Molecular Epidemiological Investigation On Francisella Tularensis In China

Tularemia is a zoonotic disease caused by the small, gram-negative highly infective bacterium, Francisella tularensis (F.tularensis) which is considered to disperse throughout the Northern Hemisphere with foci in certain parts of North America, Europe, and northern Asia. F.tularensi has long been considered a potential biological weapon because of its extreme infectivity, ease of dissemination and substantial capacity to cause illness. Actually the agent had been listed as a biological weapons agent. F. tularensis infections have been reported in over 250 animal species, including mammals and invertebrates. A wide range of arthropod vectors have been implicated in the transmission of tularemia between mammalian hosts, including fleas, deer flies, stable flies, lice, bedbugs, mosquitoes and ticks. In China, F. tularensis was isolated in Citellus major from Tongliao region, Inner Mongolia municipality as early as in 1957. The first human tularemia epidemic which was caused by contact with infected hares , was reported in Heilongjiang province in 1959. The natural foci of the disease were subsequently identified in Tibet, Heilongjiang and Xinjiang provinces. F. tularensis was also isolated from patients, ticks and rodents in the above areas. An outbreak of tularemia occured in Shandong province in 1986. All the 31 reported patients were workers in a food-processing factory, where hares were slaughtered and processed. Thereafter, there has never been any report on either tularemia cases or natural infection of F. tularensis in wild animals and arthropods. It is unknown whether the foci have become quiescent or the disease is underreported because tularemia is not the notifiable disease in China. In addition, many epidemiological characteristics such as major hosts and subspecies type of the pathogen remain unclear, and there lacks rapid identification methods to discriminate whether the disease is natural occurrence or an intended initiation.To address these crirical issues, rodents and ticks were collected from 3 former tularemia endemic areas (Xinjiang, Heilongjiang and Inner Mogonalia) and 3 un-investigated areas, among which included 2 southern provinces and Jilin province. Samples were detected for F. tularensis by nested polymerase chain reaction. Then the PCR products of all the positive specimens were amplificated by another PCR to determine the subspeciese of the agent. A total of 421 rodents were assayed by nested PCR. Twenty rodent samples tested positive, giving out overall positive rate of 4.75%. The prevalence of infected rodents varied with the geographical origin (χ2=20.898, P<0.001). The infection rate was highest (11.65%) in rodents from Jilin Province . Altogether 14 rodent-species were identified in the captured animals. F. tularensis was detected from seven species. Although E. sibiricus and M. lybicus were most likely to be infected with the positive rates of 25% and 22.2% respectively, there was no significant difference in infection rate among the 7 positive rodents species (χ2 = 11.82, df = 6, P = 0.066). Out of 1670 adult ticks examined, thirty three were positive for F. tularensis. The overall positive rate was 1.98%. The positive rates of F. tularensis in D. silvarum were 3.67% and 2.33% in I. persulatus (Table 1),while all of the 144 Haemaphysalis verticalis and 299 Boophilus microplus were PCR-negative. The difference in positive rates among different species was significant (χ2 =14.366, P = 0.002). No positive result was obtained in samples from Zhejiang and Fujian provinces. All positive samples amplified the 220-bp fragment for the ppI-helicase gene that identifies F. tularensis subsp.holarctica (GenBank accession no. EU526911). Thus, all ticks in the present study were infected with F. tularensis subsp.holarctica.Simultaneously, we have established and evaluated 3 genotyping methods, including AFLP, PFGE and MLVA to determine their suitability in the analysis of Chinese strains. The results from AFLP showed, three of 45 primer pairs could only differentiate two Tibet strains from others, but not among other strains. Sequencin of the AFLP fragments demonstrated many repeats, which made AFLP unsuitable to be used to analyse F. tularensis strain. PFGE analysis showed high discrimination abilities from the restriction digestion using BamHI and XhoI. When the bacterium was disgested by XhoI, it produced same bands among the strains which ranged from 20～100 kb. As for BamHI, it produced different bands from different strains which ranged from 27～330 kb. Twenty-five informative markers were used to reveal genetic relationships among Chinese in MLVA. The 25 variable repeat motifs displayed repeat sizes from 2 bp (Ft-M25) to 23 bp (Ft-M23) in length. A 16-bp repeat structure was previously identified within an insertion sequence element designated ISFtu1 SSTR16. In this study, markers Ft-M5 and Ft-M7-10 represented five variable copies of this 16-bp motif. Allele states at these loci were reported to be most variable among the F. tularensis isolates, but not in Chinese strains. Previous studies indicated Ft-M19 contains a 30-bp sequence present in isolates of F. tularensis subsp. tularensis, mediasiatica, and novicida but absent in isolates of F. tularensis subsp. holarctica, in this study, amplicon sizing from our strains confirmed previous findings. Ft-M3 containning 9-bp sequence revealed that each of the variants of the 9-bp repeats were arranged in tandem and that the tandem regions were always arranged in the order A, B, C, D, and E, A and E were presented in all Chinese strains. Seven genotypes were determined from Chinese strains,while the variability did not affect the encoded amino acid. 2-3 genotypes were determined in Chinese strains according to Ft- M-4,20,21,22,24 varibility, other loci showed no variability in Chinese strains. The calculated diversity value (D) reflects the discriminatory capacity of marker and is a function of individual allele frequency. In this context, each sequence at a given locus with a different number of tandem repeats is designated as a distinct allele. Ft-M3 was 0.93, 0.50 for Ft-M4, 0.62, 0.35, 0.44, 0.34, 0.52 for Ft-M6, Ft-M20, Ft-M21, Ft-M22, Ft-M24, respectively. It was notable for Ft-M1 and M25, all reported strains had 3 and 5 copies, but there were 19 and 7 copies in one Chinese strain. The sequences from these two loci revealed the changes of the encoded amino acid.Using the established MLVA in the lab, we analysed the genetic relationships among Chinese strains. Phylogenetic analysis using the sequences from 25 loci revealed 5 groups among Chinese strains, i.e. the group of Tibet, Xinjiang, Heilongjiang, Inner Mongolia and vaccine, which was concordant with the distribution pattern of tularemia in China. When distance matrix based on copy numbers of 25 loci was applied to construct the phylogenetic tree, we got same results as those from sequence analysis. The reference isolates of type A and B, the isolates from Sweden, Amerian and Japan were included to analyse the genetic relationships among Chinese strains and foreign strains. The results showed that most Chinese strains were clustered closely to the American isolates, some were closed to Sweden isolates. Some isolates even clustered with Japanses isolate which was proposed as a new subspecies by recent studies. F. tularensis from tick and rodent samples were analysed based on 6 polymorphic loci. Results showed one genotype was prevalence in positive tick samples. Cluster analysis showed there was genotype-host associations among samples but there was no geographic association.To sum it up, the study has shown the persistance of natural foci of tularemia in China and epidemiological characteristics of Chinese strains. We have also established two analysis methods, PFGE and MLVA, for rapid discrimination when disease occurred. Phylogenetic analysis have revealed the relationships not only among Chinese isolates, but with international strains.