Integrons: novel mobile genetic elements
mediating antibiotic resistance in Enterobacteria and Pseudomonas
Paul H Roy
Université Laval, Ste Foy, Québec, Canada
The problem of multiple antibiotic resistance in Gram-negative
bacteria has been somewhat overshadowed recently by the more acute problem of resistant Gram-positive bacteria,
including vancomycin-resistant enterococci (VRE) and methicillin-resistant Staphylococcus
aureus (MRSA). However, aided by some intricate genetic
mechanisms, Gram-negative bacteria have been slowly but steadily accumulating and disseminating multiple resistance
to widely used antibiotics to a point where, particularly for Pseudomonas, therapeutic choices are becoming very limited.
Plasmids, transposons, and integrons
Many resistance genes in Enterobacteriaceae and Pseudomonas are on large, transferable extrachromosomal DNA elements, called plasmids,
on which may be other mobile elements, called transposons. Some transposons code for single resistances, e.g.,
the TEM beta-lactamase (and all of its extended spectrum derivatives) on Tn3, kanamycin resistance on Tn5, and tetracycline
resistance on Tn10.
Plasmids and transposons coding multiple drug resistance often possess another genetic element, the more recently
Integrons contain one, or more, antibiotic resistance gene which is present as a mobile gene cassette, inserted
in various arrangements between two conserved DNA regions (Figure 1 +2). Integrons were discovered in the mid-1980's,
when DNA sequencing of several resistance genes revealed a common sequence in their upstream region (1-4). The
upstream region codes for a common promoter for the inserted antibiotic resistance genes and, on the opposite strand,
an integrase similar to those of temperate bacteriophages (5). The integrase has been shown to mediate gene cassette
excision and insertion via circular intermediates (6,7) (Figure 1). Thus, cassettes are in fact mobile elements,
although they do not code for genes involved in their own mobility. Their movement is allied to the integrase activity.
An interesting feature of cassettes is a palindromic sequence known as a 59-base element which occurs at the end
of the cassette. The 59-base elements are involved in the site-specific recombination reaction and differ among
cassettes, although they have similar DNA sequences at their extremities (8). Cassettes insert preferentially up
front in the integron, immediately downstream of the promoter region, which produces the highest expression. Further
away from the promoter region, expression is decreased (9). As cassettes "leap-frog", antibiotic selection
pressure can select cassette arrangements in which a given resistance is nearest the promoter and thus most strongly
expressed. In addition to the mobility of cassettes, whole integrons are also mobile elements. At least one integron,
has a complete set of four transposition genes (10).
Evolution of integrons
Integrons probably were around for a long time, evolutionarily speaking, before ever having seen an antibiotic
resistance gene. The ancestral integron comes from an organism, not necessarily related to clinical bacteria, with
a moderately G+C rich DNA. An "empty" integron has been found on plasmid pVS1 (11). In the plasmid, the
conserved regions are immediately adjacent, confirming that 59-base elements are associated with the resistance
gene cassettes themselves (Figure 2). However, no free 59-base elements, and no resistance genes identical to those
found in integrons, have yet been found in the DNA database. It remains unknown how resistance genes become associated
with 59-base elements and how they are recruited into integrons. Codon usage analysis indicates a wide variety
of origins of the genes found as cassettes.
The first cassette to arrive in integrons may have been the qacE antiseptic resistance gene (see Tn402, Figure 2a). Most integrons are descended from a version in which the qacE gene has been
truncated by the non-specific insertion of a sulfonamide resistance gene and in which two of the transposition
genes have been lost. Thus, most integrons are defective transposons. It is interesting to speculate whether the
arrival of the qacE
cassette dates from the 1920's, when acridine orange was used therapeutically and the insertion of the sulfonamide
gene from the 1930's. In contrast, codon usage analysis indicates a close correlation between, and thus a potentially
long-co-evolution of, the integrase and the transposition genes. There may have been an ancient association between
a transposon and the remnants of a temperate phage.
A second class of integrons occurs in the transposon Tn7 and its relatives (Figure 2b). Only a very limited range of cassettes are
found, including resistance to trimethoprim (dhfrI), streptothricin (sat) and streptomycin-spectinomycin (aadA1) (12). The dhfrI and aadA1 cassettes are identical to those found in the pVS1 class of integrons. Indeed
cassette may have first occurred in Tn7,
later being transferred into the pVS1 class of integrons where, in plasmid pLMO150 (13), dhfrI is now associated with sulfonamide resistance, thus limiting the usefulness
of the low-cost antibiotic co-trimoxazole. It is worthy of note that these two plasmids were isolated from strains
from Sri Lanka.
Resistance genes within integrons
Integron-borne genes include beta-lactamases (OXA, PSE, and CARB), aminoglycoside modifying enzymes (aacA4, aacC1, aadB,
and many others), trimethoprim resistance (dhfr's I, IIa, IIb, IIc, V, VII, IX, X, and XII) and chloramphenicol resistance
see below). Clinically the most important up to now have been the amikacin (aacA4) and gentamicin (aacC1,
aadB) resistance genes.
The mapping and sequencing of integrons has led to the characterization of some novel mechanisms of resistance.
gene of Tn1696,
which expresses non-enzymatic chloramphenicol resistance, codes for an integral membrane protein, likely an efflux
protein (14). The chloramphenicol acetyltransferase (Cat) of Tn2424 was, surprisingly, dissimilar to the known Cat's (which all were similar
to one another) and rather resembled another acetyltransferase family (15). However, several Cat's of the new type
have now been found, both in integrons and elsewhere.
Carbapenem resistance on an integron
Recently a novel type B beta-lactamase, which inactivates imipenem and related carbapenems, has been located on
an integron. Its emergence poses a threat for rapid dissemination both geographically and among species. The IMP-1
carbapenemase was first described as a chromosomal gene in Serratia
marcescens strain TN9106 (16). The sequence upstream of
the structural gene is the conserved sequence of an integron, although this fact was not mentioned by the authors
at that time. An identical IMP-1 gene, also just downstream of the integron conserved sequence, was found in plasmid
RDK4 in Klebsiella aerogenes
Again the identical IMP-1 gene has been cloned from another S.
marcescens strain, AKL9373 (18). Unlike S. marcescens
TN9106, IMP-1 is plasmid encoded in AK9373. In the upstream sequence in AK9373 is a gene which codes for yet another integrase! Its closest relatives are
the pVS1 integrase (60% identity) and the Tn7 integrase (45% identity). Moreover, at its downstream end the IMP-1 cassette
is followed by an aacA4
(aminoglycoside-6'-acetyltransferase) cassette. Thus it appears that this is the first example of a third class
The pVS1-class integrons, the Tn7-class
integrons, and the new integron recently isolated in Serratia may thus represent transposons which long ago obtained integrases and attachment
sites from cryptic prophages, and have fortuitously become ideal vehicles for dissemination of genes which enable
survival under strong selective pressure. Thus far they harbor antibiotic resistance genes, although open reading
frames (orf's) for proteins of unknown function suggest that integrons serve other purposes in cell survival. How
chromosomal genes acquire 59-base elements and become cassettes and whether other genes, such as virulence factors,
have the potential to become cassettes remains unknown for the present.
- Cameron FH, Groot Obbink DJ, Ackerman VP, Hall RM.
Nucleic Acids Res 1986; 14:8625-8635.
- Hall RM, Vockler C. Nucleic Acids Res 1987; 15:7491-7501.
- Ouellette M. Bissonnette L, Roy PH. Proc Natl Acad
Sci USA 1987;84:7378-7382.
- Sundström L, Rådström P, Swedberg G,
Skölf O. Mol Gen Genet 1988; 213:191-201.
- Ouellette, Roy PH. Nucleic Acids Res 1987; 15:10055.
- Collis CM, Hall RM. Mol Microbiol 1992; 6:2875-2885.
- Collis Cm, Hall RM. J Bacteriol 1992; 174:1574-1585.
- Hall RM, Brookes DE, Stokes HW. Mol Microbiol 1991;
- Collis CM, Hall RM. Antimicrob Agents Chemother 1995;
- Rådström P, Sköld O, Swedberg G, Flensburg
J. Roy PH, Sundström L. J Bacteriol 1994; 176:3257-3268.
- Bissonnette L, Roy PH. J Bacteriol 1992; 174:1248-1257.
- Sundström L, Roy PH, Sköld O. J Bacteriol
- Sundström L, Sköld O. Agents Chemother 1990;
- Bissonnette L, Champetier S, Buisson J-P, Roy Ph. J
Bacteriol 1991; 173:4493-4502.
- Parent R, Roy PH. J Bacteriol 1992; 174:2891-2897.
- Osane E, Arakawa Y. Wacharotayankun R, Ohta M, Horii
T, Ito H, Yosihimura F, Kato N. Antimicrob Agents Chemother 1994; 38:71-78.
- Sawai T. GenBank submission, accession number D29636.
- Arakawa Y, Murakami M. Suzuki K, Ito H, Wacharotayankum
R, Ohsuka S, Kato N, Ohta M. Antimicrob Agents Chemother 1995; 39:1612-1615.