Characteristics of the Mucus Layer on the Surface of the Bluegill (Lepomis Macrochirus) and the Bacterial Flora in the Mucus

Microbes Environ. Vol. 20, No. 1, 69–80, 2005 http://wwwsoc.nii.ac.jp/jsme2/

Characteristics of the Mucus Layer on the Surface of the Bluegill (Lepomis macrochirus) and the Bacterial Flora in the Mucus

TAKEAKI HASHIZUME1, CHIKAKO TAKAI1, MANAMI NAITO1 and HISAO MORISAKI1*

1 Department of Bioscience and Biotechnology, Faculty of Science and Engineering, Ritsumeikan University, 1–1–1 Nojihigashi, Kusatsu, Shiga 525–8577, Japan

(Received October 9, 2004—Accepted December 24, 2004)

The layer of mucus on the surface of bluegills (Lepomis macrochirus) captured in Lake Biwa was characterized as 1) large enough to host microbes (ca. 76 m thick), 2) a physically different environment from the surrounding lake water in viscosity and buffering capacity, and 3) chemically rich in organic substances, which may be utilized as nutrients. Based on DAPI staining and on the number of colonies formed respectively, it was found that ca. 103 times and 3 to 7 times the number of microbial cells were present in the mucus layer, as compared with the lake water. The bacterial flora of the mucus was greatly different from that of the lake water, according to a phylogenetic analysis. About 60% of the isolates from the mucus were Gram-positive. These Gram-positive isolates could be divided into two major groups. Each group consisted of strains sampled in one season, i.e., the strains sampled in July were closely related to the genus Staphylococcus, while the strains sampled in November were close to the genus Mycobacterium. In contrast, most isolates from the lake water were Gram-negative (72%); with all the strains closely related to - and -Proteobacteria sampled in July. With the exception of one strain, the Gram-positive isolates from the lake water (6 strains) were all sampled in November. Almost all of the isolates from the mucus could metabolize glucose, whereas only half of the isolates from the lake water could do the same.

Key words: mucus layer, bacterial flora, bluegill (Lepomis macrochirus), Lake Biwa, 16S rDNA sequence

Microbes inhabiting fish were reported as early as 110 The purpose of the present study was 1) to characterize years ago (the first was a Vibrio isolated from a diseased the mucus layer of fish, as a habitat for microbes, and 2) to fish5)). Since then, many kinds of microbes causing diseases reveal the characteristics of the bacterial flora in the layer in in fish have been investigated3,7,14,15,23). These, as well as relation to the properties of the mucus, in comparison with non-pathogenic microbes, seem to inhabit fish and interact the relationship between the microbes in the adjacent water with the surrounding environment. The surface of the fish is and the properties of that water. The bluegill Lepomis mac- the outermost boundary of that organism, and interacts di- rochirus was selected, because 1) bluegills, which have rectly with the outside environment. Fish are covered by a been causing a serious ecological problem by expelling in- layer of mucus, which has different characteristics from the digenous species, now account for 80–90% of the fish popu- environment outside, water; it is a place where various sub- lation in Lake Biwa19), which means that we could readily stances or possibly microbes are exchanged with the water obtain experimental animals for our study; and 2) the spread outside. However, few attempts have been made to charac- of this species over various areas of Japan makes it possible terize the microbes in mucus in connection with the proper- to investigate the environmental factors affecting the micro- ties of the mucus. bial flora on fish in further studies. In the present paper, we report the physicochemical fea- * Corresponding author; E-mail: [email protected], Tel: tures and seasonal changes of the mucus layer of the blue- 81–77–561–2767, Fax: 81–77–561–2659 gill. We also report that the bacterial flora in the mucus dif- 70 HASHIZUME et al. fer in character from those in the surrounding water, and NaCl (pH7.2). The DNB was a 100-fold dilution of the NB undergo seasonal changes. medium. NB plates and DNB plates were prepared for each sample in triplicate and incubated aerobically at 20LC for 30 Materials and Methods days, and the number of colonies appearing every day was counted. The colonies were distinguished each day by pen Sampling markings of different colors and shapes made on the back of The fish (bluegills, or Lepomis macrochirus) used in this the Petri dishes. After the incubation period, strains were study were extracted from their environment without being randomly isolated from the colonies. touched by bare hands and without disturbance of the ground at the site (34L58'30'' N, 135L54'30'' E), where the DNA extraction river Seta flows out from Lake Biwa (the largest lake in Ja- Each isolate was cultured in 10 mL of NB medium, with pan). The fish were placed in a box containing lake water shaking at 100 rpm and at 27LC, for 1 to 2 days. The cell and brought back to the laboratory within several hours. A suspension at the exponential growth-phase was centrifuged 300-mL sample of water was also obtained. The fish were (4LC, 10 min, 12000Pg). Then, the cell pellet was re-sus- kept in a state of syncope by cooling to 4LC, and then cov- pended in distilled water. The cell suspension was frozen ered with a perforated polyethylene rubber sheet (for large with liquid nitrogen and thawed three times. After thawing, fish, the perforated area was rectangular, 5 cmP3 cm, and 10 L of proteinase K (10 mg/mL) and 50 L of BL for small fish, 5 cmP2 cm). Surface mucus from each fish buffer12) (containing 40 mM of Tris aminomethane, 1 mM was gathered through the perforated area, as shown in Fig. of EDTA·2Na, 1% of Tween 20 and 0.5% of nonidetP-40) 1, with a sterilized rubber scraper. were added to 50 L of the suspension. After incubating at 60LC for 30 min, the supernatant (4LC, 10 min, 12000Pg) Total number of particles was used for amplification by PCR. The total number of particles showing bright blue fluorescence after DAPI (4',6-diamidino-2-phenylindole; 1 g/ PCR amplification and sequencing of the 16S rRNA gene mL final concentration) staining of each sample was count- The PCR mixture consisted of 0.75 units of Takara Ex- ed under an Olympus model BX50 BX-FLA epifluores- Taq, 1x Taq polymerase buffer, 200 M of dNTPs, 50 pmol cence microscope. of each primer and the extracted DNA (50 ng to 100 ng), all in a 20- L reaction mixture. The PCR primers used for am- Isolation of microbes plifying the 16S rDNA of the isolated bacteria were 25F17) The mucus samples used for the isolation of microbes (5'-AGTTTGATCCTGGCTC-3') as the forward primer and were collected on July 5 and November 7, 2002. The mucus 1510R20) (5'-GGCTACCTTGTTACGA-3') as the reverse was diluted 103-fold with sterilized water. NB or DNB agar primer, corresponding to positions 10–26 and 1495–1510, (1.5 wt%) medium (ca. 20 mL) was poured onto the diluted respectively, in the 16S rRNA gene sequence of Escheri- mucus samples (1 mL) and mixed. The NB medium con- chia coli. tained (per liter) 10 g of polypepton from Nihon Seiyaku, 10 The thermal cycling program used was as follows: 5 min g of bonito extract from Wako Pure Chemicals and 5 g of initial denaturation at 95LC, then 30 cycles at 95LC for 1 min, 52LC for 2 min, and 72LC for 2 min, and a final exten- sion for 10 min at 72LC. The PCR products were analyzed by electrophoresis on 1% Takara Agarose LO3 gels in 1PTAE. A 200-bp Bexel DNA Marker was used as the molecular weight standard, and stained with ethidium bromide (1 g/mL with 1PTAE buffer) for visualization. The PCR products were purified with a Viogene PCR-M Clean-Up System in accordance with the manufacturer’s instructions. The sequences were determined with the primer 907R24) (5'-CCGTCAATTCCTTTGAGTTT-3') and an ABI PRISM AVANT 3100 genetic analyzer (PE Biosystems, Foster Fig. 1. The surface area (a rectangle, 3 cm×5 cm) from which the City, USA). The BigDye Terminator Cycle Sequencing mucus was obtained. Ready Reaction Kit, ver. 3.1 (PE Biosystems, Foster City, Bacteria in the Mucus Layer of Bluegills 71

USA), was used in accordance with the manufacturer’s di- fish on October 18. Each sample of mucus collected was rections. centrifuged (4LC, 14000Pg, 30 min) to remove black substances. The mucus was diluted 10, 50 and 100 fold with Phylogenetic analysis distilled water. The viscosity of each diluted mucus solution Approximately 500 bp were used for the phylogenetic was determined three times with a Sibata No. 2 Ostwald vis- analysis. Unaligned sequences were submitted to the Ad- cometer at 25LC, and the results were averaged. The viscos- vanced BLAST search program at the website of the Na- ity of each mucus solution relative to that of distilled water tional Center of Biotechnology Information (NCBI), in or- was plotted against the logarithmic value of the dilution rate der to find closely related sequences. The sequences were of the mucus (e.g., log 50 for a 50-fold dilution). Then the aligned with the CLUSTAL X program (version 1.83)32). viscosity of non-diluted mucus was deduced from the linear Nucleotide positions containing ambiguous alignments and relation (confirmed beforehand in another experiment) be- gaps were omitted from the subsequent phylogenetic analy- tween the relative viscosity and the dilution rate. sis. The analysis was performed with CLUSTAL-X, using the neighbor-joining method of Saitou & Nei26). pH value of the mucus The pH values of the mucus collected from the fish and Nucleotide sequence accession numbers that of the lake water were measured with Advantec BTB All the sequences of the 16S rRNA gene determined in test paper and a Horiba model F-13 pH meter, respectively. this study have been submitted to DDBJ under the accession numbers AB192350 to AB192409 and AB196729. Buffering capacity of the mucus The mucus remaining after the pH measurement was di- Morphological and physiological traits of the isolates luted ten-fold with distilled water. To the diluted mucus so- We characterized the isolated strains by examining their lution (8 mL), an aqueous solution of 0.01 M HCl or NaOH morphological and physiological traits. The cell shape, mo- was added in 500 L increments at room temperature. The tility and Gram-staining reaction were examined for the buffering capacity of the lake water at the sampling site was cells incubated for 1 to 2 days at 20LC in NB liquid medium determined as well: 0.01-M HCl or NaOH solution was add- with shaking at 100 rpm. The isolates were also character- ed in 10 L increments. ized using oxidase, catalase and oxidation-fermentation (OF) tests. Concentrations of sugar and protein in the mucus The concentrations of sugar and protein in the mucus Mucus density were measured by the phenol-sulfuric acid assay10) and the Mucus was collected from 26, 12, 18, 8 and 20 fish (cap- Bradford assay4), respectively, as equivalent to glucose and tured on July 5, 18, and 29, October 18 and November 7, bovine serum albumin (BSA). The concentrations of sugar 2002, respectively). A 200- L volume of the mucus was and protein were measured for 26 and 20 individual fish weighed with a balance and divided by volume, in order to sampled on July 5 and November 7, 2002, respectively. obtain the mucus density. Results and Discussion Thickness of the mucus layer The mucus was collected from a defined surface area of The characteristics of the mucus layer each fish (see Sampling, described above). The weight of The thickness of the layer of mucus on the skin differed the mucus obtained from each fish was measured and divid- greatly among individual bluegills. The minimum value was ed by the mucus density (determined as described above) ca. 30 m and the maximum was ca. 120 m, with no clear and the surface area from which the mucus was collected, in relationship between the thickness and fish size, which order to determine the thickness of the mucus layer on the ranged from 9.5 cm to 17.7 cm in length. The mean value of surface of the bluegills. the thickness was about 76 m for the fish captured on 6 different occasions (Fig. 2). Although the standard devia- Viscosity of the mucus tion of the thickness was rather large, the mean value We collected mucus from 6 bluegills on July 29 (2002) seemed to decrease in samples obtained during late autumn and another two sets from 6 fish on the same day; we also (see Fig. 2). collected mucus from 18 fish on September 24 and from 8 The density of the mucus (1.038, 1.075, 1.025, 1.040 and 72 HASHIZUME et al.

lake water, indicating that the mucus has a large buffering capacity. The buffering capacity of the 10-fold diluted mucus was experimentally revealed to be equivalent to ca. 2.3 mM and 12 mM of phosphate buffer, respectively, when an acidic or alkaline solution was added. The concentration of the constituents corresponding to sugar and protein in the mucus differed significantly (p0.01, Student’s t-distribution) between the samples obtained on different occasions (see Fig. 4a, 4b, 4c and 4d). The mucus of the fish captured in summer (July) contained more organic substances (Fig. 4a and 4c) corresponding to sugar and protein (equivalent to 2.0 mg of glucose and 1.8 Fig. 2. The thickness of the mucus layer on the bluegills on various mg of BSA per milliliter of mucus), compared with that of sampling dates. The mean values and the dates for 18 (7/29), 10 the fish captured in late autumn (November), equivalent to (8/2), 10 (9/17), 18 (9/24), 6 (10/11) and 8 (10/18) fish. The bars 0.81 mg of glucose and 1.2 mg of BSA per milliliter (Fig. represent the standard diviation. 4b and 4d). June and July are the spawning months for the bluegills in Lake Biwa31); as shown in Fig. 2, the mucus 1.055 g/cm3 at 25LC for the samples taken on July 5, 18, and thickness was greater during these months than in late au- 29, October 18 and November 7, 2002, respectively) was al- tumn. Although we could not distinguish male from female most equal to that of water. However, the mucus was much bluegills, the difference in the life cycle of bluegills may more viscous than water; its viscosity relative to water was correlate with the mucus thickness as well as the concentra- between 1.49 and 1.72. tion of organic substances, as revealed in this study. The The mucus was acidic and showed a rather constant pH factors causing these changes and the function of the organ- value of ca. 6.4, regardless of the sampling dates and pH ic substances, including the possibility of their being uti- value of the lake water which differed on different dates lized as nutrients, are topics of further study. (7.4 on Aug. 2, 7.2 on Sep. 17 and 8.5 on Oct. 11). Based on the characteristics of the mucus revealed as de- It is noteworthy that the mucus displayed a buffering ca- scribed above, it can be deduced that the bacteria live in re- pacity against the addition of acidic or alkaline solution, as gions 1) large in size (several ten times greater than the cell shown in Fig. 3. The slope of the pH change, averaged for size), 2) physically different from water (more viscous) and three samples obtained on August 2, September 17 and Oc- 3) capable of maintaining a stable habitat despite outer fluc- tober 11 (2002), caused by the addition of the HCl (Fig. 3a) tuations, as exhibited by the constant pH value, the high or NaOH aq. (Fig. 3b) solution to the 10-fold diluted mucus buffering capacity of the mucus and the rich concentration was, respectively, ca. 1/47 and 1/15 of the pH change of the of organic substances, which may be utilized as nutrients.

Fig. 3. The change in the pH values of 10-fold diluted mucus and lake water samples, along with the addition of HCl (a) or NaOH (b) aq. solution. The mucus was sampled on Aug. 2 ( ), Sep. 17 ( ) and Oct. 11, 2002 ( ), and the lake water ( ) on Aug. 29, 2004. Bacteria in the Mucus Layer of Bluegills 73

Fig. 4. Concentrations of sugar (a, b; equivalent to glucose) and protein (c, d; equivalent to BSA) in the mucus sampled on two different dates. Each individual fish is represented by a different letter.

decaying cells. Bacteria in the mucus layer Figure 6 shows the change in the number of bacterial col- The numbers of particles showing pale fluorescence after onies along with the incubation time in the nutrient broth staining with DAPI were 2.70P109/mL (SE1.33) and (NB) and in a 100-fold diluted nutrient broth (DNB) agar 1.56P109/mL (SE0.23) for the samples taken in July and medium. In the mucus samples of July and November, more November 2002, respectively. The number of particles in colonies were formed in the NB medium, than the DNB me- the lake water was about three orders of magnitude lower dium (Fig. 6a and 6c), which may be due to the higher con- than that in the mucus, i.e., 3.78P106/mL (SE0.76) and centration of organic substances in the mucus. 1.51P106/mL (SE0.41) for the samples taken in July and The above results differ greatly from those observed in November, respectively. Although the number of particles various soil samples, where the number of colonies formed was greater in the mucus, most of the particles (ca. 73%) on the DNB medium was one or two orders of magnitude had vague outlines and were less than 1 m in size (see Fig. greater than that on the NB medium11,18,25). Although this 5a). In contrast, most of the particles in the lake water (ca. was the case for the water sampled at other sites in Lake 81%) had sharp outlines and were ca. 1 m in size (see Fig. Biwa (Matsuda and Hisadome; personal communication), 5b). The particles in the mucus probably consisted mostly of the difference was small for the water sample taken in No- 74 HASHIZUME et al.

Fig. 5. Mucus and lake water samples viewed by epifluorescence microscopy after DAPI staining. The mucus sample (a) was diluted 1000-fold with distilled water before staining, while the lake water (b) was not diluted.

Fig. 6. The number of colonies that appeared on the NB and DNB agar media, along with the incubation period. Both samples (a: mucus; b: lake water) were obtained on July 5, 2002 and other samples (c: mucus and d: lake water) were obtained on November 7, 2002. The mean numbers of colonies for NB ( ) and DNB ( ) are plotted. The means of SD of the numbers of colonies were as follows: mucus: 19.3 (NB) and 4.37 (DNB); lake water: 2.38 (NB) and 1.71 (DNB) (July 5, 2002). mucus: 97.6 (NB); 20.4 (DNB); lake water: 4.44 (NB) and 8.20 (DNB) (No- vember 7, 2002). vember (Fig. 6d) and even smaller for that taken in July the numbers of formed colonies between the NB and the (Fig. 6b) in the present study. The total amount of organic DNB medium. carbon in the water near the sampling site was ca. 1.5 to 2 In the mucus samples, as well as the lake water samples, times that in the water of Lake Biwa in 200228). Such water the number of colonies increased mainly in the first 10 days. quality may be one of the reasons for the small difference in It has been revealed18) that isolates that take a shorter time to Bacteria in the Mucus Layer of Bluegills 75 form visible colonies show greater growth rates18). Thus, the Kocuria (1 strain) and Microbacterium (1 strain), as shown isolates from the mucus layer and the lake water could con- in Fig. 7. A seasonal change, as found in the isolates from sist mainly of fast-growing bacteria. the mucus, was also observed in the isolates from the lake water. The strains closely related to - and -Proteobacteria Analysis of isolates through 16S rRNA gene sequencing were all sampled in July. With the exception of one strain, From the colonies formed on the NB or DNB nutrient Lo4, the Gram-positive isolates from the lake water were all agar medium, strains were isolated at random and analyzed sampled in November. through 16S rRNA gene sequencing. Phylogenetic trees Thus, the bacterial flora in the mucus and the lake water constructed using these sequences are shown in Figures 7 were very different, reflecting characteristic seasonal and 8. changes. The isolates from mucus contained a few dominant groups, i.e., strains close to the genera Staphylococcus, My- Morphological and physiological characteristics of the cobacterium and Plesiomonas. In contrast, the isolates from isolates the lake water were dispersed over more genera, with no We selected some strains randomly from the isolates be- discernible dominant group. longing to each cluster in the phylogenetic trees shown in Of the 36 strains found in the mucus, 22 were Gram-posi- Figures 7 and 8: one or two strains from the cluster consist- tive. Among these, a distinct difference was observed being of a small number of isolate(s) and three or five strains tween the July and November samples, as shown in Fig. 7. from the cluster consisting of a greater number of isolates. Most strains of July (15 out of 16 strains) were closely relat- The morphological and physiological traits of the selected to the genus Staphylococcus, whereas the strains of No- ed strains are shown in Table 1. These traits matched those vember were all close to the genus Mycobacterium (6 described in the literature1,13,16,27,29,30), for the genera to strains). As mentioned before, the organic substances equiv- which the closest relatives belonged, with the exception of alent to glucose and BSA in the mucus of the fish were ca. 2 some differences in the results of the OF test; strains Q27 times greater in concentration in the July than November and AMi17 were omitted because of their very low similari- samples. This may correlate with extensive seasonal chang- ty index. In two mucus strains (H18 and S9), no color es in the bacterial flora. More studies are necessary to con- change was observed in the OF tests, whereas they should firm these findings, as well as to reveal the factors affecting have been oxidative, according to the literature13,30). In the the bacterial flora. water samples, this type of mismatch was more frequent. The rest of the isolates from the mucus were Gram-nega- Seven (Hi5, Hi7, Hi10, Lo6, Lo24, AHi5 and ALo7) out of tive bacteria (14 strains out of 36), with no clear difference 16 strains showed no color change in the OF test, regardless between the July and November samples. Among them, of the oxidative reaction proposed in the literature13,27). The many strains (10 strains) belonged to -Proteobacteria, isolates from the water samples in the present study may while other strains (3 and 1) belonged to -Proteobacteria have been weak in glucose metabolic activity. As a result, a and the CFB group, respectively, as shown in Fig. 8. It is distinct difference appeared in glucose metabolism between noteworthy that no isolate affiliated to -Proteobacteria the isolates from mucus and water. Fifty-six percent of mu- could be found in the mucus, although this group of bacteria cus strains metabolized glucose by fermentation, whereas is often found in aquatic environments2,6,21), as well as in the 33% did so by oxidation; 11% did not utilize glucose at all. isolates from the lake water sampled in the present study, as In contrast, many water strains (47%) could not utilize glu- shown below. cose, whereas 33% metabolized glucose by oxidation and Contrary to the isolates from the mucus, most isolates (18 20% by fermentation. As revealed in this study, the mucus out of 25 strains) from the lake water were affiliated to - is rich in organic substances (equivalent to 0.81–2.0 mg of Proteobacteria (5 strains), -Proteobacteria (3 strains; one glucose per milliliter of mucus). The bacteria living in the strain AMi 17, belonging to -Proteobacteria, was omitted mucus layer probably utilize these substances, resulting in from Fig. 8 due to its incomplete sequence-determination, the high ratio of glucose-utilizing strains (89%). Although which made the construction of a phylogenetic tree impossi- the mucus layer was thin (ranging from 30 m to 120 m, ble), -Proteobacteria (7 strains) and the CFB group (3 with an average thickness of 76 m), oxygen diffusion strains), as shown in Fig. 8. The rest of the isolates were through this layer may have been deterred and oxygen may Gram-positive (7 out of 25 strains), and closely related to have been depleted as a result of bacterial metabolism in the genera Bacillus (2 strains), Mycobacterium (3 strains), certain parts away from the interface between the mucus 76 HASHIZUME et al.

Fig. 7. Phylogenetic tree illustrating the relationships between partial 16S rDNA sequences of the isolates from mucus and lake water and reference strains of Gram-positive bacteria. indicates the strains from the mucus and indicates the strains from the lake water (July 2002). indicates the strains from the mucus and indicates the strains from the lake water (November 2002). Bacteria in the Mucus Layer of Bluegills 77

Fig. 8. Phylogenetic tree illustrating the relationships between partial 16S rDNA sequences of the isolates from mucus and lake water and the reference strains of Gram-negative bacteria. indicates the strains from the mucus and indicates the strains from the lake water (July 2002). indicates the strains from the mucus and indicates the strains from the lake water (November 2002). Note that one strain, AMi 17, from the lake water, belonging to -Proteobacteria, was omitted from the figure due to an incomplete sequence-determination, which made the construction of a phylogenetic tree impossible. 78 HASHIZUME et al.

Table 1. Phylogenetic affiliations and physiological traits of isolates

Gram Cell % Strain No. Motility a Catalase Oxidase b staining shape OF test Closest relative (accession No.) Similarity mucus sample S9 rods N Nocardia corynebacterioides (AY438619) 100 F148 rods O Mycobacterium brisbanense (AY012577) 99.6 H28 rods O Mycobacterium brisbanense (AY012577) 99.8 W16 cocci F Staphylococcus warneri (L37603) 100 W21 cocci F Staphylococcus warneri (L37603) 100 W53 cocci F Staphylococcus warneri (L37603) 99.9 W17 cocci F Staphylococcus epidermidis (AY030342) 100 W20 cocci F Staphylococcus epidermidis (AY030342) 100 Q27 rods F Riemerella anatipestifer (AY612170) 93.0 F22 cocci O Chelatococcus asaccharovorans (AJ294349) 99.1 H18 rods N Sphingomonas asaccharolytica (Y09639) 98.7 W33 rods O Pseudaminobacter salicylatoxidans (AF072542) 97.4 B1 rods O Stenotrophomonas maltophilia (X95924) 100 H6 rods O Acinetobacter baumannii (X81660) 99.5 H30 rods F Plesiomonas shigelloides (X60418) 99.8 H44 rods F Plesiomonas shigelloides (X60418) 99.8 Q19 rods F Plesiomonas shigelloides (X60418) 99.8 S7 rods F Aeromonas hydrophila (X74677) 100 F89 rods F Enterobacter aerogenes (AF395913) 99.1 water sample Lo4 rods O Microbacterium sp. (AJ251194) 99.6 AMi16 cocci F Kocuria varians (AF542074) 99.5 AHi45 rods O Mycobacterium canariasense (AY255478) 99.6 ALo7 rods N Bacillus cereus (AY795568) 100 Lo23 rods O Flavobacterium sp. (AB118230) 96.7 AMi4 rods F Thioclava pacifica (AY656719) 97.5 Lo6 rods N Brevundimonas vesicularis (AJ227781) 100 Lo8 rods O Caulobacter sp. (AJ227767) 98.4 AHi5 rods N Caulobacter crescentus (AE006011) 99.8 Hi5 rods N Comamonas aquatica (AJ430346) 100 Lo18 rods O Pseudoxanthomonas mexicana (AF273082) 99.7 Hi7 rods N Acinetobacter sp. (AJ293690) 99.2 Lo24 rods N Acinetobacter sp. (AY055373) 99.9 Hi10 rods N Acinetobacter haemolyticus (AY047216) 98.7 Hi13 rods F Aeromonas hydrophila (AY686711) 100 AMi17 rods N Sterolibacterium denitrificans (AJ306683) 93.7 a O: oxidative metabolism of glucose; F: fermentative metabolism; N: did not utilize glucose. b The closest relative by sequence comparison. and the lake water, a situation already observed in some the isolates from the mucus and the lake water. These activ- biofilms8,9,22). This may have resulted in the high ratio of ities of the isolates in situ, i.e., in the mucus and the lake mucus strains (56%) exhibiting fermentative glucose metab- water, especially the motility, must be investigated in fur- olism. ther studies. There was no clear difference in the results of the catalase In conclusion, the physical and chemical properties of the and oxidase tests and in the ratio of motile strains between mucus layer on the surface of bluegills (Lepomis macrochi- Bacteria in the Mucus Layer of Bluegills 79 rus) was characterized for the first time in the present study, 13) Holt, J.G., N.R. Krieg, P.H.A. Sneath, J.T. Staley and S.T. which revealed a relation between the bacterial flora of the Williams. 1994. Bergey’s Manual of Determinative Bacteriology, 9th edn. Baltimore: Williams and Wilkins. mucus and the characteristics of the mucus. The biological 14) Hoshina, T., T. Sano and Y. Morimoto. 1958. A Streptococcus effect of the mucus layer on the bacteria remains a subject pathogenic to fish. J. Tokyo Univ. Fish. 44: 57–68. for further study. 15) Iqbal, M.M., K. Tajima and Y. Ezura. 1998. Phenotypic identification of motile Aeromonas isolated from fishes with epizootic ulcerative syndrome in Southeast Asian countries. Bull. Fac. Acknowledgements Fish. Hokkaido Univ. 49: 131–141. 16) Kampfer, P., C. Muller, M. Mau, A. Neef, G. Auling, H.J. Busse, We are grateful to Dr. S. Okunishi for his valuable com- A.M. Osborn and A. Stolz. 1999. Description of Pseudamino- ments and support of our experiments. We also thank Miss bacter gen. nov. with two new species, Pseudaminobacter salicy- M. Obayashi, Mr. S. Ikeda and Mr. K. Nakayama for their latoxidans sp. nov. and Pseudaminobacter defluvii sp. nov. Int. J. Syst. Bacteriol. 49: 887–897. help in sampling the bluegills. This study was supported in 17) Kane, M., H. Maeda, T. Fukunaga and K. Nishi. 1997. Molecular part by a Grant-in-aid for Scientific Research (no. phylogenetic relationship between strains of cyanobacterial pico- 14360203) from The Ministry of Education, Culture, plankton in Lake Biwa, Japan. J. Mar. Biotechnol. 5: 41–45. Sports, and Technology of Japan. 18) Kasahara, Y. and T. Hattori. 1991. Analysis of bacterial popula- tions in a grassland soil according to rates of development on sol- id media. FEMS Microbiol. Ecol. 86: 95–102. 19) Lake Biwa Museum. 1998. Effect of social factors on the biologi- References cal environments at small lakes connected to the large lake (Lake 1) Auling, G., H.J. Busse, T. Egli, T. El-Banna and E. Stackebrandt. Biwa) “Heisei 10 nendo Biwako hakubutsukan kyoudou kenkyu 1993. Description of the gram-negative, obligately aerobic, ni- houkoku”, Lake Biwa Museum, Kusatsu, Japan (In Japanese). trilotriacetate (NTA)-utilizing bacteria as Chelatobacter heintzii, 20) Lane, D.J. 1991. 16S/23S rRNA sequencing, p. 115–175. In E. gen. nov., sp. nov., and Chelatococcus asaccharovorans, gen. Stackebrandt and M. Goodfellow (ed.), Nucleic acid techniques nov., sp. nov. Syst. Appl. Microbiol. 16: 104–112. in bacterial systematics. J. Wiley and Sons, Chichester, United 2) Battin, T.J., A. Wille, B. Sattler and R. Psenner. 2001. Phyloge- Kingdom. netic and functional heterogeneity of sediment biofilms along en- 21) Manz, W., K. Wendt-Potthoff, T.R. Neu, U. Szewzyk and J.R. vironmental gradients in a glacial stream. Appl. Environ. Micro- Laurence. 1999. Phylogenetic composition, spatial structure, and biol. 67: 799–807. dynamics of lotic bacterial biofilms investigated by fluorescent in 3) Biosca, E.G., C. Amaro, C. Esteve, E. Alcaide and E. Garay. situ hybridization and confocal laser scanning microscopy. Mi- 37 1991. First record of Vibrio vulnificus biotype 2 from diseased crob. Ecol. : 225–237. European eel Anguilla anguilla. L. J. Fish Dis. 14: 103–109. 22) Marshall, C. Walters, III, Frank Roe, Amandine Bugnicourt, 4) Bradford, M.M. 1976. A rapid and sensitive method for the quan- Michael J. Franklin and Philip S. Stewart. 2003. Contributions of titation of microgram quantities of protein utilizing the principle Antibiotic Penetration, Oxygen Limitation, and Low Metabolic of protein-dye binding. Anal. Biochem. 72: 248–254. Activity to Tolerance of Pseudomonas aeruginosa Biofilms to 5) Canestrini, G. 1893. La malattia dominante delle anguille. Atti Ciprofloxacin and Tobramycin. Antimicrob. Agents Chemother. 47 Ist. Veneto Sci. Lett. Arti Cl. Sci. Mat. Nat. 7: 809–814 (In Ital- : 317–323. ian). 23) Muroga, K., Y. Jo and M. Gam. 1973. Studies on red spot disease 6) Crump, B.C., E.V. Armbrust and J.A. Baross. 1999. Phylogenetic of pond-cultured eel I. The occurrences of the disease in eel cul- 8 analysis of particle-attached and free-living bacterial communi- ture ponds in Tokushima prefecture in 1972. Fish Pathol. : 1–9. ties in Columbia River, its estuary, and the adjacent coastal 24) Muyzer, G., A. Teske, C.O. Wirsen and H.W. Jannasch. 1995. ocean. Appl. Environ. Microbiol. 65: 3192–3204. Phylogenetic relationships of Thiomicrospira species and their 7) Davis, H.S. 1926. A new gill disease of trout. Trans. Am. Fish. identification in deep-sea hydrothermal vent samples by denatur- Soc. 56: 156–160. ing gradient gel electrophoresis of 16S rDNA fragments. Arch. 164 8) DeBeer, D., P. Stoodley and Z. Lewandowski. 1994. Liquid flow Microbiol. : 165–172. in heterogeneous biofilms. Biotechnol. Bioeng. 44: 636–641. 25) Okunishi, S., K. Yokota and H. Morisaki. 2000. Microflora and 9) DeBeer, D., A. Schramm, C.M. Wolfaardt and D.R. korber. 1997. Mineralization Activity in The Bottom Sediment of Lake Biwa. 15 A nitrite microsensor for profiling environmental biofilms. Appl. Microbes Environ. : 1–10. Environ. Microbiol. 63: 973–977. 26) Saitou, N. and M. Nei. 1987. The neighbor-joining method: a new 4 10) Dubois, M., K.A. Gilles, J.K. Hamilton, P.A. Rebers and F. method for reconstructing phylogenetic trees. Mol. Biol. Evol. : Smith. 1956. Colorimetric Method for Determination of Sugars 406–425. and Related Substances. Anal. Chem. 28: 350–356. 27) Segers, P., M. Vancanneyt, B. Pot, U. Torck, B. Hoste, D. Dewet- 11) Gorlach, K., R. Shingaki, H. Morisaki and T. Hattori. 1994. Con- tinck, E. Falsen, K. Kersters and P. De Vos. 1994. Classification struction of eco-collection of paddy field soil bacteria for popula- of Pseudomonas diminuta Leifson and Hugh 1954 and tion analysis. J. Gen. Appl. Microbiol. 40: 509–517. Pseudomonas vesicularis Busing, Doll, and Freytag 1953 in 12) Hiraishi, A. 1992. Direct automated sequencing of 16S rDNA Brevundimonas gen. nov. as Brevundimonas diminuta comb. nov. amplified by polymerase chain reaction from bacterial cultures and Brevundimonas vesicularis comb. nov., respectively. Int. J. 44 without DNA purification. Lett. Appl. Microbiol. 15: 210–213. Syst. Bacteriol. : 499–510. 80 HASHIZUME et al.

28) Shiga Prefectural Government. 2003. White Paper on the envi- plants: proposal of Sphingomonas rosa sp. nov., Sphingomonas ronment of Shiga Prefecture “Shigaken Kankyo Hakusho”, Envi- pruni sp. nov., Sphingomonas asaccharolytica sp. nov., and Sph- ronmental Policy Division Department of Lake Biwa and Envi- ingomonas mali sp. nov. Int. J. Syst. Bacteriol. 45: 334–341. ronment, Shiga Prefecture, Otsu, Japan, (in Japanese). 31) Terashima, A. 1989. Bluegill. p. 506–511. In H. Kawanabe and 29) Stackebrandt, E., C. Koch, O. Gvozdiak and P. Schumann. 1995. N. Mizuno (ed.), The fresh water fishes in Japan “Nihonno tan- Taxonomic dissection of the genus Micrococcus: Kocuria gen. suigyo”. Yama-kei Publishers, Tokyo, Japan (In Japanese). nov., Nesterenkonia gen. nov., Kytococcus gen. nov., Dermacoc- 32) Thompson, J.D., T.J. Gibson, F. Plewniak, F. Jeanmougin and cus gen. nov., and Micrococcus Cohn 1872 gen. emend. Int. J. D.G. Higgins. 1997. The ClustalX windows interface: flexible Syst. Bacteriol. 45: 682–692. strategies for multiple sequence alignment aided by quality analy- 30) Takeuchi, M., T. Sakane, M. Yanagi, K. Yamasato, K. Hamana sis tools. Nucleic Acids Res. 25: 4876–4882. and A. Yokota. 1995. Taxonomic study of bacteria isolated from