The physical isolation of SCP1 (Kinashi and Shimaji-Murayama 1991) was followed by the cloning of segments of SCP1 in cosmids and the establishment of a restriction map (Redenbach et al. 1998), and eventually the sequencing of the plasmid (Bentley et al. 2004) and the chromosome of its host (Bentley et al. 2002). It became possible to investigate the physical basis of some of the chromosome-donating strains described above.
Integration of SCP1 Within the Chromosome
The first SCP1-chromosome chimera to be investigated was that of an NF strain 2612 (Fig. 5a: Hanafusa and Kinashi 1992; Kinashi et al. 1992). Combining that analysis with DNA sequence information (Yamasaki et al. 2000, 2001; Bentley et al. 2002, 2004), it emerged that chromosomal integration of SCP1 had involved (a) interruption of the chromosome by insertion of SCP 1 into gene SCO3464, such that the left end of SCP1 at this junction was almost intact, and (b) at the other end of SCP1, recombination between a copy of IS466 in SCP1 (IS466S; SCP1.276) and another copy in the chromosome (IS466B; SC03490) (Fig. 5a). In the course of these events, both the right arm of SCP1 and a ca. 33-kb segment of the chromosome (genes SC03465-SC03489) were lost. The chromosomal deletion includes a large series of genes involved in an undefined aspect of carbohydrate metabolism. At one end of the deleted segment is the dagA gene encoding a diffusible agarase, which Hodgson and Chater (1982) had earlier found was missing from NF strains.
An independently isolated "NF-like" strain, A634 (Vivian and Hopwood 1973, see above), was obtained in a second step from a donor strain that initially showed a unidirectional gradient of transfer. In the formation of A634,
SCP1 recombination with the chromosome again involved interactions of plasmid- and chromosome-borne copies ofIS466; but this time SCP1 became integrated in the opposite orientation. The multi-step formation of A634 is reflected by the presence of deletions of both ends of the integrated SCP1 (Fig. 5a). Because of these end deletions, it is not possible to deduce the position at which SCP1 first interacted with the chromosome: this may have been in SCO3464, as in the case of NF, but with the opposite orientation of SCP1. Yamasaki et al. (2001) hypothesised a series of subsequent events to give rise to the eventual structure of A634. Their model invokes an exchange between IS466 copies in two non-identical chromosomes—the A634 progenitor and a wild-type copy—to account for the eventual location of the left-hand junction of the chromosome and SCP1 at a copy of IS466 (SCO3469), as well as for the occurrence of a duplicated 5.4-kb chromosome segment on either side of the integrated plasmid. In A634, another transposable element, IS468, is present in two places in the integrated SCP1, whereas SCP1 usually contains no IS468 copies. The two implied IS468 transposition events must have taken place from the chromosome, which has tandemly arranged copies located in the SCP1 integration target region (IS468A, SCO3466; IS468B, SCO3467). Recombination of IS468A with one of the SCP1-located copies accounts for the deletion of part of the SCP1 TIR-L during the formation of A634. A partially alternative model for the origin of A634 is possible, based on a Campbell integration event (see earlier). Because such events are mediated by a single crossover between homologous sequences present on two different DNA molecules, they result in duplication of the common sequence, such as is seen in A634. Perhaps the lineage of A634 included a circularised SCP1 plasmid carrying this 5.4-kb segment.
It is not known what gives rise to unidirectional and bidirectional gradients of marker inheritance in crosses of the various donor strains with SCP1-recipients. The devising of models is limited by a lack of knowledge about the basic enzymology and geometry of SCP1 transfer, which is different from that of the plasmids of enteric bacteria (such as F in E. coli), and probably involves double-stranded transfer, though whether starting from one or both ends, or from an internal transfer origin, is unclear (see Hopwood 2006 and the chapter by Chen, in this volume, for more detailed reviews of these aspects).
Single Crossover Recombination Between a Linear Plasmid and the Chromosome to Generate Molecules with Heterologous Ends
Interestingly, a single exchange between SCP1 and the chromosome (albeit showing complexity at the nucleotide sequence level) can generate a viable strain (Yamasaki and Kinashi 2004). In a unidirectional donor strain 2106 (Hopwood and Wright 1976b) capable of high-frequency donation of the cysD gene (SC06098), and of donation of a gradient of markers clockwise from cysD (Fig. 2), the housekeeping genes of S. coelicolor have become split between two chromosomes: one of 7181 kb, comprising the left end of the normal chromosome (SC00001-SC06389') and the left end of SCP1 (SCP1.1-SCP1.136'), and the other of 1843 kb, fusing the right end of the normal chromosome (SC06388'-SC07846) to the right end of SCP1 (SCP1.136-SCP1.353) (Fig. 5b). The recombination event giving rise to these two complementary fusion chromosomes was an "illegitimate" event, in that there was no sequence homology at the point of crossing over and a small number of bases had been lost at one junction. Strikingly, the two hybrid chromosomes both have non-identical termini (one from SCP1, one from the chromosome), although these termini have some of the general structural features of Strep-
tomyces telomeres (though it should be noted that the SCP1 termini do have some atypical features) (see Chen 2007, in this volume).
One might wonder how replication and partitioning of the two chromosomes in 2106 could be stably achieved during various key points in mycelial growth (Fig. 6). Thus, at least one chromosome origin maintains a position closely behind growing hyphal tips, and this positioning must be established at the new tips generated by branch emergence. Also, chromosomes need to be partitioned to either side of newly formed septa, most particularly during sporulation, when each spore usually receives just one copy of the chromosome (the need for a plasmid to partition into spores would also be crucial). In fact, both of the SCP1-chromosome hybrid replicons in strain 2106 appear to be equipped with replication origins and partitioning functions. The larger of the hybrid molecules contains the entire chromosomal replication origin region, including the parAB partitioning genes (Kim et al. 2000). The smaller presumably depends on equivalent functions of SCP1: the right end of SCP1 present in this hybrid chromosome contains a segment capable of autonomous replication in an SCP1-free host (extending from within SCP1.194 to a site within SCP1.199, Redenbach et al. 1999), as well as two sets of putative partitioning genes (ORFs SCP1.138-139; SCP1.221-
222). (One or both of the latter pairs may be responsible for the observation that the deletion of the chromosomal parAB genes, which causes frequent failure of chromosome partitioning during sporulation of SCP1- strains, had little effect on a strain containing an integrated copy of SCP1 (Bentley et al. 2004).)
Other cases of exchange of a chromosome end with a linear plasmid end have been reported, demonstrating the "viability" of linear replicons with almost no TIRs. In S. lividans, this apparently resulted from recombination between chromosomal and plasmid SLP2-borne copies of a 5.3-kb transposable element (Tn4811) (resolution of a Tn4811 transposition event was another possible explanation) (Huang et al. 2003). A further example is provided by the acquisition of an S. rimosus chromosome end, including oxytetracycline biosynthetic genes, by the linear plasmid pPZG101: an event involving recombination at a homologous sequence of just four base pairs (Gravius et al. 1994; Pandza et al. 1998; see below). We note that in these cases, physical analysis of terminal structures was confined to the mainly plasmid hybrid molecules, and was not extended to the mainly chromosomal ones.
As already mentioned, SCP1 has also been found to recombine with a circular plasmid, SCP2, in an SCP1+ strain that also carried SCP2; SCP1 was seen as a ladder of bands on PFGE, with increments of ca. 30 kb caused by tandem integrated sequences mostly derived from SCP2 in a complex manner that is not yet fully elucidated, but which may be the result of a single initial crossover between copies of Tn5714 present in both elements (Kinashi et al. 1993; Bentley et al. 2004; H. Kinashi, unpublished results). This provides a model for the probable origin of linear chromosomes from a single crossover between a circular chromosome and a linear extrachromosomal replicon.
Was this article helpful?