IS1634, a Novel Insertion Element Creating Long, Variable-Length Direct Repeats Which Is Specific for Mycoplasma mycoides subsp. mycoides Small-Colony Type

Abstract

A new insertion sequence, IS1634, has been identified in Mycoplasma mycoides subsp. mycoides small-colony type (SC). IS1634 shows structural and functional similarities to IS1549 of Mycobacterium smegmatis and with it seems to form a new class or family of insertion sequences. IS1634 has a size of 1,872 bp, including two 13-bp terminal inverted repeats. It contains an open reading frame (ORF) encoding a product of 533 amino acids which shows similarity to the transposase of IS1549 and to a lesser extent to the transposases of IS elements of the IS4 family. IS1634 is present at about 30 copies in the genome of all 22 different field strains of M. mycoides subsp. mycoides SC tested. Characteristic of IS1634 are the long and variable-length direct repeats at the sites of insertion which were found to reach up to about 500 bp. IS1634 is specific to M. mycoides subsp. mycoides SC and is not present in any of the other members of the M. mycoides cluster. Neither was it found in other closely related Mycoplasma species of ruminants.

Mycoplasma mycoides subsp. mycoides small-colony type (SC) is the infectious agent of contagious bovine pleuropneumonia (CBPP), a severe, highly contagious disease of cattle causing important economic losses and large socioeconomic problems in many countries worldwide. CBPP has reemerged in a few European countries during the last few years, thus threatening the industrialized countries in which the disease was successfully eradicated during the late 1950s (5, 12, 21). Substantial efforts in research on M. mycoides subsp. mycoides SC are required to elaborate efficient methods for diagnosis, prevention, and treatment of CBPP, which are necessary to eradicate this epizootic and prevent it from further spread. Although M. mycoides subsp. mycoides SC is distinguished by its very high pathogenicity from the other members of closely related mycoplasmas of the M. mycoides cluster, which are mostly of only low epidemiological importance, it is phenotypically and antigenically very similar to these. This fact is a major hindrance to the serodiagnosis of CBPP and to detection and identification of M. mycoides subsp. mycoides SC. It has also been shown that, phylogenetically, M. mycoides subsp. mycoides SC can hardly be distinguished from most other members of the M. mycoides cluster, as revealed by sequence analysis of the rrn (16S rRNA) genes and physical genome mapping (14, 16). Moreover, all genes or DNA fragments from M. mycoides subsp. mycoides SC characterized so far have also been found with a variable degree of similarity in other members of the M. mycoides cluster (6–8, 20). A detailed genetic analysis is therefore necessary to discover the main factors that differentiate M. mycoides subsp. mycoides SC from the close relatives of the M. mycoides cluster.

Insertion sequences are transposable elements of about 800 to 2,500 bp (10) which are often present in multiple copies in bacterial genomes. They are able to induce high-frequency chromosomal rearrangements causing phenotypic changes and are commonly used as a genetic tool for the identification of bacterial pathogens. Recently, an insertion element, IS1296, was discovered in M. mycoides subsp. mycoides SC (6). Its insertion profile allowed differentiation between strains isolated from the reemerging outbreaks in Europe since 1980 and strains from the African and Australian continents collected over the last 45 years (4). IS1296 was, however, found at low copy numbers also in a few strains of M. mycoides subsp. mycoides LC and Mycoplasma sp. bovine group 7. We now report the discovery of a novel insertion sequence, named IS1634, which is highly specific for M. mycoides subsp. mycoides SC and produces unusually long and variable-length direct repeats.

Strains, growth conditions, and DNA extraction.

The mycoplasma isolates used in this study are listed in Table 1. Mycoplasmal cultures were made in a standard PPLO broth medium enriched with 20% horse serum, 2.5% yeast extract, and 1% glucose at 37°C until stationary growth phase (1). DNA extractions using guanidium thiocyanate were performed as previously described (4).

TABLE 1.

Mycoplasma strains used

Mycoplasma species	Strain	Collectionb	Origin	Yr isolated	Host	IS1634c
M. mycoides subsp. mycoides SC	PG1	NCTC		1931	Cattle/type strain	+
	2059	LPB	Spain	1984	Cattle/lung	+
	B773/125	LNV	Portugal	1991	Cattle/semen	+
	C305	LNV	Portugal	1993	Goat/lung	+
	O326	LNV	Portugal	1993	Sheep/milk	+
	PO 2	CIRAD	France	1980	Cattle/lung	+
	2022	LPB	France	1984	Cattle/lung	+
	L2	IVBBE	Italy	1993	Cattle/lung	+
	402	LPB	Italy	1990	Cattle/lung	+
	6479	LPB	Italy	1992	Cattle/lung	+
	Afadé	CIRAD	Chad	1968	Cattle	+
	Fatick	CIRAD	Senegal	1968	Cattle	+
	B17	CIRAD	Chad	1967	Zebu	+
	9050-529/1	CIRAD	Ivory Coast	1990	Cattle	+
	91130	CIRAD	Central African Republic	1991	Cattle	+
	94111	CIRAD	Rwanda	1994	Cattle	+
	95014	CIRAD	Tanzania	1995	Cattle	+
	T1/44	CIRAD	Tanzania	1952	Cattle/vaccine strain	+
	T1/Sr50	CIRAD	Tanzania	1952	Cattle/vaccine strain	+
	KH3J	CIRAD	Sudan	1940	Cattle/vaccine strain	+
	Gladysdale	AAHL	Australia		Cattle	+
	V5	AAHL	Australia	1965–1968	Cattle/vaccine strain	+
M. mycoides subsp. mycoides LCa	Y-goat	NCTC	Australia		Goat/type strain	−
	152/93	FVLP	Grand Canary		Goat	−
	LC8065	CIRAD	France		Cattle	−
	D2482/91	IVBBE	Switzerland		Goat	−
	8794-Inde	CIRAD	India		Goat	−
M. mycoides subsp. capri	PG3	NCTC			Goat/type strain	−
	N108	CIRAD	Nigeria			−
	capri L	CIRAD	France		Goat	−
	9139/11-91	CIRAD	Turkey			−
Mycoplasma sp. bovine group 7	PG50	NCTC	Australia		Cattle/reference strain	−
	B5415	NVI	Portugal		Cattle	−
	CP291	NVI	Portugal		Goat	−
	PAD3186	BVVG	India		Goat/milk	−
	FRD424	BVVG	India		Goat/milk	−
	Calf 1	NVI	Nigeria			−
	D318b	NVI	Germany			−
	C2306	NVI	Portugal			−
	D424	NVI	Germany			−
	QR1/92	NVI	Australia			−
	4055	NVI	France			−
Mycoplasma sp. serogroup L	B144P	NVI	United States			−
M. capricolum subsp. capricolum	California kid	NCTC	California		Goat/type strain	−
	173/87	CIRAD	Greece		Sheep	−
M. capricolum subsp. capripneumoniae	F38	NCTC	Kenya		Goat/type strain	−
	9081-487p	CIRAD	Oman		Goat	−
	Gabès	CIRAD	Tunisia		Goat	−
M. putrefaciens	KS1	NCTC			Goat/type strain	−
M. agalactiae	PG2	NCTC			Goat/type strain	−
M. bovis	PG45	NCTC			Cattle/type strain	−

^aLC, large-colony type. 

^bAAHL, Australian Animal Health Laboratory, Geelong, Victoria, Australia; BVVG, BVVG, Jena, Germany; CIRAD, CIRAD-EMVT, Montpellier, France; FVLP, Faculdad de Veterinaria, Universidad de Las Palmas, Las Palmas, Spain; IVBBE, Institute for Veterinary Bacteriology, University of Berne, Berne, Switzerland; LNV, Laboratorio Nacional de Veterinaria, Lisbon, Portugal; LPB, Laboratoire de Pathologie Bovine, Lyon, France; NCTC, National Collection of Type Cultures, PHLS, London, United Kingdom; NVI, National Veterinary Institute, Uppsala, Sweden. 

^cAs determined by PCR and/or Southern blot hybridization. 

Cloning, detection, and analysis of IS1634.

IS1634 was detected while studying the DNA segments proximal to copies of IS1296 of M. mycoides subsp. mycoides SC strains L2 and Afadé. Genomic HindIII fragments were cloned into the HindIII site of pBluescriptII SK⁻, and clones were screened with a specific IS1296 probe prepared as described previously (4). Fragments neighboring IS1296 were sequenced by using the oligonucleotide primers IS1296S1 (5′-CAACTGAAATATCATATTTATTTGC-3′) and IS1296S2 (5′-GGATTGTCTCCTGTACAATTCAG-3′) matching the ends of IS1296 and facing outward from the insertion element. In one clone, plasmid pJFFev7.0-L2 containing a 7.0-kb-long insert from M. mycoides subsp. mycoides SC strain L2, we detected a new insertion element only 17 bp away from IS1296. This novel insertion element has been designated IS1634 by the Plasmid Reference Center, Stanford University School of Medicine, Stanford, Calif. Determination of the entire nucleotide sequence of IS1634 was done by creation of deletion subclones by exonuclease III digestion and sequencing with oligonucleotide primers complementary to the T3 and T7 promoters flanking the cloning site of pBluescriptII SK⁻. This copy of IS1634 has a size of 1,872 bp, contains characteristic 13-bp inverted repeats at its ends, and includes a single open reading frame (ORF) spanning nearly the entire sequence between the two inverted repeats. There are several putative mycoplasmal start codons for this ORF, including GUG at nucleotides (nt) 110 and 194 and UUA at nt 182, but no AUG. Although UUA is most rarely used as a start codon in eubacterial genes, it must be noted that UUA at nt 182 is preceded by a canonical sequence for a ribosome binding site (AAAGG) 8 bp upstream (11, 13) and could be a potential start codon of this transposase, as was also observed for the closely related Mycoplasma capricolum (2). The resulting peptide of 533 amino acids has 28% identical and 45% similar positions in a 266-amino-acid overlap of the central region with the corresponding part of the putative transposase from insertion element IS1549 of Mycobacterium smegmatis (15). This indicates that the ORF represents the transposase of IS1634. There are 5 UGA_Trp codons in this transposase gene. UGA encodes tryptophan in the genus Mycoplasma but is generally used as a stop codon in prokaryotes. The A+T content of IS1634 is 73.9%, reflecting the A+T-rich genome characteristic of mycoplasmas.

IS1634 is distantly related to IS5377 of Bacillus stearothermophilus and to IS1151 of Clostridium perfringens, which are the next most closely related IS elements after IS1549 and which belong to the IS4 family of insertion elements. IS1634 shows similarity to IS5377 and to IS1151 only in the conserved C1 (integrase C′ signature domain) region of IS4 family transposases which contains the C1 signature sequence Y-(X₂)-R-(X₃)-E-(X₆)-K (17). The corresponding sequence in the transposase of IS1634 is Y-(X₂)-Q-(X₃)-E-(X₆)-K. IS1634 shows no similarity to the N3 (integrase N′ signature domain) region which is located in the N-terminal half of the transposase and which is characteristic of the IS4 family of insertion elements (17). In this domain, however, IS1634 shows significant homology (38% identical and 74% identical plus similar amino acids) to IS1549.

Different copies of IS1634 were sequenced and shown to be well conserved at the nucleotide sequence level. Minor variations were observed, with an average of 4.2 differences per 1,000 bp over the entire length of IS1634, only half of them affecting the amino acid composition of the putative transposase.

Identification of direct repeats.

To verify the correct ends of IS1634 and to study its sites of transposition, we have cloned and analyzed independent copies of this insertion element. Primers IS1634(in)L (5′-ACTTAGTCGCGAAAACACTC-3′) and IS1634(in)R (5′-AGCATAAAACAAGAAAACATAC-3′), matching the ends of IS1634 and facing inward, were designed to produce a probe for IS1634 from plasmid pJFFev7.0-L2. This probe was used to screen the two gene banks from M. mycoides subsp. mycoides SC strains L2 and Afadé. In both libraries, 30 to 50 of 300 colonies hybridized to the IS1634 probe, indicating a relatively high copy number of this insertion element. Four clones from the L2 library (plasmids pJFFev7.0-L2, pJFFev9.0-L2, pJFFev2.0-L2, and pJFFev6.0-L2) and five clones from the Afadé library (plasmids pJFFev3.9-Af, pJFFevAM5802, pJFFevAM5830, pJFFevAM2520, and pJFFevAM2516) were retained to be sequenced across the junctions of IS1634 integration, using primers ISnew(out)L (5′-TATAGCCGCTGAAAACTGAG-3′) and ISnew(out)R (5′-AAAAAACATAACTTATGAGATC-3′) facing outward from IS1634. Long stretches and segments which could not be sequenced accurately by this strategy were sequenced by the use of primers derived from sequenced segments. Analysis of the sequences obtained confirmed the ends of IS1634 and revealed that the different copies of IS1634 were flanked by direct repeats of variable length, ranging from 17 to 478 bp (Fig. 1). This sequence strategy was not successful with plasmids pJFFev6.0-L2 and pJFFevAM2520 since they both presented overlapping sequences, suggesting the presence of two (or more) IS1634 copies in the same clone.

FIG. 1.

Structure of IS1634 insertion sites. The map for the IS1634 clones (open boxes) that were analyzed shows the long, variable-length direct repeats at the site of insertion as black boxes and black arrows beneath (A to F). The plasmid pJFFevAM2516 contains one copy of IS1634 with no apparent flanking direct repeats (G). Hatched boxes represent all or part of IS1296. Grey triangles represent the inverted repeats IR-L and IR-R of IS1634 or IS1296.

The copy of IS1634 in pJFFev7.0-L2 which was found only 38 bp away from a copy of IS1296 was flanked by direct repeats of 21 bp (Fig. 1A). The plasmid pJFFev9.0-L2 contained an IS1634 that was transposed into a copy of IS1296. It had 75-bp direct repeats that were segments of IS1296 itself (Fig. 1B). The plasmid pJFFev2.0-L2, with a 2.0-kb HindIII fragment from strain L2, contained a full-length IS1634 apparently not flanked by direct repeats. This copy of IS1634 was flanked by 111- and 16-bp segments ending in HindIII sites. To determine if the apparent lack of direct repeats in this clone was due to HindIII cloning, a specific PCR amplification to detect a putative 127-bp repeat (111 plus 16 bp) was performed with primers 2.0repL (5′-CGTAGGAAAATAATACAAAAAG-3′) and 2.0repR (5′-AGGGTTACTCACTCATAAGC-3′). It produced a fragment of the expected size, indicating that the combination of the 111- and 16-bp flanking sequences forms the direct repeats (Fig. 1C). Moreover, this primer pair amplified from genomic DNA of strains PG1, L2, T1/Sr50, and T1/44 a 130-bp fragment as well as a 2.0-kb fragment containing the copy of IS1634. Strain Afadé only showed the 130-bp fragment, not the 2.0-kb fragment, indicating that the genome of Afadé lacks this particular copy of IS1634 (data not shown). The copy of IS1634 from strain Afadé that was contained on plasmid pJFFev3.9-Af presented direct repeats of 478 bp, the longest observed in this work (Fig. 1D), while pJFFevAM5830 contained direct repeats with a length of 17 bp, the shortest studied (Fig. 1E). The copy of IS1634 in pJFFevAM5802 was flanked by 70-bp-long direct repeats (Fig. 1F). One copy of IS1634 cloned from strain Afadé (pJFFevAM2516) showed no flanking direct repeats at all (Fig. 1G). This copy of IS1634 is flanked by part of an IS1296 on one side and by a DNA segment which shows no relationship to known sequences in the EMBL/GenBank database on the other side. We assume that this unusual copy of IS1634 with no apparent direct repeats might have occurred by a recombination event between a copy of IS1634 inserted into IS1296 and a closely neighboring copy of IS1634, leaving only one part of IS1296 remaining. Such a mechanism of truncation of IS1296 would also explain the few weakly hybridized bands in the IS1296 Southern blots as shown by Cheng et al. (4).

Hence, IS1634 is characterized by the presence of unusual long direct repeats which vary in length at the site of insertion, similar to that recently found for IS1549 (15). This is in contrast to all other insertion elements found so far, which present short direct repeats of constant length that normally consist of 4 to 9 bp, the most common length being 9 bp. It indicates that the putative transposases of IS1634 and IS1549, in addition to their significant homology in the central part of their sequence, have the same peculiar mechanism of creation of staggered nicks at variable distances in the target DNA sequence at the initial step of transposition, in accordance with the models proposed by Galas and Chandler (9) and by Shapiro (18). This indicates that IS1634 together with IS1549 forms a novel class or family of insertion elements which is distinct from all other known classes or families of IS elements. From the available sequence data, there is no evidence of nucleotide sequence similarity between the target sites into which the different copies of IS1634 had inserted. IS1634 transposition in the chromosome seems to occur at essentially unrelated sites.

Distribution of IS1634 in M. mycoides subsp. mycoides SC and related species.

To analyze the distribution of IS1634, we have performed Southern blot hybridization analysis with HindIII-digested genomic DNA of a large number of M. mycoides subsp. mycoides SC strains as well as many strains from the other members of the M. mycoides cluster by using the IS1634 probe. As shown in Fig. 2, all strains of M. mycoides subsp. mycoides SC originating from different countries and continents and collected over a period of more than 60 years contain about 30 copies of IS1634 and show a constant picture with only minor differences. These observations indicate that IS1634, similarly to the previously discovered IS1296 (4, 6), has a relatively low transposition frequency or that M. mycoides subsp. mycoides SC has acquired an optimal constellation of IS1634 copies on its genome for survival in its ecological niche, the infected tissues of the animal host. Since IS1634 insertions show a lower percentage of variations than IS1296 insertions do, typing of strains with this new IS element gives fewer different profiles than the previously reported typing with IS1296 (4).

FIG. 2.

Southern blot hybridization of HindIII-digested DNA of a selection of M. mycoides subsp. mycoides SC isolates. DNA digests of different M. mycoides subsp. mycoides SC strains were separated electrophoretically on a 0.5% agarose gel, transferred to a positively charged nylon membrane, and hybridized to the IS1634 probe. The strains are described in Table 1. Abbreviations: Eu, Europe; Af, Africa; Au, Australia; T, type strain; superscripts v, vaccine strain; superscript c, caprine strain; superscript o, ovine strain. Continent designations with no superscript indicate strains from cattle. Std, molecular mass standards: 23.1, 9.4, 6.6, 4.4, 2.3, and 2.0 kb. Arrows indicate particular IS1634 copies. (A) Blot results for strains from Europe in comparison to PG1. A thin arrow indicates the region of the blot at about 4.0 kb where the IS1634 pattern of strain O326 differed from the others. The 7.-kb fragment characteristic of all European strains is indicated by a horizontal dotted arrow. (B) Blot results for strains from Africa and Australia in comparison to PG1. A horizontal arrow indicates the bands at about 3.0 to 3.5 kb where African strains showed heterogeneity in the IS1634 pattern.

All other members of the M. mycoides cluster and other related mycoplasmas are devoid of IS1634. To further verify these results, PCR analysis of genomic DNA from all mycoplasma strains used in this study was performed with the IS1634-specific primers IS1634(in)L and IS1634(in)R. This analysis confirmed the presence of IS1634 in all M. mycoides subsp. mycoides SC strains and the absence of this insertion element in all other related mycoplasmas (Table 1).

The Southern blot analysis of IS1634 copies in the different strains of M. mycoides subsp. mycoides SC revealed relatively homogeneous patterns for the different strains, which represent a geographically and periodically heterogeneous collection. A particularly homogeneous picture is seen among the European strains. All except one presented the same IS1634 pattern. The only European strain that differed in its IS1634 pattern was O326, which had a different disposition of the IS1634 copies in the blot at about 4.0 kb (Fig. 2A). O326 was isolated from sheep (3). Characteristic of all European strains is a 7.0-kb fragment (Fig. 2A) which differentiates them from the African and Australian strains. African strains showed a certain heterogeneity of the IS1634 pattern, in particular at about 3.0 to 3.5 kb (Fig. 2B). Of particular interest is the lack of the 2.0-kb band in the pattern of strains Afadé and B17, both isolated from Chad. This agrees with the result of the previously mentioned PCR amplification with genomic DNA from strain Afadé with primers 2.0repL and 2.0repR, which did not give rise to the 2.0-kb product. Thus, plasmid pJFFev2.0-L2 contains the copy of IS1634, flanked by 127-bp repeats, which is absent in strains Afadé and B17. Another interesting observation is the fact that the two Australian strains tested (Gladysdale and the vaccine strain V5) demonstrate apparently identical IS1634 profiles, although they were distinguished by their IS1296 profiles (4). Furthermore, it is worthwhile noting that the three African nonpathogenic vaccine strains tested (T1/44, T1/Sr50, and KH₃J) have identical IS1634 profiles and differ from the other African strains by two fragments in the 10- to 15-kb size range, in spite of the fact that KH₃J was isolated in a different country 12 years earlier than T1/44 and T1/Sr50 and differs in its IS1296 profile (4). IS1634 typing might therefore serve as a valuable tool to detect possible locations on the genome that can attenuate virulence. Moreover, IS1634 profiles together with the IS1296 profiles can further refine the previously obtained clustering of different M. mycoides subsp. mycoides SC strains, particularly in the case of the type strain PG1, which shows an identical IS1634 pattern to the Australian field strain Gladysdale and hence seems to be most closely related to Australian strains. The finding that several copies of IS1634 are found at variable chromosomal locations among different strains, together with the fact that DNA sequence analysis of a few copies of IS1634 cloned from an African and a European strain showed only minor variations which do not affect the inverted repeats nor the essential amino acid sequence of the putative transposase, indicates that IS1634 is a functional IS element.

Concluding remarks.

M. mycoides subsp. mycoides SC, by far the most pathogenic mycoplasma in the M. mycoides cluster, is striking in its exceptionally high number of insertion elements, about 20 copies of IS1296 and about 30 copies of IS1634, representing 7% of total genomic DNA (approximately 90 kb). In view of the capacity of insertion elements to modulate gene expression (10, 19), one might speculate that M. mycoides subsp. mycoides SC has used these elements to regulate the expression of its genetic potential, resulting in a phenotype for invasion ability into host tissue being conferred to this microorganism. The fact that IS1634 is specific to M. mycoides subsp. mycoides SC, where it is found at a high copy number, makes it an optimal target for identification of M. mycoides subsp. mycoides SC and for the development of highly sensitive detection methods.

Nucleotide sequence accession number.

The EMBL/GenBank accession numbers for the nucleotide sequences of five of the copies of insertion element IS1634 analyzed are AF062493, AF101410, AF101411, AF101412, and AF101413.