Characterization of the minimal replicon of pHM300 and independent copy number control of major and minor chromosomes of Haloferax mediterranei


Correspondence: Hua Xiang, State Key Laboratory of Microbial Resources, Institute of Microbiology, Chinese Academy of Sciences, No. 1 Beichen West Road, Chaoyang District, Beijing 100101, China. Tel./fax: (86) 10 6480 7472; e-mail:


The presence of minichromosomes is very common in haloarchaea, but little is known about the coordination of replication between the major and minor chromosomes. In this study, we analyzed the replication of pHM300, a 321,908-bp minichromosome, which encodes versatile metabolism pathways in Haloferax mediterranei. The replication origin of pHM300 was predicted in the 699-bp intergenic region between the cdc6K and tbp4 gene, and the minimal replicon, consisting of an AT-rich region flanked by putative origin recognition boxes (ORBs) and the adjacent cdc6K gene, was determined by assaying for its ability to replicate autonomously in Haloarcula hispanica. Southern blot analysis indicated that the ratio of pHM300 to chromosome increased from the early exponential to middle stationary phase. The copy numbers of these minor and major chromosomes were then evaluated by real-time PCR and showed that both decreased in stationary phase. However, the decrease in the copy number of the major chromosome was a little earlier and much greater than that of pHM300, revealing that the copy number control of the minichromosome pHM300 is independent from that of the major chromosome in H. mediterranei.


The extremely halophilic archaea (haloarchaea) are members of the family Halobacteriaceae in the domain Archaea, and they thrive in high salt conditions, such as salt lakes, salterns, and salt ponds. In contrast to many other archaea, haloarchaea are easy to culture, ease for genetic manipulation, and well-developed genetic tools are available for many haloarchaeal species (Cline et al., 1989; Peck et al., 2000; Bitan-Banin et al., 2003; Allers et al., 2004; Liu et al., 2011a). Thus, haloarchaea are good systems for genetic research.

Interestingly, the presence of megaplasmids and/or minichromosomes in haloarchaea is very common (DasSarma et al., 2009; Capes et al., 2011). Currently, of the 17 haloarchaeal genomes completely sequenced, with the exception of Halorhabdus utahensis (Anderson et al., 2009) and Haloquadratum walsby (Bolhuis et al., 2006), the other 15 sequenced haloarchaeal genomes all harbor megaplasmids larger than 100 kb, and some megaplasmids have been considered as minichromosomes due to the presence of important or essential genes (Ng et al., 2000; Baliga et al., 2004; Falb et al., 2005; Pfeiffer et al., 2008; Tindall et al., 2009; Hartman et al., 2010; Roh et al., 2010; Jiang et al., 2011; Liu et al., 2011b). While many haloarchaeal megaplasmids and minichromosomes have been sequenced, few functional studies have been performed. To date, pNRC100 in Halobacterium sp. NRC-1 is the best studied one: its minimal replication origin has been determined (Ng & Dassarma, 1993), and its gene content and evolution have been investigated (Ng et al., 1998). The copy numbers of the chromosome and three extrachromosomal replicons (pHS1 to pHS3) in Halobacterium salinarum were also quantitated, finding that the copy numbers of the main chromosome and pHS1 show growth phase–dependent regulation, while pHS2 and pHS3 have low copy number that is not growth-phase regulated (Breuert et al., 2006).

Haloferax mediterranei, an extremely halophilic archaeon, exhibits versatile metabolic and physiological features, such as halocin production (Meseguer & Rodriguez-Valera, 1985, 1986), polyhydroxyalkanoate (PHA) accumulation (Lillo & Rodriguez-Valera, 1990), nitrate reduction (Martinez-Espinosa et al., 2001; Lledó et al., 2004), gas vesicle formation (Jäger et al., 2002), and extracellular polysaccharide production (Antón et al., 1988). The genome of H. mediterranei has been recently sequenced by our group (Han et al., 2012). Like other haloarchaea, this genome has multiple replicons, consisting of a 2,948,884-bp main chromosome and three large extrachromosomal replicons (the 129,210-bp pHM100, the 321,908-bp pHM300, and the 504,705-bp pHM500). Interestingly, sequence analysis reveals that the halocin gene halH4 (Cheung et al., 1997), the nar operon involved in the nitrate respiratory process (Lledó et al., 2004), the phaEC genes encoding PHA synthase (Lu et al., 2008), and the other key genes for PHA biosynthesis, that is, the phaB2 gene encoding a β-ketoacyl-CoA reductase (Feng et al., 2010) and the phaP gene encoding the predominant structure protein on the PHA granules (Cai et al., 2012), are all located on pHM300. This finding implies that pHM300 is of great significance for H. mediterranei, prompting us to investigate the basic features of the pHM300 in this genome.

In this study, we have systematically analyzed the gene content of pHM300 and investigated its replication characteristics by determining the replication origin, the minimal replicon, and the copy number during cell growth. These studies provide a more detailed perspective on the replication characteristics of haloarchaeal minichromosomes and a specific insight into the copy number control of the main chromosome and minichromosomes.

Materials and methods

Strains, plasmids, and primers

The strains and plasmids used in this study are listed in Table S1, and all primers are listed in Table S2. The Haloarcula hispanica CGMCC 1.2049 was used as the transformation host to avoid homologous recombination. The haloarchaeal strains were grown at 37 °C in AS-168 medium as described previously (Han et al., 2007). The Escherichia coli strain JM109 was used for gene cloning experiments and grown in Luria–Bertani broth at 37 °C (Sambrook & Russell, 2001). The recombinant plasmids (Table S1) were constructed in E. coli and introduced into H. hispanica using the polyethylene glycol (PEG)–mediated transformation method (Cline et al., 1989). When needed, ampicillin or mevinolin was added to a final concentration of 100 μg mL−1 (for E. coli) or 5 μg mL−1 (for H. hispanica), respectively.

Determination of minimal replication of pHM300

To determine the minimal replicon of pHM300 that is essential for autonomous replication, different fragments were amplified by PCR with the primers listed in Table S2, digested with NcoI and HindIII, and inserted into a nonreplicating Mevr-marked plasmid, pBI101 (Zhou et al., 2007), to generate a series of derivative plasmids, pHM311 to pHM319 (Table S1, Fig. 4a). These plasmids were then introduced into H. hispanica for autonomous replication assays. To discriminate between plasmid integration and autonomous replication, total DNA extracted from the transformants was supplied for Southern blot analysis. The plasmid-specific probe of the bla gene was obtained by PCR amplification with primers AmpRF/AmpRR and labeled by digoxigenin.

Determination of the ratio of the copy number of pHM300 to that of chromosome during growth

To estimate the ratio of the copy number of pHM300 to that of chromosome during growth, Southern blot analysis was conducted. The total DNA extracted from H. mediterranei at different growth stages (from 15 to 108 h) was digested with the restriction enzyme SalI and separated by electrophoresis on 0.8% agarose gels. Probes specific for pHM300 near cdc6K and specific for the chromosome near cdc6A having the same lengths (800 bp) and G+C contents (59%) were chosen, so that the chromosome-borne signal (1254 bp) and the pHM300-borne signal (2167 bp) could be clearly discriminated in hybridization (Fig. 5a). The two probes were obtained by PCR amplification with the primer pairs p300F/p300R and chrPF/chrPR (Table S2), respectively. The probes were labeled with [α-32P]-dCTP and used for Southern blot analysis. The hybridized signal levels of the pHM300 and the chromosome were quantitated using a phosphoimager (Amersham Biosciences) with imagequant 5.2 software. The ratio of the copy number of pHM300 to that of chromosome was estimated by comparing the hybridization signals of the pHM300 and the chromosome.

Determination of the copy numbers of pHM300 and chromosome during growth

To directly determine the copy numbers of pHM300 and chromosome per cell throughout the growth curve, real-time PCR method was employed as described previously (Breuert et al., 2006). The two 800-bp PCR fragments used as probes for Southern blot analysis (Fig. 5a) were respectively used as the standards for quantitation of pHM300 and chromosome copy numbers. Primer pairs q-300F/q-300R and q-chrF/q-chrR were used to detect the pHM300 and chromosome, respectively (Table S2, Fig. 5a). Every experiment was repeated twice, each in triplicate.

Nucleotide sequence accession number

The nucleotide sequence of pHM300 is deposited in GenBank under the accession number CP001870.


The complete sequence of pHM300 was obtained from the genome sequencing of the host strain, H. mediterranei CGMCC 1.2087, which was recently published (Han et al., 2012). The 321,908-bp sequence of pHM300 has an average G+C content of 57.9%, which is a little lower than that of the chromosome (61.1%). The coding region comprises 86% of the replicon, and a total of 291 open reading frames (ORFs) were identified (Fig. 1).

Figure 1.

Circular representation of pHM300. From outside to inside, the first and second circles show predicted protein-coding regions on the plus and minus strands. The third circle shows the GC skew in which regions >0 are in gold and those smaller are in purple. The fourth circle shows tRNA genes, with blue on plus strand and red on minus strand. The locations of the important genes are indicated.

pHM300 endows H. mediterranei with many specific functions

To understand the function of pHM300, the predicted 291 ORFs were analyzed using the COG database (Fig. 2). Apart from the unknown function class, the largest class of functional genes (33%) present on the pHM300 is involved in metabolism (Fig. 2). Besides the known genes involved in specific physiology and metabolism including halocin production (HFX_5264) (Cheung et al., 1997), nitrate utilization (HFX_5101-5108) (Lledó et al., 2004), and PHA biosynthesis (HFX_5215, HFX_5219-5220) (Lu et al., 2008; Feng et al., 2010; Cai et al., 2012), there are still many genes that are significant for the versatile metabolic ability of H. mediterranei. These include the four genes (HFX_5036-5039) encoding chitinases, the cob operon (HFX_5051-5054, HFX_5056-5070) encoding cobalamine biosynthesis proteins, the bio operon (HFX_5076-5079) involved in coenzyme metabolism, and the gene clusters (HFX_5040-5045) encoding siderophore biosynthesis proteins (Fig. 1). Notably, pHM300 encodes several genes likely essential for cell survival, such as a lone copy of the gene (racX, HFX_5011) encoding aspartate racemase that participates in alanine and aspartate metabolism, one of two threonyl-tRNA synthetase genes (thrS, HFX_5166) as well as one of five genes encoding tRNA-Ser (HFX_5225) in the genome (Fig. 1). These results indicated that pHM300 endows H. mediterranei with many functions, enabling this organism to adapt to its environments.

Figure 2.

Functional distribution of the 291 genes according to the COG functional classification. Each color is defined by the class definition. The number and percentage of identified proteins associated with each category are shown.

Prediction of the replication origin of pHM300

As pHM300 is of great importance for the host strain, it is necessary to clarify the mechanisms of the maintenance of this replicon, especially the replication mechanisms. Firstly, the putative replication origin was predicted based on a nucleotide disparity analysis using the Z-CURVE algorithm (Zhang & Zhang, 2005). The AT disparity curve has a single peak corresponding to the location of the Cdc6/Orc1 gene (cdc6K) on pHM300, suggesting there is a single replication origin adjacent to cdc6K, which is consistent with most archaeal replication origins (Fig. 3a). Meanwhile, the 699-bp intergenic region between cdc6K and the adjacent gene, tbp4, shows similar features to those of characterized archaeal replication origins (Grabowski & Kelman, 2003), including several AT-rich segments and multiple ORB consensus sequences (Fig. 3b and c). It is noteworthy that this putative origin shares a lot in common with the replication origin of pHV3 (Norais et al., 2007), a 437,906-bp minichromosome in Haloferax volcanii. The sequences of their ORBs are quite similar (Fig. 3b), and the two Cdc6/Orc1 proteins near the replication origin also exhibit very high levels of sequence identity (96.8%), although the pHV3 and pHM300 are quite different in the gene content (Fig. S1).

Figure 3.

Analysis of the replication origin of pHM300. (a) Nucleotide disparity curves of pHM300. Position of the cdc6k gene is indicated. The curve is calculated using the software Z Curve Plotter ( (b) Origin recognition box (ORB) elements located at the origin. Above the line are ORBs in the origin of pHM300. Below the line are ORBs in the origin of pHV3. Conserved sequences are shaded. The arrows indicate the orientation of the sequence. (c) Nucleotide sequences of the intergenic region. Arrows along the sequence indicate the ORB elements. The AT-rich region is boxed. Primer positions are indicated with bent arrows. Dashed lines indicate the open reading frames of the tbp4 and cdc6K genes. The numbers on the right indicate the nucleotide positions in pHM300.

Determination of the minimal replicon

To test the replication activity of the putative origin and to determine the minimal replicon, fragments containing different parts of the putative origin region and the adjacent cdc6K gene were cloned into the plasmid pBI101, which could not replicate in haloarchaea. Then, the derived plasmids were assayed for their autonomous replication ability in H. hispanica. The resulting plasmids pHM311, pHM312, pHM313, pHM316, and pHM317 showed high transformation efficiencies (Fig. 4a). Southern blotting with a bla gene–specific probe indicated that these five plasmid bands were consistent with their replicative plasmid forms from E. coli (Fig. 4b), suggesting these plasmids could replicate effectively in H. hispanica. In contrast, no transformants were obtained for the plasmids pHM314, pHM315, pHM318, and pHM319 (Fig. 4a), indicating that the fragments contained in these four plasmids did not have autonomous replication ability.

Figure 4.

Analysis of the minimal replicon of pHM300. (a) Determination of the minimal replicon. The genetic loci and sequence features of the replication origin are shown at the top. The fragments inserted into pBI101 for autonomous replication assays are shown by the lines below the genetic map with the names of constructed plasmids indicated on the left. The names of the primers used to amplify the fragments are shown under the end of each fragment. The replication ability (Rep) of each derived plasmids in Haloarcula hispanica is indicated by ‘+’ for yes or ‘−’ for no. (b) Southern blot analysis of the derived plasmids with a bla gene probe. Lanes 1–5 contain DNA isolated from H. hispanica transformed with plasmids (lane 1, pHM311; lane 2, pHM312; lane 3, pHM313; lane 4, pHM316; lane 5, pHM317). Lanes 6–10 contain plasmid DNA isolated from E. Coli (lane 6, pHM311; lane 7, pHM312; lane 8, pHM313; lane 9, pHM316; lane 10, pHM317). Lane M contains a DNA ladder marker.

Particularly, there are two AT-rich fragments in the putative replication origin of pHM300 (Fig. 3c and 4a), which may be involved as the duplex unwinding elements (DUEs). However, deleting the first AT-rich fragment did not abolish the replication of the derived plasmids pHM316 and pHM317 (Fig. 4a), indicating that the second AT-rich fragment is more likely to serve as the DUE of this origin. Moreover, deleting the first ORB sequence made the derived plasmid pHM318 unable to replicate. Removing the WH domain located in the C terminal of Cdc6K resulted in the derived plasmid pHM314 losing its replication ability. Furthermore, deleting the cdc6K gene also destroyed the replication ability of the resulting plasmid pHM315 (Fig 4A).

These results show that the minimal replicon of pHM300 contains only the putative replication origin region and the adjacent gene, cdc6K. The putative ORB sequences flanking the putative AT-rich DUE in the origin region are necessary for the replication activity of this origin.

The copy number control of pHM300 is independent from that of the chromosome

To explore the relationship of replication between pHM300 and the major chromosome, the copy numbers of pHM300 and the chromosome were investigated throughout the growth curve. Interestingly, Southern blot analysis indicated that the ratio of the copy number of pHM300 to that of the chromosome increased from exponential phase and is up to the maximum as the culture entered into stationary phase (time points 1–9, Fig. 5b) and then decreased (time point 10, Fig. 5b). These are likely due to that the copy number of chromosome greatly decreased from the late exponential phase (lanes 3–10, Fig. 5b), whereas the copy number of pHM300 did not exhibit such obvious changes (Fig. 5b).

Figure 5.

Determination of the copy numbers of pHM300 and chromosome during growth. (a) Schematic diagrams of the probes used for Southern blot analysis. The probes are shown as black bars with ‘a’ indicating the pHM300-specific probe and ‘b’ the chromosome-specific probe. The size of hybridization fragments is indicated as follows: 2167 bp for pHM300 and 1254 bp for the chromosome. The primers used for the real-time PCR are indicated by rent arrows. (b) Evaluation of the copy number ratio of pHM300 to chromosome throughout the growth phase of Haloferax mediterranei. Growth curve (●) was determined by monitoring the OD600 nm. The ratio (■) of the Southern blotting signals of pHM300 to chromosome was determined. The numbers (1–10) above the growth curve indicate where DNA samples were taken for Southern blotting. The inset is autoradiograph of the Southern blotting of genomic DNA probed with 32p-labeled probes a and b indicated in (a). The 10 samples (lanes 1–10) correspond to the 10 time points on the growth curve. Experiments were performed in triplicate, and the representative results were shown. (c) Determination of the copy numbers of the chromosome and pHM300 using the real-time PCR method. Three independent cultures were used to determine the copy numbers, one of the growth curve (●), the ratio of copy numbers of pHM300 to chromosome (○), the average copy numbers per cell of chromosome (□) and pHM300 (■), and their standard deviations are shown.

Although the Southern blot analysis has clearly revealed the change of the ratio of pHM300 to chromosome throughout the cell growth curve, it does not directly determine the copy numbers of both chromosomes, as two different probes were used for pHM300 and chromosome, respectively. To further understand the dynamic changes of the ratio of pHM300 to the major chromosome, the copy numbers of pHM300 and chromosome were directly determined by real-time PCR. The profile of the ratio of pHM300 to chromosome determined by real-time PCR was consistent with that yielded by quantitative Southern blotting (Fig. 5c and b). Notably, the copy number of chromosome decreased greatly at stationary phase, whereas the copy number of pHM300 was also decreased, but much less than that of chromosome (Fig. 5c). Besides, the copy number of the major chromosome is much higher than pHM300 in exponential phase and likely decreased a little earlier (Fig. 5c and b). These together contribute the increase in the ratio of pHM300 to chromosome from the early exponential to middle stationary phase (Fig. 5b and c). In short, our results indicate that the copy number control of pHM300 is independent from that of the major chromosome in H. mediterranei.


Haloarchaea usually consist of a main chromosome and several megaplasmids. Some megaplasmids have been classified as minichromosomes because they possess important or essential genes (Baliga et al., 2004; Norais et al., 2007). Because pHM300 in H. mediterranei carries essential and important genes and uses a Cdc6/Orc1-dependent origin as the main chromosome, we proposed that pHM300 should be considered as a minichromosome according to the criteria established previously (Ng et al., 1998).

Comparative analysis of the H. mediterranei genome with that of H. volcanii reveals that the main chromosomes of these two strains are highly homologous, whereas their smaller replicon (minichromosomes, megaplasmids, and/or plasmids) are quite divergent (Han et al., 2012). Interestingly, the origin of pHM300 in H. mediterranei and pHV3 in H. volcanii exhibits high similarities, while their gene content is quite different, suggesting that these two minichromosomes, respectively, may have formed after the divergence of the H. mediterranei and H. volcanii. It is likely that they captured different genes and eventually formed two quite different replicons with a homologous origin. This might be a strategy for evolution of haloarchaea to adapt to highly dynamic environments. Similar results were reported in the minichromosomes of H. hispanica and Haloarcula marismortui (Wu et al., 2012).

The cdc6k gene was indispensable for the ARS activity of pHM300 minimal replicon in H. hispanica. This is different to what is reported in H. volcanii, in which the cdc6/orc1 gene is not required for the ARS activity of the plasmid containing the origin fragment of pHV3 in the native strain (Norais et al., 2007). This discrepancy may be due to that we performed the ARS analysis in a heterologous host, H. hispanica. It suggested that H. hispanica does not contain Cdc6/Orc1 proteins that can recognize the ORB elements on pHM300 and further suggested that the origin of pHM300 was specifically recognized by the Cdc6K encoded by this minichromosome. Thus, it can be speculated that there is a corresponding relationship between Cdc6/Orc1 proteins and the ORB elements, and each origin may be preferentially recognized by its adjacent Cdc6/Orc1 protein. This would be very flexible for controlling the replication of respective replicons in the haloarchaeal cells, as different Cdc6/Orc1 proteins could be used to regulate the replication initiation of different replicons. Similar mechanism was also observed in Sulfolobus solfataricus, whose genome contains three Cdc6/Orc1 proteins. The two identified origins in S. solfataricus are recognized by different subsets of the three Cdc6/Orc1 proteins (Robinson et al., 2004).

The copy number of chromosome was always higher than that of pHM300, but the hybridized signal levels of the chromosome were lower than that of pHM300 from late exponential phase in Southern blotting (Fig. 5b), this may be due to the different efficiencies of probe labeling and hybridization. The chromosomal copy number was downregulated at the stationary phase, this was also reported for H. salinarum and H. volcanii (Breuert et al., 2006). Compared to chromosome, the copy number of pHM300 was downregulated but at much lower degree, meaning that pHM300 maintains adequate copy number during the stationary phase. This may be helpful for the host strain to adapt the environment, as many genes in pHM300, for example, the halocin gene, are required for expression in stationary phase but not in exponential phase.

In conclusion, our observations of pHM300 in this report have provided genetic details of the replication of a haloarchaeal minichromosome and the independent copy number control of the major and minor chromosome. However, the molecular mechanisms to control copy numbers and to coordinate the replication of multiple replicons in the haloarchaeal genome remain to be illustrated.


This work was supported partially by grants from the National Natural Science Foundation of China (Grant Nos. 30830004, 30925001, 31100893) and the Chinese Academy of Sciences (KSCX2-EW-G-2-4).