Molecular typing of Mycobacterium tuberculosis strains with a common two-band IS6110 pattern.

We conducted a progam of population-based molecular typing of all Mycobacterium tuberculosis isolates obtained in Alabama since 1994. Of 2,452 isolates, 1,013 (41%) had fewer than 6 bands of IS6110; 348 (14%) had a single two-band pattern (JH2). With conventional epidemiologic methods, we identified three groups of related patients with JH2 isolates. Spoligotyping and pattern of variable number of tandem repeats identified 10 molecular groups; two found by conventional methods were subdivided.

**********

To achieve the national goal of tuberculosis (TB) elimination, disease control efforts must be based on a thorough understanding of transmission patterns among the general population and among groups at particular risk (1,2). In close coordination with the Alabama Department of Public Health, we conducted a program of population-based molecular typing of all Mycobacterium tuberculosis isolates identified since 1994 in Alabama. The state contains a mixture of rural and urban settings with a stable population, and the proportion of tuberculosis cases among the foreign-born and HIV-positive groups is low (2).

The Study

Through May 2001, we typed 2,452 isolates using the IS6110 restriction fragment length polymorphism (RFLP) method and have combined the results of molecular analysis with an aggressive contact investigation program. We found that 1,013 (41%) of the IS6110 patterns had 56 bands, including 348 (14%) with a two-band RFLP pattern called JH2. Conventional epidemiologic techniques identified three groups with related JH2. Resistance to a single drug (isoniazid or streptomycin) was associated with two groups, and homelessness was associated with a large group. Results from the latter group have been previously reported (3,4).

To further characterize JH2, the most common RFLP pattern statewide, we used two polymerase chain reaction (PCR)-based methods for secondary typing methods, spacer oligonucleotide type (spoligotyping) and variable number of tandem repeats (VNTR). Isolates were selected for further molecular typing to determine which method or combination of methods could help us understand the epidemiology of this important RFLP pattern in Alabama and the extent of molecular relatedness in the JH2 cluster.

We selected 102 (29%) of the 348 JH2 pattern isolates from the Alabama genotype database that represented two-band IS6110 RFLP strains. Groups of isolates identified through conventional epidemiologic links were selected for evaluating the secondary molecular typing techniques. Fifteen isolates from northeast Alabama were obtained from a conventional epidemiologic investigation that showed a convenience store as the site of transmission (5). Twenty-four isolates from central Alabama came from a previous investigation (3,4) of a TB outbreak in a large community of homeless persons, and four isolates were from a known outbreak on a school bus in southeast Alabama. The remaining 59 isolates were selected from the general strain collection for comparison. The selection was based on availability of viable cultures for DNA extraction and the year that these outbreaks occurred. Two isolates from the nonclustered strain collection did not generate an adequate VNTR pattern and were excluded from the final analysis. Therefore, 100 isolates were available for the combined analysis.

Isolates of M. tuberculosis were cultured on Lowenstein-Jensen or 7H 11 Middlebrook plates for at least 4 weeks before DNA extraction. Chloroform-isoamyl alcohol was used to extract chromosomal DNA from isolates, and IS6110 RFLP typing was performed according to international standards (6,7). All fingerprint images were scanned into a computerized database and Whole Band Analyzer, version 4.01 (Genomic Solutions Inc., Ann Arbor, MI, USA) was used. Spoligotyping was used as a secondary typing method on the basis of the presence or absence of 43 variable spacers in the direct repeat region of M. tuberculosis (3,8-12). Spoligotyping membranes were used according to manufacturer's recommendations (Isogen Bioscience BV, Bilthoven, the Netherlands), and the spacer regions were numbered as reported previously (8). A spreadsheet was used to analyze the spoligotyping results. For national database reporting purposes, we converted the spoligotyping image into an octal digital format (13).

VNTR typing was used to further investigate strains that matched by both IS6110 and spoligotyping methods or had a known epidemiologic link. This typing yielded 86 samples. As previously described (14), we used seven sets of primers and individually amplified regions with known informative tandem repeats. Electrophoresis was conducted on the PCR products in agarose. The ethidium bromide-stained gel was imaged to determine the size of the amplicon and the number of tandem repeats in each locus. We used a DNA ladder and amplicons from strain H37Rv to determine fragment sizes, according to Frothingham and Meeker-O'Connell (14).

Results

We used spoligotyping to separate the JH2 isolates into 20 different subtypes. Four of the 20 patterns contained >2 isolates. Eighty-four of the 100 isolates remained clustered after spoligotyping. The largest spoligotyping group (designated CDC spoligo 9) contained 55 isolates. The other 3 clustered patterns were designated CDC spoligotype 3 (n=12 isolates), CDC spoligotype 545 (n=14 isolates), and CDC spoligotype 550 (n=3 isolates). Two of the clusters previously identified through contact epidemiology were subdivided.

Eighty-six strains were typed with VNTR, which generated 15 different VNTR profiles (Table), of which the largest group contained 25 isolates. Nine profiles contained more than one isolate, and six isolates were unique. The number of isolates in the groups varied from 2 to 25. The largest cluster (homeless, n=24), which was identified previously through contact epidemiology, was subdivided into 10 subtypes.

Among those isolates with both secondary typing results, we identified 10 clusters plus 9 unique isolates. Clustered cases accounted for 77 of 100 cases. The largest spoligotype cluster of 55 isolates (CDC spoligotype 9), after combination with the VNTR results, was reclassified into nine unique profiles, including seven different clusters. The largest cluster had 22 isolates. CDC spoligotype 3 was divided into a single cluster of seven isolates and five unique profiles. CDC spoligotype 545 and CDC spoligotype 550 each represented single clusters.

The two-band JH2 pattern was the most common IS6110 fingerprint found during nearly 8 years of TB genotyping surveillance in Alabama. This pattern matches National Tuberculosis Genotyping Fingerprint Pattern (NTGFP) 00016 in the database of the National Tuberculosis Genotyping Surveillance Network. 00016 was also the most common fingerprint pattern found during 5 years of the TB genotype project and was reported from all seven sentinel surveillance sites with a final frequency of 5% of all isolates.

Conclusion

Pattern JH2 represents a group of similar IS6110 patterns. The smaller band at 1.46 kb is conserved, and the larger band is located in direct repeat 24, a common insertion site. Consequently, the size of the larger band varies, depending on the number of direct repeats located upstream of spacer 24. A strain's spoligotype can predict the size of the larger band; the larger bands analyzed in our study were 4.5 kb-4.8 kb. The resolution of the gel and the Whole Band Analyzer (Genomic Solutions, Inc.) do not discriminate well between these sizes, which could result in different spoligotypes of a single IS6110 insertion pattern being considered as different RFLP patterns.

The prominence of pattern JH2 (NTGFP 00016) statewide and throughout the network suggests an older, more stable IS6110 pattern. A group of strains likely spread throughout the general population early in the TB epidemic (19th and early 20th centuries) and remain endemic in the 21st century in the United States. Although this pattern is found rarely in European isolates, some countries in Africa report a high prevalence (15).

Despite large numbers of the IS6110 pattern found in Alabama, we could define only a few distinct groups through conventional epidemiologic methods. Using VNTR and spoligotyping techniques to reexamine some of these clusters, we gained a better understanding of disease transmission among community groups. Two clusters previously classified by conventional epidemiologic methods were confirmed by this study. Both clusters were associated with unique drug resistance patterns in outbreaks involving a school bus (isoniazid resistance) and a neighborhood convenience store (streptomycin resistance). The largest cluster, which involved the homeless community, revealed multiple molecular-based subclusters not identifiable by routine epidemiologic study, drug susceptibility data, or both. These subclusters signify the actual genetic diversity of this most prevalent IS6110 pattern.

The importance of secondary typing for low-copy IS6110 strains is well accepted (9-12), as is the necessity of combining conventional methods with those of molecular epidemiology (2-4). We found that VNTR is particularly well suited as a secondary typing method for the common JH2 pattern. However, the VNTR method is more expensive and time-consuming because seven PCRs are required. VTNR may be a preferred method for TB research, but not for widespread TB surveillance. Automated systems could reduce overall costs, however. While spoligotyping was less discriminatory in this study, we believe it should be evaluated carefully for use in other settings.

Support for this project was provided by a cooperative agreement between the Centers for Disease Control and Prevention and the Alabama Department of Public Health (Agreement C10119105) and by all sites of the National Tuberculosis Genotyping and Surveillance Network. Support was also provided through the Pulmonary Disseminated Infectious Diseases in HIV training grant from (No. T32AI07441) the National Institutes of Health.

Table. Comparison of a common two-band IS6110 RFLP pattern with
spoligotyping and VNTR (a)

Spoligotyping                               VNTR

Octal (CDC designation)    No. of       VNTR results (AL designation)
                           isolates

777776777760601 (0009)        55             6,3,2,4,2,3,2 (01)
                                             6,3,2,3,2,2,2 (02)
                                             6,3,2,4,2,3,3 (03)
                                             6,3,2,3,2,2,3 (04)
                                             6,3,2,2,2,2,2 (05)
                                             6,3,2,3,2,3,2 (06)
                                             6,3,2,3,2,3,3 (07)
                                             6,3,2,4,2,2,2 (08)
                                             6,2,2,3,2,2,2 (09)
037776777760601 (0545)        14             6,3,2,3,2,4,2 (10)
777776777760771(0003)         12             6,3,2,2,2,1,2 (11)
                                            6,3,2,3,2,3,2.5 (12)
                                             6,3,2,2,2,1,3 (13)
                                             6,3,1,2,2,2,3 (14)
                                             6,3,2,3,2,2,2 (02)
                                             6,3,2,3,2,3,3 (07)
001776777760601(0550)          3             6,3,2,4,2,3,2 (01)
777776374360711(1445)          1            6,3,2,1,2,3,2.5 (15)
037766677760601(n/a)           1             6,3,2,3,2,4,2 (10)
Total                         86

Spoligotyping                VNTR    Related to conventional outbreak

Octal (CDC designation)     No. of      Store    Bus    Homeless
                           isolates

777776777760601 (0009)        22                            2
                              12                  4         4
                               8                            2
                               4                            3
                               3                            3
                               2
                               2                            1
                               1
                               1                            1
037776777760601 (0545)        14         14
777776777760771(0003)          7                            5
                               1
                               1
                               1                            1
                               1                            1
                               1
001776777760601(0550)          3
777776374360711(1445)          1                            1
037766677760601(n/a)           1
Total                         86         15       4        24

(a) RFLP, restriction fragment length polymorphism; VNTR, variable
number of tandem repeats; CDC, Centers for Disease Control and
Prevention.

References

(1.) Jasmer R, Hahn J, Small P, Daley C, Behr M, Moss A, et al. A molecular epidemiologic analysis of tuberculosis trends in San Francisco, 1991-1997. Ann Intern Med 1999;130:971-8.

(2.) Dunlap NE. The use of RFLP as a tool for tuberculosis control: utility or futility? Int J Tuberc Lung Dis 2000;4:S134-8.

(3.) Kimerling M, Benjamin W, Lok K, Curtis G, Dunlap N. Restriction fragment length polymorphism screening of Mycobacterium tuberculosis isolates: population surveillance for targeting disease transmission in a community. Int J Tuberc Lung Dis 1998;2:655-62.

(4.) Kimerling ME, Shakes CF, Carlisle R, Lok KH, Benjamin WH, Dunlap NE. Spot sputum screening: evaluation of an intervention in two homeless shelters. Int J Tuberc Lung Dis 1999;3:613-9.

(5.) Kimerling ME, Brook NL, Lok KH, Benjamin WH Jr., Dunlap NE. Mycobacterium tuberculosis at a neighborhood convenience store [abstract]. Am J Respir Crit Care Med 1998;157:708.

(6.) van Embden JD, Cave MD, Crawford JT, Dale JW, Eisenach KD, Gicquel B, et al. Strain identification of Mycobacterium tuberculosis by DNA fingerprinting: recommendations for a standardized methodology. J Clin Microbiol 1993;31:406-9.

(7.) Yang ZH, Ijaz K, Bates JH, Eisenach KD, Cave MD. Spoligotyping and polymorphic GC-rich repetitive sequence fingerprinting of Mycobacterium tuberculosis strains having few copies of IS6110. J Clin Microbiol 2000;38:3572-6.

(8.) Groenen PM, Bunschoten AE, van Soolingen D, van Embden JD. Nature of DNA polymorphism in the direct repeat cluster of Mycobacterium tuberculosis; application for strain differentiation by a novel typing method. Mol Microbiol 1993;10:1057-65.

(9.) Goyal M, Saunders NA, van Embden JD, Young DB, Shaw RJ. Differentiation of Mycobacterium tuberculosis isolates by spoligotyping and IS6110 restriction fragment length polymorphism. J Clin Microbiol. 1997;35:647-51.

(10.) Soini H, Pan X, Amin A, Graviss EA, Siddiqui A, Musser JM. Characterization of Mycobacterium tuberculosis isolates from patients in Houston, Texas, by spoligotyping. J Clin Microbiol 2000;38:669-76.

(11.) Soini H, Pan X, Teeter L, Musser JM, Graviss EA. Transmission dynamics and molecular characterization of Mycobacterium tuberculosis isolates with low copy numbers of IS6110. J Clin Microbiol 2001;39:217-21.

(12.) Cronin WA, Golub JE, Magder LS, Baruch NG, Lathan MJ, Mukasa LN, et al. Epidemiologic usefulness of spoligotyping for secondary typing of Mycobacterium tuberculosis isolates with low copy numbers of IS6110. J Clin Microbiol 2001;39:3709-11.

(13.) Dale JW, Brittain D, Cataldi AA, Cousins D, Crawford JT, Driscoll J, et al. Spacer oligonucleotide typing of bacteria of the Mycobacterium tuberculosis complex: recommendations for standardized nomenclature. Int J Tuberc Lung Dis 2001;5:216-9.

(14.) Frothingham R, Meeker-O'Connell WA. Genetic diversity in the Mycobacterium tuberculosis complex based on variable numbers of tandem DNA repeats. Microbiology 1998;144:1189-96.

(15.) Hermans PW, Messadi F, Guebrexabher H, van Soolingen D, de Haas PE, Heersma H, et al. Analysis of the population structure of Mycobacterium tuberculosis in Ethiopia, Tunisia, and The Netherlands: usefulness of DNA typing for global tuberculosis epidemiology. J Infect Dis 1995;171:1504-13.

Kerry H. Lok, * William H. Benjamin, Jr., * Michael E. Kimerling, * Virginia Pruitt, ([dagger]) Donna Mulcahy, ([dagger]) Nancy Robinson, ([dagger]) Nancy B. Keenan, ([dagger]) and Nancy E. Dunlap *

* University of Alabama School of Medicine, Birmingham, Alabama, USA; and ([dagger]) Alabama Department of Public Health, Montgomery, Alabama, USA

Dr. Lok is a postdoctoral trainee in the Pulmonary, Allergy, and Critical Care Division, of University of Alabama School of Medicine, Birmingham, Alabama. His research interests include the molecular epidemiology of tuberculosis.

Address for correspondence: Michael E. Kimerling, University of Alabama at Birmingham, RPHB 217, 1530 3rd Ave. South; Birmingham, AL 35294-0022 USA; fax: [phone]; e-mail: [e-mail]