Mycobacterium tuberculosis Lineage Influences Innate Immune Response and Virulence and Is Associated with Distinct Cell Envelope Lipid Profiles

et al. (2011) Mycobacterium tuberculosis Lineage Influences Innate Immune Response and Virulence and Is Associated with Distinct Cell Envelope Lipid Profiles. PLoS ONE 6(9): e23870. doi:10.1371/journal.pone.0023870 Mycobacterium tuberculosis Lineage Influences Innate Immune Response and Virulence and Is Associated with Distinct Cell Envelope Lipid Profiles Nitya Krishnan 0 Wladimir Malaga 0 Patricia Constant 0 Maxine Caws 0 Tran Thi Hoang Chau 0 Jenifer Salmons 0 Nguyen Thi Ngoc Lan 0 Nguyen Duc Bang 0 Mamadou Daffe 0 Douglas B. Young 0 Brian D. Robertson 0 Christophe Guilhot 0 Guy E. Thwaites 0 Ludovic Tailleux, Institut Pasteur, France 0 1 Centre for Molecular Microbiology and Infection, Imperial College London , London , United Kingdom , 2 CNRS , IPBS (Institut de Pharmacologie et de Biologie Structurale) , Toulouse, France, 3 Universite de Toulouse, UPS, IPBS, Toulouse , France , 4 Oxford University Clinical Research Unit, Ho Chi Minh City, Vietnam, 5 Hospital for Tropical Diseases , Ho Chi Minh City , Vietnam , 6 Pham Ngoc Thach Hospital for Tuberculosis and Lung Disease, Ho Chi Minh City, Vietnam, 7 National Institute for Medical Research , London , United Kingdom The six major genetic lineages of Mycobacterium tuberculosis are strongly associated with specific geographical regions, but their relevance to bacterial virulence and the clinical consequences of infection are unclear. Previously, we found that in Vietnam, East Asian/Beijing and Indo-Oceanic strains were significantly more likely to cause disseminated tuberculosis with meningitis than those from the Euro-American lineage. To investigate this observation we characterised 7 East Asian/Beijing, 5 Indo-Oceanic and 6 Euro-American Vietnamese strains in bone-marrow-derived macrophages, dendritic cells and mice. East Asian/Beijing and Indo-Oceanic strains induced significantly more TNF-a and IL-1b from macrophages than the Euro-American strains, and East Asian/Beijing strains were detectable earlier in the blood of infected mice and grew faster in the lungs. We hypothesised that these differences were induced by lineage-specific variation in cell envelope lipids. Whole lipid extracts from East Asian/Beijing and Indo-Oceanic strains induced higher concentrations of TNF-a from macrophages than Euro-American lipids. The lipid extracts were fractionated and compared by thin layer chromatography to reveal a distinct pattern of lineageassociated profiles. A phthiotriol dimycocerosate was exclusively produced by East Asian/Beijing strains, but not the phenolic glycolipid previously associated with the hyper-virulent phenotype of some isolates of this lineage. All Indo-Oceanic strains produced a unique unidentified lipid, shown to be a phenolphthiocerol dimycocerosate dependent upon an intact pks15/1 for its production. This was described by Goren as the 'attenuation indictor lipid' more than 40 years ago, due to its association with less virulent strains from southern India. Mutation of pks15/1 in a representative Indo-Oceanic strain prevented phenolphthiocerol dimycocerosate synthesis, but did not alter macrophage cytokine induction. Our findings suggest that the early interactions between M. tuberculosis and host are determined by the lineage of the infecting strain; but we were unable to show these differences are driven by lineage-specific cell-surface expressed lipids. - Large chromosomal deletions, also termed large sequence polymorphisms, classify Mycobacterium tuberculosis into six major lineages, each strongly associated with specific geographical human population [1] (figure 1). Whether bacterial lineage influences the development of tuberculosis disease is uncertain, although there is long-standing evidence that some strains of M. tuberculosis are more virulent than others [2]. Fifty years ago Mitchison and colleagues reported strains of M. tuberculosis from southern India were less virulent in guinea pigs than UK strains [3] and, more recently, others have documented a wide variation in the virulence of M. tuberculosis strains following infection of mice [4,5,6,7] and rabbits [8,9]. The significance of these findings to human disease, however, remains uncertain. Perhaps the most compelling evidence for clinically important M. tuberculosis genetic variation has derived from the study of tuberculosis outbreaks and has suggested the strains responsible cause disease through subversion of the host innate immune response [2]. For example, strain HN878 (a member of the East Asian/Beijing lineage), which caused several disease outbreaks in Texas, USA, was found to be hyper-virulent in animal models [10]. Its unusual virulence was attributed to the production of a cell wall phenolic glycolipid (PGL), which suppressed the release of tumour necrosis factor-alpha (TNF-a) and interleukin (IL) -12p40 from human monocyte-derived macrophages compared with the laboratory strain H37Rv [11]. Subsequently, the effect of PGL was found to be influenced by the background of the strain. M. tuberculosis strain, H37Rv, expressing PGL suppressed proinflammatory cytokine production from human monocytes, in Figure 1. The phylogeny of M. tuberculosis in South East Asia. Large sequence polymorphisms define six major lineages of M. tuberculosis (A) which are strongly associated with specific geographical regions (B). In Vietnam, three lineages cause the majority of disease: the East Asian/Beijing, the Indo-Oceanic, and the Euro-American (adapted from reference 1). Numercial values in the figure represent regions of deletions (RD) that define each of the lineages. doi:10.1371/journal.pone.0023870.g001 vitro, but did not display increased virulence in mice or guinea pigs [12]. Similarly, a strain responsible for an outbreak of tuberculosis amongst school children in the UK (and a member of the AfricanIndian lineage) was also associated with an anti-inflammatory phenotype, causing suppression of IL-12p40 and up-regulation of IL-10 from macrophages [13]. It is uncertain whether these phenotypes, and the mechanisms that drive them, are strain-specific or common to the broader lineages of M. tuberculosis to which they belong. A recent study of a panel of 26 global clinical isolates, with representatives from all six lineages, found considerable heterogeneity in the inflammatory response induced from human macrophages by the strains. However, strains from the modern lineages (Euro-American, East Asian/Bejing and Indian/East African) induced less inflammatory response than those from ancient lineages (Indo-Oceanic and West African) [14]. It is hypothesised that a reduced inflammatory response may confer a selective advantage to strains from the modern lineages, resulting in impaired bacterial control by the host, more rapidly progressive disease, and enhanced transmission. Previously, to investigate whether some lineages of M. tuberculosis may be more capable than others at causing disseminated tuberculosis with meningitis we perfo (...truncated)