OUP user menu

Cultivation of Walsby's square haloarchaeon

David G. Burns , Helen M. Camakaris , Peter H. Janssen , Mike L. Dyall-Smith
DOI: http://dx.doi.org/10.1111/j.1574-6968.2004.tb09790.x 469-473 First published online: 1 September 2004


The square haloarchaea of Walsby (SHOW group) dominate hypersaline microbial communities but have not been cultured since their discovery 25 years ago. We show that natural water dilution cultures can be used to isolate members of this group and, once in pure culture, they can be grown in standard halobacterial media. Cells display a square morphology and contain gas vesicles and poly-β-hydroxybutyrate (PHB) granules. The 16S rRNA gene sequence was >99% identical to other SHOW group sequences. They prefer high salinities (23–30%), and can grow with a doubling time of 1–2 days in rich media. The ability to culture SHOW group organisms makes it possible to study, in a comprehensive way, the microbial ecology of salt lakes.

  • Archaea
  • Halobacteria
  • Haloarchaea
  • Isolation
  • Salt lake
  • SHOW group
  • Cultivation

1 Introduction

Salt lakes, with near saturating salinities, are common in countries around the world, and are extreme environments that maintain dense populations of halophilic microbes, mainly Archaea belonging to the family Halobacteriaceae [19]. In 1980, Walsby [3] described prokaryotic cells having the unusual shape of thin squares or tiles, that were present in high concentration (7 × 107 cells/ml) in a hypersaline pond, or sabkha, near the Red Sea. Despite initial scepticism by others [10], subsequent work confirmed that these were indeed free-living cells, were present in hypersaline lakes around the world, and their fine structure (as shown by electron microscopy) was that of a typical prokaryote, with a cell wall structure similar to members of the extremely halophilic Archaea [1,2,48,11,12]. The edge dimensions of individual cells varied between 2 and 4 μm but they were extremely thin, around 0.15 μm, and contained characteristic gas vacuoles and storage granules of PHB that distended the cell, giving cell surfaces a lumpy appearance. Attempts in various laboratories over the past 25 years to isolate pure cultures of these organisms have been unsuccessful [1,3,5,13] but some of their characteristics have been examined using cells from natural samples, including their rRNA gene sequences, which indicate they represent a separate taxon [5]; their membrane lipids, which are of the typical archaeal type [14]; their uptake of radiolabelled growth substrates [13]; and their dominant status in the microbial communities of hypersaline lakes and saltern crystallizer ponds [2].

Other haloarchaea have since been found to have triangular and even square shaped cells, such as Haloarcula quadrata[15], but they belong to genera that are distinct from the SHOW group, can be readily cultured in the laboratory [9] and are not usually dominant members of natural haloarchaeal communities [2]. They also lack gas vesicles and are more variable in shape. In a previous study, we were able to grow members of all the major haloarchaeal groups present in a saltern crystallizer pond on solid media, except the SHOW group [1]. Since there are bacteria that do not grow as colonies on agar plates [16], we investigated the use of liquid dilution cultures, including media based on natural water, to isolate SHOW group organisms. The latter technique was recently successful in isolating the dominant marine bacterium, SAR11 [17].

2 Materials and methods

2.1 Water samples and light microscopy

Crystallizer pond water samples were collected aseptically from the Cheetham Salt Works (Geelong, Victoria; 38°09.841′ S, 144°25.274′ E) over the period 2002–2004. Cells were examined by phase contrast and fluorescence (acridine-orange stain) microscopy using a Leitz Diaplan microscope (Leica Microsystems AG, Wetzlar, Germany).

2.2 Extinction cultures

(a) MGM cultures. Crystallizer water was diluted in a ten-fold series using MGM as the diluent. Between the 10−6 and 10−10 dilutions the series was expanded to 20 tubes per dilution, and these tubes were incubated in the dark at 37 °C for 3 weeks. Most probable number (MPN) calculations were performed using the software program MPN3 [18]. (b) Natural water cultures. Ten-fold serial dilutions were made in crystallizer pond water that had been previously filtered through a 0.2 μm membrane and then autoclaved (15 min, 121 °C, 101 kPa). The dilution series was expanded to 25 tubes per dilution between 10−6 and 10−10 dilution steps and additional dilutions of 2 × 10−9 and 2 × 10−10 were also made. Tubes were incubated, stationary, in the dark at 37 °C for 4 weeks, then screened for growth using the multiplex PCR protocol DB239 (see below). Four formulations using natural saltern water supplemented with selected nutrients were also used. They contained sterile saltern water with: 5 μM amino acids and 5 μM pyruvate (Medium A); 5 μM amino acids, 5 μM pyruvate and 5 μM acetate (Medium B); 50 μM amino acids and 0.5% (wt/vol) pyruvate (Medium C); and 50 μM amino acids, 50 μM pyruvate and 50 μM acetate (Medium D).

2.3 PCR detection of haloarchaeal growth

DNA extraction of cultures has been described previously [1]. Each PCR contained 1.75 mM MgCl2, PCR buffer (Qiagen), 200 μM dNTPs, 37.5 pmol F1 forward primer, 37.5 pmol SHOWprb forward primer, 60 pmol reverse primer (1492Ra) and 2 units HotStarTaq DNA polymerase (Qiagen, USA). The volume was made up to 49 μl with distilled water, and 1 μl of DNA preparation (cell lysate or extract of DNA) was added and cycled at: 15 min/95 °C; 30 × (1 min/95 °C; 30 s/46 °C; 90 s/72 °C); 10 min/72 °C. A Haloferax volcanii lysate was used as a positive control in this reaction for the large (∼1500 bp) fragment, and the SHOW group clone, CSWFA61 (described in [1]), used as a positive control for both the larger and SHOW-specific (∼800 bp) bands. Distilled water was used as the negative control. Amplified products were resolved by electrophoresis on 1.2% agarose gels alongside Precision Molecular Mass Ruler size standards (Bio-Rad Laboratories Pty. Ltd., Australia).

2.4 DNA sequencing and RFLP

PCR products were sequenced using the ABI PRISM Big Dye system (PE Applied Biosystems) as described previously [1]. PCR products were generated as described above except using 50 pmol each of primers F1 and 519R (GWATTACCGCGGCKGCTG). The latter primer was modified from Lane et al. [19]. The product was cut out of a gel and purified using an Ultraclean GelSpin DNA purification kit (MoBio Laboratories, Inc., USA). DNA was cut with MlyI (New England Biolabs, USA) and electrophoresed on 3% agarose gels.

2.5 Primer design and phylogenetic tree reconstructions

The SHOW group specific primer (SHOWprb: ACGGCACAACAGAGACGC) design and phylogenetic tree reconstructions were performed using the Apple Mac OS X version of the ARB phylogeny package (http://www.arb-home.de/) and the16S rRNA sequence database based on that of Hugenholtz (http://rdp.cme.msu.edu/html/alignments.html), updated with haloarchaeal sequences from NCBI (http://www.ncbi.nlm.nih.gov/). 16S rRNA sequences were screened for chimeras using the Bellepheron server (http://foo.maths.uq.edu.au/~huber/bellerophon.pl). For phylogenetic tree reconstructions, a backbone tree was constructed using only sequences of >1300 nt. The alignment was exported to PAUP (Phylogenetic Analysis Using Parsimony) [26] and tree reconstruction performed using the maximum likelihood algorithm. Bootstrap values were derived from 1000 replicate tree reconstructions using PAUP (distance matrix algorithm). The tree was imported into ARB and partial sequences of cloned genes and isolates (all >400 nt) were then added to the tree using the parsimony add option within ARB.

3 Results and discussion

3.1 Extinction cultures using a conventional medium (MGM) or natural water

Crystallizer pond water samples were collected aseptically from the Cheetham Salt Works (38°09.841′ S, 144°25.274′ E) over the period 2002–2004. The chemical composition and microbial diversity of this crystallizer pond have been previously described [1]. The cell concentration of a freshly taken water sample, containing numerous SHOW cells, was estimated by microscopy, and the water was serially diluted in tubes containing 5–10 ml volumes of 23% salt water (SW) MGM (a complex medium containing 0.1% yeast extract and 0.5% peptone) [1] to dilutions such that tubes would be likely to be seeded with individual cells (extinction cultures). After 3 weeks incubation, growth was clearly visible in all tubes down to 10−7 dilution, beyond which the fraction of turbid tubes diminished rapidly. MPN calculations from these results estimated the viable count of the saltern water sample at 1.3 × 106 cells/ml. Light microscopy of 15 turbid cultures around the extinction dilution revealed they all contained cells similar to Halorubrum species, and 16S rRNA sequencing confirmed they all closely matched (>99%) Halorubrum sequences. No cells were detected by light microscopy in non-turbid cultures.

We hypothesized that SHOW group cells may have been inhibited by the peptone or yeast extract in the MGM [17], or perhaps its lower salt concentration (23%), and proceeded to test the use of natural saltern water as a medium for extinction cultures. Saltern crystallizer water (salinity ∼ 33%) was filtered, heat sterilized, dispensed into sterile dilution tubes and a fresh crystallizer pond sample diluted as in the previous extinction culture experiment. Since the natural substrate concentrations present in the water were low, screening for growth was performed using PCR instead of visible turbidity. Multiplex PCR was used to detect microbial growth using universal rRNA specific primers F1 and 1492Ra [1] (expected product ∼ 1450 nt) and the presence of SHOW group cells using a SHOW-specific primer, SHOWprb (ACGGCACAACAGAGACGC; expected product of ∼800 bp). After incubation for 4 weeks none of the tubes showed visible turbidity, but the PCR screen detected growth in tubes down to 10−9 dilution. One culture from the 2 × 10−9 dilution series gave a positive SHOW-specific band. At this dilution, only 15% of tubes contained detectable SHOW organisms as determined by PCR. The DNA sequence of the SHOW-specific PCR product was almost identical (99.8% similarity) to sequences of the SHOW group reported by others [6] and those detected in our previous study [1]. Light microscopy of this culture showed square or rectangular, gas-vacuolated cells (about 1.7–3 μm on a side), typical of the SHOW group, but these were present at only about 15% of the cell population, with the balance being smaller (1.1–1.4 μm diameter), flat, rounded cells. The sequence reaction of the universal rRNA PCR product from this culture returned a chromatogram that was predominantly a Halorubrum Group 2 sequence, with secondary minor peaks identical to that of the SHOW group organism. Subsequent passage of this culture resulted in the loss of the SHOW group organism (D. Burns, BSc (hons) thesis, 2002). The results indicated that the culture was mixed, but proved that members of the SHOW group could be grown in laboratory culture.

We attempted to improve the extinction culture method by supplementing the natural saltern water with specific nutrients (amino acids mixture (1), pyruvate and acetate, Media A-D). The addition of acetate and amino acids was based on nutrient uptake studies of Rosselló-Mora and colleagues [13], and the addition of pyruvate on our experience in maximizing growth of a slowly growing, novel group of haloarchaea (Antarctic Deep Lake group) [1]. After three weeks incubation, multiplex PCR screening of dilution cultures of the four media revealed three with SHOW specific signals. All three were isolated from Medium C (50 μM amino acids and 0.5% (wt/vol) pyruvate). By light microscopy, all three cultures showed square cells, typical of SHOW organisms. The best growing isolate, C23, was further passaged and has now been diluted by a factor of 2.46 × 1014 from the original saltern water. The cell density reaches about 107 cells/ml, at which point cultures are slightly turbid by eye. Isolate C23 can also grow in defined media (see below), and even in 23% (SW) MGM, a medium in which we were previously unable to isolate it. It does not grow on solid media. Purity and identity were confirmed by microscopy, restriction fragment length polymorphism (RFLP) of the 16S rRNA gene (using MlyI digestion of PCR products generated by consensus rRNA primers), and sequencing the 16S rRNA gene PCR product from the culture. The RFLP pattern was clean and matched the pattern expected of the SHOW group, and sequence chromatograms showed no traces of contamination. The 16S rRNA sequence of isolate C23 (GenBank Accession AY676200) closely matched SHOW clone sequences (>99% similarity).

3.2 Characteristics of SHOW isolate C23

Light microscopy of isolate C23 cultures (Fig. 1(a) and (b)) showed very thin, square (average of 2.6 μm) or rectangular (average 2.3 × 3 μm) cells with well-defined, sharp corners, prominent, phase-bright, gas vesicles and dark granules. Apart from occasional triangular cells, the culture appeared homogeneous. Division borders between cells were not easy to see, making estimates of cell sizes difficult. The gas vesicles were easily collapsed by pressure. The dark granules were shown by Nile Blue A staining [20] to be PHB (Fig. 1(c) and (d)). By negative stain electron-microscopy (Fig. 1(e)) the gas vesicles are seen as thin cylinders of variable length with cone-shaped ends (∼0.1 μm × 0.2–0.6 μm). They were often found near the periphery of the cell. Three, electron-dense PHB granules (∼0.3 μm diameter) can also be seen at the right hand corner of the cell in Fig. 1(e). All of the characteristics shown by isolate C23 are very similar to those of naturally occurring SHOW organisms, as reported originally by Walsby [3] and in later studies [1,2,48,11,12].

Figure 1

Isolate C23 cells viewed by light and electron microscopy. (a) cells examined by phase contrast microscopy, (b) the same field viewed by epifluorescence microscopy after acridine orange staining (acridine orange (10 μg/ml) with para-phenylene diamine (0.5% wt/vol) as an anti-fading agent), (c) cells fixed and stained with Nile Blue A (a poly-β-hydroxybutyrate-specific, fluorescent stain) [20], and viewed by phase contrast microscopy, (d) the same field as in (c) viewed by epifluorescence, and the image superimposed over that of (c) to show the cell borders relative to the red-fluorescing granules. Bars in panels (a)-(d) represent 5 μm, (e) uranyl acetate (2%) stained preparation examined by transmission electron-microscopy. Bar represents 1 μm.

To determine the growth characteristics, isolate C23 was inoculated, at a 1:20 dilution, into 23% (SW) MGM, and incubated at 37 °C with shaking (180 rpm). When the culture had reached late exponential phase, it was used to inoculate fresh medium (again at 1:20 dilution), and growth followed over time (Fig. 2). Cells were counted using epifluorescence microscopy (acridine orange stain) and a Thoma counting chamber. For the first 10 days the culture grew with a doubling time of 5.5 days, then changed to a much faster doubling time of around 1.5 days. In a subsequent passage of late exponential phase cells into fresh medium the doubling time remained at the faster rate (data not shown). The final cell density of the culture shown in Fig. 2 was 107 cells/ml. When MGM was supplemented with 0.5% pyruvate, final cell densities reached 108 cells/ml (data not shown). Cell pellets were strongly red in color, and in the case of pyruvate supplemented cultures, they were visibly red at stationary phase.

Figure 2

Growth of isolate C23 in 23% (SW) MGM. A culture of isolate C23, growing in 23% (SW) MGM was inoculated 1:20 into fresh medium and growth followed using a Thoma counting chamber. Cells were stained with acridine orange and viewed by epifluorescence. Vertical bars represent one standard deviation.

The full 16S rRNA gene sequence was determined, aligned with sequences of all members of the Halobacteriaceae as well as related clone sequences in GenBank, including sequences derived from our previous study of this saltern [1], and used to reconstruct a phylogenetic tree. A portion of the tree is shown in Fig. 3, with only representative sequences longer than 1300 nt included. The C23 isolate branched within the SHOW clade, a position supported by high bootstrap confidence values. Nucleotide similarity values between the C23 sequence and that of other SHOW group clone sequences such as X84084 [2,21] or CSWFA061 [1] were >99%. The next closest sequence to all three SHOW group sequences, about 91% similarity, was that of Halogeometricum borinquense ATCC 7000274T.

Figure 3

Phylogenetic tree reconstruction using 16S rRNA gene sequences of SHOW group clones, isolate C23, and closely related sequences of members of the family Halobacteriaceae. Bootstrap values are shown at the major branch points. Labels at the right show the name of the isolate or clone, the GenBank Accession code, and the number of nucleotides. Outgroup sequences not shown were those of Hrr. sodomense (Accession D13379) and Hbl. gomorrense (Accession L37444). The bar represents 0.01 substitutions per site.

The ability to culture SHOW group halobacteria is an important step towards understanding the microbial ecology of extremely halophilic environments. Salt lakes are biologically highly active environments [22] and are estimated to cover a similar surface area as fresh water systems [23]. Isolate C23 can be used as a model organism to study the structure, growth, metabolism, and genetics of these peculiar SHOW group organisms, as well as their interactions with other microbes, including viruses [24]. The isolate can be used as a host to sample the virus space of salt lakes more extensively than previously possible [25].


We thank W. Rickard and N. Taylor (Cheetham Salt Ltd., Geelong, Australia; http://www.cheethamsalt.com.au) for allowing us to sample crystallizer ponds. We thank A. Friedhuber for assistance with electron microscopy. DGB was supported by an Australian Postgraduate Award. This work was supported by grants from the Australian Research Council.


  • Editor: A. Oren


  1. [1].
  2. [2].
  3. [3].
  4. [4].
  5. [5].
  6. [6].
  7. [7].
  8. [8].
  9. [9].
  10. [10].
  11. [11].
  12. [12].
  13. [13].
  14. [14].
  15. [15].
  16. [16].
  17. [17].
  18. [18].
  19. [19].
  20. [20].
  21. [21].
  22. [22].
  23. [23].
  24. [24].
  25. [25].
  26. [26].
View Abstract