Global Study of IS6110 in a Successful Mycobacterium tuberculosis Strain: Clues for Deciphering Its Behavior and for Its Rapid Detection

Abstract

The Mycobacterium tuberculosis insertion sequence IS6110, besides being a very useful tool in molecular epidemiology, seems to have an impact on the biology of bacilli. In the present work, we mapped the 12 points of insertion of IS6110 in the genome of a successful strain named M. tuberculosis Zaragoza (which has been referred to as the MTZ strain). This strain, belonging to principal genetic group 3, caused a large unsuspected tuberculosis outbreak involving 85 patients in Zaragoza, Spain, in 2001 to 2004. The mapping of the points of insertion of IS6110 in the genome of the Zaragoza strain offers clues for a better understanding of the adaptability and virulence of M. tuberculosis. Surprisingly, the presence of one copy of IS6110 was found in Rv2286c, as was recently described for a successful Beijing sublineage. As a result of this analysis, a rapid method for detecting this particular M. tuberculosis strain has been designed.

INTRODUCTION

The airborne spread and long incubation period of tuberculosis (TB) make it difficult to prevent the transmission of the disease, which affects about 8 million people each year. Efficient molecular methods allow the genetic typing of Mycobacterium tuberculosis. In a population-based molecular study conducted in Zaragoza, Spain, from 2001 to 2004, the occurrence of a large TB outbreak involving 85 patients was detected. A change in the epidemiological pattern of TB was observed, with a single epidemic strain, named M. tuberculosis Zaragoza, causing 18.7% of all TB cases (1). The spoligotyping pattern SIT-773 presented by this strain was described in 5 cases geographically distributed in the United States and the Netherlands and was reported as an unclassified pattern in the updated database SITVITWEB (http://www.pasteur-guadeloupe.fr:8081/SITVIT_ONLINE) (2). Some strains either are prevalent in different geographic areas or are more successfully transmitted in the population, as has been described for the Beijing or Haarlem family (3, 4). Relative to isolates of other genotypes, an M. tuberculosis Zaragoza isolate achieved the highest fitness ranking in the lungs of low-dose-aerosol-infected mice (5). However, the reasons for the dominance and spread of the Zaragoza strain in our population remain unclear.

The M. tuberculosis insertion sequence IS6110 was first described in 1990 (6, 7). The IS6110 sequence in the genome of M. tuberculosis has shown the stability required for use in molecular epidemiology. In some cases, when a specific location is detected, it has been used as a target to design a multiplex PCR method that allows rapid strain identification. For example, the IS6110 located within the designated region NTF-1 (Rv3128c to Rv3129), Rv2180c, and Rv2664 to Rv2665 allowed identification of the W-multidrug-resistant Beijing strain, the GC1237 Beijing strain, and clone B0/W148 Beijing, respectively, to better control the transmission of specific epidemic strains (8–10).

Moreover, IS6110 induces loss of gene activity either by mediating deletion events (11–16) or by disrupting coding sequences and regulatory domains (17, 18). The IS6110 element itself modulates the expression of neighboring genes by acting as a promoter sequence, driving or enhancing their expression (19–22). All of these changes seem to have effects on the molecular structure of the mycobacterial genome, modifying the biology and fitness of the bacilli.

In the present work, we have characterized the successful Zaragoza strain and localized all IS6110 copies in its genome. The study offers clues for better understanding the adaptability and virulence of M. tuberculosis and, in addition, has allowed us to design an accurate multiplex PCR assay that may be used for rapid detection and surveillance of this variant of M. tuberculosis.

MATERIALS AND METHODS

Characterization of the Zaragoza strain.

The presence or absence of the TbD1 region was assessed by PCR amplification using the primers TBD1fla1-F and TBD1fla1-R (Table 1). TbD1 deletion results in an amplicon size of 400 bp, whereas the presence of the TbD1 region results in an amplicon size of 2,153 bp. Three single nucleotide polymorphisms (SNPs) described as markers of the M. tuberculosis Haarlem genotype strains, i.e., mgtC¹⁸²(Arg-His) (23), ogt⁴⁴(Thr-Ser), and ung⁵⁰¹(Leu-Leu) (24), were studied by PCR and DNA sequencing using the primers described in Table 1.

Table 1.

Primers used in this study

Gene(s)	Primer	Sequence (5′ to 3′)	Reference or source
TbD1	TBD1fla1-F	CTA CCT CAT CTT CCG GTC CA	Brosch et al. (16)
	TBD1fla1-R	CAT AGA TCC CGG ACA TGG TG	Brosch et al. (16)
mgtC	mgtC-F	CGC CTA GGC TCA AAC TGC TG	Alix et al. (23)
	mgtC-R	CAA TAC CCG GCG GAT CTA CC	Alix et al. (23)
ogt	ogt-F	CCC AGC ACC TGT GGA CCA	Olano et al. (24)
	ogt-R	ACT CAG CCG CTC GCG AGC	Olano et al. (24)
ung	ung-F	GCT GGC AAT CGT TTG G	Olano et al. (24)
	ung-R	GGC AAC AAG AAG CGA CTC	Olano et al. (24)
	Salgd	TAG CTT ATT CCT CAA GGC ACG AGC	Prod'hom et al. (25)
	Salpt	TCG AGC TCG TGC	Prod'hom et al. (25)
	ISA1	CCT GAC ATG ACC CCA TCC TTT CC	Mendiola et al. (40)
	ISA3	GAG GCT GCC TAC TAC GCT CAA CG	Mendiola et al. (40)
Rv0755A, thrV	755-F	ATC TTG ATC GCA TCG GAA GC	This work
	755-R	AAC CCG CTG AGC AGC ATC GC	This work
Rv0794c, Rv0797	795-F	CCT ATC GCC CAC CAG ACG C	This work
	795-R	ACA CCG TGC GCC TCT ACC TGC	This work
Rv1668c, Rv1869	1668-F2	ACC CGC AAT CCG CTA CGC	This work
	1668-R2	AGC ACG AAT CAT GGA CTC GGC	This work
Rv2286c, yjcE	2286-F	AGC TCA TGA CCG ATC GCT GC	This work
	2286-R	ACC GCC GCG TTG ACC TGT TTC G	This work
Rv2349c (plcC)	2349-F	AAT GCC GAC GTC ATA TCG CC	Vera-Cabrera et al. (34)
	2349-R	ACG TCA TCA ACA ACA CGC TGC	Vera-Cabrera et al. (34)
Rv2813, Rv2814 (DR region)	DR43-F	ACC CGG TGC GAT TCT GCG	This work
	DR22-R	AGA CGG CAC GAT TGA GAC	This work
Rv2823c	2823-F	AAG AGC TGT GCG GTC AGG	This work
	2823-R	AAG GTG ATC GAG GAG AAG TAC CGG C	This work
Rv3229c (desA3)	3229-F	AGC ATC TGC CGT AGG TAC CAC	This work
	3229-R	AAC ACC ATC CTT GCG ATC G	This work
MT1802, MT1814 (RvD2, RvD3)	Cut1-F	ACG TAT CCG AGA GTG TGA C	This work
	RvD3-R	CAA TCA TGC TGT TTG GC	This work
PPE38	PPE38-F	AAG TGC GGA TTT TCG GTG TGG	This work
	PPE38-R	ATC AAA CGA GGA CGC CGA GG	This work
MT2423 (RvD4)	RvD4-F2	GTT GCC GTT GTT TCC GTT ACC	This work
	RvD4-R2	GGT GGT GGT GGC GGC T	This work
MT3429, MT3430 (RvD5)	MT3429-F	GCA ATC AGA ACG TCG GTG T	This work
	RvD5-R	GTA CCC GCA CCA CCT GCT	This work
gyrB	MTUB-F	TCG GAC GCG TAT GCG ATA TC	Herrera-León et al. (27)
	MTUB-R	ACA TAC AGT TCG GAC TTG CG	Herrera-León et al. (27)
Rv2823c	MTZ-RF	CGG GGC GGT TCA TTG GTG	This work
	2823c-R2	AGG TGA TCG AGG AGA AGT ACC G	This work

Automated DNA sequencing was performed by the DNA sequencing support service of the Spanish National Cancer Research Center (Madrid, Spain), by capillary electrophoresis with an Applied Biosystems 3730 DNA analyzer using the ABI Prism BigDye Terminator system (Applied Biosystems). The sequence generated was aligned with the sequence of M. tuberculosis H37Rv (http://genolist.pasteur.fr/TubercuList).

Mapping of IS6110.

We used two different methodologies to identify the points of insertion of IS6110 in the genome of the M. tuberculosis Zaragoza strain. Ligation-mediated PCR, as described previously by Prod'hom et al. (25), was used to amplify the flanking sequences of the IS6110 of the DNA obtained from one isolate in 2004, considered the reference for the Zaragoza strain, and Mycobacterium bovis BCG, as a control. Briefly, the total genomic DNA was restricted with SalI (Boehringer Mannheim) and ligated to a linker containing a SalI restriction site. The oligonucleotides used as linkers and primers are indicated in Table 1. SalI was then used to digest the resulting template. This strategy was also applied to digested fragments of DNA separated by electrophoresis, in order to gain effective resolution with the technique. DNA was recovered in seven different size ranges through overnight electrophoresis in low-melting-point agarose gels (Bio-Rad), following the protocol described by the manufacturer (Fermentas Life Science). For PCR amplification, two sets of primers were used, (i) Salgd and ISA1, which amplified the DNA sequence next to the 5′ end of the IS6110, and (ii) Salgd and ISA3, which amplified the DNA sequence next to the 3′ end of the IS6110 (Table 1). The PCR products obtained were purified using the GFX PCR DNA gel band purification kit (GE Healthcare), to be sequenced as described above. The sequence generated was aligned with the sequences of M. tuberculosis H37Rv (http://genolist.pasteur.fr/TubercuList) and CDC1551 (http://blast.ncbi.nlm.nih.gov).

In a second approach, four isolates of the Zaragoza strain, including the 2004 reference isolate, were analyzed in the context of a high-throughput survey of in vivo IS6110 transposition in multiple M. tuberculosis genomes for the simultaneous identification of points of insertion of IS6110 with Illumina sequencing technology. This new methodology was described by Reyes et al. (26).

Based on the results obtained with the two methods, new primers were designed to verify the points of insertion of IS6110 (Table 1). Most of the expected PCR products were in the range of 1,500 to 2,000 bp if the IS6110 was present and in the range of 300 to 600 bp if the IS6110 was absent from the site of amplification. The PCR products were sequenced and aligned with the M. tuberculosis H37Rv genome with BLAST.

Design of a rapid method for detection of the Zaragoza strain.

The points of insertion of IS6110 detected in the Zaragoza strain that had not been previously reported as preferential loci for IS6110 transposition were analyzed in 100 samples from a collection of isolates in the University of Zaragoza that had been isolated between 2005 and 2008. This collection included different strains selected on the basis of their diverse IS6110 restriction fragment length polymorphism (RFLP) patterns. Separate PCRs were performed and, in cases in which the IS6110 appeared in non-Zaragoza isolates, another point of insertion was tested (Table 1). Four isolates of the Zaragoza strain were included as positive controls. The copy inserted at point 3129520/25 of the genome of strain H37Rv (GenBank accession no. NC_000962.2), in the Rv2823c gene, was selected as the specific polymorphism for detection of the Zaragoza strain.

A multiplex PCR detecting two different targets was designed for rapid identification of the M. tuberculosis Zaragoza genotype. The first target was the specific IS6110 in Rv2823c in the Zaragoza strain. For this purpose, the primer MTZ-RF, annealing to the end of IS6110 and the following sequence of the genome, and Rv2823c-R2 were designed. The second target was the gyrB gene, specific for M. tuberculosis complex strains and amplified with primers MTUB-F and MTUB-R, as described previously (27, 28) (Table 1). Multiplex PCR was performed in a total volume of 25 μl, containing 200 μM each deoxynucleoside triphosphate in 10 mM Tris-HCl, 50 mM KCl, 1.5 mM MgCl₂, 2.5 U of DNA polymerase (puReTaq Ready-to-go; GE Healthcare) in reaction buffer, 25 μM each primer, and 2 μl of template DNA. Temperature cycling conditions included 95°C for 5 min, 35 cycles of 94°C for 30 s, 65°C for 30 s, and 72°C for 90 s, and a final extension at 72°C for 10 min. The PCR products were analyzed by electrophoresis on 1.2% agarose gels. The molecular sizes of the multiplex PCR products were estimated by comparison with the migration distances of the DNA bands of a 100-bp DNA ladder. The Zaragoza strain was expected to produce two amplicons, i.e., 1,039-bp (gyrB) and 120-bp (IS6110 in Rv2823c specific for the Zaragoza strain) products. A non-Zaragoza M. tuberculosis complex isolate was expected to produce only the 1,039-bp product. A non-M. tuberculosis complex strain was expected to produce no amplicon (Fig. 1).

Fig 1.

(A) Results of multiplex PCR visualized in agarose gel electrophoresis. Lane 1, molecular size marker (100-bp DNA ladder); lane 2, the M. tuberculosis Zaragoza strain (1,039 bp and 120 bp); lane 3, H37Rv strain (1,039 bp); lane 4, negative PCR control; lanes 5, 6, 7, and 8, different clinical isolates of M. tuberculosis. (B and C) Schematic representations of the targets used for the diagnostic test, i.e., the gyrB region in the M. tuberculosis complex (B) and the position and orientation of the IS6110 insertion in the Rv2823c region in the Zaragoza strain (C). White arrow, IS6110 element; arrowhead, direction of transcription of the putative transposase; small black arrows, primers.

Verification of multiplex PCR for detection of the M. tuberculosis Zaragoza strain.

In order to probe the specificity of the multiplex PCR assay, we tested 122 positive stored cultures of M. tuberculosis isolated in the Hospital Universitario Miguel Servet in 2000. The strains were defrosted and recovered. The extraction of genomic DNA was done in two different ways. For 68 of the 122 isolates, DNA was extracted from solid cultures by the chemical lysis method (29). For the remaining 54 isolates, rapid DNA extraction was performed directly from the glycerol-stored samples. An aliquot of 200 μl was spin dried at 1,000 rpm for 15 min, and the supernatant was removed; 50 μl of Tris-EDTA (TE) buffer was added, and the mixture was heated at 100°C for 20 min. The sample was then spin dried at 1,000 rpm for 5 min, and the supernatant was used to amplify DNA by the multiplex PCR. In addition, these strains were analyzed by spoligotyping with a commercially available kit, according to the manufacturer's instructions (Ocimum Biosolutions Ltd., Hyderabad, India), as described by Kamerbeek et al. (30).

RESULTS

Classification of the Zaragoza strain.

The TbD1 region was absent from the Zaragoza strain, and the mgtC¹⁸²(Arg-His), ogt⁴⁴(Thr-Ser), and ung⁵⁰¹(Leu-Leu) SNPs characteristic of the Haarlem genotype were not present, which allowed us to classify Zaragoza as a non-Haarlem but “modern” strain of M. tuberculosis.

Localization of the points of insertion of IS6110.

The application of two different methodologies allowed us to identify the 12 points of insertion of IS6110 in the M. tuberculosis Zaragoza genome (Table 2). Eight IS6110 sites were localized by both methodologies, and four additional IS6110 sites were found by the Illumina sequencing technology. The 12 insertion points were analyzed by PCR to confirm the sites of insertion, the presence or absence of direct repeat (DR) nucleotides at these points, and the IS6110 orientation. Eleven of the 12 copies generated direct repeat sequences of 3, 4, 6, and 8 bp next to the insertion site.

Table 2.

Distribution and characteristics of the 12 points of insertion of the copies of IS6110 found in the M. tuberculosis Zaragoza strain^a

Gene(s)	Insertion point position in the Zaragoza strain	No. of positive isolates/no. testedb	Testing	Orientation	DR	Intergenic	Distance to genec
Rv0755A–thrV	850536–850539	6/30	Hot spot for strains analyzed in our control study	Direct	TGTC	Yes	13 bp to Rv0755A
							103 bp to thrV
Rv0794c–Rv0797	889029–890376		Hot spot (17, 31); insertion similar to H37Rv	Direct	GAGG	Yes	393 bp to Rv0794c
							12 bp to Rv0797
Rv1668c–Rv1669	1895651–1895654			Inverse	TAGG	Yes	308 bp to Rv1668c
							75 bp to Rv1669
MT1802–MT1814	1985524–1995905			Direct	−d	No	471 bp to MT1802
							93 bp to MT1814.2
Rv2286c	2559568–2559570	0/30		Inverse	ATC	No	758 bp to Rv2285
							135 bp to yjcE
Rv2349c (plcC)	2627492–2627494	0/30	Hot spot (48)	Direct	TCA	No	512 bp to Rv2348c
							1,289 bp to plcB
PPE38	2633977–2633979			Inverse	GCG	No	1,902 bp to plcA
							551 bp to PPE39
MT2423	2633461–2633464			Inverse	TTTC	No	1,382 bp to MT2422
							970 bp to MT2423.1
Rv2813–Rv2816c	3119660–3121887		Hot spot (17, 37)	Direct	GTCTGACG	Yes	624 bp to Rv2813
							1,738 bp to Rv2816c
Rv2823c	3129520–3129525	0/100		Inverse	TTGTGA	No	173 bp to Rv2822c
							2,250 bp to Rv2824c
Rv3229c (desA3)	3606308–3606310	0/30		Inverse	ACG	No	590 bp to Rv3228
							804 bp to Rv3230c
MT3429–MT3430	3709089–3709091			Direct	GCC	Yes	18 bp to MT3429
							197 bp to MT3430

^aResults of the analysis to detect the specificity of the IS6110 localization in 100 different clinical isolates are shown. The orientation of each IS6110 copy with regard to the genome sequence was direct or inverse. The direct repeat nucleotide duplications generated at the points of insertion of IS6110 and the distances to neighboring genes are detailed. The numbering was made with respect to the H37Rv sequence; when an insertion locus could not be mapped to H37Rv, it was referred to CDC1551.

^bDifferent clinical isolates analyzed for detection of the specificity of the localization of IS6110.

^cDistance from IS6110 to neighboring genes.

^d−, absence of direct repeat.

There were six intragenic insertions, localized in Rv2286c, plcC (Rv2349c), PPE38 (Rv2352c), Rv2823c, desA3 (Rv3229c), and MT2423 (RvD4 region). Five IS6110 insertions were localized in different intergenic regions, at positions 393, 308, 13, 18, and 594 bp upstream of the RV0794c, Rv1668c, Rv0755A, MT3429 (RvD5 region), and Rv2813 genes, respectively (Table 2). The 12th IS6110 site was located between MT1802 and MT1814 (RvD2 and RvD3 regions).

A deletion was found next to the point of insertion of a copy of IS6110 in two cases; the first was in the highly conserved 36-bp DR region within a DR (direct copy) sequence of 36 bp. There was a duplication of an 8-bp direct repeat flanking this insertion point. The analysis of the DR region indicated an 863-bp deleted region, which included the 25th to 40th interspaces (see Fig. 2 for a detailed representation of this region). The second deletion was from the RvD2 to RvD3 regions of H37Rv, which resulted in the absence of the cut1, wag22, Rv1760, Rv1761c, Rv1762c, Rv1763, Rv1764, and Rv1765c genes. The 5′ end of the IS6110 was inserted into the cut1 gene, at the same insertion point as in the CDC1551 strain but in a different orientation. This region is included in the RD152 region characteristic of the Beijing lineage.

Fig 2.

Map of the DR region of the M. tuberculosis Zaragoza strain at the point of insertion of IS6110. The numbering refers to the H37Rv genome. Black arrow, IS6110 sequence; black bars, DRs; white arrows, spacer sequences (numbered with reference to the spoligotyping results); gray shading, eight direct repeat nucleotides at the insertion point.

Selection of the Rv2823c gene as the specific target for rapid detection.

We studied the 12 insertion points in order to select a specific target to develop a method for rapid detection of M. tuberculosis Zaragoza isolates. Three locations were ruled out as they were previously described as preferential loci for IS6110 transposition, i.e., Rv0794c to Rv0797, plcC (Rv2349c), and the site in the DR region (31–34). Thus, we investigated the presence of IS6110 in the other locations (Rv0755A, Rv2286c, Rv2823c, and desA3 [Rv3229c]) in 30 isolates. Six of the isolates tested presented the IS6110 in the same nucleotide position in the Rv0755A-thrV(tRNA-Thr) intergenic region as in the Zaragoza strain. Further, 100 isolates were tested for the Rv2823c point of insertion (position 3129525 in H37Rv). All isolates showed the same size of amplified product as the control H37Rv, indicating that IS6110 was not present (Fig. 1). Subsequently, this insertion was selected to design a rapid diagnostic test. The results are shown in Table 2. In addition, the 12 IS6110 sites were present in four M. tuberculosis Zaragoza isolates used as positive controls, as expected.

Validation of multiplex PCR as a rapid diagnostic test for the Zaragoza strain.

We analyzed 122 clinical M. tuberculosis complex isolates from 2000 by multiplex PCR and spoligotyping. The results for the detection of Zaragoza isolates by the two techniques were concordant for a total of 26 samples (Table 3). In all Zaragoza isolates, the 120-bp Zaragoza-specific amplicon was produced. In only nine Zaragoza isolates, for which DNA was extracted directly from glycerol-stored samples and might have been insufficient or degraded, the 1,039-bp M. tuberculosis complex-specific amplicon of the gyrB gene was not obtained (Table 3). This rapid multiplex PCR test achieved sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV) of 100% in identifying Zaragoza isolates.

Table 3.

Results of the analysis of 122 isolates from 2000 by multiplex PCR rapid testing for detection of the M. tuberculosis Zaragoza strain and by spoligotyping

Extraction and testinga	No. of positive isolates of:
	M. tuberculosis	Zaragoza strain
Extraction of genomic DNA
Multiplex PCR of the Zaragoza strain	68	18
Spoligotyping	68	18
Rapid DNA extraction
Multiplex PCR of the Zaragoza strain	45b	8c
Spoligotyping	54	8

^aThe two tests achieved sensitivities, specificities, PPVs, and NPVs of 100% in identifying M. tuberculosis Zaragoza isolates.

^bThe absence of the gyrB amplicon in the rapid extraction samples could be due to degradation or insufficient amounts of DNA.

^cThe 1,039-bp amplicon specific for the M. tuberculosis complex was not produced in one Zaragoza isolate.

DISCUSSION

A large unsuspected TB outbreak involving 85 patients in Zaragoza, Spain, in 2001 to 2004 was attributed to the M. tuberculosis Zaragoza strain (35). The lack of the TbD1 region described here confirms that Zaragoza is a modern strain, in agreement with our previous work showing that the strain belongs to principal genetic group 3 (PGG-3) (35). Other authors found that isolates belonging to PGG-1 and PGG-2 are more frequently associated with clustered cases of TB (36), whereas the Zaragoza strain (fitted in PGG-3) was responsible for a widespread outbreak. In an attempt to better understand the Zaragoza strain, we proposed to determine the localization of all of the IS6110 integration sites in its genome. Depending on the position of integration, different alterations in the bacteria can be produced, ranging from lethality to improved fitness (20–22).

Transposition is the more common mechanism used by IS6110 to move along genomes, which leads to the generation of 3- to 4-bp direct repeats immediately flanking the IS6110 sequences (7). The direct repeat nucleotides flanking the insertion site were present in 11 of 12 copies in the strain. As far as we know, this is the first time that 6-bp and 8-bp direct repeat sequences have been described.

Similar to most M. tuberculosis complex strains, the Zaragoza strain contains a copy of IS6110 in the DR region, which is the clustered regularly interspaced short palindromic repeat (CRISPR) locus of M. tuberculosis (33, 37). Unlike H37Rv, M. bovis BCG, and M. bovis AF2122/97, the Zaragoza strain carries IS6110 in this locus inserted at a different point, 3 nucleotides from the end of a direct repeat sequence. The analysis of the DR region indicated an 863-bp deletion that is reflected in its discriminative spoligotype, which lacks amplified spacers 24 to 40 (35). Since 8-nucleotide repeats flanking the insertion were detected, it seems that this copy was the consequence of transposition; therefore, we cannot determine whether IS6110 is implicated in the process of deletion of the adjacent region (Fig. 2).

Genomic regions absent in the M. tuberculosis H37Rv genome named RvD regions were described, and some of them were apparently created by homologous recombination of two adjacent copies of IS6110 (38). Only one copy of IS6110 was not flanked by direct repeats in the M. tuberculosis Zaragoza strain. In this region, a rearrangement between two genomic regions absent in the M. tuberculosis H37Rv genome, named RvD2 and RvD3 (MT1802 to MT1814), was detected. This could be interpreted as a recombination process involving neighboring copies of IS6110 elements, which resulted in the loss of wag22, Rv1760, Rv1761c, Rv1762c, Rv1763, and Rv1764 and interruption of the cut1 gene, strongly suggesting that these genes are not essential for M. tuberculosis. This IS6110-mediated deletion was included in the RD152 region characteristic of the Beijing lineage (39). This result is in agreement with other studies indicating that strains with greater numbers of IS6110 copies have lost genomic regions (16, 22). In contrast, there is no description of these events in strains with only few copies, as all IS6110 copies are flanked by direct repeats (33, 40).

According to the distance and the orientation of IS6110 (21), IS6110 could have a transcriptional influence on neighboring genes. The possibility of such an influence on Rv0797, MT1814.2, Rv1668c, Rv2823c, and MT3430 should be analyzed.

Five of the 12 insertion points were located in intergenic regions representing 41.6% of the known sites, i.e., the DR region, Rv0755A to thrV, Rv0794c to Rv0797, Rv1668c to Rv1669, and MT3429 to MT3430 (RvD5 region). According to other authors, these results suggest that transposition is relatively more frequent in noncoding sequences of the genome (17, 18, 33).

According to the intragenic points of insertion of the IS6110, we could assume that Rv2286c, plcC (Rv2349c), PPE38 (Rv2352c), Rv2823c, desA3 (Rv3229c), and MT2423, coding for a PPE family protein (RvD4 region), are naturally inactivated in the Zaragoza strain. Among them, at least three copies of IS6110 interrupt different described operons, i.e., Rv2349c (plcB and plcC), Rv2823c (Rv2824c and Rv2819c), and Rv3229c (Rv3230c and Rv3229c) (41–43), and the potential transcriptional influence on these operons is an interesting point to be studied.

The insertion in the desA3 gene caught our attention, as it is thought to be involved in lipid metabolism. Its product is a stearoyl-coenzyme A desaturase (DesA3) implicated in the synthesis of oleic acid and considered to be essential for intracellular growth in macrophages (44, 45). In addition, DesA3 is a target of the second-line antituberculosis drug isoxyl, which was used together with isoniazid in multiple-drug therapy in the 1960s (43, 44). Isoxyl inhibits the synthesis of oleic and mycolic acids (43) and was demonstrated to be effective against various clinical isolates of multidrug-resistant strains of M. tuberculosis (46). The finding of the insertion of IS6110 in the desA3 (Rv3229c) gene suggests that isoxyl would not be effective against the Zaragoza strain. However, our preliminary studies showed that the Zaragoza strain is susceptible (data not shown), suggesting that multiple targets are used by isoxyl. More studies should be carried out to elucidate this process.

The plcABC locus, described as a probable virulence factor (47), is known as a preferential site for IS6110 transposition (48). Confirming this finding, the Zaragoza strain has three IS6110 copies, in plcC (Rv2349c), PPE38 (Rv2352c), and MT2423 (RvD4 region), within a region of around 10 kb. Rv2286c and Rv2823c are conserved hypothetical proteins with unknown functions (http://genolist.pasteur.fr/TubercuList). Interestingly, one IS6110 copy was found to be interrupting the Rv2286c gene in a recent study of a highly transmitted Beijing strain (22). IS6110 had the same orientation in the Zaragoza strain and the Beijing strain just mentioned, but the insertion points differed by 64 nucleotides (position 2559504 in the Beijing strain versus position 2559568 in the Zaragoza strain). The possibility exists that this copy of IS6110 could be affecting regulation of the transcription of the adjacent gene yjcE (Rv2287). The IS6110 location in the Rv2823c gene was specific for the M. tuberculosis Zaragoza strain, and the region was chosen to design the multiplex PCR for identification of the strain.

The use of whole-genome sequencing with last-generation systems shows difficulties in localizing the repeated sequences in the genome, such as IS6110 sequences in the M. tuberculosis complex genome. Nevertheless, knowledge of the specific locations of these insertion sequences has allowed the design of rapid techniques to identify specific genotypes of the W Beijing strain (8–10). Another objective of this study was to develop a rapid diagnostic method based on a specific point of insertion of IS6110 to identify M. tuberculosis Zaragoza isolates. The specific target chosen, IS6110 inserted in Rv2328c, was not found in any other isolates tested. We could verify its specificity in a study that analyzed the insertion points of IS6110 in 570 strains (data not shown), although another point of insertion in the Rv2823c gene was described for 9 strains belonging to the Latin American-Mediterranean 5 family (26). During the selection of the point of insertion of IS6110 to be used as the target, the Rv0755A-thrV(tRNA-Thr) intergenic region was found in the same nucleotide position in 6 other isolates. Surprisingly, these 6 different isolates were classified as family T by their spoligotyping patterns, which raises the possibility that these strains are epidemiologically or evolutionally related.

Application of the diagnostic test should make possible the rapid identification of patients who have been infected with the Zaragoza strain within a brief period. The multiplex PCR developed as part of this study could identify 26 Zaragoza isolates (21.3%) among the 122 isolates from 2000 that were analyzed, showing that the outbreak was already ongoing at that time. All of those isolates demonstrated amplification of the specific target of the M. tuberculosis Zaragoza genotype. We confirmed these results by comparison with spoligotyping results, supporting the specificity of the test. Currently, spoligotyping and IS6110 RFLP assays are performed routinely for all M. tuberculosis complex isolates in the region; therefore, the incidence of this strain is being monitored, which allows us to verify that spoligotype SIT-773 has remained specific for the Zaragoza strain in our population to 2012 (data not shown). This multiplex PCR is a quick easy technique with high specificity, which can be used without delay for a single sample in the laboratory.

We conclude that IS6110 transposition in the Zaragoza strain may be a driving force in adaptation to the human host, especially for sites that do not correspond to hot spots for other M. tuberculosis strains, including interruption of the desA3 gene (Rv3229c), which is considered a target for new drugs against tuberculosis. The similarity of the Zaragoza strain to Beijing lineage strains with respect to the deletion between the RvD2 and RvD3 regions and the insertion of IS6110 in Rv2286c, which was described for the highly transmitted Beijing strain although at a different insertion point, could provide a clue for the success of this strain in our population. Finally, the specific site found in Rv2823c, which was not present in any other strain tested, has allowed us to use it for identification of the Zaragoza strain.