Elimination of internal restriction enzyme sites from a bacterial luminescence (luxCDABE) operon.

The luxA and luxB genes of luminescent bacteria, most frequently from Vibrio spp., are widely used as reporter genes (3,10). The heterodimer (αβ) luciferase enzyme catalyzes oxidation of the reduced flavin mononucleotide (FMNH2) and a long-chain fatty aldehyde, resulting in emission of light. In vitro, dodecanal or nonanal are used as preferred exogenous aldehyde substrates for bacterial luciferases (3). In addition to the luxAB genes, the lux structural operons of luminescent bacteria also direct the synthesis of a fatty acid reductase complex (encoded by the luxCDE genes), which synthesizes tetradecanal, a natural substrate of luciferase (7). Introduction of the luxCDABE operon of a luminescent insect pathogen bacteria, Xenorhabdus luminescens, into Escherichia coli resulted in a luminescent phenotype (5,6) because of the expression of the luciferase enzyme and synthesis of the tetradecanal substrate. This indicates that luminescence of these recombinant E. coli cells is independent of the exogenous substrate. In this regard, the luxCDABE operon is a more easily applicable system than the luxAB genes, in which addition of an exogenous substrate is required for light emission. A particular advantage of using the lux operon of X. luminescens as a reporter system is that the LuxAB protein of this species is the most thermostable of all the known bacterial luciferases (3). A factor that may limit the use of the luxCDABE operon of X. luminescens in genetic engineering is the presence of several internal restriction enzyme sites. The operon can be isolated as a 6960-bp EcoRI fragment (5,6), and while it has no SacI, KpnI, SmaI, XmaI, BamHI, SalI and PstI sites, there are HindIII, SphI and XbaI sites in the operon (Figure 1). Because these sites are common in the multiple cloning sites (MCSs) of many plasmid vectors, we aimed to eliminate them from the lux operon of X. luminescens to facilitate manipulation of the operon. The sites were eliminated by site-directed mutagenesis, and we produced a lux cassette by cloning the mutant operon into the symmetrical MCS of plasmid pMTL25 (2). Site-directed mutagenesis was performed on plasmid pLITE27 containing the 6960-bp EcoRI lux fragment of X. luminescens in pUC118 (6). The seven mutagenic primers used for the elimination of the restriction sites from the lux operon (Figure 1) were designed in such a way that the base-pair exchanges do not alter the encoded amino acids of the Lux proteins. An additional primer was used to eliminate an external HindIII site flanking the lux operon in pLITE27. The elimination of an external restriction site is required for removal of the wild-type plasmid molecules during the mutagenesis process (4). The sequence of the external primer we used is: 5′-GGCCAGTGCCAACCTTGCATGC-3′, where the mutant restriction site is underlined with the exchanged nucleotide in boldface. The mutagenesis was performed according to the method described by Deng and Nickoloff (4) with two modifications. First, T4 DNA polymerase was replaced by T7 DNA polymerase (New England Biolabs, Beverly, MA, USA) in the second-strand synthesis reaction. Like T4 DNA polymerase, T7 DNA polymerase does not perform strand displacement synthesis (an important requirement in site-directed mutagenesis; Reference 1). The polymerization by T7 DNA polymerase, however, is faster and more efficient than that by T4 DNA polymerase (1). This is an advantage if the template molecule is relatively large, as is the case with pLITE27 (about 10 kb). Second, we incorporated all eight primers into a single second-strand synthesis reaction. Previously, Ray et al. reported achieving eleven point mutations in four rounds of site-directed mutagenesis (8). Following the second transformation of the mutagenesis procedure, about 200 primary transformants grown on LB/ampicillin agar plates were screened for luminescence. The plates were taken into a darkroom, and after allowing 2–3 min for the eyes to accommodate to the dark, 14 highly luminescent colonies were identified by visual examination of the plates with the naked eye. Plasmid DNA was isolated from these colonies using the alkaline lysis miniprep method (9). A HindIII diagnostic digest revealed that eight of the isolated plasmids are not cut by this enzyme. Of these eight plasmids, four that were purified from colonies displaying the same level of luminescence as the original pLITE27/ DH5α strain (DH5α from Life Technologies, Gaithersburg, MD, USA) (data not shown) were selected for further examination. Each of them was digested with EcoRI, and the 6960-bp lux fragments were gel-purified using the GLASSMAX DNA Isolation Spin Cartridge System (Life Technologies). The isolated fragments were treated with SphI, XbaI and, once more, HindIII restriction enzymes. Three fragments, obtained from plasmids designated pLITE2, pLITE5 and pLITE7, were not cut by either enzyme, indicating that all