*National Research Institute of Tuberculosis and Lung Diseases (NRITLD), Shaheed Beheshti University of Medical Sciences, Tehran, Iran, **Istituto di Microbiologia, Universita Cattolica Del Sacro Cuore, Largo Francesco vito, Roma, Italy

The World Health Organization estimates that perhaps as much as one-third of the world’s population or approximately 1.9 billion persons are or have been infected with Mycobacterium tuberculosis (MTB).^1-3 The key to controlling the spread of tuberculosis include proper case finding, rapid diagnosis of tuberculosis and prompt initiation of effective chemotherapy.^4-6 Besides, we need to learn more about epidemiological spread of diseases inside and across the borders of the country, which in turn may leads us to new interventional strategies.^7-9

Advances in molecular biology have provided powerful epidemiological tools for typing and detecting MTB deoxyribonucleic acid (DNA). ^10,11 The use of repetitive elements for fingerprinting in MTB isolates was initially described by Eisenach, et al.¹² One of the insertion elements isolated by Eisenach was IS6110, which is a mobile genetic element present exclusively in the genome of members of the MTB complex and is present in multiple copies in most members of this complex.^13,14 These properties have made IS6110 highly successful as a target for detecting MTB DNA for diagnosis of TB and fingerprinting the strains of epidemiological interest.^14,15

With availability of large number of MTB isolate from different parts of Iran in this center, we decided to perform restriction fragment length polymorphism (RFLP) based on IS6110 typing to determined transmission patterns of tuberculosis in different provinces of the country. This was a preliminary screening for future evaluation of TB transmission in Iran.

RLFP was performed on 100 suitable isolates which were received between September 1998 and February 1999 from selected laboratories of East and West Azarbaijan, Khorasan, Kerman, Kermanshah and Fars. Out of the 100 MTB isolates, 32 were from East Azarbaijan, 22 from West Azarbaijan, 25 from Khorasan, 7 from Kermanshah, 8 from Kerman and 6 from Fars provinces. They were composed of 37 drug-resistant and 63 drug-sensitive strains. From these 100 culture-positive slants, the colonies were transferred to 7H9 broth medium and were cultured grown for 2-3 weeks at 37 ºC. Thereafter, they were killed by heating in 80 ºC water a bath for 30 minutes prior to being used for DNA extraction.

Oligonucleotide primers were synthesized by means of a DNA synthesizer (Milligen / Biosearch, Istituto Di Microbiologia, Universita Cattolica Del Sacro Cuore, Largo Francesco vito, Roma, Italy).

Mycobacterial genomic DNA was digested with PvuI and electrophoretically separated on 0.8% agarose gel (Figure1). After denaturation, DNA was transferred to the nitrocellulose membrane by vaccum blotting for 1 h and hybridized with digoxigenin (dig)-labelled probe. DNA was prepared by the method described by the manufacturer. Hybridization and washing were done at 68ºC.

The IS6110-specific DNA probe of 245 bp was amplified by PCR and electrophoresed on agarose gel. Each DNA probe was electroeluted and precipitated with LiCL, dried and resuspended with TE, and labelled with dig-dUTP using a dig DNA labelling kit (Boeringer/ Mannheim, Germany). The detection of dig-labelled DNA was carried out according to the manufacturer’s recommendations.

The fingerprint patterns of the isolates were pair-wise compared both by the computer-assisted analysis Gel Compar^R and by visual observation. The distance between the two profiles was calculated according to the Dice index; the algorithm used to produce the dendrogram from these distances was the unweighed pair group method of analysis (UPGMA).

The results demonstrated that two types of banding patterns for IS6110 were present in these strains; the MTB strains (33%) had low copies of IS6110 (4-6 bands) in their chromosomal DNA and MTB strains (62%) and had high copies of IS6110 (6-14 bands). In addition, five MTB strains (5%) were without any IS6110 banding pattern.

Figure 2. The DNA fingerprinting pattern of IS6110 in MTB strains collected from Azarbaijan (East and West).

Figure 3. Computer-assisted dendrom system demonstrated that the similarity of strain was from 8% to 51%.

The pattern of DNA fingerprinting of 54 chosen culture-positive slants from Azarbaijan (East and West) showed that the strains had multiple IS6110- containing restriction fragments and each strain showed a unique pattern (Figure 2). The computer-assisted dendrogram system also demonstrated no noticeable similarity in banding pattern among the strains. The similarity among them was a minimum of 10 % and a maximum of 50 % (Figure 3).

The DNA fingerprinting of 25 culture-positive slants from Khorasan province showed a few banding patterns with less than 30% similarity among them (Figure 4). However, we found a set of identical strains, which had no epidemiological linking according to available data.

Similarly, 21 chosen culture-positive slants, received in NRITLD from intermediate laboratories of Fars, Kerman and Kermanshah provinces were used for DNA fingerprinting. The computer-assisted dendogram showed a minimum of 8% and a maximum of 51% similarity with IS6110 having multiple banding patterns (Figure 4). However, in Fars province a set of identical strains (Figure 5) was detected. Reviewing their

information, our data revealed that they were epidemiologically-linked strains.

Figure 4. The MTB strain collected from Khorasan province had few a IS6110 banding in comparison to other states.

The determination of banding patterns using RFLP based on IS6110 seems to be a useful tool for the evaluation of TB transmission. It is based on the assumption that different individuals who are infected with identical genotypes of M. tuberculosis strains are epidemiologically linked, whereas those with different genotypes are unrelated.^9,8,13 Thereby, the present study aimed to investigate the banding pattern of MTB isolates collected from patients with pulmonary tuberculosis in different geographical regions of Iran using RFLP-based IS6110 for determination of TB transmission .

Figure 5. Two identical strains of MTB in Fars province could were detected and found to be epidemiologically linked.

Our results demonstrate that the similarity of banding among MTB strains was between 8% to 51% (Figure 3). These fingerprinting patterns reflect the differences in copy number and sites of insertion in the chromosome due to duplication and/ or movement of the element in the collected MTB strains. From the epidemiologic point of view, the available data on patterns of IS6110 by computer scanning states that similarities of less than 90% of MTB strains would reflect reactivation and not transmission of the disease in the community.^8,14,15

In the present investigation, 2 sets of identical strains were detected. One set from Fars province, which had epidemiological link, and the other set from Khorasan province with no epidemiological link. Reviewing their registration files, one can see that the identical strains of Fars province were from patients staying in the same localities. One of the patients worked in a shop in this region, but the other actually was a resident there. This single finding can reconfirm that IS6110 could be a very useful tool in tracing the transmission rate among epidemiologically related tuberculosis patients. In a similar observation, Curtis, et al ²⁴ detected eight cases with identical IS6110 DNA fingerprinting, all of which were infected by patients with extensive cavitary disease and who had stayed at home for almost 8 months before seeking medical care and being diagnosed. Furthermore, Kiers and his co-worker²⁵ suggested the use of RFLP typing in international tracing of infection sources. He demonstrated that the source of MTB in infected children with tuberculous meningitis in the Netherlands was in the United Kingdom.

The major consideration for determining the usefulness of strain-typing methods are its specificity.¹¹ The specificity of the fingerprinting pattern is dependent upon the number of bands obtained.^10-12 Although IS6110 was originally described as a repetitive element present in 1-20 copies, many isolates of MTB are now known to possess one or few ^1-5 IS6110 . Accordingly, a study performed by two main National Centers for Tuberculosis in Vietnam²⁶ showed that 1.8% of their MTB strains had no IS6110 in their genomes. In addition, the percentage of strains with only one IS6110 was high, and consistent with reports of the MTB strains in India.²⁷ Similarly, our DNA fingerprinting revealed that 33% of MTB strains had a low copy number of IS6110 (1-6 band).

Although, the number of MTB strains in this study was not sufficient to judge any specific grouping or clustering in Iran, similar reports by Fomukong and his colleagues²⁸ show that low and high copy strains of MTB have a separate lineage which might be due to differences in their genotyping. The present study suggests that in these provinces in Iran, tuberculosis is characterized by the absence of obvious focuses of transmission. Two or more lineages for TB strains exist in Iran, which have to be analyzed. We suggest further studies to be performed in this field.