1 Division of Food Sciences, School of Biosciences, University of Nottingham, Sutton Bonington Campus, Loughborough, Leicestershire LE12 5RD, UK
2 School of Life and Environmental Sciences, University of Nottingham, University Park, Nottingham NG7 2RD, UK
Correspondence
Jack A. Gilbert
gilbertj{at}post.queensu.ca
![]() |
ABSTRACT |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
The GenBank accession numbers for the sequences reported in this paper are listed in Table 3.
Present address: Department of Biochemistry, Queen's University, Kingston, Ontario, Canada K7L 3N6.
![]() |
INTRODUCTION |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
To our knowledge only five bacteria have been shown to possess AFP activity, Pseudomonas putida (Sun et al., 1995; Xu et al., 1998
; Kawahara et al., 2001
), Micrococcus cryophilus, Rhodococcus erythropolis (Duman & Olsen, 1993
), Marinomonas protea (Mills, 1999
) and a Moraxella species (Yamashita et al., 2002
). P. putida was isolated from plants in the high Arctic, R. erythropolis was isolated from the mid-gut of beetle larvae from an unknown location, Mic. cryophilus was purported to have been isolated from chilled fresh pork sausages, Mar. protea was isolated from the icewater interface of Ace Lake in the Vestfold Hills of Antarctica and Moraxella sp. was isolated from the McMurdo Dry Valleys region of Antarctica. AFPs are likely to have evolved in organisms that experience regular phases of low temperature and freezing. Polar regions, with their continuous low temperatures, are environments that may be expected to force the evolution of AFPs within their populations of extremophile bacteria.
The availability of free water controls the biological activity of Antarctic organisms. The Antarctic continent has only small areas of ice-free land. These so-called Antarctic oases carry a surprising diversity of lake types. One such area is the Vestfold Hills of Eastern Antarctica, which contains over 150 lakes ranging from freshwater to hypersaline. Most of these lakes were formed by a process of isostatic uplift following the last major glaciation some 10 000 years ago (Adamson & Pickard, 1986). During uplift fjords were cut off and seawater was trapped in hollows. Those lakes having inflows and outflows were gradually flushed and evolved into freshwater systems, while those lakes in closed basins evolved into saline lacustrine systems. During evolution they suffered sublimation and ablation that altered the anion and cation ratios (Burton, 1981
). Some of the lakes underwent episodes of meromixis. Simultaneously, some freshwater lakes were formed at the ice cap edge as it retreated.
The lakes of Antarctica are dominated by microbial communities. They have few or no zooplankton, and no fish (Laybourn-Parry et al., 2002). Despite the harsh environmental conditions experienced in these lacustrine ecosystems throughout the year, their microbial communities continue to function, even during winter (e.g. Henshaw & Laybourn-Parry, 2002
). Physiological plasticity is the key to survival and is likely to involve novel biochemicals such as low-temperature enzymes, AFPs and modified phospholipids. As yet there is limited published information on this important aspect of polar microbial ecophysiology (Herbert, 1986
; Whyte & Innis, 1992
; Rotert et al., 1993
; Ray et al., 1994
; Russell & Hamamoto, 1998
; Barrett, 2001
).
The current study was undertaken to assess the occurrence of AFP-active Antarctic lake bacteria across a salinity spectrum from freshwater to hypersaline (240 ) and to determine their taxonomy by molecular techniques. We were particularly interested in relating the occurrence of AFP activity in bacterial strains to the physico-chemical conditions in which they evolved. This article provides information on a wide range of Antarctic lake environments and the occurrence of bacterial AFPs therein.
![]() |
METHODS |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
Sampling from lakes.
Water samples were obtained with a 2 l Kemmener bottle through a hole drilled in the ice with a jiffy drill (AAA International). When the lake had ice cover capable of supporting personnel and equipment, it was possible to collect samples at different depths. Water was collected from 0, 2, 4, 8 and 10 m, where the depth of oxygenated waters or the depth of the lake dictated the depth of sampling. Zero metre depth was taken as the base of the ice cover and all other depths were taken from that point. In the absence of ice cover capable of supporting personnel and equipment, only surface water samples were taken. For each sample, 4 l water was taken and stored in acid-washed (10 % HCl) polypropylene containers for subsequent inorganic chemical analysis and dissolved organic carbon (DOC) analysis. A further 100 ml water was taken into a sterilized glass Duran bottle (for microbiological analysis). Ice cores were also taken from some lakes (e.g. Ace Lake, Pendant Lake) using a Sipre corer, sliced in the laboratory and the slices thawed in sterilized lake water to prevent osmotic shock. The sampling equipment was soaked in 10 % HCl and then rinsed in deionized distilled water prior to sampling at a new site.
Bacterial culturing.
Lake water (100 µl) and melted ice cores (100 µl) were spread-plated onto Tryptic Soy Agar (TSA; Sigma), yeast agar (YSA) [all components from Sigma except where indicated (l-1): 3 g yeast extract, 3 g malt, 5 g peptone, 10 g glucose, 20 g agar (Difco)], seawater agar (SWA) [3·8 % salinity; all components from Sigma except where indicated (l-1): 1 g yeast extract, 1 g peptone, 38 g salt (Coral Life; Aquatics Online), 15 g agar (Difco)] and 1/2 seawater agar (1/2 SWA; 1·9 % salinity: same as SWA but only 19 g salt). Initially all plates were grown for 1 week at three different temperatures, 5, 15 and 25 °C. All cultures grew better at 15 °C, hence all subcultures were incubated at 15 °C for 1 week. All bacterial isolates were Gram-stained using a standard method (Rodina, 1972).
Water chemistry analysis.
Samples were stored at 1 °C until all initial processing was completed. First, 500 ml aliquots were taken for inorganic nutrient analysis, filtered through GF/F (Whatman) glass microfibre filters and analysed colorimetrically for soluble reactive phosphate (SRP, PO4-P), ammonium (NH4-N), nitrate (NO3-N) and nitrite (NO2-N) according to the method of Mackereth et al. (1988). The remaining water was filtered through a 47 mm GF/F (Whatman) microfibre filter. The filtrate was decanted into a 2 l polypropylene bottle (acid-washed, 10 % HCl), 25 ml of which was dispensed into a sterile universal tube and frozen at -20 °C for DOC analysis, which was performed using a Shimadzu TOC 5000 carbon analyser (Duisburg).
Chromosomal DNA extraction.
The CTAB method of chromosomal DNA extraction (Ausubel et al., 1999) was used on cultures grown for 1 week at 15 °C in liquid media that matched the solid media upon which they were isolated.
Amplified rDNA restriction analysis (ARDRA).
The ARDRA procedure was adapted from Vaneechoutte et al. (1995). Chromosomal DNA (1 µl) was added to a 49 µl PCR reaction mixture consisting of 21 µl sterile reverse osmosis (RO) H2O, 16 µl dNTP mixture (ABgene) (6·26 µl each dNTP made up to 1 ml in sterile H2O), 5 µl Buffer IV (ABgene), 4 µl 25 mM MgCl2 (ABgene), 1 µl universal primer 1.5 F (5'-TGGCTCAGATTGAACGCTGGCG-3') (Sigma-Genosys), 1 µl universal primer 1.5 R (5'-TACCTTGTTACGACTTCACCCCA-3') (Sigma-Genosys) and 1 µl Taq DNA Polymerase (ABgene). The PCR reaction was run on a thermal cycler (Techne; Progene) using the following programme: 10 min at 95 °C followed by the addition of ice-cold Taq polymerase (1 µl per reaction); then 30 cycles of 0·5 min at 95 °C, 1 min at 55 °C, 1·5 min at 72 °C, after which there was a single cycle of 72 °C for 8 min. The PCR reaction amplified a 1500 bp 16S rDNA amplicon.
PCR products were digested overnight with AluI or HpaII (BRL GibcoInvitrogen). The reaction conditions comprised 12 µl sterile H2O, 1 µl HpaII or 1 µl AluI, 2·0 µl buffer (10x React 1; BRL Gibco) and 5·0 µl PCR product, incubated overnight at 37 °C. The digests were separated by electrophoresis on a 2 % agarose gel (3 : 1 Nu-Sieve) in 1x TAE buffer (Sambrook et al., 1989) at 20 V.
The restriction digest images were imported into the ImageMaster 1D elite V3.01 gel analysis software (Amersham Pharmacia Biotech and Non-Linear Dynamics) and patterns were compared using the Dice similarity coefficient and UPGMA cluster analysis (Michener & Sokal, 1957).
16S rDNA sequencing.
The 16S rRNA gene was amplified using PCR (see ARDRA section), purified using the QIAquick PCR purification kit (Qiagen) and the nucleotide sequence of the gene was determined by the Biopolymer Synthesis and Analysis Unit, Nottingham University. Where satisfactory results were not obtained from the purified PCR product, it was cloned using the TOPO TA Cloning kit (Invitrogen) prior to sequencing.
Crude cellular lysate extraction protocol.
The protocol for preparing crude cellular lysates was adapted from Johnston et al. (1982). Isolates were incubated at 15 °C for 1 week in 250 ml broth cultures (based on original isolation medium) and then cold-acclimatized at 4 °C for 1 week. The cultures were then centrifuged (15 500 g, 4 °C, 10 min; Beckman JA-14 fixed angle rotor) and the cell pellet was resuspended in 1 ml ice-cold protease inhibitor buffer (25 mM Tris/HCl, pH 7·0, 1 mM EDTA, 1 mM PMSF in ethanol, 2 µg Pepstatin A ml-1 in methanol), to which was added 500 mg 106 µm or finer glass beads (Sigma). To break the cells, this suspension was shaken using a Mini Beadbeater (Biospec Products) for 1 min at 5000 r.p.m. and then chilled on ice for 1 min (repeated five times). After mixing the lysate with 500 µl ice-cold protease inhibitor buffer the liquid layer (supernatant) was transferred to a clean sterile ice-cold tube and frozen at -20 °C. The protein concentration was determined using the Bio-Rad Protein Assay Kit.
High-throughput AFP protocol (HTAP).
Sucrose (50 µl, 60 %) and lysate extract (50 µl) were aliquoted into a 96-well microtitre plate. Fish AFP III (1 mg ml-1 in 30 % sucrose; A/F Protein Inc.) was used as a positive control and 30 % sucrose as a negative control. Thirty percent sucrose was used to standardize solute concentrations between different cellular lysate preparations. Lysate extract of Escherichia coli JM109 was used as a negative control in preliminary development assays. The microtitre plate was flash-frozen at -70 °C for 10 min then placed on a cooling plate (CamLab; Fryka KP 281) for recrystallization. The cooling plate was pre-cooled to -6 °C and stored in a cold room with an ambient temperature of -6 °C. The microtitre plate was viewed after 5 days, 15 cm from the surface of a light box in the cold room, so observation could occur at -6 °C to prevent temperature change within the sample altering ice-crystal size. Wells with AFP-active solutions appeared turbid, whereas the AFP-inactive solutions appeared clear; activity was recorded and a digital photograph was taken of each plate. Treatment of samples with Proteinase K (2 mg ml-1, 37 °C, 1 h) to destroy protein resulted in loss of recrystallization inhibition (RI) activity, indicating that activity was the result of a protein.
Splat assay.
The protocol was modified from Knight et al. (1988). Lysate extract solutions were standardized (with respect to solute concentration) by adding 50 µl 60 % sucrose to 50 µl lysate extract. The resulting solution (5 µl) was placed between two circular 16 mm diameter, numbered coverslips, which were blotted dry to remove any excess fluid. The coverslips were flash-frozen in 2,2,4-trimethylpentane pre-cooled to -70 °C (using dry ice) for 2 min, then placed in a -6 °C solvent bath [2,2,4-trimethylpentane pre-chilled to -6 °C using a Haake C water bath circulator and two Haake temperature control units (PG40 and F4)] and held for 1 h to allow recrystallization. The coverslips were viewed in the bath using an EF L 20/0.30 160/0-2 objective on a Leitz Dialux 20 EB stage. The observed crystal shape, size and density were recorded and related to the level of AFP activity using the Splat scoring system (see Table 1
).
|
![]() |
RESULTS |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
|
Water temperature in Ace Lake and Pendant Lake varied only slightly throughout the year, with surface water temperature fluctuating between -2 and +1 °C. Temperature was shown to steadily increase with depth throughout the year. A temperature anomaly was noted in both lakes during September with a relatively high temperature of 1415 °C at 10 m (data not shown). In Triple Lake, the depth profile for temperature was virtually uniform, but the temperatures were relatively warm in summer (
3 °C) and extremely cold in winter (-14 °C). Oval Lake had the coldest water temperature recorded during the summer, -1·9 °C.
Ace Lake, a meromictic lake, showed a distinct salinity profile in the mixolimnion (the monimolimnion was not sampled during this study) throughout the year with an increase in salinity with increasing depth, with the exception of September and November when the water column was virtually isohaline. Pendant Lake also showed an increase in salinity with depth on all sampling dates except November when a slight decrease in salinity with increasing depth was recorded. The mean salinity was shown to increase from January to November in both lakes. The salinity profile for Triple Lake was virtually isohaline; however, there was generally a small increase in salinity with depth. Oval Lake had a surface salinity of 34 .
Inorganic nutrients and DOC concentrations are crucial to controlling heterotrophic bacterial growth. Ace Lake and Pendant Lake both had nitrate concentrations below the level of detection during the 1999/2000 summer and very low or below the level of detection during July, but which increased by September. During November, Pendant Lake nitrate concentration increased again but the concentration in Ace Lake decreased to below the detection threshold. Triple Lake showed an increase in concentration from January to July. Nitrate was below the level of detection in Oval Lake during the summer. Nitrite concentrations in all four lakes were high during the 1999/2000 summer, but decreased considerably through the winter and remained low through to November. Ammonium ion concentration profiles were similar in Ace, Pendant and Triple Lakes, with low concentrations during the first summer followed by a peak in July. The concentration then decreased to near the level of detection in September and remained low in November, except in Ace Lake which had its highest concentrations during September. Oval Lake had a similar ammonium ion concentration to the other lakes during the summer.
SRP showed a similar profile to ammonium ion with concentrations higher in January to March than September to November, and peaking in July. This was indicative of all four lakes.
Concentration profiles for DOC were similar for both Ace Lake and Pendant Lake, with a relatively constant level throughout the sampling season. However, concentrations were lower in July than September, and November had the lowest concentration for the year. Triple Lake concentrations were highest in January, lowest in July and all showed a steady increase in concentration from July to November. Oval Lake showed a higher DOC concentration than either Ace or Pendant Lake, yet a lower concentration than Triple Lake, indicating a salinity dependent scale for DOC concentration.
Bacterial isolation and characterization
Eight hundred and sixty six bacterial cultures were isolated from the 38 lakes sampled. All bacterial isolates grew more densely at 15 °C, than at 5 or 25 °C; this putatively suggests a psychrophilic classification (based on Herbert, 1986). Approximately 87 % of bacterial isolates were Gram-negative, of which
98 % had rod morphologies that varied in length and type, from straight to club-shaped to long, thin curved rods. The remaining bacterial isolates were coccoid; these included all Gram-positive isolates.
HTAP development
The assay was based on the principle that a snap-frozen (rapidly frozen at -70 °C) sample is made up entirely of very small ice crystals. When a non-AFP solution is recrystallized, the lack of ice-growth inhibition means that surviving crystals increase in size at the expense of the others, whereas solutions containing AFPs show a limited increase in crystal size (Knight et al., 1988). The highly dense, small crystals resulting from AFP activity cause extensive light refraction so that the solution appears opaque as opposed to the transparent non-AFP solutions. Fish AFP III (1 mg ml-1) was used as a positive control. Crude cellular lysate extract from E. coli strain JM109 and sterile RO H2O were used as negative controls. It was found that 100 µl solutions (50 µl LE+50 µl 60 % sucrose) in 96-well microtitre plates provided the most effective, practical guide for rapid visual identification of AFP activity.
Serial dilutions of fish AFP III (1 : 10, 1 : 100 and 1 : 1000) were compared against the standard 1 mg fish AFP III ml-1 solution. A decrease in AFP concentration produced a decrease in RI activity (indicated by a decrease in opacity of solution), confirming a quantitative relationship between degree of opacity and AFP concentration, and that the minimum concentration tested, 1 µg ml-1, was detectable when visualized against a negative control.
The assay was applied to all 866 isolates from the Antarctic lake environments. Of these, 187 (21·6 %) proved to be positive for AFP activity. The level of activity could not be quantified for the samples as with a Splat assay, but were recorded as either active or non-active. Those with reduced transparency (increased opacity) or colouration which was difficult to confirm were classed as positive, and those extracts with refraction equal to or less than the negative control (30 % sucrose) were classed as negative. Of the 187 bacteria identified as AFP-active using the HTAP assay, only 12 were Gram-positive.
Splat assessment of the HTAP positive isolates
The 187 presumptive AFP producers were analysed using the Splat assay to confirm AFP activity. Nineteen of the bacteria were confirmed to have some level of AFP activity (an activity of 3 or more) (Table 1), all of which were Gram-negative. Ten of these 19 isolates showed a Splat score of 4 or 5 (Table 1
), indicating high AFP activity, comparable with fish AFP III (1 mg ml-1). These strains were isolated from Ace Lake, Triple Lake, Pendant Lake, Pendant Ice cores and Oval Lake. The greatest number of AFP-active isolates was cultured from Ace Lake, possibly due to a higher sampling frequency in this lake compared with the others (Table 1
).
Those extracts which proved negative using the HTAP assay were assessed using the Splat assay to demonstrate that there were no false negative results. One hundred of the HTAP negative isolates were also shown to be negative for AFP activity using the Splat assay (results not shown). This indicated that the HTAP assay did not produce false negative results.
There was no direct relationship between concentration of protein in the crude cellular extracts, prepared from each putative AFP-active isolate, and the Splat score (Table 1). This might indicate that the specific activity of the AFPs varied between bacterial strains, and that high total cellular protein concentrations did not necessarily correspond to high AFP activity. The conditioned growth media for two of the isolates (302 and 124) demonstrated no activity with either the HTAP or Splat assays, suggesting that the putative AFP is not expressed extracellularly. However, the conditioned growth media for the remaining isolates was not assessed in this investigation.
ARDRA of AFP-active bacterial strains
A comparison was made of the HpaII and AluI digests of the 1500 bp 16S rDNA amplicons for the 19 AFP-active isolates (Table 1) and Mar. protea (Mills, 1999
), an AFP-active
-proteobacterium which was isolated from Ace Lake during 1996/1997 (Fig. 1
a, b). The latter was included because eight AFP-active isolates were cultured from Ace Lake. Analysis of the HpaII ARDRA patterns (Fig. 1a
) showed that five of the isolates formed a 100 % similarity group with the Mar. protea isolate, indicating that this was the likely identity of these strains. Strains 302 and 33, 39 and 86, and 47 and 492 also showed 100 % similarity, suggesting that these were identical species. The remaining strains showed a much lower association, suggesting a more distant relationship.
|
16S rDNA sequencing
Sequence data for the 16S rDNA genes were entered into the GenBank database and each sequence assigned an accession number (Table 3). The closest published relative for each isolate was determined by comparison with the database using the FASTA and BLAST search programs (Altschul et al., 1990
).
|
Isolates 492 and 47 showed more than 99 % similarity to Stenotrophomonas maltophilia (AJ293464). Their similarity shown by ARDRA also supported their being related species. However, the sequence data for isolate 492 only comprised 408 bp and so were not sufficient to make a definitive identification. Identification of isolate 47 was based on 1182 bp and was therefore sufficient for putative affiliation.
Isolate 494 was identified as a Sphingomonas strain. However, this identification was based on 153 bp which, as with isolate 492, was considered too small an amplicon for confident identification. This isolate produced the most AFP-active protein extract of the 19 isolates (Table 1), therefore it was considered important to identify it to at least the genus level. Isolate 494 showed an identical relationship with two other bacterial species using ARDRA (data not shown). The amplicons from these two bacterial strains, M3C203B-B (AF395031) and SIA181-1A1 (AF395032), were also sequenced which showed their hosts to be members of the genus Sphingomonas, confirming that isolate 494 was probably a member of the Sphingomonas genus. All other strains with insufficient sequence length (<800 bp continuous sequence; Table 1
) could not be considered to be unequivocally identified.
![]() |
DISCUSSION |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
Identification of AFP activity within the total cellular protein samples using the HTAP assay was sometimes hindered by colouration, caused by pigmentprotein complexes, such as cytochromes or carotenoids (found extensively in Antarctic bacteria as protection against UV damage) (George et al., 2001), which prevented the level of refraction from being observed. Degradation of these pigmented proteins may have also caused AFP inactivation, although this was not determined here. This artefact meant that samples with refraction inhibitory colouration were recorded as positive for AFP activity and then assessed using the Splat assay at a later date. Due to the large sample size and uncertainty of location of pigment within the cell, fractionation of the solution to remove pigmentation was not performed. The HTAP assay has a long recrystallization time (5 days) compared with the Splat assay (1 hour), which provided the optimal level of contrast between positive and negative samples. For an experienced observer this time can be reduced to 24 hours. At 5 days, there was no variation between the scores of an experienced observer and a novice.
Assessment of AFP activity in the HTAP-positive isolates using the Splat assay suggested that only 2·1 % of the cultured bacteria were AFP-active; however, the cultured biota undoubtedly does not reflect the true bacterial diversity. This would indicate that AFP activity is a rare phenomenon in culturable Antarctic lake bacteria; however, as the 866 bacterial isolates have not yet been characterized, replication of species within the collection could lead to underestimation of species richness. It is also widely accepted that the majority of bacterial diversity is not presently amenable to culture. The presence of other cold adaptation techniques, which mainly involve structural adaptations in proteins and membrane lipids to maintain functionality at cold temperatures (Fields, 2001), indicates other strategies could have been implemented for growth and multiplication in the extreme cold. It is also possible that some bacteria may be genotypically AFP-positive, but phenotypically negative because of inappropriate induction conditions or protein degradation/inactivation during extraction. If this were the case, then there could be a very high proportion of false negatives within the 866 bacteria cultivated. Knowledge of the genes which encode AFPs should provide alternative strategies for investigation. However, to date, bacterial AFP genes remain uncharacterized. Another potential difficulty is that AFP proteins, and hence the genes from which they are derived, are often non-homologous (Jia & Davies, 2002
). Thus such an approach could still lead to an underestimation of the presence of AFP genes in a bacterial community.
The evolution of the brackish and saline lacustrine systems drives the evolution of the microbial communities within the lakes. Changes in salinity, stratification of the water column, natural eutrophication and the geological context of the lake all influence the species composition and trophic structure of the microbial community. The saline lakes of the Vestfold Hills are marine-derived. Through time the original marine bacterial communities have undergone significant changes. The invasion and establishment of aerobiological propagules (Marshall, 1996) and periodic marine incursions (Burton, 1981
) have probably acted to a small degree to change the microbial community composition during its evolution; however, changes in the salinity of the lake and accompanying changes in temperature regimen and stratification patterns probably imposed the greatest impact.
During the current study it was noted that the saline lakes usually had higher inorganic nutrient and DOC concentrations than the freshwater lakes (Table 2). The high concentrations of DOC in the extremely hypersaline lakes is thought to be due to a slow turnover of the DOC pool, in the absence of significant microbial decomposition resulting from reduced bacterial abundance (Hand, 1980
). As one moves along the salinity continuum of the Vestfold Hills lakes the number of eukaryotic species (algae, protozoa) in the community declines (Perriss & Laybourn-Parry, 1997
). Consequently, the grazing pressure on bacteria also diminishes.
AFP-active bacteria were demonstrated in isolates from Ace Lake, Pendant Lake, Triple Lake and Oval Lake, with Ace Lake and Pendant Lake containing the greatest number. The lakes share a similar salinity (Table 2), but Ace Lake is meromictic. In contrast Oval Lake (34
), which is also meromictic, provided only one AFP strain. Triple Lake is a hypersaline system (170
) and 5 of the total 19 AFP-active bacteria were isolated from this lake. Thus bacterial AFPs appear to have evolved under or been selected by quite different lake conditions. It has been suggested that AFPs might arise within drastically different ecosystems (Davies & Sykes, 1997
; Barrett, 2001
). The bacteria in the lakes of the Vestfold Hills are all exposed to some form of cold stress throughout the year, and in the saline lakes they suffer severe stress in winter. While temperatures are low in the freshwater lakes and the range is small, usually only a few degrees over the year, but the temperature range in the saline lakes can be very significant (Table 2
). For example, in Deep Lake the annual temperature can range from 11 °C in summer to -18 °C in winter (Kerry et al., 1977
). No AFPs were found in the freshwater isolates. Bacteria in the saline lakes face much greater cold and freezing stress and our data indicate that AFPs appear to be limited to isolates from these ecosystems. Fish AFPs have so far only been discovered in marine fish (Fletcher et al., 2001
), presumably because they are exposed to temperatures below the freezing point of their cellular fluid due to the higher solute concentration of their environment. Freshwater fish will never experience such temperatures and hence never require a thermal hysteresis active protein. This could explain why no freshwater AFP-active bacteria were isolated.
Due to the extremely low temperatures experienced by these bacteria it is not considered likely that the AFP would act as a freezing point depressant. We therefore suggest that RI activity plays a more crucial role in protecting the bacteria from recrystallization damage. This is also observed in plant AFPs where low thermal hysteresis activity was demonstrated, suggesting RI activity is important in protecting the plants during freeze/thaw cycles (Meyer et al., 1999; Pudney et al., 2003
). This is also shown by Raymond & Fritsen (2001)
in Antarctic cyanobacterial mats, which have negligable thermal hysteresis activity but demonstrate RI activity. An absence of activity in assayed conditioned growth medium suggests that the AFPs from at least two of the bacterial strains are expressed either intracellularly or in the periplasmic space. This is contrary to previous reports of most bacterial AFPs, which suggest extracellular secretion (Sun et al., 1995
; Kawahara et al., 2001
; Yamashita et al., 2002
).
Relatedness between AFP-active isolates was determined using ARDRA with two restriction enzymes (Fig. 1). Whilst the Mar. protea and Pseudoalteromonas groupings were maintained with both endonucleases, the relatedness of other isolates varied between the two restriction digestions, demonstrating the need to base such studies on more than one set of restriction patterns. Such variations have been noted previously by Jawad et al. (1998)
who showed that variation in clustering occurred when five enzymes were used to analyse 32 strains of Acinetobacter spp. It has been recommended that additional phenotypic tests be used to reconcile such variations (Vaneechoutte et al., 1995
).
The 19 bacterial isolates that demonstrated AFP activity were found predominantly within the -Proteobacteria, with one isolate from the
-Proteobacteria. All the closest relatives of these isolates have been shown to be common in marine and/or Antarctic ecosystems. They have also all been shown to have demonstrated some level of psychrophily, with the exception of Idiomarina loihiensis, which to date has only been isolated from the hydrothermal fluids of a submarine volcano in Hawaii (Ivanova et al., 2000
). The extreme differences in the respective environments of isolate 53 and I. loihiensis suggest that they are separate species, as to date no single species is known to survive across the entire temperature range found in the biosphere (Fields, 2001
). It is possible that many of the non-culturable bacteria in the lakes also have AFP proteins. As molecular biology technology progresses with respect to the isolation of environmental DNA and ability to screen it for genes of interest, we may be able to develop a much more accurate picture of AFP distribution among polar saline lake bacterial communities.
To date, all characterized AFP-active bacteria have been shown to belong to the - and
-Proteobacteria (Duman & Olsen, 1993
; Sun et al., 1995
; Xu et al., 1998
; Mills, 1999
; Kawahara et al., 2001
; Yamashita et al., 2002
, ibid.), except Rhodococcus erythropolis which belongs to the Actinobacteria (high G+C Gram-positive class of the phylum Firmicutes) (Duman & Olsen, 1993
) and the cyanobacteria Nostoc and Phormidium from Antarctic cyanobacterial mats (Raymond & Fritsen, 2001
). Thus whilst AFP activity is predominantly found among the Proteobacteria, which is the dominant group in marine/marine-derived ecosystems, it is also found outside this group, and hence is potentially widespread amongst Domain Bacteria taxonomic groups. Based on the data presented here, it is, at present, difficult to unravel the evolutionary forces that have lead to the acquisition of functional AFP activity in microbes of the Vestfold Hills lakes. Characterization and purification of the AFPs from these strains is currently being undertaken.
![]() |
ACKNOWLEDGEMENTS |
---|
![]() |
REFERENCES |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
Altschul, S. F., Gish, W., Miller, W., Myers, E. W. & Lipman, D. J. (1990). Basic local alignment search tool. J Mol Microbiol 215, 403410.
Ausubel, F. M., Brent, R., Kingston, R. E., Moore, D. D., Seidman, J. G., Smith, J. A. & Struhl, K. (editors) (1999). Short Protocols in Molecular Biology. New York: Wiley.
Barrett, J. (2001). Thermal hysteresis proteins. Int J Biochem Cell Biol 33, 105117.[CrossRef][Medline]
Bridson, E. Y. (1998). The Oxoid Manual, 8th edn. Basingstoke: Oxoid.
Burton, H. R. (1981). Chemistry, physics and evolution of Antarctic saline lakes. Hydrobiologia 82, 339362.
Davies, P. L. & Sykes, B. D. (1997). Antifreeze proteins. Curr Opin Struct Biol 7, 828834.[CrossRef][Medline]
Duman, J. G. (1994). Purification and characterization of a thermal hysteresis protein from a plant, the bittersweet nightshade Solanum dulcamara. Biochim Biophys Acta 1206, 129135.[Medline]
Duman, J. G. (2001). Antifreeze and ice nucleator proteins in terrestrial arthropods. Annu Rev Physiol 63, 327357.[CrossRef][Medline]
Duman, J. G. & Olsen, T. M. (1993). Thermal hysteresis protein activity in bacteria, fungi, and phylogenetically diverse plants. Cryobiology 30, 322328.[CrossRef]
Fields, P. A. (2001). Review: protein function at thermal extremes: balancing stability and flexibility. Comp Biochem Physiol A 129, 417431.
Fletcher, G. L., Hew, C. L. & Davies, P. L. (2001). Antifreeze proteins of teleost fishes. Annu Rev Physiol 63, 359390.[CrossRef][Medline]
George, A. L., Murray, A. W. & Montiel, P. O. (2001). Tolerance of Antarctic cyanobacterial mats to enhanced UV radiation. FEMS Microbiol Ecol 37, 91101.[CrossRef]
Griffith, M., Ala, P., Yang, D. S. C., Hon, W. & Moffatt, B. A. (1992). Antifreeze protein produced endogenously in winter rye leaves. Plant Physiol 100, 593596.
Hand, R. M. (1980). Bacterial populations of two saline Antarctic lakes. In Biogeochemistry of Ancient and Modern Environments. Proceedings of the Fourth International Symposium on Environmental Biogeochemistry (ISEB). Edited by P. A. Trudinger & M. R. Walter. Canberra: Australian Academy of Science.
Henshaw, T. & Laybourn-Parry, J. (2002). The annual patterns of photosynthesis in two large, freshwater, ultra-oligotrophic Antarctic lakes. Polar Biol 25, 744752.
Herbert, R. A. (1986). The ecology and physiology of psychrophilic microorganisms. In Microbes in Extreme Environments. Special Publication of the Society for General Microbiology. Edited by R. A. Herbert & G. A. Codd. London: Academic Press.
Ivanova, E. P., Romanenko, L. A., Chun, J. & 7 other authors (2000). Idiomarina gen. nov., comprising novel indigenous deep-sea bacteria from the Pacific Ocean, including descriptions of two species, Idiomarina abyssalis sp. nov. and Idiomarina zobellii sp. nov. Int J Syst Evol Microbiol 50, 901907.[Abstract]
Jawad, A., Snelling, A. M., Heritage, J. & Hawkey, P. M. (1998). Comparison of ARDRA and recA-RFLP analysis for genomic species identification of Acinetobacter spp. FEMS Microbiol Lett 165, 357362.[CrossRef][Medline]
Jia, Z. & Davies, P. L. (2002). Antifreeze proteins: an unusual receptorligand interaction. Trends Biochem Sci 27, 101106.[CrossRef][Medline]
Johnston, K., Somon, F. J. & Ellar, D. J. (1982). Teichoic acid and lipid metabolism during sporulation of Bacillus megaterium KM. Biochem J 202, 459467.[Medline]
Kawahara, H., Li, J., Griffith, M. & Glick, B. R. (2001). Relationship between antifreeze protein and freezing resistance in Pseudomonas putida GR12-2. Curr Microbiol 43, 365370.[CrossRef][Medline]
Kerry, K. R., Grace, D. R., Williams, R. & Burton, H. R. (1977). Studies on some saline lakes of the Vestfold Hills, Antarctica. In Adaptations within Antarctic Ecosystems. Proceedings of the Third SCAR Symposium on Antarctic Biology. Edited by G. A. Llano. Washington, DC: Smithsonian Institution.
Knight, C. A., Hallett, J. & Devries, A. L. (1988). Solute effects on ice recrystallisation: an assessment technique. Cryobiology 25, 5560.[Medline]
Laybourn-Parry, J., Quayle, W. & Henshaw, T. (2002). The functional dynamics and evolution of Antarctic saline lakes in relation to salinity and trophy. Polar Biol 25, 542552.
Mackereth, F. J. H., Heron, J. & Falling, J. I. (1988). Water Analysis. Freshwater Biological Association Scientific Publication 36. Ambleside: Freshwater Biological Association.
Marshall, W. A. (1996). Biological particles over Antarctica. Nature 383, 680.[CrossRef]
Meyer, K., Keil, M. & Naldrett, M. J. (1999). A leucine-rich protein of carrot that exhibits antifreeze activity. FEBS Lett 447, 171178.[CrossRef][Medline]
Michener, C. D. & Sokal, R. R. (1957). A quantitative approach to a problem in classification. Evolution 11, 130162.
Mills, S. V. (1999). Novel biochemical compounds from Antarctic microorganisms. PhD thesis, Nottingham University.
Perriss, S. J. & Laybourn-Parry, J. (1997). Microbial communities in the saline lakes of the Vestfold Hills (Eastern Antarctica). Polar Biology 18, 135144.[CrossRef]
Pudney, P. D., Buckley, S. L., Sidebottom, C. M. & 7 other authors (2003). The physico-chemical characterization of a boiling stable antifreeze protein from a perennial grass (Lolium perenne). Arch Biochem Biophys 410, 238245.[CrossRef][Medline]
Ray, M. K., Sitaramamma, T., Ghandhi, S. & Shivaji, S. (1994). Occurrence and expression of cspA, a cold shock gene, in Antarctic psychrophilic bacteria. FEMS Microbiol Lett 116, 5560.[CrossRef][Medline]
Raymond, J. A. & DeVries, A. L. (1977). Adsorption inhibition as a mechanism of freezing resistance in polar fishes. Proc Natl Acad Sci U S A 74, 25892593.[Abstract]
Raymond, J. A. & Fritsen, C. H. (2001). Semipurification and ice recrystallisation inhibition activity of ice-active substances associated with Antarctic photosynthetic organisms. Cryobiology 43, 6370.[CrossRef][Medline]
Rodina, A. G. (1972). Methods in Aquatic Microbiology. Baltimore: University Park Press.
Rotert, K. R., Toste, A. P. & Steiert, J. G. (1993). Membrane fatty acid analysis of Antarctic bacteria. FEMS Microbiol Lett 114, 253258.[CrossRef][Medline]
Russell, N. J. & Hamamoto, T. (1998). Psychrophiles. In Extremophiles: Microbial Life in Extreme Environments, pp. 2545. Edited by K. Horikoshi & W. D. Grant. Chichester: Wiley-Liss.
Sambrook, J., Fritsch, E. F. & Maniatis, T. (1989). Molecular Cloning: a Laboratory Manual, 2nd edn. Cold Spring Harbor, NY: Cold Spring Harbor Laboratory.
Sun, X., Griffith, M., Pasternak, J. J. & Glick, B. R. (1995). Low temperature growth, freezing survival and production of antifreeze protein by the plant growth promoting rhizobacterium Pseudomonas putida GR12-2. Can J Microbiol 41, 776784.[Medline]
Vaneechoutte, M., Dijkshoorn, L., Tjernberg, I., Elaichouni, A., De Vos, P., Claeys, G. & Verschraegen, G. (1995). Identification of Acinetobacter genomic species by amplified ribosomal DNA restriction analysis. J Clin Microbiol 33, 1115.[Abstract]
Whyte, L. G. & Innis, W. E. (1992). Cold shock proteins and cold acclimation proteins in a psychrotrophic bacterium. Can J Microbiol 38, 12811285.
Worrall, D., Elias, L., Ashford, D., Smallwood, M., Sidebottom, C., Lillford, P., Telford, J., Holt, C. & Bowles, D. (1998). A carrot leucine-rich-repeat protein that inhibits ice recrystallisation. Science 282, 115117.
Xu, H., Griffith, M., Patten, C. L. & Glick, B. R. (1998). Isolation and characterization of an antifreeze protein with ice nucleation activity from the plant growth promoting rhizobacterium Pseudomonas putida GR12-2. Can J Microbiol 44, 6473.[CrossRef]
Yamashita, Y., Nakamura, N., Omiya, K., Nishikawa, J., Kawahara, H. & Obata, H. (2002). Identification of an antifreeze lipoprotein from Moraxella sp. of Antarctic origin. Biosci Biotechnol Biochem 66, 239247.[CrossRef][Medline]
Received 30 June 2003;
revised 29 September 2003;
accepted 30 September 2003.