Detection and identification of ciprofloxacin-resistant Yersinia pestis by denaturing high-performance liquid chromatography

Abstract

Denaturing high-performance liquid chromatography (DHPLC) has been used extensively to detect genetic variation. We used this method to detect and identify Yersinia pestis KIM5 ciprofloxacin-resistant isolates by analyzing the quinolone resistance-determining region (QRDR) of the gyrase A gene. Sequencing of the Y. pestis KIM5 strain gyrA QRDR from 55 ciprofloxacin-resistant isolates revealed five mutation types. We analyzed the gyrA QRDR by DHPLC to assess its ability to detect point mutations and to determine whether DHPLC peak profile analysis could be used as a molecular fingerprint. In addition to the five mutation types found in our ciprofloxacin-resistant isolates, several mutations in the QRDR were generated by site-directed mutagenesis and analyzed to further evaluate this method for the ability to detect QRDR mutations. Furthermore, a blind panel of 42 samples was analyzed by screening for two mutant types to evaluate the potential diagnostic value of this method. Our results showed that DHPLC is an efficient method for detecting mutations in genes that confer antibiotic resistance.

FIG. 1.

DHPLC peak profiles obtained from Y. pestis strain KIM5 and ciprofloxacin-resistant mutants of KIM5. All peak profiles depicted are from the first lot of samples and were generated from the same column, which was the column used to generate the peak profiles depicted in Fig. 2. (A) KIM5 homoduplex; (B) KIM5-mutation type M1 heteroduplex; (C) KIM5-mutation type M2 heteroduplex; (D) KIM5-mutation type M3 heteroduplex; (E) KIM5-mutation type M4 heteroduplex; (F) KIM5-mutation type M5 heteroduplex.

FIG. 1.

DHPLC peak profiles obtained from Y. pestis strain KIM5 and ciprofloxacin-resistant mutants of KIM5. All peak profiles depicted are from the first lot of samples and were generated from the same column, which was the column used to generate the peak profiles depicted in Fig. 2. (A) KIM5 homoduplex; (B) KIM5-mutation type M1 heteroduplex; (C) KIM5-mutation type M2 heteroduplex; (D) KIM5-mutation type M3 heteroduplex; (E) KIM5-mutation type M4 heteroduplex; (F) KIM5-mutation type M5 heteroduplex.

FIG. 1.

DHPLC peak profiles obtained from Y. pestis strain KIM5 and ciprofloxacin-resistant mutants of KIM5. All peak profiles depicted are from the first lot of samples and were generated from the same column, which was the column used to generate the peak profiles depicted in Fig. 2. (A) KIM5 homoduplex; (B) KIM5-mutation type M1 heteroduplex; (C) KIM5-mutation type M2 heteroduplex; (D) KIM5-mutation type M3 heteroduplex; (E) KIM5-mutation type M4 heteroduplex; (F) KIM5-mutation type M5 heteroduplex.

FIG. 2.

DHPLC peak profiles of QRDR mutations generated by site-directed mutagenesis. All peak profiles depicted are from the first lot of samples and were generated from the same column, which was the column used to generate the peak profiles depicted in Fig. 1. (A) KIM5-mutation type M6 heteroduplex; (B) overlay of KIM5-M4 mutation type heteroduplex (tan) and KIM5-M6 mutation type heteroduplex (red); (C) KIM5-M7 mutation type heteroduplex; (D) KIM5 homoduplex (right), KIM5-M8 mutation type heteroduplex (left) and KIM5-M9 mutation type heteroduplex (middle).

FIG. 2.

DHPLC peak profiles of QRDR mutations generated by site-directed mutagenesis. All peak profiles depicted are from the first lot of samples and were generated from the same column, which was the column used to generate the peak profiles depicted in Fig. 1. (A) KIM5-mutation type M6 heteroduplex; (B) overlay of KIM5-M4 mutation type heteroduplex (tan) and KIM5-M6 mutation type heteroduplex (red); (C) KIM5-M7 mutation type heteroduplex; (D) KIM5 homoduplex (right), KIM5-M8 mutation type heteroduplex (left) and KIM5-M9 mutation type heteroduplex (middle).

FIG. 3.

Blind panel DHPLC peak profiles. All peak profiles depicted are from the first lot and second lot of samples and were generated from a column different from the column used for Fig. 1 and 2. (A) Overlay of blind panel KIM5-M1 mutation type reference heteroduplex (purple) and blind panel sample 6 (KIM5-M1 mutation type heteroduplex) (blue); (B) overlay of blind panel KIM5-M1 mutation type reference heteroduplex (purple) and blind panel sample 28 (KIM5-M2 mutation type heteroduplex) (blue); (C) overlay of blind panel KIM5-M8 mutation type reference heteroduplex (green) and blind panel sample 7 (KIM5-M9 mutation type heteroduplex) (gray); (D) overlay of KIM5-M8 mutation type reference heteroduplex (green) and blind panel sample 9 (KIM5-M4 mutation type heteroduplex) (black).

FIG. 3.

Blind panel DHPLC peak profiles. All peak profiles depicted are from the first lot and second lot of samples and were generated from a column different from the column used for Fig. 1 and 2. (A) Overlay of blind panel KIM5-M1 mutation type reference heteroduplex (purple) and blind panel sample 6 (KIM5-M1 mutation type heteroduplex) (blue); (B) overlay of blind panel KIM5-M1 mutation type reference heteroduplex (purple) and blind panel sample 28 (KIM5-M2 mutation type heteroduplex) (blue); (C) overlay of blind panel KIM5-M8 mutation type reference heteroduplex (green) and blind panel sample 7 (KIM5-M9 mutation type heteroduplex) (gray); (D) overlay of KIM5-M8 mutation type reference heteroduplex (green) and blind panel sample 9 (KIM5-M4 mutation type heteroduplex) (black).