Since the beginning of the mass production of antibiotics in the 1930ıs, antibiotic resistant strains of bacteria have been on the rise. Bacteria can evolve so quickly that they become resistant to the antibiotics faster than we can make new derivatives. Within recent years we have seen many old-time diseases reemerge with multi-drug resistances, making treatment very difficult and costly. With the rise of multi-drug resistant bacteria, scientists are trying to understand how so many different strains of bacteria could have acquired these genes so quickly. Pathogenisity islands are found in many virulent strains of bacteria, but their mode of transfer is largely unknown. Phages and transposons are also able to transfer antibiotic resistances to many bacterial strains. Within the last 25 years, new elements called conjugative transposons, plasmids, and related elements have been discovered. These elements often share many of the same general properties, but also have distinct differences that set them apart from the other elements. In 1999, Hochhut and Waldor proposed the term CONSTIN for all elements that have a conjugal mode of transfer, are self-transmissible to the recipient (not requiring host genes), and integrate into the host chromosome (2). The term constin, however, is still relatively new and is not yet widely used. The following review traces the history of the original CONSTIN, the SXT element discovered in Vibrio cholerae.
Until recently, Vibrio cholerae O1, also known as El Tor, was the only serogroup known to cause epidemic cholera. However, in early 1993, a large epidemic that broke out in India was cause by a novel Vibrio cholerae strain. Strain O139 had three important differences from the El Tor strain. First, it had a novel O139 serogroup antigen. Second, there was a deletion for most of the O1 serogroup antigen. Third, the new strain had acquired four antibiotic resistance genes including sulfamethoxazole and trimethoprim (collectively called SXT), streptomycin, and furazolidone. It was believed that a horizontal transfer of the O139 antigen to El Tor followed by a recombination event must have occurred. However, this did not explain how El Tor could have acquired the antibiotic resistance genes (1).
To address this question they selected for antibiotic resistance from a cosmid library. The results showed that sulfamethoxazole, trimethoprim and streptomycin were linked to each other, but furazolidone was not. Careful analysis also indicated that the SXT genes where found on a novel genetic element (1). To prove that this element could transfer the SXTR genes to a broad host range, Hochhut and Waldor selected for SXTR in V. cholerae O1 and E.coli SXTS recipients after conjugation with a SXTRV. cholerae donor. No plasmid was ever recovered and pulsed-field gel electrophoresis analysis of the recipient DNA indicated that the element was approximately 62 Kb and had a preferential integration site in the chromosome (1).
Knowing that the SXT element had a preferential integration site somewhere in the host chromosome, the next question was where in the chromosome was this site. Sequencing the ends of the element mapped the insertion site to the 5ı end of the prfC gene. prfC encodes peptide release factor 3 which functions in the termination of protein synthesis. This gene is found in many organisms, possibly indicating why the element has such a broad host range. Integration of the SXT element does not disrupt function of RF3, but does provide it with a novel 5ı end. The target sequence of the SXT element is 17 bp long (see below) with possible cut sites between the 6-8 bp and 9-14 bp. Integration results in sequence duplication at the ends of the element (2).
To determine if excision or integration of SXT was recA dependant, conjugation experiments with recA minus donor or recipient strains were preformed. The results indicated that recA minus recipients had a 20-50 fold decrease transmission of SXT. However, if the donor was recA minus, there was no transmission of SXT element. Therefore, it was concluded that recA was either need for excision from the chromosome or in transfer of the element to recipient (2).
When conjugative transposons are transferred to another host, they excise from the chromosome and form an extra chromosomal circular form. To determine if the SXT element also had a circular intermediate, a PCR primer that would only recognize the joined ends of the element was made. Indeed, a circular intermediate was found. When SXT excises form the chromosome, it doesnıt leave a copy of the element behind. Excision was not to be recA dependant, because a recA strain still contained the circular intermediate (2). Therefore, recA must be required for transfer of the element to the recipient.
A gene (int) with high sequence similarity to the integrase family of recombinases was found near the 5ı end of the element. A mutation in int resulted in no formation of the circular intermediate, and consequently no transfer of the SXT element. If int was supplied on a plasmid, transfer of SXT element was restored. Therefore, it was determined that int had function in the excision or circularization of the element as well as integration of SXT into the host chromosome.
Many conjugative transposons can mobilize plasmids from the host cell. To test if the SXT element could also mobilize plasmids, Hochhut and Waldor monitored the frequency of transfer of a plasmid containing a KanR cassette. It was found that this plasmid was transferred, but the frequency of transfer was reduced approximately100 fold that of SXT transfer. In addition, transfer of Kan and SXT were independent of each other. Therefore, it was concluded that the transfer of SXT and the plasmid where independent functions. An int minus donor was found to transfer the plasmid at the same frequency. Therefore, it was concluded that conjugative function, which is required for mobilization of the plasmid, is not dependent on the expression of the int gene product. In addition it was determined that excision and circularization of the SXT element is not needed for expression of the genes required for conjugal transfer of plasmids (3).
They also wanted to determine if the SXT element was able to transfer chromosomal DNA from the donor to recipient. To test this, Tn10 was inserted up and downstream of the SXT element. It was found that Tn10 insertions upstream of SXT element where not transferred. In contrast, Tn10 insertions downstream of the element were transferred. The transfer of the chromosomal DNA was found to be dependant on the presence of the SXT element in the donor stain and the DNA was transfer independently and in a directional manor from the SXT element, much like Hfr. The integration of the DNA was recA dependant and was mediated through homologous recombination (3).
Although a lot of scientific data has been accumulated on the SXT element, there is a lot more to be learned. Very little is understood about the mechanisms of integration, excision, or transfer of the element. Under standing these mechanisms and any regulation of these mechanisms will help shed light on how horizontal gene transfer occurs.
History has shown us that bacteria are more than capable of defeating us at our war against the spread of epidemic diseases. Understanding how bacteria horizontally transfer genes is one piece of the larger puzzle that could eventually turn the table in this war. Knowledge is power.
Written by Lisa Mueller
Microbiology Department, University of Illinois-Urbana, Champaign
This site is maintained by Stanley Maloy, please send comments, suggestions, or questions to
Last modified May 10, 2000