• Antarctica;
  • Psychrotrophic;
  • Food-borne;
  • Bacterial growth inhibition;
  • Bacterial diversity


  1. Top of page
  2. Abstract
  3. 1Introduction
  4. 2Materials and methods
  5. 3Results
  6. 4Discussion and conclusions
  7. References

The aim of this study was to identify Antarctic microorganisms with the ability to produce cold-active antimicrobial compounds with potential for use in chilled food preservation. Colonies (4496) were isolated from 12 Antarctic soil samples and tested against Listeria innocua, Pseudomonas fragi and Brochothrix thermosphacta. Thirteen bacteria were confirmed as being growth-inhibitor producers (detection rate 0.29%). When tested against a wider spectrum of eight target organisms, some of the isolates also inhibited the growth of L. monocytogenes and Staphylococcus aureus. Six inhibitor producers were psychrotrophic (growth optima between 18 and 24 °C), halotolerant (up to 10% NaCl) and catalase-positive; all but one were Gram-positive and oxidase-positive. The inhibitors produced by four bacteria were sensitive to proteases, suggesting a proteinaceous nature. Four of the inhibitor–producers were shown to be species of Arthrobacter, Planococcus and Pseudomonas on the basis of their 16S rRNA gene sequences and fatty acid compositions. It was concluded that Antarctic soils represent an untapped reservoir of novel, cold-active antimicrobial-producers.


  1. Top of page
  2. Abstract
  3. 1Introduction
  4. 2Materials and methods
  5. 3Results
  6. 4Discussion and conclusions
  7. References

The production of antimicrobial compounds such as bacteriocins by lactic acid bacteria is widely recognized and there is much interest in developing novel applications for these natural agents in the food, cosmetic and pharmaceutical industries [1]. However, the antagonistic properties of cold-loving organisms have not been investigated as extensively as those of the mesophiles. Unlike the inhibitors produced by mesophiles, the antimicrobials produced in cold environments need to function at low temperatures for the organisms to gain a competitive advantage during their growth cycle. Such cold-active antimicrobial compounds may be exploitable in industrial applications including chilled-food preservation.

Antarctica is a rocky continent almost completely covered by a massive ice sheet, with regions of cold desert soils that have little free water and temperatures that rarely rise above freezing. The marine soils are free of snow and ice in the summer months and receive inputs from the ocean and animal life [2]. In coastal areas seal and penguin rookeries may contribute significant quantities of organic material to soils, known as ornithogenic soils; although they are high in nutrients, such soils are marked by rapid freeze–thaw cycles, which are more lethal than a permanently cold environment and hence restrict life. Microbial survival and growth in Antarctic soils are limited not only by low temperatures but also by low aw and osmotic stress [3]. The transition from winter to summer involves a freeze–thaw phase, which is critical for the onset of microbial activity. The microorganisms need to be efficient at rapidly switching their metabolism on and off according to prevailing conditions. In view of the severe environmental conditions, it could be argued that the production of extracellular antimicrobial compounds would be a particular advantage in order to reduce inter-species competition.

The aim of the present study was to identify Antarctic microorganisms with the ability to produce cold-active antimicrobial compounds with potential for use in chilled-food preservation. Both classical and molecular methods were used to characterize and identify a small number of inhibitor producers with potential for use at low temperatures. Since the extent of microbial species and diversity in soil varies with the location and is influenced by salinity, pH and the availability of moisture and nutrients, a number of soils from different Antarctic sites have been screened.

2Materials and methods

  1. Top of page
  2. Abstract
  3. 1Introduction
  4. 2Materials and methods
  5. 3Results
  6. 4Discussion and conclusions
  7. References

2.1Bacteria and their cultivation

Antarctic bacteria were maintained in dilute tryptone soya broth/agar (DTSB/DTSA, one-tenth the strength of TSB/TSA); marine broth/agar (MB/MA); and Luria–Bertani broth/agar (LBB/LBA) containing tryptone (10 gl−1), yeast extract (5 gl−1) and NaCl (5 gl−1) with agar (15 gl−1) in solid media. Indicator bacteria used for detecting antimicrobial activity were maintained on tryptone soya broth/agar (TSB/TSA) and nutrient broth/agar (NB/NA). Overnight cultures for antimicrobial testing were prepared using Listeria innocua NCTC 10528 and NCRZ 4202, L. monocytogenes NCTC 11984, Brochothrix thermosphacta NCTC 2304 and Pseudomonas fragi (Leatherhead Food International, UK), all grown at 30 °C; and Escherichia coli O157:H7 (attenuated strain, Unilever Research, UK), Salmonella enterica sv. Enteritidis (13244WT University of Bath, UK) and Staphylococcus aureus NCTC 10788 grown at 37 °C. All microbiological media were from Oxoid (Basingstoke, England) except MB/MA, which was from Difco (Detroit, USA).

2.2Isolation of Antarctic bacteria and detection of inhibitor production

Soil samples were collected aseptically in Antarctica from the locations given in Table 1 (72° 19S to 77° 83S; 160° 55E to 170° 16E) by one of the authors (NJR) during the 1995/1996 field season and were stored at −80 °C. For isolations the soil samples were defrosted and 1 g of each was inoculated into 50 ml DTSB, MB and LB. The samples were incubated at 10 °C with shaking for 24 h. Those turbid samples, indicating growth, were diluted, plated on their respective agars and incubated at 10 °C for 5 days or longer as necessary. Colonies were counted and their morphological characteristics recorded. Master plates were made by randomly picking over 4000 colonies from the three types of agar. Five of the soil samples were also incubated at 5 °C for up to a month using the same method.

Table 1.  Numbers of bacteria cultured from Antarctic soils incubated on dilute tryptone soy agar (DTSA), marine agar (MA) and Luria–Bertani agar (LBA) at 10 °C for 5 days
LocationDescriptionAltitude (m)Bacterial numbers (logcfug−1)
  1. NG, no growth.

  2. aIncubated for 8 weeks.

Nutrient-rich, ornithogenic
Cape HallettDisused penguin rookerySea level>9.06.8>9.0
Edmonson Point (a)Penguin rookery507.37.56.7
Kay IslandEntrance to snow petrel nesting site∼809.36.06.3
Cape RussellExposed guano-covered ridge∼6008.38.38.3
Low altitude (<500 m)
Edmonson Point (b)Maritime soilSea levelNGa7.0NGa
Lake HoarePolygon-crack silt in dry valley∼1504.33.7a8.9
Harrow PeaksLight gravel2006.57.98.0
Crater CirceRaised lake beach/pond5006.06.05.8
High altitude (>500 m)
Battleship PromontoryScree beneath rockface with endoliths1000NGa9.0>9.0
Mount McGeeCoarse gravel soil14104.84.85.0
Mount RittmannMoist clay silt/soil2000NGa>9.08.3
Mount MelbourneNorth flank, exposed volcanic soil26504.0a<3.0a6.3

Inhibitor production was determined using the deferred antagonism procedure of Kekessy and Piguet [4]. Replica plates were made from the master plates using a Multipoint Inoculator (Sigma Chemical Co., UK). Replica plates were prepared using the same agar as that used for isolating the organisms on the master plates (e.g. Marine Agar was used to prepare replica plates for organisms isolated on Marine Agar master plates). The replica plates were overlaid with soft TSA containing one of eight indicator organisms. Zones of clearance surrounding the producer colony following incubation at 10 °C (or 15 °C when Staph. aureus, E. coli O157:H7 and S. Enteritidis were used) for 4/5 days indicated the presence of an antagonistic agent.

2.3Sensitivity of inhibitors to enzymes

The enzymes trypsin (EC in 67 mM sodium phosphate buffer (pH 7.6) and pronase E (EC in 50 mM phosphate buffer (pH 7.5) were used to determine whether the inhibitory substances produced by the Antarctic bacteria were proteinaceous. Catalase (EC in 50 mM phosphate buffer (pH 7.0) was used to eliminate inhibition by hydrogen peroxide. Lipase (EC in 50 mM phosphate buffer (pH 7.7) and α-amylase (EC in phosphate buffer (pH 6.9) were used to determine the presence of lipid or glycogen components in the inhibitory substances. Final concentration of all enzymes was 25 mg ml−1. Deferred antagonism assays were performed by placing a 5 μl aliquot of the enzyme solution next to the fully grown producer-cell spot on an agar plate and incubating at 37 °C (or 20 °C for α-amylase) for 1 h before overlaying the agar with the target organism. Sensitivity of the inhibitor to the enzyme was indicated when the zone of inhibition around the producer colony was altered, giving a “half-moon” appearance.

2.4Characterisation and identification of Antarctic bacteria

Bacteria were identified initially on the basis of their Gram reaction, morphology (by light microscopy), plus their catalase and oxidase reactions tested according to Roberts [5]. The cardinal growth temperatures (i.e. minimum, optimum and maximum) were determined on DTSA or MA according to the method of White et al. [6]. Briefly, a purpose-built vertical temperature gradient incubator (range 3.3–40 °C) with 2.5 °C increments between the shelves was used. The time taken for the appearance of colonies visible to the naked eye was used as the measure of growth. Agar plates were monitored daily for signs of growth. Results were plotted as growth rate in days−1 versus temperature of growth. Salt gradient plates (1–10% NaCl) were made in 25 cm×25 cm square Petri dishes according to Venables et al. [7]. Bacterial isolates were streaked from high to low salt concentrations on the plates and incubated at 4, 10, 22, 30 and 37 °C for 7–10 days. Growth was scored as the distance in cm reached by visible colonies across the gradient.

The fatty acid composition of selected Antarctic bacteria was determined by gas chromatography according to the method of Kates et al. [8]. Cultures were grown on DTSA or MA for 5–7 days at 10 °C. Fatty acid methyl esters (FAME) were prepared with 2.5% (v/v) sulphuric acid in dry methanol and analysed using a Hewlett–Packard 5880A gas chromatograph fitted with a Supelco 2380 polar fused silica capillary column and a Spectraphysics SP420 integrator. The carrier gas was nitrogen (flow rate 1 mlmin−1) with detection using flame ionisation. Detection and injection temperatures were 250 °C. Oven temperatures were set initially at 130 °C for 10 min and then programmed to rise at 4 °Cmin−1 to 250 °C, which was held for a further 15 min. Individual peaks were identified by comparison of retention times with standards (14:0, 16:0, 18:0, 16:1 and 18:1) and confirmed by sequential hydrogenation with platinum dioxide/H2 gas (to reveal unsaturated components) and bromination with bromide/N2 gas (to identify cyclopropane fatty acids).

Selected bacterial isolates were also subjected to partial sequencing of their 16S rRNA gene. Briefly, DNA was extracted from exponential phase cultures using phenol–chloroform extraction according to the method of Sambrook and Maniatis [9]. A set of primers was designed based on probe sequences made from environmentally derived and taxon-specific 16SrRNA sequences [10,11]. PCR reactions were performed in a Techne Progene PCR thermocycler as follows: 94 °C for 2 min, denaturation at 94 °C for 30 s, annealing at 45 °C for 45 s and extension at 72 °C for 2 min, for 30 cycles. The PCR product was ligated into the pGem-Teasy vector and transformed in CaCl2-competent E. coli (Promega, Madison, WI, USA). Plasmid DNA of six clones was isolated and sequenced (MWG Biotech) using an Automated Biosystems Sequencer. The 16S rRNA gene sequences obtained ranged in size from 920 to 940 bp. Sequences were identified using the Ribosomal Database Project (RDP) 16S rRNA gene sequence database and BLAST searching of the NCBI database. To further analyse the relationships of these organisms, the percent sequence similarities of the 16S rRNA genes were calculated using the CLUSTAL V programme of the MegAlign package from DNA Star Software, Editseq. (DNA Star Inc., USA).


  1. Top of page
  2. Abstract
  3. 1Introduction
  4. 2Materials and methods
  5. 3Results
  6. 4Discussion and conclusions
  7. References

3.1Isolation of bacteria from Antarctic soils

Three types of media were used for the isolation of bacteria from Antarctic soils: dilute tryptone soya broth and agar to isolate bacteria from nutrient-deficient soils, marine broth and agar to isolate bacteria with high salt requirements, and Luria–Bertani broth and agar to isolate fastidious bacteria. Two temperatures were used, 10 and 5 °C, in order to isolate both psychrotrophic and psychrophilic organisms. Table 1 shows the viable bacterial counts obtained from twelve soil samples on the three different media after 5 days at 10 °C. Ornithogenic soils are rich in nitrogen and phosphorus due to the deposition of guano by penguins: one would anticipate, therefore, that these soils would be capable of supporting good microbial growth. As shown in Table 1, all four ornithogenic soil samples in this study (penguin rookery at Edmonson Point, Cape Russell, Kay Island and Cape Hallet) yielded relatively high bacterial counts (6–9 logcfug−1) on all media. Very high counts (10 logcfug−1) determined in ornithogenic soil by direct microscopic counting have been reported previously but only about 20% of these could be cultured [10]. Total counts obtained by microscopy may be misleadingly high due to the presence of dead cells and other debris.

Maritime soil samples from Edmonson Point gave no growth on DTSA or LBA, whereas on MA counts of more than 7 logcfug−1 were obtained. This indicated that a halotolerant group of organisms dominated in this region and is a reflection of the maritime nature of the soil, which had been collected at sea level. Maritime soils often have additional nutrients available from coastal macrofauna, penguins and other seabirds and seals. Lake Hoare samples gave low counts on both DTSA and MA (3–4 logcfug−1); however, a high count of 8 logcfug−1 was obtained on LB agar suggesting that the microflora was nutritionally fastidious. The Harrow Peaks and Crater Circe samples produced medium-level (6–8 logcfug−1) counts on all three agars (Table 1).

The Battleship Promontory sample supported the growth of lichens beneath the rock face, hence providing nutrients for a microbial population. In this sample, no bacterial growth was detected on DTSA but very high numbers (9 logcfug−1) were observed on MA and LBA. The Mount Rittmann sample, consisting of moist clay/silt, also yielded no growth on DTSA but high numbers (8–9 logcfug−1) on MA and LBA. It is interesting to note that the apparently halophilic populations from these two samples were from sites that were 1000–2000 m above sea level. It is often the case in Antarctic soils that with freezing of water, there is an increase in solute concentration in the soil resulting in creation of a hypersaline micro-environment [11]. High altitude soils in the Trans-Antarctic mountains provide the least hospitable environment for microbial growth since temperatures rise above freezing for only brief periods each year. Mount McGee soil at over 1000 m above sea level contained low levels of organisms (4–5 logcfug−1) on all three media.

The numbers of viable bacteria isolated from the Antarctic soils after incubation at 5 °C for one month were much lower (3–4 logcfug−1 and often <3 logcfug−1) than at 10 °C. It has been observed previously that it is difficult to isolate true psychrophiles even at lower temperatures [6]. This is probably because of the occasional high ground temperature and the fact that the samples in this study were all collected in the summer months. True psychrophiles are more often found in consistently cold environments such as the Southern Ocean [12].

3.2Antimicrobial screening

As shown in Table 2, 4496 Antarctic colonies were picked at random and screened for inhibitor production against the psychrotrophic P. fragi, B. thermosphacta and L. innocua. The former two species are common causes of food spoilage and L. innocua was chosen as a model for the food-borne pathogen L. monocytogenes. B. thermosphacta was sensitive to 51 of the inhibitors produced, whereas P. fragi was sensitive to only one inhibitor. Notably, soil samples from sites at an altitude of over 1400 m (Mt. McGee, Mt. Melbourne and Mt. Rittmann) yielded no inhibitor producers. Crater Circe seemed to be the best source of inhibitor producers, yielding 42 strains, whereas Lake Hoare provided 14 producers. When the 74 putative producers obtained at 10 °C were re-tested against the three target organisms to confirm inhibition, 13 strains were positive. The loss of inhibitor production following the second screening may have been due to an overcrowding of colonies on the plates in the primary screening, making it difficult, at times, to assess true zones of inhibition. Furthermore, the genes for inhibitors may be plasmid encoded and their loss on subculturing could account for the loss of ability to produce the inhibitor compound on re-screening [13].

Table 2.  Identification of inhibitor–producers at 10 °C against three food-borne bacteria (Brochothrix thermosphacta, Listeria innocua and Pseudomonas fragi
LocationNo. of colonies tested for inhibitor productionNo. of inhibitor–producers identified againstTotal no. of inhibitor–producers identified
  B. thermosaphactaL. innocuaP. fragi 
Nutrient-rich, ornithogenic
Cape Hallett4361001
Edmonson Point (a)2880000
Kay Island5762002
Cape Russell4080000
Low altitude (<500 m)
Edmonson Point (b)2160101
Lake Hoare41286014
Harrow Peaks5283216
Crater Circe5762913042
High altitude (>500 m)
Battleship Promontory3128008
Mount McGee3840000
Mount Rittmann1920000
Mount Melbourne1680000
Grand total44965122174

A total of 576 colonies isolated at 5 °C were also tested for antimicrobial production; however, only four inhibitor producers were detected and all four strains were active against B. thermosphacta only. On re-screening, none of the strains produced zones of inhibition against the same target organisms. The dearth of inhibitor producers at 5 °C suggests that psychrophilic organisms may have developed other survival strategies, e.g. dormancy at very cold temperatures.

3.3Host spectrum of inhibitor-producing bacteria

Table 3 shows the effect of the 13 inhibitor–producers obtained at 10 °C against the three initial target bacteria, plus another five spoilage/pathogenic bacteria: E. coli O157:H7, L. innocua 4202, L. monocytogenes, S. Enteritidis and Staph. aureus. None of the Antarctic bacteria inhibited E. coli O157:H7 or S. Enteritidis. The latter two are Gram-negative bacteria that are often more inherently resistant to microbiocidal substances, including most bacteriocins, which is due to the protective outer membrane [14]. Only one isolate (HPG8) inhibited the Gram-negative P. fragi, whereas all except one (HPF5) inhibited B. thermosphacta with a zone radius of up to 22 mm.

Table 3.  Zones of inhibition formed by 13 inhibitor-producing bacteria isolated from Antarctic soils against eight food-borne bacteria at 10 °C (unless otherwise indicated)
Soil sourceIsolate no.Radius of zone of inhibition (mm) against
  B. thermos-phactaL. innocua 10528L. innocua 4202L. monocy-togenesP. fragiStaph. Aureus (at 15 °C)
  1. Note. No zones of inhibition were formed against Escherichia coli O157:H7 or Salmonella enterica sv. Enteritidis by any of the Antarctic isolates at 15 °C. Isolates shown in boldface were selected for further study.

Nutrient-rich, ornithogenic
Cape HallettCHF812520500
Low altitude (<500 m)
Lake HoareLHA8535000
Harrow PeaksHPC22800002
Crater CirceCrCB2320400010

3.4Characterisation of inhibitor-producing bacteria

Six inhibitor producers were selected for further study on the basis of their host spectrum and size of inhibition zone. As shown in Table 4, all six inhibitor producers except for the isolate from Crater Circe (CrCD21) were Gram-positive. All of the isolates were catalase positive and with the exception of HPG8, all were oxidase positive. All of the six inhibitor-producing bacteria were pigmented, and it was noted that many of the original 4496 colonies isolated were pigmented, e.g. yellow, orange, red and pink. The predominance of pigmented isolates in Antarctic collections has been observed previously [15,16]. Armstrong [17] suggested that these pigments might have a role in the regulation of membrane fluidity at cold temperatures. Pigments may also have a role in protecting the organisms from UV radiation that is very strong in snowy mountains even below ice cover [3]. Many of the colonies produced extracellular slime, which is often associated with psychrotrophic organisms growing in cold habitats that are also subjected to frequent or continuous desiccation.

Table 4.  Morphology and growth characteristics of six inhibitor-producing bacteria from Antarctic soils
Soil sourceIsolateGram stainOxidase reactionCatalase reactionMorphologyaGrowth optimumGrowth in salt (1–10% NaCl)c at
      Temp (°C)Mediumb4 °C10 °C22 °C30 °C37 °C
  1. Y, yes; N, no. Growth was tested in concentrations of salt ranging from 1% to 10% .

  2. aWhen grown at 10 °C.

  3. bThree media tested: dilute tryptone soy agar (DTSA), marine agar (MA) and Luria–Bertani agar (LBA). No isolates producing antimicrobial were obtained from LBA.

  4. cGrown in optimum medium (marine broth for Cape Hallett and Crater Circe isolates; dilute tryptone soy broth for Harrow Peaks isolates).

Cape HallettCHF8+++Orange colonies, single cocci24MAY (10%)Y (10%)Y (10%)NN
Crater CirceCrCD21++Cream colonies, short fat curved rods21MAY (5%)Y (5%)Y (10%)NN
 CrCB23+++Cream colonies, long thin rods21MAY (5%)Y (10%)Y (10%)Y (5%)N
Harrow PeaksHPG8++Orange colonies, cocci in pairs and tetrads, slime18DTSAY (5%)Y (10%)Y (10%)NN
 HPH17+++Yellow colonies, cocci in long chains, slime18DTSAY (5%)Y (5%)Y (5%)NN
 HPC22+++Yellow colonies, cocci in tetrads, slime18DTSAY (10%)Y (10%)Y (10%)NN

The optimum temperatures for growth shown in Table 4 were determined from Figs. 1(a) and (b). At 22 °C, which was close to the optimum growth temperature for the six inhibitor producers, all bacteria except HPH 17 tolerated up to 10% (w/v) salt. In Antarctica, salts are derived from a variety of sources including rock weathering, sea spray, seawater evaporation and atmospheric aerosols. In this study, HPC22 and CHF8 were the most salt-tolerant bacteria as they tolerated 10% NaCl at temperatures as low as 4 °C. Isolate CHF8 came from a disused penguin rookery, where the soils have high levels of salts due to penguin nasal secretions and ammonium salts derived from the decomposition of penguin guano. The salt tolerance limits obtained in this study need to be treated with some caution as the gradient plate technique may not be as reliable after 3–4 days of incubation due to diffusion of salt through the agar.


Figure 1. Growth temperature profiles of Antarctic inhibitor-producing soil isolates cultivated in dilute tryptone soy agar (HPC22, HPH17 and HPG8) and marine agar (CrCB23, CrCD21 and CHF8).

Download figure to PowerPoint

Of the six inhibitor–producers studied, only one (CrCB23) grew above 30 °C (Fig. 1(b)). The lowest temperature that could be achieved in the incubator used in this study was 3.3 °C, so temperatures below this value could not be tested. The maximum growth temperature for HPG8 was 24 °C, whereas for the remainder of the strains it was 28 °C. The optimum growth temperature for the three Harrow Peaks bacteria was 18 °C. At 10 °C (temperature at which antimicrobial assays were performed), HPG8 and HPC22 had the fastest growth rate of the six bacteria. The optimum growth temperature for CrCB23 and CrCD21 (both from Crater Circe) was 22 °C, although the latter isolate did not grow above 28 °C. All of the strains fitted the description by Morita (1975) for psychrotrophic bacteria, in that their maximum temperature of growth was above 20 °C but they could also grow at temperatures close to 0 °C.

Following growth at 10 °C for 7 days in shake-flask cultures, only one of the six bacteria (HPG8) secreted its inhibitor into cell-free supernatant (CFS). The cell-free supernatants of the remaining five isolates failed to inhibit B. thermosphacta in the deferred antagonism assay.

3.5Sensitivity of inhibitors to enzymes

The sensitivity of the inhibitors to enzymes was tested in order to gain insight into their chemical structure. None of the inhibitors produced by the six strains shown in Table 4 were sensitive to treatment with catalase, lipase or α-amylase for 1 h, indicating that the active moiety was not hydrogen peroxide, a lipid or a glycan, respectively. The inhibitors from CHF8, HPC22 and HPH17 were sensitive to both trypsin and pronase; those from CrCB23 and CrCD21 were resistant to trypsin but sensitive to pronase; and that from HPG8 was resistant to both trypsin and pronase. These results suggested that five of the six inhibitors may have been proteinaceous in nature.

3.6Identification of inhibitor producers

Fatty acid profiles of microorganisms are species-specific and can help in the identification of a particular strain. The fatty acid composition of five Antarctic inhibitor producers is shown in Table 5. The fatty acids of the three Harrow Peaks bacteria were similar and typical of Gram-positive soil bacteria such as Arthrobacter [6]. The major fatty acid for these three bacteria was aC15:0 but there were large differences in the amount of aC15:1 present: 20% in HPG8 and less than 1% in HPH 17 and HPC22. The fatty acid composition of CHF8 was also typical of Gram-positive cocci of the family Micrococcaceae that includes the genus Planococcus[18]. In contrast, the fatty acid composition of CrCD21 was typical of many Gram-negative bacteria, including the family Pseudomonadaceae that contains the genus Pseudomonas[19].

Table 5.  Fatty acid composition of five inhibitor-producing bacteria from Antarctic soils
Fatty acidComposition (wt%)
  1. Data represent means of at least three replicate determinations. Note. 0.5% as Tr; nd, not detected.


Sequence analyses of 16S rRNA from five of the inhibitor producers revealed that four bacteria could be assigned to known genera: HPG8 and HPH17 to Arthrobacter, CHF8 to Planococcus and CrCD21 to Pseudomonas. Alignments were carried out using MegAlign (DNA Star) software and phylogenetic trees were constructed using the Clustal V program (Fig. 2). The sequence from one of the strains, HPC22, had less than 50% similarity to any of the sequences in the database and so is not shown in Fig. 2. The phylogenetic analyses of the Harrow Peaks isolates HPH17 and HPG8 showed that they had a 99.2% similar identity. This would suggest that they belonged to the same species; however, they had different phenotypic characteristics. For example, their maximum growth temperatures differed by 4 °C and HPG8 was more halotolerant than HPH 17, which did not grow in 10% NaCl. Furthermore, HPG8 produced an antimicrobial different to that of HPC22. However, because most species descriptions of Arthrobacter are based on single strains, the range of variation within species is not known [20,21]. It is possible, therefore, that some of the variation in the phenotypic features of these two bacteria may be strain-, rather than species-specific.


Figure 2. Phylogenetic trees derived from 16S rRNA gene sequence data recovered from four Antarctic inhibitor-producing soil isolates: (a) CHF8; (b) HPG8 and HPH17; and (c) CrCD21. Clostridium thermocellum was used to root the trees.

Download figure to PowerPoint

4Discussion and conclusions

  1. Top of page
  2. Abstract
  3. 1Introduction
  4. 2Materials and methods
  5. 3Results
  6. 4Discussion and conclusions
  7. References

In this study of bacteria isolated from Antarctic soils, we found that no single growth medium yielded more organisms than another, and counts varied greatly between the origin of soil and the type of agar medium used. The predominance of psychrotolerant isolates is a common finding in permanently cold environments. Although the temperature of most Antarctic soils seldom rises above 10 °C, exposed soils may reach temperatures as high as 20–25 °C during the summer. Therefore, psychrotolerant bacteria with growth optima at 18–24 °C are able to take full advantage of the short time available for rapid growth when the soil is warm. The soil samples in this study were collected during the summer, which could account for the predominance of psychrotolerant bacteria isolated.

The detection rate for antimicrobial production (% inhibitor producers) in this study was 0.29%. This is comparable to detection rates reported in the literature, although they depend partly on the criteria used for selection. In a study [22] based on a total of 663,533 colonies isolated from dairy and meat sources, a detection rate of 0.20% was reported for bacteriocin producers using direct plating methods. When a further 83,000 colonies from fish and vegetable sources were isolated using enrichment procedures, a detection rate of 3.4% was reported for bacteriocin producers. Others [13] have reported detection rates of 0.35% for bacteriocin producers following screening of 48000 lactic acid bacteria from food sources. These studies were focused on the isolation of antimicrobial peptides from mesophilic microorganisms with generally recognized as safe (GRAS) status. In studies focused on obtaining bacteria with general antimicrobial activities, the detection rates have been much higher. For example, Hentschel et al. [23] isolated over 200 bacteria from Mediterranean sponges and found that 11.3% had antimicrobial activity.

Maximum likelihood analyses of HPG8 and HPH17 16S rRNA gene sequences with those from Arthrobacter spp. obtained from the RDP database placed the Antarctic isolates within the Arthrobacter genus. A. psychrolactophilus was identified as the nearest neighbour with percentage similarities of 94.6% and 94.3% for HPH17 and HPG8, respectively. Stackebrandt and Goebel [24] have proposed that organisms with < 97% similarity could be considered different species. They have also emphasised that DNA–DNA hybridisation remains the optimal method for confirming extent of relatedness. When 16S rRNA gene sequence homology was below 97.5%, these authors found that the two organisms may have no more than 60% similarity in their DNA. The 16S rRNA gene sequence similarity of HPG8 and HPH17 with A. psychrolactophilus was less than this proposed threshold, suggesting that HPH17 and HPG8 may be new species of the genus Arthrobacter. Furthermore, the sequence similarities between these two microorganisms and their nearest Arthrobacter species (A. psychrolactophilus) was significantly less than seen between many other known species of Arthrobacter[25,26]. For example, the reported similarities between A. globiformis and A. polychromogenes and A. oxydans are 96.2% and 96.0%, respectively, and the similarity between A. flavus and A. agilis is 97.9%. DNA–DNA hybridisation studies would be necessary to substantiate the case for the designation of HPG8 and HPH17 as new species.

Analysis of the 16S rRNA gene sequence from the Cape Hallet isolate, CHF8, placed it within the Planococcus genus. Sequence alignments showed that CHF8 had 93.3% sequence similarity with Pl. kocurii, significantly lower than the 97.5% threshold suggested by Stackebrandt and Goebel [24] discussed above. On this basis, isolate CHF8 is not a strain of Pl. citreus (90.9%) or Pl. kocurii (93.3%) but is a distinct species.

Maximum likelihood analyses of the 16S rRNA gene sequence from strain CrCD21 and those of Pseudomonas species placed the isolate from Crater Circe within the Pseudomonas genus. A high % similarity of 99.4% was shown with an unidentified “gamma proteobacterium”; however, the fact that only an incomplete sequence of the rRNA gene from the latter was available in the database precluded confirmation of identity. Comparison of CrCD21 with the full rRNA sequences (>1400 bp) from identified Pseudomonas spp. showed that this isolate had 96.4%, 96.2% and 96.1% similarities with P. veronii, P. migulae and P. mandelii, respectively. The latter strains, originally isolated from mineral waters and identified as fluorescent pseudomonads, were grouped into three phenotypic subclusters, XIIIb, XVa and XVc. They were subsequently characterised at the genotypic level and identified as novel species of Pseudomonas[27,28]. High 16S rRNA gene similarities (>99%) were reported for these new species; however, the results were not supported by high DNA–DNA hybridisation levels, which were <36%[27,28].

The fatty acid composition of the Antarctic inhibitor–producers in this study added further support to the genus designations obtained above. The main fatty acid of the Harrow Peaks bacteria was identified as aC15:0, which has also been reported for other Arthrobacter spp. including A. psychrolactophilus[25]. The isolate CHF8 had a complex fatty acid profile and not all of the fatty acids could be identified. The main fatty acid was aC15:0, which is a major component of Gram-positive cell membrane lipids, is indicative of psychrotolerance and has been observed in psychrotolerant Listeria spp. [29]. The major cellular fatty acids of planococci are branched-iso and anteiso-branched fatty acids with aC15:0 being the predominant cellular fatty acid [18]. The profile fitted well with the DNA sequence, which placed this isolate within the genus Planococcus.

The Crater Circe isolate, CrCD21, had a high content of C16:0, C16:1 and C18:1, typical of Pseudomonas species [19], including those that grow at low temperatures [6]. From the results of 16S rRNA gene analysis this isolate was identified as Pseudomonas. No published fatty acid profiles are available for P. veronii, P. migulae or P. mandelli, the species most closely matching the 16S rRNA gene sequence of CrCD21. The fatty acid composition of CrCD21 closely matches that of another psychrotrophic Pseudomonas sp. CL1-1 isolated by White et al. [6] also from Antarctic soil. However, they are not the same species because, unlike CrCD21, CL1-1 does not form cyclopropane fatty acids.

In the natural environment, antibiotics and other secondary metabolites serve multiple functions related to the survival of the microorganism [30]. There are three reports in the literature of arthrobacters producing antimicrobial compounds. Kamigiri et al. [31] found that an Arthrobacter sp. isolated from Indonesian soil produced a quinolone antibiotic. This organism was classified using phenotypic characteristics, including the identification of lysine in the peptidoglycan cell wall. The antibiotic was purified and characterised; however, no subsequent work has been reported. Carnio et al. [32] investigated the anti-listerial properties of microorganisms isolated from French and German red-smear cheeses and reported a single strain of Arthrobacter sp. showing antagonism against listeriae by testing 53 physiological characteristics, including cellular fatty acid composition. Hentschel et al. [23] isolated Arthrobacter spp. capable of antimicrobial production from the Mediterranean sponges Aplysinia aerophoba and Aplysinia cavernicola. One of the antimicrobial-producing strains, Arthrobacter sp. SB95, was identified using 16S rRNA gene sequence analysis and found to belong to the same species as a marine Arthrobacter isolate MB8-13 that had been recovered from Antarctic sea ice. While 16S rRNA gene sequence analysis did not place Arthrobacter sp. HPH17 in the same species as the marine isolate MB8-13, it has previously been reported that different bacteria can produce the same bacteriocin [33].

In the present study, Arthrobacter sp. HPH17, Planococcus sp. CHF8 and Pseudomonas sp. CrCD21 were found to produce proteinaceous compounds sensitive to proteases; the latter two strains produced compounds active against Listeria spp. and B. thermosphacta. No reports of bacteriocin or other antimicrobial production by Planococcus sp. have been published to date. Antimicrobial activity has often been associated with Pseudomonas spp., e.g. Sano et al. [34] have reported the production of narrow-spectrum bacteriocins known as pyocins by P. aeruginosa. Similarly, Laue et al. [35] and Parret and DeMot [36] have identified bacteriocins from P. fluorescens.

In conclusion, this study has shown that bacteria isolated from the Antarctic environment are an untapped source of novel antimicrobial compounds, which may be exploitable in food, therapeutic and health applications in the future. Further work is needed to characterize the chemical structure of the antimicrobial compounds and to verify the activity of the inhibitors in real applications such as chilled foods.


  1. Top of page
  2. Abstract
  3. 1Introduction
  4. 2Materials and methods
  5. 3Results
  6. 4Discussion and conclusions
  7. References
  • [1]
    Natural Antimicrobials for the Minimal Processing of Foods (Roller, S., Ed.), p. 306. Woodhead Publishing and CRC Press, Cambridge, UK and Boca Raton, FL, USA, 2003
  • [2]
    Wynn-Williams, D.D. (1990) Ecological aspects of Antarctic microbiology. Adv. Microbial Ecol. 11, 71146.
  • [3]
    W.F. Vincent (1988Microbial Ecosystems of Antarctica. Cambridge University Press, Cambridge, MA. 1–6 153–176.
  • [4]
    Kekessy, A.M., Piguet, D.B. (1970) New method for detecting bacteriocin production. Appl. Microbiol. 20, 282283.
  • [5]
    D. Roberts (1995) Practical Food Microbiology (D. Roberts, W. Hooper, M. Greenwood, Eds.) Public Health Laboratory Service, 191–202.
  • [6]
    White, L.P., Wynn-Williams, D.D., Russell, N.J. (2000) Diversity of thermal responses of lipid composition in the membranes of the dominant culturable members of an Antarctic fellfield soil bacterial community. Ant. Sci. 12, 386393.
  • [7]
    Venables, W.A., Wimpenny, J.W., Ayres, A., Cook, S.M., Thomas, L.V. (1995) The use of two-dimensional gradient plates to investigate the range of conditions under which conjugal plasmid transfer occurs. Microbiology 141, 27132718.
  • [8]
    M. Kates, E.L. Pugh, G. Ferrante (1984) Regulation of membrane fluidity by lipid desaturases. (Eds. M. Dates, A. Manson), Membrane Fluidity, Biomembranes (vol. 12) Plenum Press, New York. 379–395.
  • [9]
    I. Sambrook, E.F. Fritsch, T. Maniatis (1989) Molecular Cloning: A Laboratory Manual. Cold Spring Harbor Laboratory, Cold Spring Harbor, New York.
  • [10]
    Bowman, J.P., Cavanagh, J., Austin, J.J., Sanderson, K. (1996) Novel Psychrobacter species from Antarctic ornithogenic soils. J. Syst. Bacteriol. 46, 841848.
  • [11]
    H.S. Vishniac (1993) Soil microgeology. (Ed. E.I. Friedmann), Antarctic Microbiology. Wiley–Liss, New York. 297–341.
  • [12]
    Morita, Y. (1975) Psychrophilic bacteria. Bacteriol. Rev. 39, 144167.
  • [13]
    Vaughan, E.E., Caplice, E., Looney, R., O'Rourke, N., Coveny, H., Daly, C., Fitzgerald, G.F. (1994) Isolation from food sources, of lactic acid bacteria that produced antimicrobials. J. Appl. Bacteriol. 76, 118123.
  • [14]
    Helander, I., von Wright, A., Matilla-Sandholm, T-M. (1997) Potential of lactic acid bacteria and novel antimicrobials against Gram-negative bacteria. Trends Fd. Sci. Technol. 8, 146150.
  • [15]
    Rotert, K.R., Toste, A.P., Steiert, J.G. (1993) Membrane fatty acid analysis of Antarctic bacteria. FEMS Microbiol. Lett. 114, 253258.
  • [16]
    Pearce, D.A., van der Gast, C.J., Lawley, B., Ellis-Evans, J.C. (2003) Bacterioplankton community diversity in a maritime Antarctic lake, determined by culture-dependent and culture-independent techniques. FEMS Microbiol. Ecol. 45, 5970.
  • [17]
    Armstrong, G.A. (1997) Genetics of eubacterial carotenoid biosynthesis: a colourful tale. Annu. Rev. Microbiol. 51, 629659.
  • [18]
    W.M. O'Leary, S.G. Wilkinson (1988) Gram-positive bacteria. (Eds. C. Ratledge, S.G. Wilkinson), Microbial Lipids (vol. 1) Academic Press, London. 117–201.
  • [19]
    S.G. Wilkinson (1988) Gram-negative bacteria. (Eds. C. Ratledge, S.G. Wilkinson) Microbial Lipids (vol. 1) Academic Press, London. 299–488.
  • [20]
    R.M. Keddie, D. Collins, D. Jones (1986) Genus Arthrobacter. (Eds. P.H.A. Sneath, N.S. Mair, M.E. Sharpe, J.G. Holt), Bergey's Manual of Systematic Bacteriology (vol. 2) Williams and Wilkins, Baltimore. 1288–1301.
  • [21]
    Keddie, R.M., Jones, D. (2002) The Genus Arthrobacter. In: The Prokaryotes (on line edition). Springer, New York, accessed 17/03/02. Available from: <>
  • [22]
    Coventry, M.J., Gordon, J.B., Wilcock, A., Harmark, K., Davidson, B.E., Hickey, M.W., Hiller, A.J., Wan, J. (1997) Detection of bacteriocins of lactic acid bacteria isolated from foods and comparison with pediocin and nisin. J. Appl. Microbiol. 83, 248258.
  • [23]
    Hentschel, U., Schmid, M., Wagner, M., Fieseler, L., Gernert, C., Hacker, J. (2001) Isolation and phylogenetic analysis of bacteria with antimicrobial activities from the Mediterranaen sponges Aplysina aerophoba and Aplysina cavernicola. FEMS Microbiol. Ecol. 35, 305312.
  • [24]
    Stackebrandt, E., Goebel, B.M. (1994) Taxonomic note: a place for DNA–DNA reassociation and 16S rRNA gene sequence analysis in the present species definition in bacteriology. Int. J. Syst. Bacteriol. 44, 846849.
  • [25]
    Loveland-Curtze, J., Sheridan, P.P., Gutshall, K.R., Brenchley, J.E. (1999) Biochemical and phylogenetic analyses of psychrophilic isolates belonging to the Arthrobacter subgroup and description of Arthrobacter psychrolactophilus, sp. nov. Arch. Microbiol. 171, 355363.
  • [26]
    Crocker, F.H., Fredrickson, J.K., White, D.C., Ringelberg, D.B., Balkwill, D.L. (2000) Phylogenetic and physiological diversity of Arthrobacter strains isolated from unconsolidated subsurface sediments. Microbiology 146, 12951310.
  • [27]
    Elomari, M., Caroler, L., Hoste, B., Gillis, M., Izard, D., Leclerc, H. (1996) DNA relatedness among Pseudomonas strains isolated from natural mineral waters and proposal of Pseudomonas veronii sp. nov. Int. J. Sys. Bacteriol. 46, 11381144.
  • [28]
    Verhille, S., Baida, N., Dabboussi, F., Izard, D., Leclerc, H. (1999) Taxonomic study of bacteria isolated from natural mineral waters: proposal of Pseudomonas jessenii sp. nov. and Pseudomonas mandelii sp. nov. Syst. Appl. Microbiol. 22, 4558.
  • [29]
    Annous, B.A., Becker, L.A., Bayles, D.O., Labeda, D.P., Wilkinson, B.J. (1997) Critical role of Anteiso-C15:0 fatty acid in the growth of Listeria monocytogenes at low temperatures. Appl. Environ. Microbiol. 63, 38873894.
  • [30]
    Vining, L.C. (1990) Functions of secondary metabolites. Annu. Rev. Microbiol. 44, 395427.
  • [31]
    Kamigiri, K., Tokunaga, T., Shibazaki, M., Setiavan, B., Rantiatmodjo, R.M., Morioka, M., Suzuki, K-I. (1996) YM-30059, a novel quinolone antibiotic produced by Arthrobacter sp. J. Antibiot. 49, 823825.
  • [32]
    Carnio, M.C., Eppert, I., Scherer, S. (1999) Analysis of the bacterial surface ripening flora of German and French smeared cheeses with respect to their anti-listerial potential. Int. J. Food Microbiol. 47, 8997.
  • [33]
    Jack, R.W., Tagg, J.R., Ray, B. (1995) Bacteriocins of Gram positive bacteria. Microbiol. Rev. 59, 171200.
  • [34]
    Sano, Y., Matsui, H., Kobayashi, M. (1993) Molecular-structures and functions of pyocin-S1 and pyocin-S2 in Pseudomonas aeruginosa. J. Bacteriol. 175, 29072916.
  • [35]
    Laue, R.E., Jiang, Y., Chabra, S.R., Jacob, S., Stewart, G.S.A.B., Hardman, A., Downie, J.A., O'Gara, F., Williams, P. (2000) The biocontrol strain Pseudomonas fluorescens F113 produces the rhizobium small bacteriocin, N-(3-hydroxy-7-cis-tetradecenoyl) homoserine lactone, via HdtS, a putative novel N-acylhomoserine lactone synthase. Microbiology 146, 24692480.
  • [36]
    Parret, A.H.A., De Mot, R. (2002) Bacteria killing their own kind: novel bacteriocins of Pseudomonas and other gamma-proteobacteria. Trends Microbiol. 10, 107112.