Rapid Genotyping of Shigella sonnei by Use of Multiplex High-Resolution Melting

LETTER

Shigella sonnei, an emerging global cause of shigellosis, consists of four distinct lineages, and the current pandemic involves several geographically associated, multidrug-resistant clones that belong to lineage III (1–3). A typing scheme based on high-resolution melting (HRM) of six chromosomal single nucleotide polymorphisms (SNPs) has been described to identify all lineages/sublineages (4). HRM detects SNPs in small DNA fragments based on the melting patterns (5, 6). HRM involves minimal labor, with a rapid turnaround time and a decreased risk of PCR carryover contamination (7). A sensitive intercalating fluorescent dye, EvaGreen, may yield robust fluorescence intensities and strong sharp melting peaks upon melting of the PCR amplicons (8–10). Based on the SNPs summarized in Table 1, we have developed a multiplex HRM assay for identification of lineages I, II, and III using the EvaGreen dye and a Rotor-Gene 6000 instrument (Qiagen).

TABLE 1.

SNPs used for multiplex-HRM typing of S. sonnei in this study

Gene	SNP positiona	Nucleotide in:
		Ancestor	Derived sequence	Lineage I	Lineage II	Sublineage IIIa	Sublineage IIIb	Sublineage IIIc
kduD	3155111	T	C	C	T	T	T	T
deoA	4803842	G	A	G	A	G	G	G
emrA	2973337	T	C	C	C	T	T	T
fdx	2760031	G	A	G	G	G	A	A
menF	2455693	A	G	A	A	A	A	G

^aThese SNPs were extracted using S. sonnei strain Ss046 as the reference as described by Holt et al. (1) and Sangal et al. (4).

Ten strains (2 for each lineage/sublineage) were from a collection held at the Institut Pasteur. Only one lineage IV strain is reported so far (1), which is not included in the study. Primers were designed to amplify fragments that contain three SNPs within kduD, deoA, and emrA for typing 3 main lineages in a first set run and to amplify fragments that contain SNPs within fdX and menF for typing sublineages IIIa, IIIb, and IIIc in a second run. Primer sequences are summarized in Table 2.

TABLE 2.

Primers used for multiplex-HRM typing of S. sonnei in this study

Target of primers	Direction	Sequence
kduD	Forward	5′-CGACGGCGAAACACTTTATC-3′
	Reverse	5′-CGCGTATAAGAAGGCACACG-3′
deoAD	Forward	5′-GGAGATGCTTATCTCCGGCAAA-3′
	Reverse	5′-AGTCGGTTGGGCCTTTT-3′
emrA	Forward	5′-TGCCACCGAAGTACGTAACG-3′
	Reverse	5′-CATCCACCCACATATTGGTG-3′
fdx	Forward	5′-CAAAGCCTGGGACTGGA-3′
	Reverse	5′-CATGGTTGATAGTGTAACGC-3′
menF	Forward	5′-TATTCTCGCGCTGGTTTTTA-3′
	Reverse	5′-GCTTTTCTTGGCTCTTCACC-3′

In the first run, we set up a 20-μl multiplex real-time PCR mixture containing 10 μl SensiMix with EvaGreen dye (Bioline), 20 ng of template DNA, and 2.5 μl of each primer of kduD (100 μM), 1 μl of each primer of deoA (100 μM), and 1 μl of each primer of emrA (50 μM). In the second set run, the 20-μl reaction mix was the same, except that the primers used were for fdX (10 μM) and menF (10 μM), with a quantity of 2 μl and 1 μl, respectively. The PCR was programmed as follows: the mixture was held at 95°C for 10 min and then subjected to 35 cycles of 95°C for 20 s, 60°C for 20 s, and 72°C for 22 s. HRM was performed between 80 and 90°C, with a stepwise increase of 0.05°C. The melting temperatures (T_m) were determined by the negative derivative of the decrease in fluorescence over the increase in temperature (df/dt) for each SNP using the proprietary software (version 1.7.34).

The first set run clearly separated lineages I, II, and III with distinctive melting curves, and the T_m of each allele was at least a half degree apart from that of the other alleles (Fig. 1A). The second set run distinguished the sublineages IIIa, IIIb, and IIIc with distinctive melting curves (Fig. 1B). Table 3 summarizes T_m for all alleles. To assess the reproducibility, we blindly ran all 10 samples of various lineages randomly under the same conditions, which revealed the same results with a ±0.05°C difference in T_m. Sequencing of five SNPs independently verified the reliability of the T_m data.

FIG 1.

Multiplex-HRM curves of two set runs. (A) Multiplex-HRM curves of three SNPs in kduD, deoA, and emrA, respectively (from left to right), which separates lineages I, II, and III (strains 1263, 55623, and 54184, respectively); (B) multiplex-HRM curves of two SNPs in fdX (left) and menF (right), respectively, which separates sublineages IIIa, IIIb, and IIIc (strains 54184, 2073, and 6224, respectively). deg., °C.

TABLE 3.

Summary of T_m for each of the five SNPs used for multiplex-HMR

Lineage	Tm (°C) (±0.05°C) for:
	kduD	deoA	emrA	fdX	menF
Main lineages
I	83.0	85.1	87.7
II	82.5	84.8	87.7
III	82.5	85.2	87.3
Sublineages
IIIa				82.5	85.7
IIIb				82.5	86.0
IIIc				82.9	85.7

A simple and robust genotyping report is important for rapid identification of the emerging multidrug-resistant clones of lineage III (1). The multiplex-HRM approach that we present here can be very useful in this regard.