The role of the 20,922-Da RecX protein and its interference with RecA activity were analyzed in Streptomyces lividans. The recX gene is located 220 bp downstream of recA. Transcriptional analysis by reverse transcriptase PCR demonstrated that recX and recA constitute an operon. While recA was transcribed at a basal level even under noninducing conditions, a recA-recX cotranscript was only detectable after induction of recA following DNA damage. The recA-recX cotranscript was less abundant than the recA transcript alone. The recX gene was inactivated by gene replacement. The resulting mutant had a clearly diminished colony size, but was not impaired in recombination activity, genetic instability, and resistance against UV irradiation. Expression of an extra copy of the S. lividans recA gene under control of the thiostrepton-inducible tipA promoter was lethal to the recX mutant, demonstrating that RecX is required to overcome the toxic effects of recA overexpression. Since inactivation of the recX gene did not influence transcription of recA, the putative function of the RecX protein might be the downregulation of RecA activity by interaction with the RecA protein or filament.

The glutamine synthetase II (GSII, encoded by glnII) activity detectable in crude extracts from Streptomyces coelicolor is low compared to the activity of glutamine synthetase I (GSI, encoded by glnA) and to that of GSII from S. viridochromogenes. We have identified and sequenced a 3.9-kb BglII-BamHI fragment carrying the glutamine synthetase II gene (glnII) from S. coelicolor. Besides glnII, this region contains four ORFs (orf1-orf4). While homologues of orf1 and orf2 were also found in the glnII region of the S. viridochromogenes chromosome, this was not the case for orf3 and orf4, which encode a putative hydrolase and a transcriptional regulator (Ptr) of the MarR family, respectively. High-resolution S1 nuclease mapping showed that the S. coelicolor glnII gene is expressed from two overlapping promoters. The first comprises a vegetative promoter sequence and the second contains sequence elements that are recognized by Esigma31. Similar promoter structures were found upstream of the S. viridochromogenes glnII gene. The involvement of ptr in glnII regulation was studied by gel retardation assays. Recombinant Ptr interacted with the upstream region of ptr, but not with the promoter region of glnII. A ptr gene replacement mutant (S. coelicolor IP) was also constructed. RT-PCR analysis of RNA from wild-type S. coelicolor and the IP mutant demonstrated that expression of orf3 depends on Ptr. Thus, the difference in gene organization between S. coelicolor and S. viridochromogenes is not responsible for the difference in GSII activity.

In contrast to recA of other bacteria, the recA gene of Streptomyces lividans has been described as indispensable for viability (G. Muth, D. Frese, A. Kleber, and W. Wohlleben, Mol. Gen. Genet. 255:420-428, 1997.). Therefore, a closer analysis of this gene was performed to detect possible unique features distinguishing the Streptomyces RecA protein from the well-characterized Escherichia coli RecA protein. The S. lividans recA gene restored UV resistance and recombination activity of an E. coli recA mutant. Also, transcriptional regulation was similar to that of E. coli recA. Gel retardation experiments showed that S. lividans recA is also under control of the Streptomyces SOS repressor LexA. The S. lividans recA gene could be replaced only by simultaneously expressing a plasmid encoded recA copy. Surprisingly, the recA expression plasmid could subsequently be eliminated using an incompatible plasmid without the loss of viability. Besides being UV sensitive and recombination deficient, all the mutants were blocked in sporulation. Genetic complementation restored UV resistance and recombination activity but did not affect the sporulation defect. This indicated that all the recA mutants had suffered from an additional mutation, which might allow toleration of a recA deficiency.

Gene transfer is a basic requirement for optimizing bioactive natural substances produced by an increasing number of industrially used microorganisms. We have analyzed quantitatively horizontal gene transfer from Escherichia coli to Actinomycetes. The efficiencies of DNA transfer of four different systems were compared that consist of conjugative and mobilizable plasmids with a broad-host range. Three novel binary vector set-ups were constructed based on: (i) the IncQ group of mobilizable plasmids (RSF1010), (ii) IncQ-like pTF-FC2 and (iii) pSB102 that belongs to a new class of broad-host-range plasmids. The established system based on the IncP␣ group of conjugative plasmids served as the reference. For all plasmids constructed, we confirmed the functional integrity of the selected transfer machineries by intrageneric matings between E. coli strains. We demonstrate that the transfer systems introduced in this study are efficient in mediating gene transfer from E. coli to Actinomycetes and are possible alternatives for gene transfer into Actinomycetes for which the IncP␣-based transfer system is not applicable. The use of plasmids that integrate into the recipients' chromosomes compared to that of plasmids replicating autonomously is shown to allow the access to a wider range of hosts.