Genomic and phenotypic insights into Pseudocolwellia antarctica sp. nov., a novel psychrotolerant bacterium with symbiotic potential from Antarctic zooplankton Authors: Min Kyo Kim, Seung Ju Oh, Bo Min Lee, Jun Lee, Chung Yeon Hwang Published: 2026 ABSTRACT Zooplankton in the Southern Ocean represent unique microhabitats for microbial communities that contribute significantly to polar biogeochemical cycles. In this study, we performed a polyphasic and genomic characterization of two Gram-negative, psychrotolerant bacterial strains, HL-MZ7T and HL-MZ19, isolated from Antarctic zooplankton. Phylogenetic and phylogenomic analyses clearly placed these isolates within the genus Pseudocolwellia, forming a sister clade to Pseudocolwellia agarivorans QM50T. The complete genomes of HL-MZ7T (4.5 Mbp) and HL-MZ19 (4.6 Mbp) exhibited DNA G + C contents of 36.0% and 35.9%, respectively. Average nucleotide identity (ANI) and digital DNA-DNA hybridization (dDDH) values confirmed that both strains represent a novel genomic species, distinct from P. agarivorans QM50T (79.5% ANI and 24.0% dDDH). Comparative genomic analysis revealed a rich repertoire of genes associated with cold environmental adaptation and host-microbe interactions. Notably, we identified biosynthetic gene clusters of aryl polyenes and homoserine lactones, alongside genes involved in surface adhesion and nutrient acquisition, suggesting a specialized symbiotic potential within the zooplankton micro-niche. Phenotypically, the strains are characterized as strictly aerobic, with optimal growth at 15–20 °C and 3.0% (w/v) sea salts. These findings expand the genomic and ecological understanding of zooplankton-associated bacteria in polar marine ecosystems, and support the proposal of Pseudocolwellia antarctica sp. nov.Supplementary InformationThe online version contains supplementary material available at 10.1038/s41598-026-53601-5. Introduction Zooplankton play a pivotal role in marine food webs, contributing substantially to the biodiversity and stability of marine ecosystems1,2. Beyond their role as primary consumers, zooplankton serve as specialized microhabitats, hosting diverse microbial communities both internally and externally that provide unique ecological niches for symbiotic microorganisms, including members of the orders Burkholderiales, Pseudomonadales, and Flavobacteriales3–5. These host-associated microbiota are known to influence zooplankton physiology, behavior, and reproduction, thereby exerting a profound impact on broader biogeochemical processes5,6. In polar regions, where low temperatures and limited nutrient availability can constrain metabolic activity7, zooplankton-associated microbiota are likely to play particularly important ecological roles. Recent studies on Antarctic krill (Euphausia superba) have shown that their gut microbiota harbor diverse hydrolytic enzymes capable of degrading complex organic substrates, indicating a key role in host nutrient acquisition and organic matter turnover in polar ecosystems8,9. Despite the increasing recognition of their ecological relevance, detailed taxonomic and functional characterizations of culturable bacteria associated with Antarctic zooplankton remain limited. The genus Colwellia was originally described to encompass psychrophilic, facultatively anaerobic, and chemoorganotrophic bacteria10,11 that are widely distributed across diverse cold marine environments. Representative isolates have been reported from Arctic marine sediments (e.g., Colwellia psychrerythraea), Arctic sea ice (e.g., Colwellia polaris), and Antarctic sea ice (e.g., Colwellia demingiae, Colwellia hornerae, and Colwellia rossensis)11–14. Members of this lineage are well adapted to low-temperature conditions through physiological and genomic traits such as membrane modifications, metabolic versatility, and the production of extracellular enzymes and protective compounds10–12. These adaptations not only support survival under cold conditions but also contribute to carbon and nutrient cycling in marine ecosystems, particularly through the degradation of organic matter in cold environments. Within this genus, Colwellia agarivorans, isolated from temperate coastal seawater, was distinguished by its agar-digesting ability15. Recent advancements in genome-based phylogenetics and comparative genomic indices, such as average nucleotide identity (ANI) and digital DNA-DNA hybridization (dDDH), have refined our understanding of this lineage. These analyses revealed that C. agarivorans is phylogenetically distinct from other members of Colwellia, leading to its reclassification as Pseudocolwellia agarivorans and the establishment of the genus Pseudocolwellia16. However, since its proposal, the genus Pseudocolwellia has remained poorly characterized, with only a single validly described species (i.e., P. agarivorans)16. Nevertheless, publicly available genome data indicate the presence of two additional, yet undescribed strains, Pseudocolwellia sp. AS88 and Pseudocolwellia sp. G5-9, isolated from temperate coastal environments in the United States and the Republic of Korea, respectively. Although genome sequences of these strains are available in public databases, their physiological characteristics, genomic features, and ecological roles remain largely unexplored. Consequently, the ecological distribution, genetic diversity, and functional potential of this genus, especially its adaptation to host-associated lifestyles, are largely unknown. As part of an effort to elucidate the diversity and ecological functions of bacteria associated with marine zooplankton, we isolated 24 bacterial strains from zooplankton collected in Antarctic seawaters. Preliminary 16S rRNA gene sequence analysis indicated that two of these strains, HL-MZ7T and HL-MZ19, were closely affiliated with the genus Pseudocolwellia. In this study, we provided a comprehensive description of these two strains using a polyphasic taxonomic approach. Furthermore, by employing whole-genome sequencing, we investigated the genomic features related to their lifestyle and identified functional gene clusters potentially involved in host-microbe interactions. This work not only expands the known diversity of the genus Pseudocolwellia but also provides new insights into microbial adaptation and potential symbiosis in the Antarctic pelagic ecosystem. Methods Isolation and cultivation Two strains, HL-MZ7T and HL-MZ19, were isolated from zooplankton collected in Marian Cove, King George Island, Antarctica, in January 2024. Zooplankton samples were obtained near the surface using horizontal tows conducted between 62.12°S, 58.51°W and 62.12°S, 58.46°W with an 80 μm mesh, 30 cm diameter plankton net (Daihan ChemLab, Republic of Korea). Environmental parameters were recorded in situ using portable instruments. Water temperature (1.4 °C) and salinity (34.5 psu) were measured using a YSI 30 (YSI Inc., USA), while pH (7.79) was determined using a Uni pH Testa (Trans Instruments, Singapore). The collected samples were filtered through a 100 μm sieve (Daihan Scientific, Republic of Korea), washed with 0.2 μm-filtered, autoclaved seawater, and stored in filtered, autoclaved seawater at 4 °C. After transport to the laboratory, isolation and cultivation procedures were carried out based on previously described approaches for bacteria associated with zooplankton17,18, with a minor modification, omitting digestive track dissection due to the small size of the zooplankton samples. Aliquots (1 mL) were centrifuged at 1,000 × g for 10 min, and the supernatant was discarded to remove free-living bacteria. The pellets were washed by resuspension in 1 mL of 0.2 μm-filtered, autoclaved seawater, followed by vortexing and centrifugation. The pellets were homogenized using a sterile pestle, and aliquots (50 µL) were spread onto marine agar (MA; Difco) plates. Plates were incubated at either 5–15 °C for 7 days. Strains HL-MZ7T and HL-MZ19 were isolated at 15 °C and 5 °C, respectively, and purified by repeated streaking on fresh MA plates. They were subsequently stored in 20% (v/v) glycerol at − 80 °C for preservation. 16S rRNA gene and genome sequencing The 16S rRNA genes of strains HL-MZ7T and HL-MZ19 were amplified by PCR. DNA was extracted by the boiling method as described by Englen & Kelley19. Briefly, fresh colonies were suspended in 10 µL of nuclease-free water, boiled at 90 °C for 10 min, and cooled at 4 °C for 25 s. The extracted DNA was amplified using AccuPower® PCR PreMix (Bioneer, Republic of Korea) with the bacterial universal primers 27F and 1492R20, and purified using ExoSAP-IT (Thermo Fisher Scientific, USA). The PCR products were directly sequenced using the same primers (27F and 1492R) on an Applied Biosystems sequencer (ABI 3730XL) at Macrogen (Seoul, Republic of Korea). The 16S rRNA gene sequences of strains HL-MZ7T and HL-MZ19 were assembled using CodonCode Aligner version 11.0.1 (CodonCode Co., USA). Genomic DNA of strains HL-MZ7T and HL-MZ19 was extracted using the DNeasy Blood & Tissue kit (Qiagen, Germany). For sequencing, libraries were prepared using the MinION Mk1C sequencer (Oxford Nanopore Technologies; ONT) with the Native Barcoding Expansion kit 1–12 (EXP-NBD104; ONT) and the Ligation Sequencing kit (SQK-LSK109; ONT). Raw reads were base-called using Guppy version 6.5.7 and MinKNOW version 23.7.12, assembled with Flye version 2.9.121, and polished using Medaka version 1.7.2 (https://github.com/nanoporetech/medaka). Genome assemblies were visualized using Proksee22. The genome size and DNA G + C content were determined using QUAST version 5.2.023, and genome coverage was calculated using SAMtools version 1.1724. Genome completeness and contamination were estimated using CheckM2 version 1.1.025. The overall genome relatedness index values were assessed by calculating ANI using the USEARCH algorithm26, average amino acid identity (AAI) using EzAAI version 1.2.327, and dDDH values using the Genome-to-Genome Distance Calculator (GGDC) version 3.028. The genome sequences of strains HL-MZ7T and HL-MZ19 were annotated using the NCBI Prokaryotic Genome Annotation Pipeline (PGAP)29. Secondary metabolite biosynthetic gene clusters (BGCs) were identified using antiSMASH version 7.030. Core biosynthetic genes within each predicted BGC were analyzed using BLASTP against the NCBI database to infer putative functions and sequence similarity. Metabolic pathway analysis was performed using BlastKOALA31. Comparative genomic analysis was conducted among strains HL-MZ7T, HL-MZ19, P. agarivorans QM50T, and 13 type strains from the related genera Colwellia and Cognaticolwellia using Anvi’o v832, following a published pangenomic workflow33. The pangenome was constructed based on gene clustering using sequence similarity, allowing classification into core, accessory, and unique gene clusters. To assess cold adaptation potential, genes associated with low-temperature adaptation in members of the family Colwelliaceae were compiled from a previous study11. Representative amino acid sequences were used as queries in BLASTP searches against the genomes of strains HL-MZ7T, HL-MZ19, P. agarivorans QM50T, Pseudocolwellia sp. AS88, and Pseudocolwellia sp. G5-9. Homologs were identified based on sequence similarity using an E-value threshold of ≤ 1 × 10⁻²⁰, as described previously11. Phylogenetic and phylogenomic analyses The 16S rRNA gene sequences of each strain were retrieved from their genome sequences using ContEst16S34. The 16S rRNA gene sequences of strains HL-MZ7T (1,375 bp; PV190323) and HL-MZ19 (1,369 bp; PV190325), determined by the Sanger method, were identical to the corresponding sequences retrieved from the genome sequences of strains HL-MZ7T (1,547 bp; PV190324) and HL-MZ19 (1,547 bp; PV190326). The genome-derived 16S rRNA gene sequences were analyzed using BLAST searches against the EzBioCloud35 and GenBank databases36. The 16S rRNA gene sequences of strains HL-MZ7T and HL-MZ19, together with those of related genera Pseudocolwellia, Cognaticolwellia, and Colwellia, as well as the outgroup Idiomarina abyssalis KMM 227T were aligned using EzEditor237 based on secondary structure information (Table S1). Phylogenetic analyses were conducted using MEGA 11. The neighbor-joining (NJ) tree, a pairwise-distance based method38, was generated using the Jukes-Cantor model39 with pairwise deletion and uniform rates options. The maximum-likelihood (ML) tree, which estimates the tree with the highest likelihood under a substitution model40, was constructed using the General Time Reversible model (GTR)41 with a gamma distribution with invariant sites (G + I) method. The maximum-parsimony (MP) tree, which infers phylogeny by minimizing the number of evolutionary changes42, was reconstructed using the Subtree-Pruning-Regrafting (SPR) method43. To evaluate the robustness of the reconstructed trees, bootstrap analyses with 1,000 replications were conducted. To determine the phylogenomic position of strains HL-MZ7T and HL-MZ19, genome sequences of these strains and related species from the genera Pseudocolwellia, Cognaticolwellia, and Colwellia were retrieved from the National Center for Biotechnology Information (NCBI) GenBank database (Table S1). Phylogenomic analyses were performed based on the Genome Taxonomy Database (GTDB) taxonomy using GTDB-Tk version 2.4.044. A concatenated alignment of 120 conserved marker genes was used for phylogenomic inference. Phylogenomic trees were reconstructed using NJ, ML, MP methods implemented in MEGA 1145. The NJ tree was generated using the Jones-Taylor-Thornton model46 with a pairwise deletion option and uniform rates. The ML tree was constructed using the LG model47 and gamma distribution with invariant sites option, as recommended by the ML fitting results. The MP tree was constructed using the SPR search method with search level 1, and 10 initial trees were generated by random addition. The robustness of the phylogenomic trees was evaluated through bootstrap analyses with 1,000 replications. Phenotypic characterization To investigate the phenotypic characteristics of strains HL-MZ7T and HL-MZ19, physiological, morphological, and biochemical analyses were conducted in parallel with the reference strain P. agarivorans KCTC 52273T, obtained from the Korean Collection for Type Cultures (KCTC). All experiments were performed in duplicate under controlled conditions. Growth parameters, including temperature, salinity, and pH tolerance, were determined as follows. To determine the temperature range for growth, each strain was cultivated on MA and monitored for colony formation at 0–40 °C (in 5 °C intervals, including 37 °C) under aerobic conditions for two weeks, with monitoring at 1–3 day intervals. Salt tolerance was evaluated in ZoBell broth supplemented with sea salt (Sigma-Aldrich) at concentrations of 0–15% (w/v), as all strains grew only in the presence of sea salts rather than NaCl. The pH tolerance range was determined in marine broth adjusted to pH 4.5–12.0 (in 0.5 pH unit increments) using the following buffering systems: MES (pH 4.5–6.5), MOPS (pH 7.0–7.5), AMPD (pH 8.0–9.5), and CAPS (pH 10.0–12.0). Growth under different salinity and pH conditions was monitored by measuring optical density at 600 nm using a SPECTROstar Nano spectrophotometer (BMG Labtech, Germany) for two weeks, with monitoring at 1–3 day intervals at the optimal temperature (i.e., 15 °C for strains HL-MZ7T and HL-MZ19, and 30 °C for P. agarivorans KCTC 52273T; all phenotypic and chemotaxonomic features were assessed at these respective optimum temperatures of each strain). Oxygen requirements were determined by incubating strains on MA with or without 0.1% (w/v) potassium nitrate as an alternative electron acceptor48 at the optimum temperature. Anaerobic growth was assessed using an anaerobic jar containing AnaeroPack (Mitsubishi Gas Chemical), with observations recorded over two weeks. Photobacterium alginatilyticum KCTC 52365T, a known facultative aerobe, was used as a positive control under identical conditions at 30 °C49. Gram staining was performed using a commercial Gram Staining kit (Sigma-Aldrich), following the manufacturer’s instructions. Motility was assessed using semi-solid MA (0.4% agar, w/v)50 and the hanging-drop methods51. Cell morphology and size were examined using transmission electron microscope (TEM; JEM 2100, JEOL, Japan) following standard sample preparation procedures52. For enzymatic and biochemical characterization, oxidase activity was determined using 1% (w/v) tetramethyl-p-phenylenediamine solution (Sigma-Aldrich) on a colony-streaked filter paper and observing color change53, while catalase activity was determined by adding 3% (v/v) hydrogen peroxide solution (Sigma-Aldrich) and observing bubble formation54. Hydrolysis of agar, casein, gelatin, starch, Tween 20, Tween 40, Tween 60, Tween 80, and urea was evaluated on MA55, whereas hydrolysis of cellulose, chitin, DNA, hypoxanthine, L-tyrosine, and xanthine was performed according to established methods56. Aesculin hydrolysis and reduction of nitrate to nitrite were assessed following standard procedures57. Sole carbon source utilization was tested in an artificial seawater medium containing 0.4% (w/v) carbon sources, with growth monitored by OD600 measurements at the optimum temperature for two weeks58. Additional enzyme activities, assimilation, and acid production were determined using API test kits (bioMérieux), including API ZYM, API 20E, and API 50CH. All API assays were performed at the optimum temperature using cell suspensions prepared in artificial seawater supplemented with 3% (w/v) sea salts. Chemotaxonomic characterization For chemotaxonomic analysis, strains HL-MZ7T and HL-MZ19 were cultivated on MA at 15 °C for 4 days, and P. agarivorans KCTC 52273T was cultured at 30 °C for 2 days, corresponding to the exponential growth phase of each strain. Cellular fatty acids were extracted, saponified, and methylated to form fatty acid methyl esters, and analyzed by gas chromatography according to the Microbial Identification System protocol59 using the RTSBA database version 6.21 at the Korean Culture Center of Microorganisms (KCCM). Polar lipids were extracted from lyophilized cells using a chloroform-methanol solvent system and separated by two-dimensional thin-layer chromatography (TLC)60. Lipid components were visualized using appropriate detection reagents to identify major lipid classes61 at the KCCM. The isoprenoid quinone composition of strain HL-MZ7T was determined by extraction with organic solvents, followed by analysis using high-performance liquid chromatography (HPLC)62 at the KCCM. Results and discussion Phylogenetic and phylogenomic inference The complete 16S rRNA gene sequence similarity between strains HL-MZ7T and HL-MZ19 was 99.6%. Strains HL-MZ7T and HL-MZ19 showed the highest sequence similarity to P. agarivorans QM50T (99.0% and 98.6%, respectively), followed by Cognaticolwellia aestuarii CGMCC 1.6965T (97.5% and 97.8%) and Cognaticolwellia beringensis NB097-1T (97.3% and 97.6%). Phylogenetic analyses based on 16S rRNA gene sequences revealed that strains HL-MZ7T and HL-MZ19 formed a robust monophyletic clade within the genus Pseudocolwellia (Fig. 1). This clade was clearly distinct from a neighboring lineage comprising Pseudocolwellia sp. AS88, Pseudocolwellia sp. G5-9, and P. agarivorans QM50T, indicating that the two isolates represent a phylogenetically independent lineage within the genus. This phylogenetic relationship was consistently supported across NJ, ML, and MP trees (Fig. S1). The topology of the tree also resolved the genus Pseudocolwellia as a distinct lineage within the family Colwelliaceae, clearly separated from other related genera such as Colwellia and Cognaticolwellia. Fig. 1Neighbor-joining tree based on 16S rRNA gene sequences of strains HL-MZ7T, HL-MZ19, and the members of the genus Pseudocolwellia, Cognaticolwellia, and Colwellia. Idiomarina abyssalis KMM 227T was used as an outgroup. Only bootstrap values above 60% (1,000 replicates) are shown at the branching points. Filled circles indicate that nodes were also recovered in the maximum-likelihood and maximum-parsimony trees. Bar, 0.01 substitutions per position. Neighbor-joining tree based on 16S rRNA gene sequences of strains HL-MZ7T, HL-MZ19, and the members of the genus Pseudocolwellia, Cognaticolwellia, and Colwellia. Idiomarina abyssalis KMM 227T was used as an outgroup. Only bootstrap values above 60% (1,000 replicates) are shown at the branching points. Filled circles indicate that nodes were also recovered in the maximum-likelihood and maximum-parsimony trees. Bar, 0.01 substitutions per position. Phylogenomic analysis based on concatenated marker genes revealed that members of the genus Pseudocolwellia formed a well-supported and distinct clade within the family Colwelliaceae (Fig. 2, Fig. S2). Within this clade, strains HL-MZ7T and HL-MZ19 formed a robust monophyletic lineage, clearly separated from their closest relative, P. agarivorans QM50T. This lineage was most closely associated with Pseudocolwellia sp. AS88, whereas Pseudocolwellia sp. G5-9 clustered together with P. agarivorans QM50T in a distinct subclade. The clear separation of HL-MZ7T and HL-MZ19 from the only validly described species, P. agarivorans, indicates that these strains represent a genomically distinct lineage within the genus. This separation was strongly supported by high bootstrap value (100%), confirming the robustness of the phylogenomic placement. Fig. 2Maximum-likelihood tree based on genomic sequences of 120 concatenated marker genes of strains HL-MZ7T, HL-MZ19, and members of the genera Pseudocolwellia, Cognaticolwellia, and Colwellia. Idiomarina abyssalis KMM 227T was used as an outgroup. Only bootstrap values above 60% (1,000 replicates) are shown at the branching points. Filled circles indicate that nodes were also recovered in the neighbor-joining and maximum-parsimony trees. Bar, 0.10 substitutions per position. Maximum-likelihood tree based on genomic sequences of 120 concatenated marker genes of strains HL-MZ7T, HL-MZ19, and members of the genera Pseudocolwellia, Cognaticolwellia, and Colwellia. Idiomarina abyssalis KMM 227T was used as an outgroup. Only bootstrap values above 60% (1,000 replicates) are shown at the branching points. Filled circles indicate that nodes were also recovered in the neighbor-joining and maximum-parsimony trees. Bar, 0.10 substitutions per position. The overall tree topology demonstrated that the genus Pseudocolwellia formed a phylogenetically coherent and distinguishable group from other closely related genera such as Colwellia and Cognaticolwellia. The phylogenomic tree provided a more refined resolution of evolutionary relationships compared to the 16S rRNA gene-based phylogenetic tree, which often lacks sufficient resolution for species-level discrimination63,64. This refined phylogenetic relationship further corroborated the reliable classification of strains HL-MZ7T and HL-MZ19 as a novel species within the genus Pseudocolwellia. Genomic characteristics The complete genomes of HL-MZ7T and HL-MZ19 were assembled into single circular chromosomes (Fig. S3). Strain HL-MZ7T had a genome size of 4.5 Mbp with a DNA G + C content of 36.0%, while HL-MZ19 had a genome of 4.6 Mbp with a DNA G + C content of 35.9% (Table S1). Sequencing depths were 113.0× and 118.0×, respectively. Genome completeness and contamination were 100% and 0.1%, respectively, meeting the minimal standards for genome-based taxonomy65. Pairwise genome comparisons between HL-MZ7T and HL-MZ19 showed ANI, AAI, and dDDH values of 99.1%, 99.2%, and 92.4%, respectively (Table 1), all of which exceed the accepted species delineation thresholds (ANI, 95–96%; AAI, 95%; dDDH, 70%)66,67, confirming that the two strains belong to the same species. In contrast, comparisons between HL-MZ7T/HL-MZ19 and P. agarivorans QM50T showed ANI values of 79.5%, AAI values of 82.0–82.1%, and dDDH value of 24.0%, all far below the accepted thresholds for species delineation. Similar results were obtained when compared with Pseudocolwellia sp. AS88 and Pseudocolwellia sp. G5-9, with ANI, AAI, and dDDH values ranging from 79.8 to 87.4%, 82.2–91.7%, and 24.3–33.3%, respectively, further supporting the genomic distinctiveness of HL-MZ7T and HL-MZ19. Notably, the relatively higher similarity values observed with Pseudocolwellia sp. AS88 suggest a closer phylogenomic relationship within the genus, yet still below species-level thresholds, indicating that these strains likely represent genomically distinct species. Table 1Average nucleotide identity (ANI), average amino acid identity (AAI), and digital DNA-DNA hybridization (dDDH) values among five strains of Pseudocolwellia species. Strains: 1, HL-MZ7T; 2, HL-MZ19; 3, Pseudocolwellia agarivorans QM50T; 4, Pseudocolwellia sp. AS88; 5, Pseudocolwellia sp. G5-9.StrainANI / AAI (%)123451–99.1 / 99.279.5 / 82.187.4 / 91.779.8 / 82.2292.4–79.5 / 82.087.3 / 91.679.8 / 82.2324.024.0–79.5 / 82.391.2 / 93.8433.233.324.2–79.8 / 82.3524.324.344.224.3– dDDH (%) Table 1Average nucleotide identity (ANI), average amino acid identity (AAI), and digital DNA-DNA hybridization (dDDH) values among five strains of Pseudocolwellia species. Strains: 1, HL-MZ7T; 2, HL-MZ19; 3, Pseudocolwellia agarivorans QM50T; 4, Pseudocolwellia sp. AS88; 5, Pseudocolwellia sp. G5-9.StrainANI / AAI (%)123451–99.1 / 99.279.5 / 82.187.4 / 91.779.8 / 82.2292.4–79.5 / 82.087.3 / 91.679.8 / 82.2324.024.0–79.5 / 82.391.2 / 93.8433.233.324.2–79.8 / 82.3524.324.344.224.3– dDDH (%) Average nucleotide identity (ANI), average amino acid identity (AAI), and digital DNA-DNA hybridization (dDDH) values among five strains of Pseudocolwellia species. Strains: 1, HL-MZ7T; 2, HL-MZ19; 3, Pseudocolwellia agarivorans QM50T; 4, Pseudocolwellia sp. AS88; 5, Pseudocolwellia sp. G5-9. The antiSMASH analysis of strains HL-MZ7T and HL-MZ19 identified 4 BGCs, including aryl polyene and homoserine lactone clusters (Table S2). Aryl polyenes are structurally analogous to bacterial carotenoids and function as antioxidants, protecting bacteria from oxidative stress68. These BGCs are frequently found in bacteria that live in symbiosis with eukaryotic hosts, where they are implicated in mediating microbial interactions69,70. In addition, the presence of a homoserine lactone BGC suggests the involvement of quorum sensing mechanisms in both strains. Homoserine lactone-based quorum sensing is a key bacterial communication system, particularly in host-associated environments71. This signaling mechanism regulates various physiological processes, including biofilm formation, virulence, and nutrient exchange72, thereby facilitating interactions with hosts in marine environments by modulating surface attachment, resource acquisition, and microbial consortia stability73. Comparative genomic analysis revealed that strains HL-MZ7T, HL-MZ19, and P. agarivorans QM50T shared 2,615 core gene clusters, whereas strains HL-MZ7T and HL-MZ19 possessed 232 and 260 strain-specific gene clusters, respectively (Fig. 3). Furthermore, strains HL-MZ7T and HL-MZ19 shared 1,226 core gene clusters with 14 type strains from related genera (Pseudocolwellia, Cognaticolwellia, and Colwellia) and harbored 643 unique gene clusters not found in these other taxa (Fig. 4). Notably, strains HL-MZ7T and HL-MZ19 contained distinct gene clusters associated with surface adhesion (pilF, pilY1, and wza), nutrient acquisition (cda1, fecA, phnD, sdaC, and tauD), cell protection (adhE, fabG, glpC, and mauG), and nitrogen cycling (narK, nirK, nnrS, and norB) (Table S3). In addition, several strain-specific gene clusters were observed between HL-MZ7T and HL-MZ19, including genes related to lipopolysaccharide biosynthesis (rfaJ), iron acquisition (fecR), and nitric oxide reduction (norB), suggesting potential intraspecific functional variation. These genomic characteristics are commonly observed in zooplankton-associated bacteria from temperate and tropical aquatic environments5,74,75. In particular, the identified gene clusters are linked to host colonization and nutrient acquisition processes. Adhesion-related genes may facilitate attachment to external and internal host surfaces, whereas genes involved in the degradation of complex organic substrates such as chitin, taurine, and algal-derived polysaccharides likely support utilization of zooplankton-derived organic matter5,74. Furthermore, the presence of genes associated with carbon, nitrogen, and phosphorus cycling suggests that these strains may contribute to nutrient transformation and organic matter turnover within host-associated microenvironments5,74,75. These functional traits reflect adaptation to host-associated environments characterized by nutrient-rich conditions, intense microbial interactions, and rapid turnover of organic matter3–5. Taken together, the genomic features of psychrotolerant strains HL-MZ7T and HL-MZ19 indicate their ecological relevance in host-microbe interactions and support their potential roles in biogeochemical cycling in Antarctic zooplankton-associated ecosystems76. Fig. 3Pangenome profiles of strains HL-MZ7T, HL-MZ19, and Pseudocolwellia agarivorans QM50T. Layers represent each genome, and the darker areas represent the presence of gene clusters. The outer colored bars indicate core or unique gene clusters. Pangenome profiles of strains HL-MZ7T, HL-MZ19, and Pseudocolwellia agarivorans QM50T. Layers represent each genome, and the darker areas represent the presence of gene clusters. The outer colored bars indicate core or unique gene clusters. Fig. 4Pangenome profiles of strains HL-MZ7T, HL-MZ19, and 13 type strains from the related genera Pseudocolwellia, Cognaticolwellia, and Colwellia. Layers represent each genome, and the darker areas represent the presence of gene clusters. The outer colored bars indicate core or unique gene clusters. Pangenome profiles of strains HL-MZ7T, HL-MZ19, and 13 type strains from the related genera Pseudocolwellia, Cognaticolwellia, and Colwellia. Layers represent each genome, and the darker areas represent the presence of gene clusters. The outer colored bars indicate core or unique gene clusters. Comparative analysis of cold adaptation-related genes showed that strains HL-MZ7T and HL-MZ19 share core cold-responsive functions with other Pseudocolwellia strains, including cold-shock proteins (cspA, cspB, and cspD) and stress-related enzymes (Table S4). However, notable differences were observed in genes associated with membrane adaptation and oxidative stress response. In particular, fatty acid cis/trans isomerase was detected only in HL-MZ7T and HL-MZ19 and was absent in the other Pseudocolwellia strains isolated from temperate environments, suggesting a potential role in maintaining membrane fluidity under low-temperature conditions. Similarly, catalase-peroxidase genes (katE, katG) and organic hydroperoxide resistance protein were consistently present in the Antarctic strains, whereas their occurrence was less consistent among temperate strains, indicating enhanced oxidative stress mitigation capacity. These patterns suggest that, while core cold-response mechanisms are conserved within the genus11, Antarctic strains may possess additional adaptive features that enhance membrane stability and oxidative stress resistance under persistently cold conditions. In polar marine systems, psychrotolerant bacteria are known to remain metabolically active and contribute to the degradation and recycling of organic matter despite low-temperature constraints7. In zooplankton-associated environments, such as those observed in Antarctic krill, cold-adapted symbiotic bacteria have been shown to retain metabolic activity even at near-freezing temperatures, indicating their capacity to persist and function under extreme conditions9. Therefore, the presence of cold adaptation-related genes in HL-MZ7T and HL-MZ19 supports their potential to remain metabolically active under low-temperature conditions and highlights their ecological relevance in Antarctic zooplankton-associated habitats. Morphological, physiological, and biochemical characteristics The morphological, physiological, and biochemical characteristics are summarized in Table 2 and the species descriptions. Cells of strain HL-MZ7T were found to be Gram-negative, rod-shaped (0.8–2.4 μm in length and 0.5–0.8 μm in width), and motile with a single polar flagellum (Fig. S4). Growth of strains HL-MZ7T and HL-MZ19 was observed at temperatures ranging from 0 to 25 °C, with an optimal range of 15–20 °C. Both strains grew in sea salt concentrations of 1.5–7.0% (w/v), with optimal growth at 3.0%. The pH range for growth was 6.5–7.5 for strain HL-MZ7T and 6.5–8.0 for strain HL-MZ19, with optimal growth occurring at pH 6.5. Strains HL-MZ7T and HL-MZ19 showed optimal growth at lower temperatures and pH compared to P. agarivorans KCTC 52273T. However, they exhibited a wider growth range for salinity, indicating greater tolerance under varying salt concentrations. These physiological traits are consistent with adaptation to cold and fluctuating salinity conditions typical of Antarctic marine ecosystems7. Such tolerance may be advantageous in zooplankton-associated microenvironments, where physicochemical conditions can vary due to host activity and surrounding seawater dynamics. This adaptability likely facilitates persistence and niche differentiation within host-associated habitats at the species level4,5. Table 2Differential characteristics of strains HL-MZ7T, HL-MZ19, and a closely related strain. Strains: 1, HL-MZ7T; 2, HL-MZ19; 3, Pseudocolwellia agarivorans KCTC 52273T. +, Positive; ‒, Negative; ND, Not determined.Characteristic123Growth conditionsTemperature range (optimum, °C)0–25 (15–20)0–25 (15–20)0–35 (30)1Sea salt range (optimum, %, w/v)1.5–7.0 (3.0)1.5–7.0 (3.0)2.0–4.0 (3.0)2pH range (optimum, pH)6.5–7.5 (6.5)6.5–8.0 (6.5)6.5–8.0 (7.0)3 Hydrolysis of:Agar, gelatin, hypoxanthine, xanthine––+ Enzymatic activity (API ZYM) of:α-Chymotrypsin, trypsin––+4 Activity (API 20E) of:Gelatin hydrolysis––+ Assimilation (API 20E) of:D-Glucose++–D-Mannitol–++D-Sucrose––+ Acid production (API 50CH) from:D-Arabinose, gentiobiose, L-xylose++–Aesculin++– N-Acetyl-glucosamine–+–2-keto-Gluconate, 5-keto-gluconate, inulin, lactose, melibiose, salicin––+Maltose, starch, sucrose––+4 Utilization as sole carbon source of:Inositol, potassium gluconate++–D-Mannose, N-acetyl-D-glucosamine–+–D-Maltose, mannitol––+ DNA G + C content (%) 36.035.935.85 1Temperature range of 4–37 °C with an optimum at 28–30 °C15. 2Salinity range of 1.0–6.0% (w/v, NaCl) with an optimum at 2.0–3.0% (w/v, NaCl)15. 3pH range of 6.5–8.5 with an optimum of 7.015. 4Negative15. 535.7 mol% (HPLC)15. Table 2Differential characteristics of strains HL-MZ7T, HL-MZ19, and a closely related strain. Strains: 1, HL-MZ7T; 2, HL-MZ19; 3, Pseudocolwellia agarivorans KCTC 52273T. +, Positive; ‒, Negative; ND, Not determined.Characteristic123Growth conditionsTemperature range (optimum, °C)0–25 (15–20)0–25 (15–20)0–35 (30)1Sea salt range (optimum, %, w/v)1.5–7.0 (3.0)1.5–7.0 (3.0)2.0–4.0 (3.0)2pH range (optimum, pH)6.5–7.5 (6.5)6.5–8.0 (6.5)6.5–8.0 (7.0)3 Hydrolysis of:Agar, gelatin, hypoxanthine, xanthine––+ Enzymatic activity (API ZYM) of:α-Chymotrypsin, trypsin––+4 Activity (API 20E) of:Gelatin hydrolysis––+ Assimilation (API 20E) of:D-Glucose++–D-Mannitol–++D-Sucrose––+ Acid production (API 50CH) from:D-Arabinose, gentiobiose, L-xylose++–Aesculin++– N-Acetyl-glucosamine–+–2-keto-Gluconate, 5-keto-gluconate, inulin, lactose, melibiose, salicin––+Maltose, starch, sucrose––+4 Utilization as sole carbon source of:Inositol, potassium gluconate++–D-Mannose, N-acetyl-D-glucosamine–+–D-Maltose, mannitol––+ DNA G + C content (%) 36.035.935.85 1Temperature range of 4–37 °C with an optimum at 28–30 °C15. 2Salinity range of 1.0–6.0% (w/v, NaCl) with an optimum at 2.0–3.0% (w/v, NaCl)15. 3pH range of 6.5–8.5 with an optimum of 7.015. 4Negative15. 535.7 mol% (HPLC)15. Differential characteristics of strains HL-MZ7T, HL-MZ19, and a closely related strain. Strains: 1, HL-MZ7T; 2, HL-MZ19; 3, Pseudocolwellia agarivorans KCTC 52273T. +, Positive; ‒, Negative; ND, Not determined. 1Temperature range of 4–37 °C with an optimum at 28–30 °C15. 2Salinity range of 1.0–6.0% (w/v, NaCl) with an optimum at 2.0–3.0% (w/v, NaCl)15. 3pH range of 6.5–8.5 with an optimum of 7.015. 4Negative15. 535.7 mol% (HPLC)15. Strains HL-MZ7T and HL-MZ19 exhibited strictly aerobic growth. Notably, P. agarivorans KCTC 52273T, previously described as facultatively anaerobic15, also showed no observable growth under anaerobic conditions in this study. P. alginatilyticum KCTC 52365T, employed as a facultative aerobic control strain, grew anaerobically under identical conditions, confirming the validity of the test. These observations suggest that under the experimental conditions used in this study, P. agarivorans KCTC 52273T may be strictly aerobic, similar to the newly described strains HL-MZ7T and HL-MZ19. The two novel strains HL-MZ7T and HL-MZ19 showed distinct biochemical characteristics compared to the reference strain P. agarivorans KCTC 52273T. Strains HL-MZ7T and HL-MZ19 were negative for the hydrolysis of agar, gelatin, hypoxanthine, and xanthine, whereas the reference strain was positive for these substrates. Compared to the reference strain, strains HL-MZ7T and HL-MZ19 assimilated D-glucose, but not D-sucrose. Additionally, these strains produced acid from aesculin, D-arabinose, gentiobiose, and L-xylose, but not from keto-gluconate, inulin, lactose, maltose, melibiose, salicin, starch, and sucrose. In carbon source utilization assays, strains HL-MZ7T and HL-MZ19, unlike the reference strain, uniquely utilized inositol and potassium gluconate as sole carbon sources, whereas they were unable to utilize D-maltose and mannitol. Despite their shared differences from the reference strain, strains HL-MZ7T and HL-MZ19 also showed intraspecific variation. Strain HL-MZ19 assimilated D-mannitol, which was not assimilated by strain HL-MZ7T (Table 2). Strain HL-MZ19 also showed acid production from N-acetyl-glucosamine, in contrast to strain HL-MZ7T. In addition, strain HL-MZ19 uniquely utilized D-mannose and N-acetyl-D-glucosamine, which were not utilized by strain HL-MZ7T. This intraspecific variation may enhance ecological differentiation and functional diversification among closely related strains77. In heterogeneous microbial environments, such as zooplankton-associated microhabitats, these differences may be associated with the ability of closely related strains to occupy distinct functional roles. Chemotaxonomic characteristics The cellular fatty acid compositions of strains HL-MZ7T, HL-MZ19, and the reference strain P. agarivorans KCTC 52273T are summarized in Table 3. The major cellular fatty acids (> 10%) of strains HL-MZ7T and HL-MZ19 were C16:0 (12.9% and 14.3%) and C16:1ω9c (11.1% and 11.5%), with a dominance of summed feature 3 (C16:1ω7c and/or C16:1ω6c), comprising 52.1% and 49.6% of total fatty acids, respectively. The proportion of summed feature 3 was notably higher than that observed in P. agarivorans KCTC 52273T (30.2%). This predominance of unsaturated fatty acids may contribute to maintaining membrane fluidity under low-temperature conditions, as commonly observed in cold-adapted bacteria78. The fatty acid C14:0 was detected in strains HL-MZ7T and HL-MZ19, but was absent in P. agarivorans KCTC 52273T, while iso-C14:0 was detected only in the reference strain. Table 3Cellular fatty acid composition for strains HL-MZ7T, HL-MZ19, and a closely related strain. Strains: 1, HL-MZ7T; 2, HL-MZ19; 3, Pseudocolwellia agarivorans KCTC 52273T. Strains HL-MZ7T and HL-MZ19 were grown on marine agar at 15 °C for 4 days and the P. agarivorans KCTC 52273T was grown on marine agar at 30 °C for 2 days. Fatty acids accounting for 0.5% or more in one of the strains are given. The dominant fatty acids (> 10%) are shown in bold type. –, Not detected.Characteristic123Saturated fatty acidsC10 : 01.21.40.5C12 : 00.70.61.7C14 : 02.12.3–C16 : 0 12.9 14.3 30.5 C17 : 00.50.51.3C18 : 00.91.01.5Unsaturated fatty acidsC15 : 1ω8c0.70.61.5C16 : 1ω9c 11.1 11.5 5.7C17 : 1ω8c1.51.42.6C18 : 1ω7c2.12.21.4C18 : 1ω9c1.91.91.3Branched fatty acidsiso-C14 : 0––3.2iso-C16 : 00.20.21.0Hydroxy fatty acidsC10 : 0 3-OH0.80.71.9C11 : 0 3-OH0.50.51.4C12 : 0 3-OH6.56.58.9C16 : 0 3-OH0.70.80.9iso-C15 : 0 3-OH1.31.50.6Summed feature3 (C16 : 1ω7c / C16 : 1ω6c) 52.1 49.6 30.2 Table 3Cellular fatty acid composition for strains HL-MZ7T, HL-MZ19, and a closely related strain. Strains: 1, HL-MZ7T; 2, HL-MZ19; 3, Pseudocolwellia agarivorans KCTC 52273T. Strains HL-MZ7T and HL-MZ19 were grown on marine agar at 15 °C for 4 days and the P. agarivorans KCTC 52273T was grown on marine agar at 30 °C for 2 days. Fatty acids accounting for 0.5% or more in one of the strains are given. The dominant fatty acids (> 10%) are shown in bold type. –, Not detected.Characteristic123Saturated fatty acidsC10 : 01.21.40.5C12 : 00.70.61.7C14 : 02.12.3–C16 : 0 12.9 14.3 30.5 C17 : 00.50.51.3C18 : 00.91.01.5Unsaturated fatty acidsC15 : 1ω8c0.70.61.5C16 : 1ω9c 11.1 11.5 5.7C17 : 1ω8c1.51.42.6C18 : 1ω7c2.12.21.4C18 : 1ω9c1.91.91.3Branched fatty acidsiso-C14 : 0––3.2iso-C16 : 00.20.21.0Hydroxy fatty acidsC10 : 0 3-OH0.80.71.9C11 : 0 3-OH0.50.51.4C12 : 0 3-OH6.56.58.9C16 : 0 3-OH0.70.80.9iso-C15 : 0 3-OH1.31.50.6Summed feature3 (C16 : 1ω7c / C16 : 1ω6c) 52.1 49.6 30.2 Cellular fatty acid composition for strains HL-MZ7T, HL-MZ19, and a closely related strain. Strains: 1, HL-MZ7T; 2, HL-MZ19; 3, Pseudocolwellia agarivorans KCTC 52273T. Strains HL-MZ7T and HL-MZ19 were grown on marine agar at 15 °C for 4 days and the P. agarivorans KCTC 52273T was grown on marine agar at 30 °C for 2 days. Fatty acids accounting for 0.5% or more in one of the strains are given. The dominant fatty acids (> 10%) are shown in bold type. –, Not detected. The polar lipids of strain HL-MZ7T consisted of phosphatidylethanolamine (PE), phosphatidylglycerol (PG), and one unidentified aminolipid (AL) (Fig. S5). While its major polar lipids were similar to those of P. agarivorans KCTC 52273T, the absence of an unidentified lipid (L) and an unidentified phospholipid (PL) distinguished it from the reference strain. The primary respiratory quinone identified in strain HL-MZ7T was ubiquinone 8 (Q-8), which is a common feature in the family Colwelliaceae10. Conclusion This study describes Pseudocolwellia antarctica sp. nov. as a novel psychrotolerant species, supported by comprehensive phylogenomic and phenotypic evidence. The identification of biosynthetic gene clusters for aryl polyenes and quorum sensing signals, along with specialized surface-adhesion and nutrient acquisition genes, highlights the genomic basis for its adaptation and symbiotic potential within the zooplankton micro-niche. In addition, the presence of cold adaptation-related genes indicates that this species is capable of maintaining metabolic activity under low-temperature conditions. Taken together, these findings not only expand the known diversity of the genus Pseudocolwellia but also highlight the potential ecological importance of zooplankton-associated microbiota in Southern Ocean biogeochemical cycles, particularly in relation to organic matter transformation and nutrient cycling. These findings provide a basis for future studies on host-microbe interactions and functional roles of zooplankton-associated bacteria in polar ecosystems. Protologue Emended description of the genus Pseudocolwellia The genus description is given by Liu et al.16 with the following modifications. Cells are strictly aerobic and require sea salts for growth. The major cellular fatty acids are summed feature 3 (C16 : 1ω7c and/or C16 : 1ω6c) and C16:0. The major polar lipids are phosphatidylethanolamine (PE) and phosphatidylglycerol (PG). The genomic DNA G + C content is 35.8–36.0%. Description of Pseudocolwellia antarctica sp. nov. Pseudocolwellia antarctica (ant.arc’ti.ca. L. fem. adj. antarctica, referring to the geographical origin of the type strain). Cells are Gram-negative, strictly aerobic, approximately 0.5–0.8 μm in width, 0.8–2.4 μm in length, and motile by means of a polar flagellum. Colonies on MA plates at 15 °C are cream-colored, round, and convex after 4 days. Growth occurs at 0–25 °C (optimum, 15–20 °C) in the presence of 1.5–7.0% (w/v) sea salts (optimum, 3.0%), and at pH 6.5–7.5 (optimum, pH 6.5). Positive for oxidase, catalase, and nitrate reduction. Aesculin, DNA, Tween 20, Tween 40, Tween 60, Tween 80, and urea are hydrolyzed, but casein, cellulose, chitin, gelatin, hypoxanthine, starch, L-tyrosine, and xanthine are not. Inositol, potassium gluconate, and L-proline are utilized as sole carbon sources, but acetate, L-arabinose, D-glucose, L-lysine, malic acid, D-maltose, mannitol, pyruvate, raffinose, and trisodium citrate are not. The utilization of D-mannose and N-acetyl-D-glucosamine as sole carbon sources is variable among strains. According to the API ZYM test, positive for acid phosphatase, alkaline phosphatase, cystine arylamidase, esterase (C4), esterase lipase (C8), leucine arylamidase, naphthol-AS-BI-phosphohydrolase, and valine arylamidase, but negative for α-chymotrypsin, α-fucosidase, α-galactosidase, β-galactosidase, α-glucosidase, β-glucosidase, β-glucuronidase, lipase (C14), N-acetyl-β-glucosaminidase, α-mannosidase, and trypsin. According to the API 20E test, positive for the assimilation of amygdalin, D-glucose, sodium pyruvate, and L-tryptophane, but negative for L-arabinose, L-arginine, gelatin, inositol, L-lysine, D-melibiose, 2-nitrophenyl-β-D-galactopyranoside, L-ornithine, L-rhamnose, sodium thiosulfate, D-sorbitol, D-sucrose, trisodium citrate, and urea. The assimilation of D-mannitol is variable depending on strains. According to the API 50CH test, positive for the acid production of aesculin, D-arabinose, cellobiose, fructose, galactose, gentiobiose, glucose, glycerol, mannitol, D-xylose, and L-xylose, but negative for adonitol, amygdalin, L-arabinose, D-arabitol, L-arabitol, arbutin, dulcitol, erythritol, D-fucose, L-fucose, 2-keto-gluconate, 5-keto-gluconate, gluconate, glycogen, inositol, inulin, lactose, maltose, D-lyxose, mannose, melezitose, melibiose, methyl-α D-glucoside, methyl-α D-mannopyranside, methyl-β D-xylopyranside, raffinose, rhamnose, ribose, salicin, sorbitol, sorbose, starch, sucrose, D-tagatose, trehalose, D-turanose, and xylitol. The acid production of N-acetyl-glucosamine is variable depending on strains. The major cellular fatty acids are summed feature 3 (C16 : 1ω7c and/or C16 : 1ω6c), C16:0, and C16 : 1ω9c. The polar lipids consist of phosphatidylethanolamine (PE), phosphatidylglycerol (PG), and one unidentified aminolipid (AL). The major respiratory quinone is ubiquinone 8 (Q-8). The type strain HL-MZ7T (= KCCM 90591T = JCM 37948T) and sister strain HL-MZ19 (= KCCM 90592 = JCM 37949) were isolated from zooplankton in Antarctica. The GenBank/EMBL/DDBJ accession numbers for the genome-derived 16S rRNA gene sequences of strains HL-MZ7T and HL-MZ19 are PV190324 and PV190326, respectively. The accession numbers for the complete genome sequences of strains HL-MZ7T and HL-MZ19 are CP187926 and CP187927, respectively. The genomic DNA G + C content calculated from the genome sequences of the strains HL-MZ7T and HL-MZ19 is 36.0% and 35.9%, respectively. Supplementary Information Below is the link to the electronic supplementary material. Supplementary Material 1 Supplementary Material 1