Arthrobacter agilis strain Ant-EH-1 was isolated from cold, dry surface soils collected from Elephant Head, Ellsworth Land, Antarctica (79°49.106’S 83°18.139 W). The average summer atmospheric temperature in Elephant Head is −10.3 °C, with a yearly average of −20.3 °C (McKay et al., 2019). Surface soils where A. agilis Ant-EH-1 was isolated from (“Site 1,” 0–10 cm depth), warm above 0 °C for only a few hundred hours during the year (an estimated ~500 h based on “Site 3” located 0.3 km away). Moisture content of the soils is less than 0.5%. Total organic carbon and nitrogen content is low (<0.07 and 0.007%, respectively; Wood et al., 2024).

Dry soil from Elephant Head was added to 1.5 mL of liquid media Reasoner’s 2A broth (R2B), incubated for 1 week at 15 °C, and then spread plated onto Reasoner’s 2A agar (R2A). The plate was incubated at 15 °C for an additional week. Pink-coloured colonies of A. agilis Ant-EH-1 were streaked for isolation and growth was characterized on R2A agar and in R2B liquid media at −10, −5, 0, 5, 15, 25, 30, and 37 °C. Growth in liquid media was measured via optical density (OD) at 600 nm using a spectrophotometer. Growth rate was calculated as the change in OD over time during the exponential phase of growth. Growth was also characterized on half strength and 1/10th strength R2A as well as on R2A agar plates amended with NaCl (5, 8, and 10%) at 15 °C to examine the salt tolerance of the isolate.

Acetate mineralization by isolate A. agilis Ant-EH-1 was evaluated by a radiorespiration assay using radiolabeled acetate (1,2-¹⁴C) as a carbon substrate. Microcosms were set up in 20 mL serum vials in triplicate with triplicate autoclaved negative controls. Each 5 mL volume microcosm contained: 4860 uL R2B media; 20 uL 1,2-¹⁴C acetic acid (0.043 μCi (~95,000 disintegrations per minute, dpm)); 20 μL of unlabelled acetic acid (3.75 M); and 100 uL of 5.45×106 CFU/mL of A. agilis Ant-EH-1 liquid culture in R2B. Each microcosm also contained a vial of 0.5 mL 1 M potassium hydroxide (KOH) as a carbon dioxide trap. Sterile media (100 μL R2B) was added to negative controls to give the same final volume to all incubations. Microcosms were incubated at 30, 25, 15, 5, 0, and −5 °C. Measurements of KOH radioactivity (correlating with CO₂ released) were taken periodically by liquid scintillation spectrometry on a Beckman Coulter (CA, USA) LS 6000SC and percent mineralization calculated as in Wood et al. (2024) and Goordial et al. (2016a).

Arthrobacter agilis Ant-EH-1 was grown on R2A at 15 °C for 1 week. Isolated colonies were suspended in 750 μL of Powerbead solution from the Qiagen DNeasy PowerLyzer PowerSoil kit and DNA extraction followed manufacturers protocol. DNA was eluted with 100 μL of DNAse-free water. Library preparation was completed using the Oxford Nanopore Rapid Sequencing Kit (SQK-RAD004) for use with the flongle flow cell following manufacturers protocol (Oxford Nanopore Technologies). A. agilis Ant-EH-1 DNA was loaded into three flongle flow cells and three replicate 24-h sequencing runs were carried out. High accuracy base calling was used for the first run and fast base calling was used for subsequent runs.

Sequence data from three sequencing reactions were concatenated together for assembly and analysis. Assembly was performed using Canu (v 2.2) (Koren et al., 2017). Genes were annotated using Prokka (v 1.14.6) (Seemann, 2014) and GhostKOALA (Kanehisa et al., 2016). Completeness of metabolic pathways was visualized with KEGG Decoder (Graham et al., 2018). Additional gene prediction and functional annotation was performed within the Integrated Microbial Genomes (IMG) platform developed by the Joint Genome Institute, Walnut Creek, CA, USA (Markowitz et al., 2009). The complete genome sequence of strain A. agilis Ant-EH-1 is available for public access on the Joint Genome Institute Integrated (JGI) Microbial Genomes & Microbiomes (IMG) under Gold Study ID: Gs0160646. The 16S rRNA gene was amplified using PCR (primers 27F – 5’-AGAGTTTGATCCTGGCTCAG-3′ and 1492R – 5’-TACGGYTACCTTGTTACGACTT – 3′) and Sanger sequenced at The Centre for Applied Genomics, Toronto, Canada. The Classifier tool of the Ribosomal Database Project (RDPII) v 2.13 (Wang et al., 2007) and the NCBI GenBank database were used to identify the isolate. The sequence can be found on NCBI GenBank under accession number OQ383637.

Three A. agilis Ant-EH-1 liquid cultures grown in R2B were harvested for visualization with scanning electron microscopy (SEM): (1) growth at −5 °C for 16 months with visible flocculation (OD₆₀₀ = 1.13); (2) growth at 5 °C for 16 months with no flocculation (OD₆₀₀ = 1.57); (3) growth at 25 °C for 1 month with flocculation (OD₆₀₀ = 1.4). Cultures were grown at −5 °C and 5 °C for extended periods to reflect the incubation period of the acetate mineralization assay and obtain sufficient biomass for SEM. Cultures were diluted in R2B to an OD₆₀₀ of 1.0, and 1.5 mL of each diluted culture was centrifuged at 4,000 rpm for 4 min. Cultures without flocculation did not produce a visible pellet after initial centrifugation and were additionally centrifuged at 10,000 rpm for 4 min. Harvested cells were washed twice in phosphate buffer (35 mM K₂HPO₄ and NaH₂PO₄) with centrifugation for 4 min at 4,000 rpm for cultures with flocculation and 10,000 rpm for cultures without flocculation. Washed pellets were resuspended in 400 μL of phosphate buffer, of which 200 μL was placed on a carbon planchet and incubated for 30 min to allow cell adhesion. After adhesion, planchets were gently submerged in 0.075% ruthenium red and 2.5% glutaraldehyde in 100 mM HEPES pH 7.3 for 30 min to fix cells, then washed once in 100 mM HEPES pH 7.3 and twice in MilliQ water. Cells were then dehydrated by submerging the planchets sequentially in 50, 70, 80, 90, 100%, and a second round of 100% ethanol for 10 min each, followed by drying with a Denton DCP-1 Critical Point Dryer (NJ, USA). Planchets were then mounted on pin stubs with double-sided carbon tape and coated in gold with a Denton Desk V TSC sputter coater. Cells were imaged on a FEI Quanta FEG 250 SEM (OR, USA) at the University of Guelph Molecular and Cellular Imaging Facility (ON, CA).

Arthrobacter agilis strain Ant-EH-1 grew on R2A media forming bright pink, circular colonies on plates, and pink colouration in liquid R2B media (Figure 1) and was capable of growth from −5 °C to 30 °C on solid and liquid media. Optimum (fastest) growth rate based on optical density occurred at 25 °C (Figure 1) in R2B media, a common nutrient media for growth of oligotrophic organisms. This optimum temperature is similar to other Arthrobacter isolates from polar environments (Supplementary file 1). A. agilis Ant-EH-1 was capable of growth at lower nutrient concentrations (two and ten-fold diluted R2A), and maintained its viability in culture after 1 year (e.g., could be re-streaked from R2A plates incubated at −5 °C for 1 year) indicating its ability to survive long periods of time at sub-zero temperatures without the input of new nutrients. A. agilis Ant-EH-1 was previously observed to be halotolerant, capable of growth in media supplemented with up to 8% NaCl (Wood et al., 2024), like many other cryophilic organisms capable of sub-zero growth (Goordial, 2021). Halotolerance is thought to be needed by microbiota for activity and survival in permafrost, as solutes are concentrated into brine veins at subfreezing temperatures.

Heterotrophic respiration of acetate was detected at all temperatures tested from −5 °C to 30 °C (Figure 2A). By day 100 of the incubation, 5 °C surpassed all warmer temperatures tested in total percent acetate mineralized. Temperatures below 5 °C demonstrated lag phases that were 50–80 days in length compared with warmer temperatures which had short to no lag phase observed. Total acetate respired in incubations at warmer temperatures (>5 °C) plateaued quickly (day 15–40) compared to 5 °C which did not plateau by the end of the experiment after 466 days (Figure 2A), inset. The highest cumulative percent mineralization of acetate occurred at 5 °C (45%), compared with all other temperatures tested (Figure 2B). The second highest acetate mineralization was observed at 0 °C with 5.6% mineralization.

Arthrobacter genera have been cultivated from several Antarctic environments, and many are capable of growth at cold temperatures (0 to 5 °C) (Pindi et al., 2010; Chen et al., 2005; Reddy et al., 2000; Dsouza et al., 2015; Gupta et al., 2004; Wang et al., 2009; Reddy, 2002; Vodickova et al., 2022; Fong et al., 2001; Aislabie et al., 2013; Shen et al., 2021). Arthrobacter agilis Ant-EH-1 was closely related (similarity based on 16S rRNA gene) to other Arthrobacter from cold environments, including Antarctica (Supplementary Figure S2). To the best of our knowledge this is the first time that cell division has been documented via optical density for an Arthrobacter species at sub-zero temperatures as low as −5 °C. Prior studies stopped cultivation attempts at 4, 0 °C or −1 °C (Supplementary file 1), thus it is possible these isolates could be capable of cell division at lower temperatures. One prior study confirmed that an Arthrobacter species was capable of activity (measured via carbon dioxide production) down to −17 °C (Panikov and Sizova, 2006). A. agilis Ant-EH-1 had a maximum growth rate at 25 °C based on optical density measurements but showed significantly higher levels of activity at 5 °C based on radiorespiration assays. A. agilis Ant-EH-1 could be classified as a eurypsychrophile based on its ability to grow from −5 °C to 30 °C with a maximum growth rate above 20 °C (Raymond-Bouchard et al., 2018; Cavicchioli, 2016). Eurypsychrophiles can grow across a broad range of temperatures (temperature max > 30 °C, min below 0 °C) and have optimum growth rates around 20 °C (Raymond-Bouchard et al., 2018; Cavicchioli, 2016). However, fast growth rate at warmer temperatures does not necessarily indicate that these conditions are preferred by A. agilis Ant-EH-1. As suggested by Cavicchioli (2016) fast growth is not always better, especially in oligotrophic conditions such as those that exist in Elephant Head soils. Differences in activity and OD measurements suggests that A. agilis Ant-EH-1 employs different growth strategies at different temperatures. At warm temperatures it shows a rapid increase in activity and OD which plateaus quickly. At colder temperatures A. agilis Ant-EH-1 shows a slower initial increase in activity but then is able to sustain an active population for longer. Future transcriptomic analysis could help to determine if there is a different growth strategy employed at cold temperatures versus the slowed rates being a result of slowed kinetic reactions. This was demonstrated previously in transcriptomic studies of Psychrobacter sp. which found that it shifted from a fast-growing state at warmer temperatures (6–22 °C) to a resource efficiency state at cold temperatures (<4 °C) via downregulation of genes involved in energy metabolism (e.g., electron transport chain, TCA cycle) and biosynthesis (e.g., amino acid, nucleotide, ribosome, peptidoglycan synthesis) and upregulation of RNases and peptidases indicating a growth control response (Bergholz et al., 2009).

As the A. agilis Ant-EH-1 genome contained genes associated with cold-adaptive membrane and cell wall modifications, cultures grown at −5, 5, and 25 °C were visualized with SEM to identify potential temperature-dependent changes to the cell envelope. At all temperatures, cultures contained aggregated cells coated in extracellular polymeric substance (EPS)-like material (Figure 3). EPS have an array of functions in mediating cell--environment interactions, including formation of biofilms and aggregates and protection against environmental stressors (Costa et al., 2018). They are proposed to act as a cryoprotectant by limiting ice crystal formation and lowering the freezing point of water, among other potential mechanisms (De Maayer et al., 2014). Elevated EPS production at low and sub-zero temperatures has been observed in psychrophilic and psychrotolerant bacteria (Marx et al., 2009; Caruso et al., 2018), including A. agilis strain L77, a psychrotroph isolated from a subglacial lake (Singh et al., 2016).

Cells grown at 5 °C and −5 °C had nodule-like features partially (5 °C) or completely (−5 °C) covering the cell envelope (Figure 3), which may indicate cell wall modifications associated with cold adaptation. For example, dense nodular encrustations were observed in the psychrophile Planococcus halocryophilus Or1 grown at −15 °C as a result of peptidoglycan accumulation and calcium carbonate biomineralization (Mykytczuk et al., 2016). The genes associated with these accumulations in P. halocryophilus, peptidoglycan synthase (ftsI) and carbonic anhydrase (cab), were both identified in the A. agilis Ant-EH-1 genome. While the prominence of these nodules at −5 °C compared to 5 °C suggests a role in cold adaptation, it is unconfirmed whether these features were exclusive to A. agilis Ant-EH-1 grown at cold temperatures as thick EPS matrices at 25 °C may have obscured surface features. Additional transcriptomic and microscopic analyses are warranted to further investigate potential cold adaptive strategies.

The Arthrobacter agilis Ant-EH-1 complete draft genome is 3,764,186 bp in length with 68.04% GC content. There were 5,935 protein encoding genes predicted, of which 3,236 were assigned a predicted function and the remaining annotated as hypothetical proteins (Table 1). Consistent with other psychrophiles and cryophiles the organism had genomic traits associated with cold adaptation, stress response, and DNA repair. While many of the genes and pathways are found in non-extremophilic organisms as well, these traits have been identified as being important in cryophiles and they are often present in multiple copies, or with redundant pathways as described below.