OUP user menu

Pre-exposure to UV irradiation increases the transfer frequency of the IncJ conjugative transposon-like elements R391, R392, R705, R706, R997 and pMERPH and is recA+ dependent

Barry M. McGrath , John A. O'Halloran , J. Tony Pembroke
DOI: http://dx.doi.org/10.1016/j.femsle.2005.01.013 461-465 First published online: 1 February 2005


The enteric conjugative transposon-like IncJ elements R391, R392, R705, R706 and pMERPH, all demonstrated increased conjugative transfer upon UV irradiation. The transfer frequency increased on average from its basal rate of 10−5 to 10−3 per recipient, upon pre-exposure to UV irradiation. However, the transfer frequency of R997, which was higher than the other IncJ elements at 10−3 per donor, showed a smaller increase. This effect was shown to be recA+ dependent in all cases. Using PCR primers directed outwards from the ends of the integrated R391 element it was observed that a circular intermediate of the element forms within the host, which has been proposed to be a transfer intermediate. Using real-time PCR, it was determined that the amount of the circular intermediate produced increased substantially upon UV irradiation.

Key words
  • IncJ elements
  • Conjugative transposon
  • R391
  • R392
  • pMERPH
  • R705
  • R706
  • R997
  • UV induced conjugative transfer

1 Introduction

The IncJ group of enterobacterial mobile genetic elements, of which R391 (originally isolated in South Africa) is the prototype, comprises a small, globally restricted group of enteric conjugative transposon-like elements [1,2], which integrate at a unique site, the prfC gene in the Escherichia coli chromosome [3,4]. Only five phenotypically distinct members of this group have been isolated worldwide. These include the R391 group (consisting of five other elements: R748, R749, R706, R705 and R392, all encoding resistance to kanamycin and mercury) from Proteus and Providencia spp. isolated at a single site in South Africa [2,5], R997 from Proteus mirabilis in India [6], pMERPH from Shewanella putrefaciens in the UK [7] and pJY1 from Vibrio cholerae in the Philippines [8]. The SXT element, isolated from V. cholerae in India has been shown to share significant sequence homology with R391 indicating that SXT may be yet another member of the IncJ group [9]. Indeed, there have been several recent reports of R391/SXT-like elements emerging in epidemic strains of Vibrio in Vietnam and adjoining countries which are responsible for drug resistance emergence [10,11].

The complete nucleotide sequence of the 89-kb R391 element has revealed it to be a conjugative transposon lacking a detectable rep function and possessing the characteristics of a conjugative transposon. R391 has been shown to be a mosaic structure of phage-like integrase and regulatory elements, plasmid-like transfer and pili modules, transposon-like sequences and interestingly a number of DNA repair-like genes [12]. Indeed all the IncJ like elements thus far tested encode an unusual UV-sensitising function [2,13] when expressed in E. coli, which may be related to these repair associated genes. Analysis of R391, R392, pMERPH, R997, R705 and R706 has revealed that these elements share extensive sequence homology at their ends, contain similar integrase genes and all integrate as conjugative transposon-like elements into the prfC gene [4]. We show that many of the IncJ elements demonstrate up to a two-log increase in transfer frequency upon UV irradiation, similar to a recently reported phenomenon associated with SXT, a phenomenon that is recA+ dependent [14].

2 Materials and methods

2.1 Bacterial strains, elements and culture media

The genotype and source of bacterial strains used are listed in Table 1. Media was supplemented with appropriate antimicrobial agents (kanamycin, 30 μg ml−1; streptomycin, 50 μg ml−1; ampicillin, 50 μg ml−1; tetracycline, 30 μg ml−1; and mercuric chloride, 20 μg ml−1).

View this table:

Genotype, source and phenotype of strains and mobile genetic elements used

2.2 Conjugative transfer

Conjugation of the IncJ elements to recipient hosts was performed as described [1,13,15]. For UV irradiation, donor cells were irradiated with a Griffith and George lamp emitting wavelengths of 254 nm, with dose monitoring via a UV radiometer (Ultra-Violet Products, San Gabriel, CA, USA). During irradiation, cells were constantly agitated to prevent shielding by dead cells. After irradiation cells were conjugated and transfer monitored by counting donors and recipients on antibiotic or heavy metal selective nutrient or minimal media [1]. Survival and % killing was determined as described [13].

2.3 PCR amplification

DNA amplification and primer design was carried out using standard techniques [16]. Template DNA for amplification of circular forms of R391 was prepared as described [4] and PCR amplification carried out using the BioTaq PCR system (New England Biolabs, Hertfordshire, UK). Verification of PCR products was performed using SequiTherm Excel II DNA Sequencing Kit-LC (Epicentre Technologies, Madison, WI, USA), and fluorescent DNA primers (MWG Biotech, Milton Keynes, London, UK), labelled at the 5′ end with the dye IRD-800. Sequencing was performed using a Licor Inc. (Lincoln, NE, USA) long read IR4200 sequencer.

2.4 Real-time PCR

Cultures harvested at an optical density of 0.3, measured at 600 nm, were collected, resuspended in M9 minimal salts and after starvation for 30 min, were irradiated, collected, washed and allowed to recover for 30 min at 37 °C. Preparations were resuspended in 100 μl PCR-grade water, boiled for 10 min, centrifuged and the supernatant used as real-time PCR template. Real-time PCR was performed using the LightCycler FastStart DNA Master SYBR Green I kit (Roche, Bell Lane, Lewes, East Sussex BN7 1LG, UK) according to manufacturers instructions. Real-time PCR of the circular intermediate was performed using the primers prfCLE3 (5′-TTGTACACACTTTCCGAG-3′) and prfCRE3 (5′-GTACACACTTTCCGAGGTTACG-3′). The following temperature profile was used: denaturation for 1 cycle at 95 °C for 10 min and amplification for 5 cycles at 95 °C for 10 s, 54 °C for 5 s, 72 °C for 20 s, and for 27 cycles at 95 °C for 10 s, 49 °C for 5 s, and 72 °C for 20 s (temperature transition, 20 °C/s). Melting curve analysis was performed from 65-95 °C (temperature transition, 0.2 °C/s) and specificity of the PCR reaction was verified by ethidium bromide staining on a 2% agarose gel.

3 Results and discussion

3.1 UV enhances the transfer rate of IncJ elements

Table 2a illustrates the transfer rates of IncJ elements from un-irradiated AB1157 donor strains to CV601 (recA+) and NK5148 (recA+) recipients, and from CV601 donor strains to AB2463 (recA) and HB101 (recA) recipient strains. Table 2b indicated that transfer to and from recA strains occurred at approximately the same rates as to and from wild-type recipients. This indicated that transfer and establishment of the elements is independent of functional RecA. Results shown are the mean of 10 determinations in all cases. It can be seen that all IncJ elements transfer at approximately 10−5 per recipient with the exceptions of R997, which transfers at a rate of 10−3 per recipient and pMERPH, which is lower at 10−6 per recipient. The increased transfer rate of R997 may reflect some mutational alteration in the genetic control circuit of R997 such that the element may be de-repressed for transfer relative to the other elements. This proposal of de-repressed transfer of R997 is supported by the observation of increased pilus formation associated with R997 hosts [17].

View this table:

Transfer rates (per recipient) of IncJ elements from un-irradiated AB1157 hosts to CV601 and NK5148 (recA+) and from un-irradiated CV601 hosts to recA recipient strains HB101 and AB2463 after 2 h broth mating

View this table:

Transfer rates (per recipient) of IncJ elements from un-irradiated AB2463 (recA) hosts to recA+ and recA recipients

Analysis of the effect of pre-exposure of donor cells to UV irradiation on the transfer rate of R391 revealed that there was a two-log enhancement (Table 3), while analysis of other IncJ elements indicated that this enhancement also occurred in the case of R392, R705, R706 and pMERPH, with only minor enhancement in transfer rate occurring in the case of R997. Pre-exposure was carried out at a dose rate that gave rise to a 10% and 20% kill of donor populations. We noted in the case of pre-exposure of R391 hosts (Table 3) that initially the transfer rate increased dramatically and at higher dose rates began to decrease. This may be reflective of an increasing population of dead cells, which may be unable to act as donors. We tested the possibility that this phenomenon might be associated with the cells SOS response by analysing the transfer rate from a recA background. Direct comparison at particular UV dose rates was not possible given the UV sensitivity of recA strains. Our analysis was carried out utilising a UV dose that gave rise to 10% and 20% killing in both recA+ and recA strains, to normalise the relative proportion of cells killed in each population. It can be seen that at these levels of UV survival the transfer rates were not enhanced in a recA background indicating that the phenomenon is recA dependent (Table 3). Indeed several recA strains were analysed in addition to those presented in Table 3, which confirmed this effect, to eliminate the possibility that this phenomena was strain specific (data not shown).

View this table:

The effect of pre-exposure of donor strains to UV irradiation on the transfer rates of IncJ elements from AB1157 (recA+) and AB2463 (recA) to a recA+ recipient NK5148

3.2 Pre-exposure to UV irradiation enhances the formation of a circular intermediate of the IncJ R391 element

It has been proposed that a circular intermediate is involved in transfer of conjugative transposon-like elements such as R391 [1,4]. Although R391 integrates as a linear element into the prfC gene, a circular form can be detected which has been proposed to be a transfer intermediate. Sequence analysis of R391 has not revealed any rep-like function indicating that R391 is unable to replicate autonomously [12]. PCR primers were designed based on sequence analysis of the ends of known IncJ-like elements, pMERPH, R391, R997 and R392 [4]. These primers were designed to point towards the right and left junctions in the integrated form at the integration site [4]. Circularisation of the IncJ element would result in the primers now pointing towards each other and resulting in an amplicon of 542-bp in length in the case of R391 following PCR. Using the primers prfCRE3 and prfCLE3 an amplicon was detected in un-irradiated R391-containing hosts and upon irradiation it was observed that this product was present in larger amounts (Fig. 1(b)). Sequence analysis of the 542-bp fragment confirmed that it was the circularised junction fragment [4]. Real-time PCR was utilised to demonstrate this increase in circular intermediate in the case of R391 upon UV irradiation (Fig. 1(a)). The data indicates that a circular intermediate of R391 does form and that the quantity of this intermediate is increased upon pre-exposure of donor cells to UV irradiation at least in the case of R391. Quantification of the amount of R391-circular DNA formed based on this real-time PCR data indicates that the level is increased on average 4500 times relative to that of the un-irradiated host. Although sequence analysis has indicated that R391 is unlikely to have an autonomous replicating circular form [12], this data provides evidence that a circular intermediate does form even in un-irradiated cells, which may be a transfer intermediate. This data in conjunction with the increased transfer rate observed upon pre-exposure to UV (Tables 2a and 2b) indicates that the control of the formation of circular intermediate may be a key step in the observed transfer enhancement.


(a) Quantification of circular intermediate following UV irradiation via real-time PCR of the circular junction amplicons: 1, diluted standard used to relate samples to the standard curve; 2, un-irradiated sample; 3, sample irradiated for 100 s; 4, sample irradiated for 200 s; 5, sample irradiated for 300 s; 6, sample irradiated for 400 s; 7, negative control (no template). (b) A 2% agarose gel of real-time PCR products. Lanes: 1 and 9, molecular weight marker (Hyperladder II); 2, diluted standard used to relate samples to the standard curve; 3, un-irradiated sample; 4, sample irradiated for 100 s; 5, sample irradiated for 200 s; 6, sample irradiated for 300 s; 7, sample irradiated for 400 s; 8,negative control. (The dose rate in (a) and (b) is the time in seconds at a UV exposure of 10 μW cm−2.)

3.3 A suggested mechanism for UV induced transfer enhancement

UV irradiation of donor strains containing R391, R392, R705 and R706 resulted in a two-log increase in transfer rate and a lesser increase in the case of pMERPH and R997. This increased transfer was shown to be recA+ dependent indicating that it may be associated with the host cells SOS response. A similar phenomenon has recently been demonstrated in the case of the SXT element [14], where it was demonstrated that the drug resistant Vibrio determinant SXT showed elevated conjugation levels upon SOS induction. The recent publication of the complete sequence of R391 [12] (Gene Bank Accession number AY090559) has revealed the presence of a putative pro-phage cI like repressor (orf 96) similar at the amino acid level to the CI repressor of lambda. In post irradiation SOS induction of the bacteriophage lambda, the CI repressor is cleaved by the RecA protease, leading to induction of CI repressible genes and hence induction. An attractive possibility is that a similar phenomenon occurs with R391 and other IncJ elements. The orf 96/CI-like repressor may in fact control transfer, and indeed other element associated functions in such a way that low levels of transfer can occur under normal circumstances, but when cleavage of the repressor occurs there is a rate enhancement of some two logs. In high UV environments such as leaf surfaces, soil or water surfaces or environments exposed to inducing mutagens, such induced transfer would have a net population advantage should the element transfer advantageous survival or drug resistance traits.

3.4 Enhanced environmentally induced transfer may have survival advantages

The UV inducibility of transfer observed in the case of R391 and other IncJ elements and recently reported for SXT [14] is similar to the enhanced survival demonstrated by phage upon UV induction. It is plausible that the trait may be a genetic carry over associated with the mosaic structure of R391 and the other IncJ-like elements, being composed of phage- and plasmid-like sequences [12]. The ability to inducibly transfer traits within stressed environments would bestow evolutionary advantage for the donor population and may be an important mechanism in the selection of such elements. This would be obvious in the case of antibiotic resistance determinants, which appear to be the selective determinants in the emergence of R391/SXT-like elements in recent Vibrio epidemics [10,11]. However in the case of IncJ-like elements, such as pMERPH, there appears to be no antibiotic resistance determinant present, which suggests that the evolutionary advantage bestowed by increased transfer may be something else. We are currently identifying those sequences associated with R391 that are responsible for this rate enhancement and bestowal of the proposed selective advantage.


Part of this work was supported by EU concerted action BIO4-CT-0099 on ‘Mobile Genetic Elements Contribution to Bacterial Adaptability and Diversity’ (MECBAD) and by an Enterprise Ireland Basic Research Award to T.P.


  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].
View Abstract