Susceptibility of Tetracyclines to Tet(A) Resistance Is Independent of Interdomain Loop Sequence

LETTER

Tigecycline, the latest member of the tetracycline family to enter clinics, is reported to be poorly recognized by the most prevalent Gram-negative tetracycline-specific efflux mechanisms, namely, the Tet(A) and Tet(B) efflux pumps (1). These structurally related tetracycline-specific efflux pumps are encoded by genes on plasmids and transposons in a wide variety of bacteria and are proton-driven antiporter members of the major facilitator superfamily of transporters (2). Others have reported Salmonella species isolates with reduced tigecycline susceptibility and attributed this to the presence of a variant Tet(A) sequence associated with transposon Tn1721 (3, 4). Tuckman et al. suggested that the change from Ser-Phe-Val of the previously reported plasmid pRP1 sequence to Ala-Ser-Phe at residues 201 to 203 in Tn1721-associated Tet(A) resulted from a double frameshift within the tet(A) gene, affecting substrate specificity (4, 5). These residues are located within the pump's largest cytoplasmic loop between helices 6 and 7, connecting the α and β domains (4, 6). Here, we directly compare the substrate specificities of the two Tet(A) cytoplasmic loop sequences by expressing tet(A) genes encoding each sequence in an isogenic Escherichia coli strain background.

Cloning of Tn1721-associated tet(A) and lacZ (negative-control) structural genes under the control of an arabinose-inducible promoter in pBAD-Myc-His (Invitrogen, Carlsbad, CA) was described previously (7). For construction of pBAD-tet(A)-SFV, encoding the pRP1-associated tet(A) cytoplasmic loop sequence, a DNA sequence encoding Ser-Phe-Val was introduced into the tet(A) gene of pBAD-tet(A)-ASF by PCR using a forward primer (7) containing a 5′ NcoI site at the translation start site (GCCGGCCTGTCCCATGGTGAAACCCAACAGACCCCTGATCGTAATTCTGAG) and a reverse primer (CGGCGCCCGGGCCCACCGAACGAAGCTGAGCGGGTTGAGAGCCTCCCGGCGTAACGGCC). The reverse primer encoded Ser-Phe-Val located immediately upstream from a unique ApaI site. The PCR product was cut with NcoI and ApaI, and the 640-bp fragment was cloned to replace the corresponding Ala-Ser-Phe fragment in pBAD-tet(A)-ASF, producing pBAD-tet(A)-SFV. All genes were sequenced, transformed into E. coli DH10B cells, and grown in medium containing ampicillin (50 μg/ml).

Induction with increasing concentrations of arabinose produced comparable levels of reduced susceptibility to tetracycline, doxycycline, minocycline, and tigecycline for DH10B[pBAD-tet(A)-ASF] and DH10B[pBAD-tet(A)-SFV] cells, demonstrating that these two tet(A) genes produced proteins with similar substrate specificities under identical expression conditions in an isogenic strain background (Table 1). Controls showed that expression of lacZ had no impact on tetracycline resistance, and arabinose induction of all strains had no effect on visual growth and susceptibility to a nontetracycline antibiotic, ceftriaxone.

Table 1.

Susceptibility of E. coli DH10B cells expressing tet(A) variants^a

Antibiotic	MIC (μg/ml)
	DH10B cells with no plasmid and 0% ara	LacZ control with pBAD-lacZ		Tn1721-associated Tet(A) sequence with pBAD-tet(A)-ASF				pRP1-associated Tet(A) sequence with pBAD-tet(A)-SFV
		0% ara	1% ara	0% ara	0.01% ara	0.1% ara	1% ara	0% ara	0.01% ara	0.1% ara	1% ara
Tigecycline	0.031	0.031	0.063	0.063	0.13	0.25	1	0.063	0.063	0.5	2
Tetracycline	2	2	2	16	32	128	256	16	16	64	256
Minocycline	0.5	0.5	0.5	0.5	0.5	2	8	0.5	0.5	2	8
Doxycycline	2	2	2	2	4	8	32	2	4	8	32
Ceftriaxone	0.13	0.13	0.13	0.063	0.13	0.13	0.13	0.063	0.13	0.13	0.13

^aIn vitro susceptibility testing was done according to CLSI guidelines (8) with cation-adjusted Mueller-Hinton broth, except as noted. MIC assays for strains containing plasmids were run in the presence of ampicillin (50 μg/ml, except for DH10B cells with no plasmid) and arabinose (0.1 to 1%), as noted for the induction of tet(A) expression, where appropriate. Strains containing plasmids were preinduced with arabinose at room temperature for 30 min prior to assay plate inoculation. ara, arabinose.

As noted initially by Hentschke et al. (3) and in our own database searches, Ala-Ser-Phe at residues 201 to 203 is the most common Tet(A) sequence in clinical isolates, comprising >95% of the Tet(A) entries in GenBank, while the Ser-Phe-Val sequence at residues 201 to 203, present in the first reported Tet(A) in pRP1, was reported in only a single isolate. Thus, the pRP1-associated Tet(A) sequence appears to be a low-incidence variant. In conclusion, and in contrast to earlier reports (3, 4), the interdomain cytoplasmic loop sequence difference between Tn1721- and pRP1-associated Tet(A) sequences has no effect on substrate specificity when tested in an isogenic background. These findings suggest that other factors are responsible for the reduced tigecycline sensitivity of some tet(A)-positive clinical isolates.