Previous Article | Next Article ![]()
Journal of Clinical Microbiology, April 2003, p. 1600-1608, Vol. 41, No. 4
0095-1137/03/$08.00+0 DOI: 10.1128/JCM.41.4.1600-1608.2003
Copyright © 2003, American Society for Microbiology. All Rights Reserved.
Department of Immunology and Medicine, Armed Forces Research Institute of Medical Sciences, Bangkok,1 Institut de Recherche pour le Développement, UR34, Mahidol University at Salaya, Nakorn Prathom, Thailand,4 Unité des Rickettsies, CNRS UMR 6020, Faculté de Médecine, Marseille, France,2 Laboratory of Public Health Entomology, Harvard School of Public Health, Boston, Massachusetts,3 Bao Loc General Hospital, Bao Loc, Lam Dong Province, Vietnam5
Received 2 May 2002/ Returned for modification 6 November 2002/ Accepted 23 January 2003
|
|
|---|
|
|
|---|
Eight tick-borne rickettsioses with distinct species as agents have definitively been described throughout the world, including Rickettsia rickettsii (in the Americas), Rickettsia sibirica (in Asia), Rickettsia conorii including different strains (in Europe, Asia, and Africa), Rickettsia australis (in Australia), Rickettsia honei (in the Flinders Island, Australia), Rickettsia japonica (in Japan), Rickettsia africae (in sub-Saharan Africa and the West Indies), and Rickettsia slovaca (in Europe) (16, 19). Furthermore, Astrakhan fever rickettsia and Israeli tick typhus rickettsia, both of which are closely related to R. conorii, are known as agents of rickettsioses in Astrakhan and Israel, respectively (19). Finally, human infections due to Rickettsia aeschlimannii in Africa (18), Rickettsia helvetica and "Rickettsia mongolotimonae" in Europe (4, 5), and "Rickettsia heilongjiangii" in Asia (30) have recently been reported.
Although ehrlichioses have been recognized as infectious diseases only in animals for a long time, they are now known to be important emerging zoonoses in people. Three human ehrlichioses have been reported since 1991. They include (i) human monocytic ehrlichiosis due to Ehrlichia chaffeensis in the United States, (ii) infections due to Ehrlichia ewingii in the United States, and (iii) human granulocytic ehrlichiosis due to Anaplasma phagocytophilum (formerly named human granulocytic ehrlichia or Ehrlichia phagocytophila), which occur both in the United States and in Europe (2, 3).
In Asia, tick-borne SFG rickettsioses and ehrlichioses have been poorly studied. SFG rickettsioses have been reported from Thailand, for example, but to date the cases have been confirmed solely by general SFG serology (24). The etiologic agent(s) has never been specifically identified by isolation or molecular characterization. Two SFG rickettsiae have been identified in ticks in Thailand, including (i) Rickettsia honei and its strain, Thai tick typhus Rickettsia strain TT-118, and (ii) a new rickettsia of unknown pathogenicity, "Rickettsia thailandii sp. nov." (10, 19). Their roles as agents of human diseases in Thailand are not known.
Ehrlichioses of veterinary importance are known to occur in Thailand, mainly canine ehrlichiosis due to Ehrlichia canis, which is transmitted by the brown dog tick (Rhipicephalus sanguineus). Coinfection with three Ehrlichia species has also been reported in dogs (26). Although human ehrlichioses have been suspected in Thailand on the basis of serological data (7), no known human pathogens have been identified from patients or in ticks that bite humans.
Tick-borne diseases including SFG rickettsioses and ehrlichioses have been suspected to occur among local Karen, Mon, Burmese, and Thai rural residents living in the central part of the Thai-Myanmar border region (Sangkhlaburi District, Kanchanaburi Province, Thailand) [R. S. Miller, P. McDaniel, S. Nedek, N. Thanoosingha, N. Buathong, S. Sriwichai, A. Weld, S. R. Telford III, and C. Wongsrichanalai, Program Abstr. 49th Annu. Meet. Am. Soc. Trop. Med. Hyg., Am. J. Trop. Med. Hyg. 62(Suppl. 3):469-470, 2000]. Recently, new molecular methods have enabled the development of useful, sensitive, and rapid tools for the detection and identification of tick-borne pathogens in arthropods including ticks (25). Accordingly, in an effort to identify the possible etiologic agents for SFG rickettsioses and ehrlichioses affecting humans in these sites, we analyzed ticks collected from peridomestic or wild animals in the Sangkhlaburi District for evidence of rickettsial infections. Additional ticks collected in Vietnam were included in this work.
(These results were presented in part at the 3rd International Conference on Emerging Infectious Diseases, 24 to 27 March 2002, Atlanta, Ga.)
|
|
|---|
![]() View larger version (42K): [in a new window] |
FIG. 1. Map showing the locations where ticks were collected in Thailand and Vietnam and the results of the molecular detection of rickettsial DNA (in the following order: bacteria/tick species/number of ticks tested/number of positive ticks/host).
|
The DNA extracted from the ticks was also screened as described previously with primers EHR16SR and EHR16SD (Bioprobe Systems), which amplify a 345-bp fragment of the 16S rRNA gene of bacteria within the family Anaplasmataceae, including the genera Anaplasma, Ehrlichia, Neorickettsia, and Wolbachia (17). To amplify the main part of the 16S rRNA gene, tick DNA samples that were found to be positive with the primers described above were subjected to a second PCR with primers EHR16SR and EHR16SD coupled with universal primers fD1 and rp2, as described previously (8). Distilled water and Anaplasma phagocytophilum DNA were used in each test as negative and positive controls, respectively. After electrophoresis the amplification products were visualized on 1% agarose gels stained with ethidium bromide and examined by UV transillumination. A DNA molecular weight marker (marker VI; Boehringer Mannheim, Mannheim, Germany) was used to estimate the sizes of the products.
Gene sequencing and phylogenetic analysis. The PCR products were purified with QIAquick PCR purification kits (QIAGEN GmbH, Hilden, Germany), and DNA sequencing was performed by use of the fluorescence-labeled dideoxynucleotide technology in an ABI 3100 automated DNA sequencer (Perkin-Elmer, Applied Biosystems Division). Sequences were assembled and edited with AutoAssembler software (version 1.4; Perkin-Elmer). Multiple-sequence alignments with the corresponding sequences were performed by use of the ClustalW program (28). Phylogenetic and molecular evolutionary analyses were conducted by using MEGA software (version 2.1) (12). Phylogenetic trees were inferred from the multiple-sequence alignments, after the removal of all gaps, by the neighbor-joining method (MEGA software, version 2.1). The distance matrix was calculated by use of Kimura-2 parameters. Five hundred bootstrap replicates were used to estimate the reliabilities of the nodes on the phylogenetic trees.
Nucleotide sequence accession numbers. The nucleotide sequences of the 16S rRNA genes of the bacteria of the family Anaplasmataceae used for phylogenetic studies are available in GenBank under the following accession numbers: AF303467 for Anaplasma platys (formerly Ehrlichia platys), AF286699 for A. platys detected in Thailand, M73224 for A. phagocytophilum (formerly Ehrlichia phagocytophila), U03775 for Anaplasma bovis (formerly Ehrlichia bovis), AF283007 for Anaplasma centrale, M60313 for Anaplasma marginale, AF318945 for Anaplasma ovis, AF069758 for Ehrlichia ruminatium, M73222 for E. chaffeensis, U15527 for Ehrlichia muris, M73227 for E. ewingii, M73221 for E. canis, AF311967 for Ehrlichia sp. strain ERm58, AF311968 for Ehrlichia sp. strain EHt224; AF414399 for Ehrlichia sp. strain Tibet; AF179630 for Wolbachia pipientis; U12457 for Neorickettsia helminthoeca; M73225 for Neorickettsia sennetsu (formerly Ehrlichia sennetsu); M21290 for Neorickettsia risticii (formerly E. sennetsu), U11021 for R. rickettsii; and M21789 for Rickettsia prowazekii. The nucleotide sequences of the 16S rRNA genes of the bacteria found in this study have been deposited in GenBank under the following accession numbers: AF497576 for Anaplasma sp. strain AnDa465; AF497580 for Anaplasma sp. strain AnAj360; AF497579 for Anaplasma sp. strain AnHl446; AF497581 for Ehrlichia sp. strain EBm52; AF497578 for Ehrlichia sp. strain EHh324; AF497577 for Ehrlichia sp. strain EHh317; AF497583 for Eubacterium sp. strain Hw124, and AF497582 for Eubacterium sp. strain Hw191.
The nucleotide sequences of the citrate synthase gene (gltA) of the following rickettsiae used for comparison and phylogenetic studies are deposited in GenBank under the indicated accession numbers: Rickettsia parkeri, U59732; R. sibirica, U59734; "R. mongolotimonae," U59731; strain S, U59735; R. africae, U59733; R. conorii Seven, U59730; R. rickettsii, U59729; Astrakhan fever rickettsia, U59728; Israeli tick typhus rickettsia, U59727; R. honei strain RB, AF018074; Thai tick typhus rickettsia, U59726; R. slovaca, U59725; R. japonica, U59724; Rickettsia rhipicephali, U59721; R. montanensis U74756; Rickettsia massiliae, U59719; Bar29, U59720; R. aeschlimannii, U59722; R. helvetica, U59723; Rickettsia sp. strain IRS4, AF141906; Rickettsia sp. strain IRS3, AF140706; R. australis, U59718; Rickettsia akari, U59717; Rickettsia typhi, U59714; Rickettsia canadensis, 59713; AB bacterium, U59712; R. prowazekii, U59715; R. bellii, U59716; "Rickettsia hulinii," AF172943; "R. heilongjiangii," AF178034; Rickettsia sp. strain DnS14, AF120028; Rickettsia sp. strain RpA4, AF120029; Rickettsia sp. strain DnS28, AF120027; Rickettsia sp. strain RDa420, AF497584; and Rickettsia sp. strain RDla440, AF497585. The nucleotide sequences of the gltA genes found in this study have been deposited in GenBank under the following accession numbers: AF497584 for Rickettsia sp. strain RDa420 and AF497585 for Rickettsia sp. strain RDla440.
|
|
|---|
|
View this table: [in a new window] |
TABLE 1. Characteristics of ticks collecteda
|
Three 16S rRNA sequences of Anaplasma spp. were identified (Fig. 1). They included (i) AnDa465 (1,016 bp), which was obtained from 3 of 20 (15%; 95% CI, 0 to 30.6) Dermacentor auratus nymphs collected from dogs in Thailand and which was closely related (99.3% similarity) to the sequence of A. platys (formerly named E. platys); (ii) AnAj360 (955 bp), which was obtained from 16 of 54 (29.6%; 95% CI, 17.4 to 41.8%) adult Amblyomma javanense ticks collected from a pangolin and which showed 97.9% similarity with A. phagocytophilum and 97.8% similarity with A. platys and A. bovis (formerly E. bovis); and (iii) AnHl446 (1,014 bp), which was identified in 3 of 8 (37.5%; 95% CI, 4 to 71%) female Haemaphysalis lagrangei ticks collected from a bear from Thailand and which showed 99.6% similarity with A. bovis, 97.9% similarity with A. phagocytophilum, and 96.5% similarity with A. platys.
Three 16S rRNA sequences of Ehrlichia spp. were identified (Fig. 1). They included (i) EBm52 (1,380 bp), which was obtained from 24 of 109 (22%; 95% CI, 14.2 to 29.8%) Boophilus microplus ticks collected from cattle from Thailand and which was shown to be closely related (99.9% similarity) to a sequence identified in B. microplus ticks collected from cattle in Tibet (GenBank accession number AF414399) and also to be closely related (99.6% similarity) to two ehrlichial DNA sequences detected in cattle ticks from Africa (Erm58 and Eht224; GenBank accession numbers AF311967 and AF311968, respectively); (ii) EHh324 (902 bp), which was detected in 1 of 19 (5.3%; 95% CI, 0 to 15.3%) Haemaphysalis hystricis ticks collected from wild pigs in Vietnam and which was closely related to E. chaffeensis (99.4% similarity); and (iii) EHh317 (902 bp), which was also detected in 1 of 19 H. hystricis ticks from Vietnamese wild pigs (the tick was different from the tick positive for EHh324) but which appeared to be closely related to Ebm52 and the related sequences described above (99% similarity).
Furthermore, two 16S rRNA sequences designated Hw124 and Hw191 were identified in 2 of 55 (3.6%; 95% CI 0 to 8.5%) Haemaphysalis wellingtoni nymphs collected from chickens in Thailand (Fig. 1). Compared with the sequences available in GenBank, the sequences of both Hw124 and Hw191 appeared to differ from those of all known bacteria. The most closely related sequences available in GenBank had been deposited under the name "endosymbiont of Acanthamoeba sp." (GenBank accession number AF069962; 93% similarity when 790-bp sequences were compared) and the name "Eubacterium ZI-8" (GenBank accession number AJ292457; 96% similarity when 536-bp sequences were compared).
By a neighbor-joining analysis (Fig. 2) based on the alignments of 960 bp of the 16S rRNA genes, AnDa465 clustered with A. platys sequences. AnHl446 and AnAj360 were placed within the Anaplasma clade as well. The bacteria from which these sequences originated were temporarily called Anaplasma sp. strain AnDa465, Anaplasma sp. strain AnHl446, and Anaplasma sp. strain AnAj360, respectively. EBm52, EHh324, and EHh317 were shown to belong to the Ehrlichia clade; and the bacteria from which these sequences originated were temporarily called Ehrlichia sp. strain EBm52, Ehrlichia sp. strain EHh324, and Ehrlichia sp. strain EHh317, respectively. Furthermore, this analysis suggested that strains Hw124 and Hw191 may originate from a new group of bacteria. The organisms from which these sequences originated were temporarily called Eubacterium sp. strain Hw124 and Eubacterium sp. strain Hw191, respectively.
![]() View larger version (23K): [in a new window] |
FIG. 2. Phylogenetic tree based on studies of 960 sites of the 16S RNA genes of bacteria of the genera Anaplasma, Ehrlichia, Neorickettsia, and Wolbachia and drawn by using MEGA software (version 2.1) (12). The distance matrix was calculated by using Kimura-2 parameters. Trees were obtained by the neighbor-joining method. The numbers at the nodes are the proportions of 500 bootstrap resamplings that support the topology shown. The bacteria detected in this work are highlighted.
|
![]() View larger version (25K): [in a new window] |
FIG. 3. Phylogenetic tree based on studies of 1,035 sites of the citrate synthase genes of bacteria of the genus Rickettsia, drawn as described in the legend to Fig. 2. The bacteria detected in this work are highlighted.
|
|
|
|---|
The 16S rRNA sequences of the Anaplasma spp. and Ehrlichia spp. identified in this work were not identical to any sequence deposited in GenBank. Although a failure to match a sequence to one in GenBank is all too often taken to imply novelty, we prefer to remain conservative about such a conclusion. We provide designations to refer to the organisms that are represented by the DNA sequences that were found in ticks but do not suggest that these necessarily comprise novel entities. Within the family Anaplasmataceae, there is at present no consensus as to the degree of 16S rRNA gene dissimilarity which should be evident to distinguish two bacterial species as opposed to that which should be evident to represent natural genetic variation. A recent study suggested that 0.5% divergence in the 16S RNA gene sequence of bacteria within this family could be considered a cutoff (29); however, it seems prudent to await phylogenetic studies based on sequences of other genes before such a recommendation is accepted.
In this work, we identified Anaplasma sp. strain AnDa465 in D. auratus ticks from Thailand. According to the level of 16 sRNA gene sequence similarity (99.3%) and to our phylogenetic analysis, Anaplasma sp. strain AnDa465 appeared to be closely related to A. platys, the agent of asymptomatic to mild infectious cyclic thrombocytopenia in dogs. A similar sequence had also recently been detected in dogs from Thailand (26). As suggested in the latter work, we believe that the few differences noticed in the 16S rRNA gene sequences compared to those available in GenBank might be due to sequencing or PCR error, but they may possibly be due to variations in the sequences of strains of the same species as well. We have recently detected A. platys isolates from a dog and a tick in the Democratic Republic of Congo, and although the 16S rRNA gene sequences of the isolates showed some differences from the A. platys sequences deposited in GenBank, additional analyses based on groESL and gltA gene phylogenies suggest that they are a strain of A. platys (Y. O. Sanogo, unpublished data). Our findings thus confirm the presence of A. platys in Thailand. This bacterium has recently been detected in Japan in the brown dog tick (R. sanguineus), which could serve as a vector (9). For the first time, D. auratus was implicated here as a potential vector of canine infectious cyclic thrombocytopenia, although it is possible that the ticks that yielded the A. platys amplification products could have been feeding on bacteremic dogs.
Two more Anaplasma spp. were detected in this work. Anaplasma sp. strain AnAj360 was detected from A. javanense ticks from a pangolin. Two specimens of this animal were screened for ticks, but only ticks collected from one of them were positive. The sequence of this Anaplasma strain presented less than 98% similarity with those of other Anaplasma spp., and it could represent a new species. On the other hand, Anaplasma sp. strain AnHl446 was detected from H. lagrangei ticks from a bear. It appeared to be closely related to A. bovis according to the level of 16S rRNA gene sequence similarity (99.6%) and phylogenetic analysis; in the phylogenetic tree, it is grouped with A. bovis with 100 as a bootstrap value. Thus, Anaplasma sp. strain AnHl446 could be a strain of A. bovis, although as stated above, this cannot be definitely assumed.
Ehrlichia sp. strain EBm52 was obtained from B. microplus ticks collected from cattle. This tick species is well known as the vector of A. marginale, an intraerythrocytic pathogen that causes bovine anaplasmosis (13). This agent is distributed worldwide, and it was recently detected in B. microplus ticks collected from cattle in Myanmar (P. Parola, unpublished data). However, Ehrlichia sp. strain EBm52 was shown to be closely related to Ehrlichia sp. strain Tibet (99.9% sequence similarity) identified in B. microplus ticks collected in Tibet (29). The latter ehrlichia was recently presented as a new species within the genus Ehrlichia, based on phylogenetic analyses of the 16S rRNA gene (29). Furthermore, both Ehrlichia sp. strain Tibet and Ehrlichia sp. strain EBm52 were shown to be closely related to two ehrlichiae which we detected in African ticks, including Rhipicephalus muhsamae ticks from Mali (Ehrlichia sp. strain Erm58) and Hyalomma truncatum ticks from Niger (Ehrlichia sp. strain Eht224) (15). The gene sequences of all four of these ehrlichiae were detected in ticks removed from cattle. Given the high degree of genetic similarity, these organisms may represent strains of the same species, all of which are associated with cattle. Their zoonotic or veterinary potential remains to be described. However, B. microplus rarely bites people (if ever), and the transmission of Ehrlichia sp. strain EBm52 to humans by B. microplus is unlikely.
Although E. chaffeensis and A. phagocytophilum seroreactivities have previously been reported in humans in Thailand [7; Miller et al., Program Abstr. 49th Annu. Meet. Am. Soc. Trop. Med. Hyg., Am. J. Trop. Med. Hyg. 62(Suppl. 3):469-470, 2000], we failed to detect these known agents of human ehrlichioses in that country. However, two Ehrlichia spp. were detected in ticks from Vietnam. Ehrlichia sp. strain EHh317 was detected in H. hystricis ticks collected from wild pigs and clustered with Ehrlichia sp. strain EBm52 and Ehrlichia sp. strain Tibet. The zoonotic potentials of these entities remain undescribed, but because H. hystricis ticks are known to feed on humans, human exposure to Ehrlichia sp. strain EHh317 might confound epidemiological surveys for evidence of infection with known Ehrlichia spp. Serological cross-reactivity among Ehrlichia spp. is well known, and it may be that human exposure to EHh317-like agents may give rise to a response detectable with E. chaffeensis antigen. This is also particularly applicable to Ehrlichia sp. strain EHh324: it was detected in another specimen of H. hystricis ticks from Vietnam and is closely related to E. chaffeensis, the agent of cases of human monocytic ehrlichioses occurring in the United States. Indeed, the 16S rRNA gene sequence of Ehrlichia sp. strain EHh324 was shown to have 99.4% similarity with that of E. chaffeensis, and the strain was shown to cluster with E. chaffeensis in our phylogenetic tree. Thus, Ehrlichia sp. strain EHh324 could be a strain of E. chaffeensis. Although human monocytic ehrlichiosis has not yet been described in Asia, our work and that of others (1, 22) suggest the potential for its existence, particularly where ticks such as H. hystricis (which feed on peridomestic animals as well as humans) are common.
Two Rickettsia spp. have been detected in this work, including Rickettsia sp. strain RDa420 and Rickettsia sp. strain RDla440, from D. auratus ticks and pools of Dermacentor larvae from Sangkhlaburi, Thailand, respectively. Rickettsia sp. strain RDla440 was found to be closely related to Rickettsia sp. strain DnS14 and Rickettsia sp. strain RpA4S14, which are within the SFG rickettsiae. These rickettsiae of unknown pathogenicity have only recently been detected in Russia from several species of the genus Dermacentor and from Rhipicephalus pumilio ticks (21, 23). On the other hand, Rickettsia sp. strain RDa420 was found to be very different from all the known SFG rickettsiae: it was not possible to amplify DNA by PCR with primers CS1d and CS890r, which are known to amplify most of the SFG rickettsiae with only some exceptions (for example, R. akari and R. australis). The strain most closely related to RDa420 was found to be R. bellii, a rickettsia of unknown pathogenicity which is no longer considered to belong to the SFG of the genus Rickettsia, although its taxonomic position is disputed (19). This is the first description of these rickettsiae, and therefore, their epidemiological importance has yet to be determined; but both were detected in the tick species that readily bite humans.
Finally, two bacteria including Eubacterium sp. strain Hw124 and Eubacterium sp. strain Hw191 were detected in H. wellingtoni nymphs collected from chickens in Thailand. They may represent novel bacteria. Clearly, the specificity of our broad-spectrum primers for ehrlichia-like bacteria (primers EHR16SR and EHR16SD) is not absolute, although they were designed to amplify a 345-bp fragment of the 16S rRNA gene specific for the members of the family Anaplasmataceae and have successfully been used as epidemiologic tools in Africa and Japan (8, 15, 17). Our first hypothesis regarding these sequences was to consider them nonspecific amplification products as a result of a PCR or sequencing error. However, the sequences are related to those previously deposited in GenBank as an "endosymbiont of Acanthamoeba sp." (6) and "Eubacterium ZI-8" (the sequence available in GenBank is from an unpublished work), respectively. Furthermore, we have recently analyzed the sequences of PCR products obtained with primers EHR16SR and EHR16SD from ticks collected in the United States and Italy, and new sequences that clustered together in a clade including Eubacterium sp. strain Hw124 and Eubacterium sp. strain Hw191 were obtained (Y. O. Sanogo, unpublished). Thus, we hypothesize that the microorganism from which the sequences originated might represent a new group of bacteria associated with ticks. Further studies are needed, however, and in particular, isolation and polyphasic characterization of these bacteria would be required for rigorous testing of this hypothesis.
In conclusion, PCR assays and sequence analysis of PCR products have enabled us to provide further information on the epidemiology of tick-associated bacteria in Thailand and Vietnam, where little information on the subject exists. Bacteria closely related to animal or human pathogens as well as bacteria of unknown pathogenicities have been detected in this work. However, DNA detection does not imply a transmission competence for the tick vectors concerned, because they could have been removed from bacteremic animals. Also, because of the limited study period and the seasonal variations in tick activity, other tick species could be prevalent during the other half of the year. It would be of particular interest to determine whether R. sanguineus, Amblyomma testudinarium, and Ixodes spp. are also prevalent in Sangkhlaburi, Thailand. Indeed, two rickettsiae (including the pathogen R. honei) are known to be associated with Ixodes granulatus ticks in another location in Thailand (10). A new rickettsia has also recently been detected in A. testudinarium ticks from central Thailand (J.-P. Gonzalez, unpublished data). These findings should stimulate further investigations of the epidemiology of SFG rickettsioses and ehrlichioses in that part of the world.
This work was supported by the U.S. Department of Defense Global Emerging Infections Surveillance Program (DoD-GEIS) and NIH grant AI 39002. This work is a part of the postdoctoral research project of Philippe Parola, who had been supported during different periods by Fondation Bayer Santé, Ministère Français des Affaires Etrangères (Programme Lavoisier), Fondation pour la Recherche Médicale, Assistance Publique- Hôpitaux de Marseille, Institut de Recherche International Servier, Association des Professeurs de Pathologie Infectieuses et Tropicales, and the European Society of Clinical Microbiology and Infectious Diseases.
The views of the authors do not purport to reflect the position of the U.S. Army or the U.S. Department of Defense. The funding agencies take no responsibility for the data and the views expressed in this article.
|
|
|---|
This article has been cited by other articles:
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Copyright © 2009 by the American Society for Microbiology. For an alternate route to Journals.ASM.org, visit: http://intl-journals.asm.org | More Info»