through this 4 primer RAPD analysis much be considered
qualitatively. Robust RAPD PCR analysis of population genetics can be performed with
20 or more primers under optimal conditions, to give more accurate estimations of genetic
similanty. The ratios obtained from the four primer study have been used to assign trends
only. not true indexes of genomic similarity.
lo verify that the egg sheath isolates were the same species as isolates from the AN
gland. as had been indicated for the BO isolate by the RAPD fingerprint analysis, IDNA
probes specific to the new ly identified AN gland symbiont were shown to hybridize to both
the BO and the OG bacteria. The negative control probe to Shewanella putrefaciens did not
hybridize to either isolate, indicating that these isolates were more closely related to the neu
AN gland species, than they were to its close relative in the same genus. This observation
supported the hypothesis that the species found in the AN gland was also found in the ege
sheaths, and further indicated that the OG and BO isolates were closely related on the
species level, with genomic variation due possibly to strain type. The Wisolate from the
egg sheath did not hybridize to the probe for the newly identified AN gland symbiont.
indicating that this isolate was a different species than the BO and OG isolates.
Conclusive evidence for the species homology of the BO, OG and AN gland
symbiotic bacteria was provided by the amplification and sequencing of the l6s ribosomal
DNA gene. The 16s ribosomal DNA gene is used for sequence analysis because it contains
several regions of highly conserved sequence useful for obtaining alignments, yet contains
sufficient sequence variability in other regions of the molecule to serve as a phylogenetic
chronometer (Brock, 1996). Sequence analysis of the 16s genes of the three types of egg
sheath isolates showed that the BO and OG morphologies were closely related on the
species level, as had been indicated by the whole cell hybridization study. Both isolates
had almost identical rDNA sequences when compared to each other and to the sequence of
à previous AN gland isolate. A homology search of the Genbank database produced no
closer match to the BO and OG isolates' 16S rDNA sequences, than the percent similarity
between the egg sheath isolates and the sequence of the AN gland symbiont, leading to the
conclusion that the BO. OG and newly identified AN gland symbiont are of the same
species.
The presence of tw o morphologically different strains of the same species of
symbiont in some of the egg sheaths leads to speculation on the diversity among strains of
the AN gland symbionts. Competition between strains of Vibrio fisheri has been
documented in the light organ of the pacific squid E. scolopes (Lee. K.H., 1991). The
light organ of this squid is analogous to the AN gland of the loliginid squid how ev er it been
specialized to house bioluminescent bacteria and to emit light as a counter-predatory
measure. The studies showed that multiple strains of V. fisheri were capable of colonizing
the light organ and were able to do so when competing strains were absent. In the presence
of à dominant strain, even light organs previously colonized by non-dominant bacteria
quickly became saturated with the dominant strains to the exclusion of the less dominant
strains.
Since BO is the only strain of the newly identified Shewanella species that has been
observed in colonies obtained from AN gland tissue, it is possible that this is the dominant
strain within the species and prevents the growth of the OG strain in vivo. Perhaps
sometimes several of the OG bacteria are able to survive in the AN gland despite the
selective pressure for the BO, and are successfully vertically transmitted into the egg sheath
where, free of the selective pressure, they are able to establish an equal population as the
BO strain. This would explain why only some of the egg sheaths contained the OG strain
while all of them contained the BO and Wstrains, and would further explain why no OG
colonies had been observed during investigations of the AN gland. This hypothesis could
be investigated by plating AN gland tissue homogenate on marine agar plates containing the
antibiotic cefazolin to which the OG isolate is resistant but the BO isolate and W isolate are
not. If enough OG bacteria reside at low levels in the AN gland, they may be observable
on such a culture. However, this method may not be able to detect the presence of only a
few bacteria in the AN gland and another method of selecting for the OG strain would have
to be devised.
The third morphology of bacteria isolated from the egg sheath were the Wisolates
which RAPD RCR analysis had indicated were the same genotype as Wisolates from the
AN gland. The Wisolates were shown not to be of the same species as the BO and OG by
the whole cell hybridization study. and this was confirmed by 16S FDNA sequence
analysis. Previous studies of the AN gland's association with bacteria by electron and
fluorescent microscopy (Lai. 1994) had indicated the possibility that two distinct
morphologies of bacteria were present in the AN gland at varying relative frequency based
on the age of the squid. Whole cell images of the Wisolates roughly match the sectioned
images presented by Lai. The two morphologies described were a rod shaped bacterium
which would correspond to the BO and OG isolates from this study, and a smaller rounder
morphology, which would correspond to the Wisolate from this study. The different
morphologies in Lai's study were recognized to possibly be due to differences in the
bactenum's orientation during sectioning. However, differences in the maturity levels of
the squid examined also correlated to differing morphologies of bacteria.
A homology search through Genbank for the Wisolate's sequence presented two
new species of the genus Roseobacter as the nearest relatives based on 165 rDNA sequence
with a percentage of identity of 86.6% and 84.4%. These percentages of TDNA homology
are too low to assign genus or species verification. However, it is interesting to note that
Roseobacter algicola, the first identity suggested, is a newly identified marine symbiont of
the dinoflagellate Prorocentrum lima (Lafay, 1995). And the second identity suggested is
an unidentified species of a Roseobacter symbiont of the gall tissue of the marine red alga
Prionitis lanceolata (Ashen, Goff, 1996).
Studies by Hena Barbieri (1996) have indicated that the accessory nidamental gland
of the Atlantic squid Loligo pealei is inhabited by a diverse population of bacteria. Twenty
morphologically distinct colonies were isolated from the L. pealei AN gland. Using rRNA
sequence analysis. one of these morphological types was most closely related to the genus
Roseobacter with 93% identity to Roseobacter denitrificans. It is possible that the W
bactera is a new species within the genus Roseobacter, a member of the alpha sub-group
of the proteobacter.
Conclusion
This study has provided evidence to support the theory that the squid Loligo
opdlescens vertically transmits a reproductive gland bacterial symbiont to the egg sheaths of
its embryos by demonstrating that the bacteria isolated from freshly spaw ned egg sheaths
were of the same newly identified species as isolates from the AN gland. Diversity within
this species was indicated by the presence of a morphologically different strain in 7 out of
the 15 egg sheaths examined, indicating possible competition between strains for
colonization of the gland, analogous to the competition observed in the light organ of E.
scolopes. These sheath isolates were shown to express potentially symbiotic activities such
as the production of hydrogen sulfide, protease activity and antifungal activity. Also
indicated was the presence of a second previously unidentified symbiont of the AN gland
and egg sheaths, which appears be vertically transmitted as well.
Acknowledgments
Iwould like to thank Melissa Kaufman for her guidance, generosity and patience in
teaching me everything I needed to know for this project. Thanks also to Professor Epel,
and his lab. Nancy, Beth and Lisa. for making me feel so welcome this quarter.
Literature Cited
Ashen, JB, Goff L. "Molecular-identification of a bacterium associated with gall
formation in the marine red alga Prionitis-lanceolata" Journal of Phycology v. 32(2) pp.
286-297. 1996.
Barbieri, Elena. "New Evidence for Bacterial Diversity in the Accessory Nidamental Gland
of the Squid (Loligo pealei)." Biological Bulletin v 191 pp. 316-317. 1996.
Biggs, James and David Epel. "Egg Capsule Sheath of Loligo opalescens Berry: Structure
and Association with Bacteria." The Journal of Experimental Zoology. v.259 pp. 263.
267. 1991.
Bloodgood, Robert A. "The squid accessory nidamental gland: ultrastructure and
9 —
association with bacteria Tissue and Cell v. 9(2) pp. 197-208. 1977.
Brock. Biology of Microorganisms 7th ed. New York : Prentice Hall. pp. 618. 1996.
Hoerner, Michelle. "Analysis of Putrescine in the egg case of Loligo opalescens
Potential Roles in Predator Deterrence. " Hopkins Manne Station Final Papers 175H.
1996.
Kaufman, Melissa R. Personal communication.
Lafay, B, et al. "Roseobacter-algicola sp-nov, a new marine bacterium isolated from the
phycosphere of the toxin producing dinoflagellate Porcentrum-lima" Intemational Joumal
of Systematic Bacteriology. v. 45(2) pp. 290-296. 1995.
Lai, Cindy J. "The Association between symbiotic bacteria and the accessory nidamental
gland of squid Loligo Opalescens." Hopkins Marine Station Final Papers 175H.1994.
Ruby, Edward G. "Lessons from a cooperative, bacterial-animal association: the librio
fischeri-Euprymna scolopes Light Organ Symbiosis." Annual Review of Microbiology. v.
50 pp. 591-624. 1996.
Table 1. Age since spawning for each egg sheath and ethanol treatment times of
whole sheath and section
sheath
time since
Ethanol
umber
spawning
treatment
times for
sheath:section

3h5
2:0 min
« 3hrs
2:0 mir
trial 2
« 3hrs
2:2 min
« 3hrs
2:2 min
« 3hrs
2:0 min
trial
-1 day
2:2 min
-1 day
2:4 min
1 day
2:6 min
1day
2:0 min
1 day
10
2:0 min
trial 4
+-7 days
2:2 min
12
-7 days
2:4 min
13
-7 days
2:6 min
-7 days
114
2:8 min
115-1
-7 days
2:0 min
115-2
2:2 min
U15-3
2:4 min
L1s-4
2:6 min
TABLE 2. Average CFU per gram of sheath tissue.
(total EtOH treatment times in parentheses)
sheath
B0
6
number
11(2)
1.20E404
2.245405
2(2)
1.43E402
5.37E404
3(4)
4.67E402
1.67E402
1.07E404
4(4)
1.79E402
O.OOE400
1.20E403
95(2)
5.O0E405
5.71E404
3.14E405
16(4)
4.00E403
1.07E40
8.00E403
17(6)
1.50E403
7.00E402
2.10E403
8(8
4.00E402
6.63E403
1.60E404
9(2
1.40E405
2.445404
3.78E404
110(2
2.45E405
5.50E404
2.00E404
11(4)
1.20E404
2.35E403
112(4)
2.00E40.
2.50E405
113(4)
O.OOE4OC
6.00E402
114(2)
1.O0E402
9.00E403
115-1(2)
1.50E405
8.60E404
115-2(4)
5.30E404
5.90E404
115-3(6
O.OOE400
4.00E404
115-4(8)
O.00E400
O.OOE+00
Table 3. Activity Assays
protease activity
produces H2S
antifungal activity
prown-
prange
YES
VES
YES
orange¬
green
YES
VES
not tested
white
NO
NO
NO
Table 4. Antibiotic sensitivity assay
ANTIBIOTIC RESISTANCE
bacterial isolates from egq sheath
Antibiotic
white
brown-orange orange green
Ampicillin
Bacitrancin
Cefamanoole
Cefazolin
Ciprofloxacin
Chloramphenicol
Clindamycin
Erythromycin
Kanamycin
Moxalactam
Neomycin
Nalidixic acid
Penicillin
Polymyxin B
Rifampin
Streptomycin
Sulfisoxazole
Sulfamethoxazole/trimethoprim
Tetracycline
Tobramycin
NO RESISTANCE
Table 5. Results of Biolog identification search.
BIOLOG ASSAY
Bacterial isolate
Proposed ldentity
Shewanella putrefaciens A
18 Brown-orange
J8 Orange-green
Shewanella putrefaciens B
No 1D
92-1 White
Shewanella putrefaciens B
CDC group EF-4
Psychrobacter immobilis
Psychrobacter immobilis
92-2 White
CDC group EF-4
CDC GTOUD EF4
91 White
Vibrio Cyclosites
similarity
0.917
0.798
0.145
0.135
0.134
0.651
0.0
0.649
0.062
Table 6a.
Ratio of similarity between banding patterns obtained
by Random Amplified Polymorphic DNA (RAPD) PCR
Analysis of bacterial isolates
ISPATCC
8026
18910
L8 B0
primer
SPATCC
0.57143
0.6
S926
0.75
0.6
SD010
0.5
J8 B0
1806
J8 W.
12 W
primer 2
SPATCC
8910
8926
8 B0
SPATCC
O.44444
0.5
0.66667
S926
0.75
0.75
SD010
0.75
J8 B0
1806
38 W
12 W
primer 3
SPATCC
8926
18910
8 B0
SPATCC
0.4
0.5
8026
O.83333
0.66667
S0010
0.66667
J8 B0
1806
18 W
32 W
primer 4
SPATCC
8026
8910
J8 B0
SPATCC
O.42857
8026
0.90909
0.7692:
S0010
0.66667
18 80
1806
18 W
12 W
SPATCC
average
8026
S910
J8 80
SPATCC
O.36786 0.44167
0.3432
5026
0.81061
0.69647
SDQ10
0.64583
J8 80
J8 0G
J8 W
12 W
D806
0.4
0.75
O.33333
0.25
1806
O.44444
0.28571
0.25
O.5714:
806
O.3076.
0.42857
1806
O.36364
O.54545
0.5
J8OG
0.35202
0.47222
0.27083
O.4375
8 W.
18 W.
0.3333:
O.18182
O.36364
O.36364
O.54545
8 W.
0.16667
0.26667
O.1666.
0.25
O.16667
18 W.
O.18182
-2-
O.53333
0.6666
J8 W
0.22727
0.14949
0.17677
0.38232
O.4596
12 W.
0.22222
0.28571
O.44444
12 W
O.30768
O.16667
O.3333:
O.33332
0.9375
2 w.
0.16667
0.26667
O.16667
0.25
0.16667
2 W.
O.16667
0.16667
O.15385
0.375
0.61538
0.92308
12 W
0.16026
0.20556
0.23489
0.35060
0.32051
0.95353
Table 6b. Ratio of similarity of banding patterns between amplification fragments of
egg sheath isolates J2W and J8W and the banding pattern of AN gland isolate MK84.
K84 vs 12W
NK84 vs 18W
primer 1
nO DNA
0.66
primer 2
0.9
primer 3
0.92
0.94
primer
0.75
23
average
0.8275
10.95553

Table 8. Results of sequence homology search through Genbank
Close relatives egg sheath isolates.
FASTA Homology search
% similarity
orange-green
parophillic Shewanella
(strain 5501)
92.6
Shewanella benthica
91.9
hewanella sp. strain
DB172F
91.8
Shewanella sp. strain
DB172R
91.6
Shewanella strain
DB 6906
91.3
Shewanella strain
D5512
90.8
Shewanella strain
90.7
DB6705
Shewanella alga
89.6
89.3
B. gelidimarina
FASTA Homology search
6 similarity
white
Roseobacter algicola
86.6
Vs.
P. lanceolata gall
bymbiont/Roseobacter
84.6
Octadecobacter arcticensis
84.4
Marinosulfonas
methylotrophus
84.5
UNIP a-Proteobacterium
84.4
Silicibacter lacuscaerulensis
83.4
Paracoccus alcaliphilus
82.5
Paracoccus aminophilus
81.8
Paracoccus aminovorans
81.8
Figure Legends
Averaged ethanol treatment effects on colony formation units per gram.
Figure 1.
The cfus/g decrease linearly as the length of the Ethanol treatment times
increase.
Averaged ethanol treatment effects on cfus/g of the three morphologies
Figure 2.
isolated in trials 2 and 3.
Figure 3.
3a) Protease plate. Zone of clearing around BO and OG isolates indicate
proteolytic activity. No zone of clearing is present around white isolate.
3b)TSI plate. Black precipitate in bacterial patch indicated the production of
hydrogen sulfide. BO and ÖG both have a black precipitate, the white
colony does not.
Electron Micrographs of BO, OG and Wisolates.
Figure 4.
a) BO bacteria in log phase. Cell is dividing by binary fission.
b)W bacteria in stationary phase. Note F pilus for DNA exchange/mating
protruding from the side of the bacterium.
e)OG bacteria with single polar flagellum. Hair-like lines extending from the
surface of bacteria are most likely staining artifacts.
Two RAPD gels showing amplification patterns with 2 different primers.
Figure 5.
Similar genomic finger prints shown among the BO isolates, lanes 4-7 and
among the white isolates lanes 10-13. Different finger prints obtained
for the OG isolates compared to the BO isolates.
lanes:
8-J8 ÖG: ÖG from sheath J8
1- size standard
9- J8 ÖG: ÖG from sheath J8
2- No DNA control
10-JI W: W from sheath J1
3-Ecoli
4- SQ26: Sequence of AN isolate
11-J2 W: W from sheath J2
5- SQ1O: AN gland isolate
12-J8 W: W from sheath J8
6-JI BO: BO from sheath J
13-MK84: WAN gland isolate
7- J8 BO: BO from sheath J8
Figure 6.
Graph of RAPD analysis ratios of similarities among pairwise comparisons
of isolates. Note that the BO isolates have a high ratio of similarity when
compared to the previous sheath and AN gland isolates SQ26 and SQ1O.
The W bacteria have a high ratio of similarity when compared to MK84. a
previous white isolate from the AN gland.
Figure 7.
Photos of whole cell hybridization for BO isolate.
Note positive hybridization signals for the EUB and SQ26 probe and
negative hybnidization signals for the SPN and no probe treatments.
Figure 1.
Ethanol treatment effects on
colony formation units per gram
120000
oo000
80000
60000
40000
20000

E
EE
2min
4min
6+min
Ethanol treatment time
W
E80
Figure 2. Ethanol treatment effects on the three morphologies isolated in trials 2 and 3.
ETHANOL EFFECTS ON TRIALS 2
AND 3 INCLUDING ORANGE
GREEN ISOLATI
200000
180000
160000
140000
120000
100000
80000
60000
40000
nW
20000
BO
o
E
4min
2min
6 min
Ethanol treatment time
o

5
a.
8 a
Figure 4. a)
Brown-orange
Figure 4. b)
White
Figure 4. c)
Figure 4. d)
Figure 4.e)
Figure 5.
ratio of similarity between RAPD PCR banding pattems
— Noo
J8W -J2W
MK84-J8W
MK84-12W
S026-5910
J8B0-S926
J8B0-S910
J80G-S026
J8W-1806
SQ1O-SPATCC
J8B0-J806
J8B0-J8W
SPATCC-S926
J8B0-12W
J8OG-SPATCC
SPATCC-J8B0
J80G-12W
J80G-SQ10
J2W-S910
SATCC-8
S926-12w
J8W-S910
J2W-SPATCC
J8W-S926
Figure 7.
Brown-orange
Rrown-orange
Brown-orange
Brown-orange
No Probe
EUB
S926
SPN