Introduction

The genus Algoriphagus, affiliated to the family Cyclobacteriaceae (Nedashkovskaya and Ludwig 2011) of the phylum Bacteroidetes, was proposed by Bowman et al. (2003). At the time of writing, present genus comprises 27 species with validly published names (http://www.bacterio.net/uw/Algoriphagus.html; Euzéby 1997) and two species with invalid names. Algoriphagus ratkowskyi, the type species of Algoriphagus (Bowman et al. 2003) and A. antarcticus (Van Trappen et al. 2004) have been isolated from polar habitats. On the other hand, A. aquimarinus (Nedashkovskaya et al. 2004) originated from Japanese seawater, A. marincola (Yoon et al. 2004; Nedashkovskaya et al. 2007), A. jejuensis (Lee et al. 2012) and A. namhaensis (Oh et al. 2012) isolated from Korean seawater and ‘A. shivajiensis’ (Kumar et al. 2013) was described from a backwater sample collected at India. A. chordae and A. winogradskyi were isolated and characterized from marine alga in Japan (Nedashkovskaya et al. 2004); A. vanfongensis (Nedashkovskaya et al. 2007) was isolated from coral in Vietnam and A. machipongonensis (Alegado et al. 2013) was characterized from a colonial choanoflagellate. The species such as A. yeomjeoni (Yoon et al. 2005a) and A. locisalis (Yoon et al. 2005b) were isolated from Korean marine solar saltern, whereas A. hitonicola (Copa-Patiño et al. 2008) was isolated from a Lagoon in Spain. Several Algoriphagus species have been isolated from Korean tidal flat/sediment that include A. lutimaris (Park et al. 2010), A. chungangensis (Kang et al. 2013), ‘A. boseongensis’ (Park et al. 2014) and A. taeanensis (Kim et al. 2014) and three reclassified species A. halophilus (Yi and Chun 2004; Nedashkovskaya et al. 2004), A. mannitolivorans and A. ornithinivorans (Yi and Chun 2004; Nedashkovskaya et al. 2007). In addition, A. faecimaris (Li et al. 2011) and A. zhangzhouensis (Yang et al. 2013) have been isolated from coastal and mangrove sediments in China, respectively.

Algoriphagus species have also been found to share non-marine habitats. For example, A. aquaeductus (Rau et al. 2012) was isolated from a freshwater pipe in Spain, A. aquatilis (Liu et al. 2009) was isolated from a freshwater lake in China, and a reclassified species A. alkaliphilus (Tiago et al. 2006; Nedashkovskaya et al. 2007) was isolated from alkaline water collected at Portugal. Species such as A. terrigena (Yoon et al. 2006) and A. olei (Young et al. 2009) were isolated from soil samples collected at Korea and Taiwan, respectively. In addition, a reclassified species A. boritolerans (Ahmed et al. 2007; Nedashkovskaya et al. 2007) was originally isolated from boron-contaminated soil, Turkey. Algoriphagus species are Gram-stain negative, motile/non-motile coccobacilli, which produce characteristic non-diffusible pink or red pigments. In this study, we describe polyphasic taxonomic characterization of a xylan-degrading bacterial strain resembling Algoriphagus, designated CC-PR-82T, isolated from coastal surface seawater of Taiwan.

Materials and methods

Isolation, storage, reference strains and culture conditions

Strain CC-PR-82T was isolated from surface seawater collected at coastal Pingtung, Taiwan. The sample was subjected to a standard dilution-to-extinction plating technique using marine agar 2216 (MA; BD Difco) at 30 °C for 48–96 h. The colony of strain CC-PR-82T was isolated, purified and preserved in marine broth (MB; BD Difco) supplemented with 20 % (v/v) glycerol at −80 °C. The taxonomic investigations were carried out according to the previously published guidelines (Tindall et al. 2010). Four type strains of Algoriphagus species such as A. mannitolivorans DSM 15301T (=IMSNU 14012T), A. olei BCRC 17886T (=CC-Hsuan-617T), A. aquatilis NBRC 104237T (=A8-7T) and A. ratkowskyi DSM 22286T (=LMG 21435T) were used for direct comparative taxonomic analysis. All strains (except for A. ratkowskyi DSM 22286T, which was cultivated at 25 °C) were cultured by using NA (Himedia) or NB for 48 h at 30 °C, unless specified otherwise.

Molecular systematics

The genomic DNA of strain CC-PR-82T was isolated by using UltraClean™ Microbial Genomic DNA Isolation Kit (MO BIO, USA) according to manufacturer’s instructions. The partial 16S rRNA gene was amplified via PCR as described in Shahina et al. (2013). Gene sequencing was performed by using a Bigdye terminator kit (Heiner et al. 1998) and an automatic DNA sequencer (ABI PRISM 310, Applied Biosystems, CA, USA) (Watts and MacBeath 2001). Sequence fragments were then assembled using the Fragment Assembly System program from the Wisconsin Package (GCG 1995). Sequence similarity values were computed using the BLAST search (Altschul et al. 1990) program of EzTaxon-e (Kim et al. 2012). Sequence data were analyzed by MEGA 5 (Molecular Evolutionary Genetics Analysis, version 5.0; Tamura et al. 2011), after multiple alignment by Clustal_X (Thompson et al. 1997). The distance matrix method (distance options according to the Kimura two-parameter model; Kimura 1980) including clustering by neighbor–joining (Saitou and Nei 1987), a discrete character-based maximum-parsimony (Fitch 1971) and maximum-likelihood (Felsenstein 1981) methods were used. The topologies of the trees were evaluated by using bootstrap resampling method based on 1,000 replications (Felsenstein 1985).

Phenotypic, physiological and biochemical analysis

The following phenotypic tests were carried out exclusively on novel strain. Colonies were examined for morphological features such as colony appearance, size, shape, texture and pigmentation. Presence of endospore was assessed by phase-contrast microscopy (model A3000, Zeiss) after malachite-green staining (Smibert and Krieg 1994) of the cells grown on MA for 7 days. Cell morphology including presence of flagella was determined by placing the cells (1–2 days old) on a carbon-coated copper grid followed by staining with 0.2 % uranyl acetate for 5–10 s, brief air-drying and observation under a transmission electron microscope (JEOL JEM-1400). Gram staining was performed according to Murray et al. (1994). The requirement for NaCl was tested on R2A agar (BD Difco) supplemented with 0–10 % (w/v) NaCl (at 1 % intervals). The pH range (4.0–10.0, at 1.0 pH unit intervals) for growth was determined in MB that was prepared using 100 mM acetate (pH 4–5), 100 mM NaH2PO4/Na2HPO4 (pH 6–8) and 100 mM NaHCO3/Na2CO3 (pH 9–10) buffers. Growth at 10, 20, 25, 30, 37, 40, 45, 50 and 55 °C was tested in MB after 72 h of incubation. Anaerobic growth was tested using MA or MA supplemented with 0.1 % (w/v) KNO3 by incubating the culture plates in an anaerobic chamber (COY, USA). Saline suspension of strain was inoculated in API 50 CH strip (bioMérieux) and Biolog GN2 MicroPlate according to the manufacturers’ instructions.

The following phenotypic tests were carried out on all five strains. Growth was tested on nutrient agar (NA, Himedia), tryptic soy agar (TSA, BD difco) and R2A (BD difco) agar. Activity of catalase and oxidase, and hydrolysis of starch, egg yolk, Tweens 20 and 80, casein (skim milk), colloidal chitin, carboxymethylcellulose and xylan were tested as described previously (Hameed et al. 2013) using NA as basal media. DNase activity was assessed using DNase test agar (HiMedia). Bacterial saline suspensions were inoculated in API 20 NE and API ZYM according to the manufacturers’ instructions. Results were recorded after 48 h of incubation at 30 °C.

Chemotaxonomic analyses

For chemotaxonomic purpose, all strains (except for A. ratkowskyi DSM 22286T, which was cultivated at 25 °C) were cultivated on NA at 30 °C for 2–3 days. For whole-cell fatty acid analysis, samples were harvested during exponential growth phase, and subjected to saponification, methylation and extraction as described previously (Kämpfer and Kroppenstedt 1996) followed by gas chromatography (model 7890A, Agilent). Peaks were automatically integrated, and fatty acid names and percentages were determined using the microbial identification standard software package MIDI (version 6) (Sasser 1990) and the database RTSBA6. Respiratory quinones of strain CC-PR-82T were extracted according to Minnikin et al. (1984) and analyzed by RP-HPLC according to Collins (1985) with minor modifications (Shahina et al. 2013). Polar lipids of all five strains were extracted and analyzed by two-dimensional TLC with appropriate detection reagents for aminolipids, glycolipids and phospholipids (Embley and Wait 1994). Authentic phosphatidylcholine and phosphatidylseine standards were purchased from Sigma-Aldrich. The spot of phosphatidylcholine was confirmed by spraying the TLC plates with Dragendorff’s reagent. Genomic DNA of strain CC-PR-82T was subjected to thermal denaturation followed by enzymatic digestion into nucleosides as described previously (Mesbah et al. 1989). The resultant nucleoside mixture was separated and quantified by RP-HPLC according to Mesbah et al. (1989) with minor modifications (Shahina et al. 2013).

Results and discussion

The amplified 16S rRNA gene of strain CC-PR-82T contained 1,511 nucleotides, which has been deposited in GenBank/EMBL/DDBJ under the accession number KF732813. The sequence analysis in EzTaxon-e server revealed that strain CC-PR-82T shared high level of sequence similarity to Algoriphagus species and was most-closely related to A. mannitolivorans IMSNU 14012T (97.0 %) followed by ‘A. boseongensis’ BS-R1T (96.5 %), A. olei CC-Hsuan-617T (95.6 %), A. aquatilis A8-7T (95.6 %), A. marincola DSM 16067T (95.5 %), A. ornithinivorans IMSNU 14014T (95.5 %), ‘A. shivajiensis’ NIO-S3T (95.3 %) and A. aquaeductus T4T (95.1 %). Other species of Algoriphagus shared 94.9–92.9 % similarities, whereas A. ratkowskyi LMG 21435T (type species of Algoriphagus) shared lowest (92.6 %) level of sequence similarity with novel strain. In the neighbor-joining phylogenetic tree, the strain CC-PR-82T formed a distinct phyletic lineage associated with A. mannitolivorans IMSNU 14012T and ‘A. boseongensis’ BS-R1T (Fig. 1). Similarly, the phylogenetic association between strain CC-PR-82T and other Algoriphagus species was conserved in the trees generated using maximum-parsimony (Fig. S1) and maximum-likelihood (Fig. S2) algorithms.

Fig. 1
figure 1

Neighbor-joining phylogenetic tree based on 16S rRNA gene sequences showing the taxonomic position of strain CC-PR-82T within the family Cyclobacteriaceae. Bootstrap values (>70 %) based on 1,000 replications are shown at the nodes. Cyclobacterium marinum DSM 745T was used as an outgroup. Bar 0.05 substitutions per nucleotide position

Full size image

DNA–DNA hybridization (DDH) assay was performed between strain CC-PR-82T and A. mannitolivorans DSM 15301T (due to 97 % 16S rRNA gene sequence identity) by using DIG DNA labeling and detection kit (cat. no. 11 093 657 910; Roche Diagnostics) according to the manufacturer’s protocol. The obtained DDH values were 29.0 and 33.4 % (reciprocal using A. mannitolivorans DSM 15301T probe). These values were well below 70 % cutoff that has been prescribed for species distinction (Tindall et al. 2010), which clearly indicated that these two strains are not related at species-level.

The morphological characteristics of strain CC-PR-82T are shown in Fig. 2. Cells of strain CC-PR-82T were observed to be oval or rod-shaped with varying cell size and devoid of flagella and appendages. Cells were not filamentous. The features that distinguished the isolate from reference Algoriphagus strains are given in Table 1.

Fig. 2
figure 2

Transmission electron micrograph of a cell of strain CC-PR-82T. Cells were cultivated on MA for 24–48 h at 30 °C. Bar 0.5 μm

Full size image
Table 1 Differential phenotypic characteristics of Algoriphagus species
Full size table

The major (>5 % of total) fatty acids in strain CC-PR-82T were identified as iso-C15:0 (25.2 %), C16:1 ω6c and/or C16:1 ω7c (summed feature 3, 14.9 %), iso-C16:0 (10.1 %), iso-C17:1 ω9c and/or C16:0 10-methyl (summed feature 9, 5.8 %), anteiso-C15:0 (5.7 %) and C17:1 ω6c (5.7 %). The fatty acids such as iso-C15:0, anteiso-C15:0 and summed feature 3 were also found in major amounts in other four reference strains (Table 2). The qualitative and quantitative differences in terms of several fatty acids that distinguished these five strains of Algoriphagus are shown in Table 2.

Table 2 Whole cell fatty acid profiles (%) of Algoriphagus species
Full size table

Strain CC-PR-82T was found to contain menaquinone-7 (MK-7) as the sole respiratory quinone. Members of the genus Algoriphagus may produce MK-7 as predominant (Yi and Chun 2004; Young et al. 2009; Nedashkovskaya et al. 2004, 2007; Kang et al. 2013; Park et al. 2014) or sole respiratory quinone (Liu et al. 2009).

Polar lipids of strain CC-PR-82T included major amounts of an unidentified aminolipid (AL1) and an unidentified lipid (L2) and moderate amounts of phosphatidylethanolamine (PE), phosphatidylserine (PS) and an unidentified aminolipid (AL2) (Fig. 3). In addition, trace amounts of an unidentified phospholipid (PL), an unidentified lipid (L1) and an unidentified glycolipid (GL) were also found. The polar lipid profile of strain CC-PR-82T was almost in line with that of reference strains as they also consistently showed polar lipids such as PE, AL1, AL2, L1 and L2. A TLC spot designated AL1 (labeled as APL1 in Young et al. 2009, L1 in Oh et al. 2012 and L2 in Park et al. 2014) was poorly stained with ninhydrin reagent indicating the presence of amino-moiety. However, in our study, AL1 did not get stained with zinzadze’s reagent indicating absence of phospho-moiety. A TLC spot designated APL by Park et al. (2014) in A. ratkowskyi DSM 22686T must be PS as it can be stained by both amino- and phospho-specific stains. Strain CC-PR-82T was clearly distinguished particularly from A. mannitolivorans DSM 15301T by exclusively having PS, PL and GL and lacking phosphatidylcholine (PC) that was found in the latter. Identification of PC and PE in A. mannitolivorans DSM 15301T as well as in A. ratkowskyi DSM 22686T was in line with the data reported earlier (Park et al. 2014; Oh et al. 2012). Furthermore, we report the presence of PS for the first time and confirm heterogeneous distribution of PS, PC, PL and GL in Algoriphagus species.

Fig. 3
figure 3

Polar lipid profiles of Algoriphagus species as determined by two-dimensional thin-layer chromatography: strain CC-PR-82T (a), A. mannitolivorans DSM 15301T (b), A. olei BCRC 17886T (c), A. aquatilis NBRC 104237T (d) and A. ratkowskyi DSM 22686T (e). Total polar lipids were visualized by spraying the TLC plate with 10 % ethanolic molybdatophosphoric acid. PE phosphatidylethanolamine; PS phosphatidylserine, PC phosphatidylcholine, L1–3 unidentified lipids; AL1–2 unidentified aminolipids; GL unidentified glycolipid; PL unidentified phospholipid

Full size image

The genomic DNA G+C content of strain CC-PR-82T was determined to be 42.3 mol%, which is within the range reported from other reference strains (Table 1, Yi and Chun 2004; Young et al. 2009; Liu et al. 2009; Bowman et al. 2003).

Taken together, the present polyphasic data unambiguously indicate that strain CC-PR-82T is a novel species of the genus Algoriphagus, for which the name Algoriphagus taiwanensis sp. nov. is proposed. In addition, on the basis of new data obtained in this study, emended descriptions of A. mannitolivorans, A. aquatilis, A. olei and A. ratkowskyi are also proposed.

Emended description of Algoriphagus mannitolivorans Yi and Chun (2004) emend. Nedashkovskaya et al. (2007)

The species description is given by Yi and Chun (2004) and Nedashkovskaya et al. (2007). In addition, polar lipids included major amounts of PE, two unidentified aminolipids (AL1–2) and an unidentified lipid (L2) and moderate amounts of PC and trace amounts of an unidentified lipid (L1).

Emended description of Algoriphagus olei Young et al. (2009)

The species description is according to Young et al. (2009) with following modifications/amendments. Polar lipids included major amounts of PE, an unidentified aminolipid (AL1) and an unidentified lipid (L2) and moderate amounts of an unidentified aminolipid (AL2) and trace amounts of PS, an unidentified phospholipid (PL) and an unidentified lipid (L1).

Emended description of Algoriphagus aquatilis Liu et al. (2009)

The species description is according to Liu et al. 2009 with following amendments. Polar lipids included major amounts of PE, an unidentified aminolipid (AL1) and an unidentified lipid (L2) and moderate amounts of PC and an unidentified aminolipid (AL2) and trace amounts of two unidentified lipids (L1 and L3).

Emended description of Algoriphagus ratkowskyi Bowman et al. (2003) emend. Nedashkovskaya et al. (2004, 2007)

The description is according to Bowman et al. (2003) and Nedashkovskaya et al. (2004, 2007) with following amendments. Polar lipids included major amounts of PE, two unidentified aminolipids (AL1–2) and an unidentified lipid (L2) and trace amounts of PS, PC, an unidentified phospholipid (PL) and an unidentified lipid (L1).

Description of Algoriphagus taiwanensis sp. nov.

Algoriphagus taiwanensis (tai.wa.nen′sis. N. L. fem. adj. taiwanensis of or pertaining to Taiwan, where the type strain was isolated).

Cells are Gram-stain negative, strictly aerobic, nonspore-forming, mesophilic, 0.7–1.6 μm in length and 0.6–0.7 μm in diameter. Cells lack appendages and usually present singly or paired. On marine agar, after 1–2 days of incubation at 30 °C, colonies are 1–3 mm in diameter, circular with regular margins, convex and produce reddish non-diffusible pigments. Growth occurs at 20–40 °C (optimum, 30–37 °C), at pH 6.0–11.0 (optimum, pH 7.0–8.0) and on R2A agar supplemented with 0–2 % NaCl (optimum, 1 %). Cells grow well on marine agar, nutrient agar and R2A agar; however, no growth is observed on TSA. Catalase- and oxidase-positive. Positive for the hydrolysis of casein, starch, egg yolk and xylan; weakly positive for hydrolysis of Tween 20 and negative for the hydrolysis of DNA, colloidal chitin, carboxymethylcellulose and Tween 80. In the API 20 NE strip, positive for hydrolysis of esculin and p-nitrophenyl-β-D-galactopyranoside; negative for nitrate reduction, indole production and fermentation of d-glucose; activity of arginine dihydrolase and urease; hydrolysis of gelatin; assimilation of d-glucose, l-arabinose, d-mannose, D-mannitol, N-acetyl-glucosamine, d-maltose, potassium gluconate, capric acid, adipic acid, malic acid, trisodium citrate and phenylacetic acid. In the API ZYM strip, positive for the activity of alkaline phosphatase, esterase (C 4), esterase lipase (C 8), leucine arylamidase, valine arylamidase, cystine arylamidase, trypsin, α-chymotrypsin, acid phosphatase, naphthol-AS-BI-phosphohydrolase, β-galactosidase, α-glucosidase, β-glucosidase, N-acetyl-β-glucosaminidase and α-mannosidase; negative for the activity of lipase (C 14), α-galactosidase, β-glucuronidase, and α-fucosidase. In the Biolog GN2 MicroPlate, positive for the oxidation of α-cyclodextrin, glycogen, Tween 40, d-arabitol, d-fructose, l-fucose, d-galactose, gentiobiose, α-d-glucose, m-inositol, α-d-lactose, lactulose, maltose, d-mannitol, β-methyl-d-glucoside, d-psicose, d-raffinose, l-rhamnose, d-sorbitol, sucrose, d-trehalose, turanose, acetic acid, citric acid, d-galacturonic acid, d-gluconic acid, d-glucosaminic acid, α-keto butyric acid, α-keto glutaric acid, d,l-lactic acid, quinic acid, succinic acid, glucuronamide, l-alaninamide, d-alanine, l-alanine, l-alanyl glycine, l-asparagine, l-aspartic acid, l-glutamic acid, hydroxy-l-proline, l-ornithine, l-proline, d-serine, l-serine, l-threonine, glycerol; negative for the oxidation of the remaining Biolog GN2 substrates. In the API 50 CH strip, positive for acid production only from esculin ferric citrate; negative for acid production from remaining 48 substrates. In ATB PSE 5 antibiotic strip, sensitive to ampicillin/sulbactam, ticarcillin-pyo, ticarcillin/clavulanic acid, ticarcillin/clavulanic acid-pyo, piperacillin/uretdopen, piperacillin-pyo, piperacillin/tazobactam, piperacillin/tazobactam-pyo, imipenem, meropenem, ciprofloxacin, colistine,and cortrimoxazole; resistant to ticarcillin, cefepime, ceftazidime, amikacin, gentamicin and tobramycin. The major fatty acids are iso-C15:0, C16:1 ω6c and/or C16:1 ω7c, iso-C16:0, iso-C17:1 ω9c and/or C16:0 10-methyl, anteiso-C15:0 and C17:1 ω6c. Polar lipids included major amounts of an unidentified aminolipid (AL1) and an unidentified lipid (L2) and moderate amounts of PE, PS and an unidentified aminolipid (AL2). In addition, trace amounts of an unidentified phospholipid (PL), an unidentified lipid (L1) and an unidentified glycolipid (GL) are also found. The sole respiratory quinone is MK-7. The DNA G+C content of the type strain is 42.3 mol%.

The type strain is CC-PR-82T (=JCM 19755T =BCRC 80746T), isolated from surface seawater off coastal Pingtung, Taiwan. The GenBank/EMBL/DDBJ accession number for the 16S rRNA gene sequence of strain CC-PR-82T is KF732813.