- Email: [email protected]
Versatile persistence pathways for pathogens of animals and plants
Research Update
much weight. Such a statement is an oversimplification. By only focusing on the differences between the studies, such a claim does not acknowledge the many common conclusions arrived at by O’Quinn et al. [2] and ourselves, such as the susceptibility of C. elegans to different Burkholderia species and the requirement for living bacteria for maximal virulence (just to name a few examples). Woods also incorrectly states that our study reported that Burkholderia thailandensis is avirulent in C. elegans. In Fig. 1a of our report [1], we clearly show that B. thailandensis can kill nematodes, similar to the results reported by O’Quinn et al. [2]. Although our reports do differ on the extent of nematode killing by B. mallei, this apparent difference could result from the use of distinct assay media in the two studies. As we also showed, the environment of the host–pathogen interaction plays an important role in determining the absolute level of nematode toxicity (Fig. 3 in [1]). Addressing this issue is a potential subject for future research. Woods also dismisses as irrelevant our observation that the killing of C. elegans by B. pseudomallei might involve a secreted toxin, as ‘there is no clinical evidence either in melioidosis or glanders that a toxin is important in the pathogenesis of these diseases’. Given our current lack of knowledge concerning these microorganisms, we believe that this might be a somewhat premature conclusion. First, there are several prior reports in the literature that B. pseudomallei is capable of producing a potent cytotoxin [3,4]. Second, to rely solely on clinical evidence, which provides only a snapshot of the disease at the point of examination, carries the potential risk of missing more subtle effects that might not be easily detectable in
TRENDS in Microbiology Vol.10 No.11 November 2002
complex mammalian systems but which could also be instrumental to disease pathogenesis. A good example of this is found in the work of Ausubel and colleagues, who initially used C. elegans to discover that phenazines are important virulence factors in Pseudomonas aeruginosa, and subsequently proved that phenazines also function as multihost pathogenicity factors [5]. Before this work, the role of phenazines in P. aeruginosa virulence had been, from an examination of the available clinical evidence, at best highly controversial, and thus their importance might not have been discovered if not for the use of an alternative host. Compared with higher organisms, there are some notable differences in the interaction of Burkholderia spp. with C. elegans (e.g. B. thailandensis is nematocidal but avirulent towards humans). These differences do not automatically negate the usefulness of C. elegans as an animal host for B. pseudomallei, so long as the results from the C. elegans studies are taken in the proper context and subsequently validated. For example, we identified five potential virulence genes in B. pseudomallei by initially screening for bacterial mutants attenuated in virulence towards C. elegans, and then validated these preliminary findings by showing that the mutants also exhibited reduced virulence in mammalian hosts [1]. These results demonstrate that the simplicity of handling and powerful genetics available in C. elegans offer certain experimental advantages that make it a strong complement to other animal infection models, particularly in the conduct of high-throughput genetic screens. Burkholderia research is still relatively young and unravelling the intricacies of
485
how these organisms cause human morbidity and mortality is an important question. Although C. elegans should not be used as the sole animal model for the study of B. pseudomallei pathogenesis, its use as a model animal host constitutes one promising tool that can be used to complement other existing methodologies to elucidate these complex diseases further. References 1 Gan, Y-H. et al. (2002) Characterization of Burkholderia pseudomallei infection and identification of novel virulence factors using a Caenorhabditis elegans host system. Mol. Microbiol. 44, 1185–1197 2 O’Quinn, A.L. et al. (2001) Burkholderia pseudomallei kills the nematode Caenorhabditis elegans using an endotoxin-mediated paralysis. Cell. Microbiol. 3, 381–393 3 Haase, A. et al. (1997) Toxin production by Burkholderia pseudomallei strains and correlation with severity of melioidosis. J. Med. Microbiol. 46, 557–563 4 Haussler, S. et al. (1998) Purification and characterization of a cytotoxic exolipid of Burkholderia pseudomallei. Infect. Immun. 66, 1588–1593 5 Mahajan-Miklos, S. et al. (1999) Molecular mechanisms of bacterial virulence elucidated using a Pseudomonas aeruginosa–Caenorhabditis elegans pathogenesis model. Cell 96, 47–56
Patrick Tan* Cellular and Molecular Research, National Cancer Centre/Defence Medical Research Institute, 11 Hospital Drive, Singapore 169610, Republic of Singapore. *e-mail: [email protected] Yunn-Hwen Gan Dept of Biochemistry, Faculty of Medicine, National University of Singapore, 10 Medical Drive, Singapore 117597, Republic of Singapore.
Published online: 04 October 2002
Versatile persistence pathways for pathogens of animals and plants Danny Vereecke, Karen Cornelis, Wim Temmerman, Marcelle Holsters and Koen Goethals The glyoxylate cycle and the glycine cleavage system are part of conserved metabolic pathways involved in the chronic persistence of microorganisms in animal hosts. In the chromosome of the plant pathogen Rhodococcus fascians, the vic locus has been identified as a http://tim.trends.com
region containing genes essential for persistence inside induced leafy galls. Sequence analysis showed that this 18-kb locus is syntenic with chromosomal regions of Mycobacterium species that encompass the ‘persistence’ loci of these mammalian pathogens.
Hence, the ability to switch diet inside the host appears to be governed by ‘persistence’ enzymes that are conserved between pathogens of animals and plants. Published online: 26 September 2002
0966-842X/02/$ – see front matter © 2002 Elsevier Science Ltd. All rights reserved. PII: S0966-842X(02)02457-5
Research Update
486
TRENDS in Microbiology Vol.10 No.11 November 2002
MI 66 20
67 20 68 20
69 20
70 20
71 20
72 20
73 20 74 20 5 7 20 6 7 20 7 7 2078 2079 20 80 20 81 20
82 20
ML
Rf 12
gc
gc
v
1kb
a
B
v
15 16 17 13 14 ORF5 vicA ORF3 ORF2 ORF1
B1
11
B
s
pg
9 10
H
c
se
A2
A2
Gene 0 1 2 3 4 5 6 7 8
c
gl
gu
Mt c 43 18
c 42 18 c 41 18 c 40 18 c 39 18 c 38 18 c 37 18
c 36 18
c 35 18
34 18 c 33 18
32 18
31 18 0 3 18 29 18 8 2 18 7 2 18 26 18 5 2 18 4 2 18 3 2 18 2 2 18
21 18
Rv
TRENDS in Microbiology
Fig. 1. Overview of the genes located in the syntenic regions of the chromosomes of Mycobacterium tuberculosis (Mt), Mycobacterium leprae (Ml) and Rhodococcus fascians (Rf). ORFs that occur in all three bacteria are in red, those found only in the mycobacteria in yellow, only in R. fascians in blue, those only in M. tuberculosis in green, and those that are absent in M. leprae in brown; the pale colored ORFs (2080, 2079, 2078, 2071, 2068 and 2067) in M. leprae and gene 10 in R. fascians represent pseudogenes; triangles on the R. fascians sequence indicate the location of mutations.
An important feature of the lifestyle of inter- and intracellular pathogens is the acquisition of nutrients from their host. For intracellular pathogens of mammals, a two-carbon diet through the glyoxylate cycle is crucial for virulence and persistence. The persistence of the bacterium Mycobacterium tuberculosis and the yeast Candida albicans inside macrophages depends on the activity of the glyoxylate shunt of the tricarboxylic acid (TCA) cycle to allow growth in an environment where fatty acids are probably the main carbon source [1–3]. Here, we show that the phytopathogenic bacterium Rhodococcus fascians has similarly adopted the glyoxylate shunt as a persistence pathway in infected plant tissues. Persistence of pathogens of mammals
The glyoxylate shunt is a bypass of the TCA cycle that permits gluconeogenesis starting from acetyl-coenzyme A (acetyl-CoA), which is generated following fatty acid catabolism [4]. The shunt is composed of isocitrate lyase, which cleaves isocitrate to succinate and glyoxylate, and malate synthase, which condenses glyoxylate and acetyl-CoA to malate. As such, it circumvents the CO2-generating steps of the TCA cycle and converts one molecule each of acetyl-CoA and isocitrate to two C4 compounds that can be fed into biosynthetic processes. Mycobacterial persistence occurs after the emergence of the adaptive immune response. The intracellular environment of activated macrophages becomes more hostile and bacterial growth is http://tim.trends.com
restrained [3]. The pathogen adapts to the nutrient deprivation by shifting its metabolism towards the degradation of fatty acids. Expression of the isocitrate lyase gene is induced after phagocytosis in activated macrophages and its deletion results in reduced virulence, coinciding with a loss of persistence during lung infection of mice [1]. The reliance on a fatty acid diet is further illustrated by the relative abundance of M. tuberculosis genes encoding proteins involved in fatty acid oxidation [5]. Using whole-genome microarray analysis, the genes of the glyoxylate cycle have been shown to be induced in phagocytosed Saccharomyces cerevisiae cells [2]. In the related fungal pathogen C. albicans, both the malate synthase and isocitrate lyase genes are induced upon contact with macrophages [2]. When the isocitrate lyase gene is deleted, the virulence of C. albicans is reduced, suggesting that the interior of a macrophage is a glucose-deficient environment. These observations show that for these bacterial and fungal pathogens of mammals persistence and virulence rely on the ability to switch diet within host cells. Furthermore, they suggest that an active glyoxylate cycle has widespread significance in the strategies used by these pathogens. Persistence of the plant pathogen Rhodococcus fascians
The genus Rhodococcus is closely related to Mycobacterium and includes the species Rhodococcus equi, a facultative
intracellular pathogen of macrophages in different animals [6]. The phytopathogenic species R. fascians infects a wide range of plants, provoking the formation of leafy galls consisting of masses of shoot buds that are suppressed for further growth [7]. Pathogenesis relies on a linear plasmid carrying virulence genes involved in synthesizing signal compounds that initiate cell division in plant cortical tissues, leading to the formation of shoot meristems [8,9]. Upon epiphytic colonization, endophytic forms are found in the intercellular spaces of gall tissues and, sometimes, inside plant cells [10]. The persistent presence and activity of this bacterial subpopulation is essential for the growth and maintenance of the leafy gall [11,12]. We have characterized a chromosomal locus (vic) that contributes to virulence: an insertion mutant was strongly affected in gall formation and the few buds formed were weakly suppressed for further growth. The insertion mutation maps to a gene homologous to malate synthase and although the vic mutant grew equally well in rich medium and defined glycerol medium, it failed to grow in a medium with acetate as sole carbon source. The mutant also lacked traceable malate synthase activity, and accumulated glyoxylate in acetate medium. The expression of the malate synthase gene was induced by extracts of plant and gall tissues. Growth of the mutant ceased rapidly when confronted with gall extracts, but not with extracts from uninfected plants. In symptomatic plant tissues the number of vic mutant bacteria was significantly lower than that of wild-type bacteria. These results suggest that there is a shift in the diet of R. fascians during plant infection that requires specific metabolic reactions involving vic-encoded biochemistry. The absence of an active malate synthase in the vic mutant results in the accumulation of glyoxylate when bacteria are grown on gall extracts. Glyoxylate accumulation interferes with the metabolism of the bacteria and, ultimately, their viability, with a reduced virulence phenotype as a consequence [12]. Syntenic regions comprise ‘persistence’ loci
Interestingly, the malate synthase gene is part of a locus carrying 14 genes that
Research Update
TRENDS in Microbiology Vol.10 No.11 November 2002
Table 1. Syntenic genes in Mycobacterium tuberculosis, M. leprae and Rhodococcus fascians M. tuberculosis (bp, aa) M. leprae (bp, aa) Rv1821 (2424, 808)
b
R. fascians (bp, aa)
ML2082 (2337, 778)
Rv1822 (627, 209)
ML2081 (621, 206)
Gene 1 (606, 202)
Rv1823 (921, 307) Rv1824 (363, 121) Rv1825 (876, 292) Rv1826 (402, 134)
ML2080 (927, PG) ML2079 (325, PG) ML2078 (692, PG) ML2077 (399, 132)
Gene 2 (855, 285) Gene 3 (441, 147) Gene 4 (732, 244) Gene 5 (423, 141)
Rv1827 (486, 162) Rv1828 (741, 247) Rv1829 (492, 164) Rv1830 (675, 225) Rv1831 (255, 85) Rv1832 (2823, 941)
ML2076 (489, 162) ML2075 (756, 251) ML2074 (495, 164) ML2073 (696, 231)
Gene 6 (498, 166) Gene 7 (738, 245) Gene 8 (474, 158) Gene 9 (678, 226) Gene 10 (315, PG) Gene 11 (2853, 951) Gene 12 (1137, 379)
Rv1833c (858, 286) Rv1834 (864, 288) Rv1835c (1884, 628) Rv1836c (2031, 677) Rv1837c (2223, 741) Rv1838c (393, 131) Rv1839c (261, 87) Rv1840c (1545, 515)
a
c
Function
secA2
Gene 0 (627, 209)
ML2072 (2859, 952)
Gene
487
ML2071 (803, PG)
ML2070 (2202, 733) ML2069 (2196, 731)
ORF5 (1715, 571) vicA (2174, 724)
Rv1841c (1035, 345)
ML2068 (948, PG)
ORF3 (915, 305) ORF2 (1144, 381)
Rv1842c (1365, 455)
ML2067 (1303, PG)
ORF1 (1350, 450)
Rv1843c (1437, 479)
ML2066 (1437, 478)
Preprotein translocase subunit (Class III.D: protein and peptide secretion) mutT2 Homologue of MutT/NudiX-family protein of Brucella melitensis (Q8YII2) pgsA2 CDP-diacylglycerol-glycerol-3-P-3-phosphatidyltransferase (Class I.H.3: lipid biosynthesis) Conserved hypothetical protein (Class V) Membrane protein (Class II.C.5: cell envelope) Unknown hypothetical protein (Class VI) gcvH Glycine cleavage system protein H (Class I.C.1: general central intermediary metabolism) Conserved hypothetical protein (Class V) Conserved hypothetical protein (Class V) Conserved hypothetical protein (Class V) Conserved hypothetical protein (Class V) Unknown hypothetical protein (Class VI) gcvB Glycine cleavage system protein P (Class I.C.1:general central intermediary metabolism) Homologue of putative integral membrane protein of Streptomyces coelicolor (AL353872) Class IV.I: miscellaneous phosphatases, lyases and hydrolases Conserved hypothetical protein (Class V) Conserved hypothetical protein (Class V) Conserved hypothetical protein (Class V) glcB Malate synthase (Class I.B.4: glyoxylate bypass) Conserved hypothetical protein (Class V) Conserved hypothetical protein (Class V) PE_PGRS Class IV.C.1.b: PE family, PGRS subfamily Unknown hypothetical protein Putative integral membrane protein (Class II.C.5: cell envelope) Putative integral membrane protein (Class II.C.5: cell envelope) guaB1 Inosine-5'-monophosphate dehydrogenase (Class I.F.1: purine ribonucleotide biosynthesis)
a
Abbreviations: aa, amino acid(s); bp, base pair(s); PG, pseudogene. The sequence of the region of the R. fascians chromosome is submitted under accession number AJ301559. c The protein classes are according to the functional classification of the gene products of the Mycobacterium genomes (http://www.sanger.ac.uk/Projects/M_tuberculosis/ and http://www.sanger.ac.uk/Projects/M_leprae/). b
are highly conserved both in sequence and organization with the corresponding genomic regions of M. tuberculosis and Mycobacterium leprae (Fig. 1). Also comparable to M. tuberculosis, the malate synthase gene of R. fascians is not linked to the previously cloned isocitrate lyase gene; its gene product has high sequence identity (55%) with malate synthase G, which is involved in glycolate utilization, and a lower identity (20%) with malate synthase A of Escherichia coli [5,12–14]. In addition to genes encoding hypothetical and membrane-targeted proteins, the syntenic region also carries the genes comprising the glycine cleavage system (Table 1), which is responsible for the oxidative decarboxylation of glycine, generating http://tim.trends.com
CO2 and NADH. This pathway is thought to be involved (as glycine synthase in the reverse reaction) in the maintenance of the NAD pool under O2-limiting conditions in M. tuberculosis [15] and is required for chronic persistence of the intracellular pathogen Brucella abortus [16]. Mutagenesis of the R. fascians gene encoding the glycine dehydrogenase P homologue results in a marked decrease in virulence. This gene is not transcribed in a medium containing glycine as the sole nitrogen source, as has been found for the genes of E. coli and S. cerevisiae [17, 18], but, similar to the malate synthase gene, expression is strongly induced by extracts of infected plant tissues (D. Vereecke et al., unpublished).
What are the nutrient sources for R. fascians in leafy galls?
In congruence with Mycobacterium, the vic locus together with isocitrate lyase can be involved in the specific metabolism of fatty acids in leafy gall tissues. However, the physiology of the niche that is occupied by R. fascians – arial plant parts that are active in photosynthesis – could suggest other gall-derived nutrients. From our data, it is conceivable that the vic locus functions in the metabolism of glycine, glycolate and/or glyoxylate, in a pathway that does not require isocitrate lyase. Interestingly, these compounds are shuttled between several cellular organelles of C3 plants during a process known as photorespiration [19,20]. Assuming that photorespiration is high in leafy galls, these intermediates could be
488
Research Update
produced in sufficiently large amounts to be scavenged by R. fascians and metabolized by the vic-derived enzymes. Future research
Several questions remain to be answered to demonstrate further the analogy between the strategies used by R. fascians and M. tuberculosis and their respective hosts. What is the relationship between the metabolic adaptation and the intra- and intercellular forms of R. fascians? Are there leafy gall-abundant fatty acids that can sustain growth of R. fascians? Is the photorespiration level higher in leafy galls? Are the photorespiration intermediates used by R. fascians? If so, how are they obtained by the intercellular population? What is the role of isocitrate lyase in these processes, and hence in virulence? In conclusion, it is apparent that R. fascians and its close relative M. tuberculosis have adopted similar metabolic pathways that are linked to persistence in animal hosts or particular plant tissues. The integration of the corresponding biochemistry into the host’s metabolism might differ between pathogens of plants and animals, but relies on common genetic grounds, thus illustrating the versatility of these persistence pathways in adaptation to different host environments. Acknowledgements
We thank Wilson Ardiles for sequencing the vic locus, Martine De Cock for help with the manuscript, and Rebecca Verbanck for artwork. W.T. and K.C. are indebted to the Vlaams Instituut voor de
TRENDS in Microbiology Vol.10 No.11 November 2002
Bevordering van het WetenschappelijkTechnologisch Onderzoek in de Industrie and the ‘von Humbolt Gesellschaft’ for a postdoctoral fellowship, respectively. References 1 McKinney, J.D. et al. (2000) Persistence of Mycobacterium tuberculosis in macrophages and mice requires the glyoxylate shunt enzyme isocitrate lyase. Nature 406, 735–738 2 Lorenz, M.C. and Fink, G.R. (2001) The glyoxylate cycle is required for fungal virulence. Nature 412, 83–86 3 Höner zu Bentrup, K. and Russell, D.G. (2001) Mycobacterial persistence: adaptation to a changing environment. Trends Microbiol. 9, 597–605 4 Kornberg, H.L. (1966) The role and control of the glyoxylate cycle in Escherichia coli. Biochem. J. 99, 1–11 5 Cole, S.T. et al. (1998) Deciphering the biology of Mycobacterium tuberculosis from the complete genome sequence. Nature 393, 537–544 6 Hondalus, M.K. (1997) Pathogenesis and virulence of Rhodococcus equi. Vet. Microbiol. 16, 257–268 7 Goethals, K. et al. (2001) Leafy gall formation by Rhodococcus fascians. Annu. Rev. Phytopathol. 39, 27–52 8 Crespi, M. et al. (1992) Fasciation induction by the phytopathogen Rhodococcus fascians depends upon a linear plasmid encoding a cytokinin synthase gene. EMBO J. 11, 795–804 9 de O. Manes, C-L. et al. (2001) De novo cortical cell division triggered by the phytopathogen Rhodococcus fascians in tobacco. Mol. Plant–Microbe Interact. 14, 189–195 10 Cornelis, K. et al. (2001) The plant pathogen Rhodococcus fascians colonizes the exterior and interior of the aerial parts of plants. Mol. Plant–Microbe Interact. 14, 599–608 11 Vereecke, D. et al. (2000) The Rhodococcus fascians–plant interaction: morphological traits and biotechnological applications. Planta 210, 241–251 12 Vereecke, D. et al. (2002) Chromosomal locus that affects the pathogenicity of Rhodococcus fascians. J. Bacteriol. 184, 1112–1120 13 Vereecke, D. et al. (1994) Cloning and sequence analysis of the gene encoding isocitrate
14
15
16
17
18
19
20
lyase from Rhodococcus fascians. Gene 145, 109–114 Höner zu Bentrup, K. et al. (1999) Characterization of activity and expression of isocitrate lyase in Mycobacterium avium and Mycobacterium tuberculosis. J. Bacteriol. 181, 7161–7167 Wayne L.G. and Lin K-Y. (1982) Glyoxylate metabolism and adaptation of Mycobacterium tuberculosis to survival under anaerobic conditions. Infect. Immun. 37, 1042–1049 Hong P.C. et al. (2000) Identification of genes required for chronic persistence of Brucella abortus in mice. Infect. Immun. 68, 4102–4107 Stauffer, L.T. and Stauffer G.V. (1994) Characterization of the gcv control region from Escherichia coli. J. Bacteriol. 176, 6159–6164 Sinclair, D.A. et al. (1996) Specific induction by glycine of the gene for the P-subunit of glycine decarboxylase from Saccharomyces cerevisiae. Mol. Microbiol. 19, 611–623 Husic, D.W. et al. (1987) The oxidative photosynthetic carbon cycle or C2 cycle. CRC Crit. Rev. Plant Sci. 5, 45–100 Kozaki, A. and Takeba, G. (1996) Photorespiration protects C3 plants from photooxidation. Nature 384, 557–560
Danny Vereecke Wim Temmerman Marcelle Holsters* Koen Goethals Dept of Plant Systems Biology, Flanders Interuniversity Institute for Biotechnology, Ghent University, K.L. Ledeganckstraat 35, B-9000 Gent, Belgium. *e-mail: [email protected] Karen Cornelis Present address: Entwicklungsgenetik, Zentrum für Molekularbiologie der Pflanzen, Universität Tübingen, Auf der Morgenstelle 3, D-72076 Tübingen, Germany.
Letters to Trends in Microbiology Would you like to respond to any of the issues raised in this month’s Trends in Microbiology? Letters to the Editor can have a maximum length of 750 words and can include up to ten references and one figure or table. If you are interested in contributing a letter, please contact the Editor ([email protected]) with a brief outline. Letters commenting on Trends in Microbiology articles are always sent to the appropriate authors for review and to give them the opportunity to reply.
Please note: submission does not guarantee publication.
http://tim.trends.com
much weight. Such a statement is an oversimplification. By only focusing on the differences between the studies, such a claim does not acknowledge the many common conclusions arrived at by O’Quinn et al. [2] and ourselves, such as the susceptibility of C. elegans to different Burkholderia species and the requirement for living bacteria for maximal virulence (just to name a few examples). Woods also incorrectly states that our study reported that Burkholderia thailandensis is avirulent in C. elegans. In Fig. 1a of our report [1], we clearly show that B. thailandensis can kill nematodes, similar to the results reported by O’Quinn et al. [2]. Although our reports do differ on the extent of nematode killing by B. mallei, this apparent difference could result from the use of distinct assay media in the two studies. As we also showed, the environment of the host–pathogen interaction plays an important role in determining the absolute level of nematode toxicity (Fig. 3 in [1]). Addressing this issue is a potential subject for future research. Woods also dismisses as irrelevant our observation that the killing of C. elegans by B. pseudomallei might involve a secreted toxin, as ‘there is no clinical evidence either in melioidosis or glanders that a toxin is important in the pathogenesis of these diseases’. Given our current lack of knowledge concerning these microorganisms, we believe that this might be a somewhat premature conclusion. First, there are several prior reports in the literature that B. pseudomallei is capable of producing a potent cytotoxin [3,4]. Second, to rely solely on clinical evidence, which provides only a snapshot of the disease at the point of examination, carries the potential risk of missing more subtle effects that might not be easily detectable in
TRENDS in Microbiology Vol.10 No.11 November 2002
complex mammalian systems but which could also be instrumental to disease pathogenesis. A good example of this is found in the work of Ausubel and colleagues, who initially used C. elegans to discover that phenazines are important virulence factors in Pseudomonas aeruginosa, and subsequently proved that phenazines also function as multihost pathogenicity factors [5]. Before this work, the role of phenazines in P. aeruginosa virulence had been, from an examination of the available clinical evidence, at best highly controversial, and thus their importance might not have been discovered if not for the use of an alternative host. Compared with higher organisms, there are some notable differences in the interaction of Burkholderia spp. with C. elegans (e.g. B. thailandensis is nematocidal but avirulent towards humans). These differences do not automatically negate the usefulness of C. elegans as an animal host for B. pseudomallei, so long as the results from the C. elegans studies are taken in the proper context and subsequently validated. For example, we identified five potential virulence genes in B. pseudomallei by initially screening for bacterial mutants attenuated in virulence towards C. elegans, and then validated these preliminary findings by showing that the mutants also exhibited reduced virulence in mammalian hosts [1]. These results demonstrate that the simplicity of handling and powerful genetics available in C. elegans offer certain experimental advantages that make it a strong complement to other animal infection models, particularly in the conduct of high-throughput genetic screens. Burkholderia research is still relatively young and unravelling the intricacies of
485
how these organisms cause human morbidity and mortality is an important question. Although C. elegans should not be used as the sole animal model for the study of B. pseudomallei pathogenesis, its use as a model animal host constitutes one promising tool that can be used to complement other existing methodologies to elucidate these complex diseases further. References 1 Gan, Y-H. et al. (2002) Characterization of Burkholderia pseudomallei infection and identification of novel virulence factors using a Caenorhabditis elegans host system. Mol. Microbiol. 44, 1185–1197 2 O’Quinn, A.L. et al. (2001) Burkholderia pseudomallei kills the nematode Caenorhabditis elegans using an endotoxin-mediated paralysis. Cell. Microbiol. 3, 381–393 3 Haase, A. et al. (1997) Toxin production by Burkholderia pseudomallei strains and correlation with severity of melioidosis. J. Med. Microbiol. 46, 557–563 4 Haussler, S. et al. (1998) Purification and characterization of a cytotoxic exolipid of Burkholderia pseudomallei. Infect. Immun. 66, 1588–1593 5 Mahajan-Miklos, S. et al. (1999) Molecular mechanisms of bacterial virulence elucidated using a Pseudomonas aeruginosa–Caenorhabditis elegans pathogenesis model. Cell 96, 47–56
Patrick Tan* Cellular and Molecular Research, National Cancer Centre/Defence Medical Research Institute, 11 Hospital Drive, Singapore 169610, Republic of Singapore. *e-mail: [email protected] Yunn-Hwen Gan Dept of Biochemistry, Faculty of Medicine, National University of Singapore, 10 Medical Drive, Singapore 117597, Republic of Singapore.
Published online: 04 October 2002
Versatile persistence pathways for pathogens of animals and plants Danny Vereecke, Karen Cornelis, Wim Temmerman, Marcelle Holsters and Koen Goethals The glyoxylate cycle and the glycine cleavage system are part of conserved metabolic pathways involved in the chronic persistence of microorganisms in animal hosts. In the chromosome of the plant pathogen Rhodococcus fascians, the vic locus has been identified as a http://tim.trends.com
region containing genes essential for persistence inside induced leafy galls. Sequence analysis showed that this 18-kb locus is syntenic with chromosomal regions of Mycobacterium species that encompass the ‘persistence’ loci of these mammalian pathogens.
Hence, the ability to switch diet inside the host appears to be governed by ‘persistence’ enzymes that are conserved between pathogens of animals and plants. Published online: 26 September 2002
0966-842X/02/$ – see front matter © 2002 Elsevier Science Ltd. All rights reserved. PII: S0966-842X(02)02457-5
Research Update
486
TRENDS in Microbiology Vol.10 No.11 November 2002
MI 66 20
67 20 68 20
69 20
70 20
71 20
72 20
73 20 74 20 5 7 20 6 7 20 7 7 2078 2079 20 80 20 81 20
82 20
ML
Rf 12
gc
gc
v
1kb
a
B
v
15 16 17 13 14 ORF5 vicA ORF3 ORF2 ORF1
B1
11
B
s
pg
9 10
H
c
se
A2
A2
Gene 0 1 2 3 4 5 6 7 8
c
gl
gu
Mt c 43 18
c 42 18 c 41 18 c 40 18 c 39 18 c 38 18 c 37 18
c 36 18
c 35 18
34 18 c 33 18
32 18
31 18 0 3 18 29 18 8 2 18 7 2 18 26 18 5 2 18 4 2 18 3 2 18 2 2 18
21 18
Rv
TRENDS in Microbiology
Fig. 1. Overview of the genes located in the syntenic regions of the chromosomes of Mycobacterium tuberculosis (Mt), Mycobacterium leprae (Ml) and Rhodococcus fascians (Rf). ORFs that occur in all three bacteria are in red, those found only in the mycobacteria in yellow, only in R. fascians in blue, those only in M. tuberculosis in green, and those that are absent in M. leprae in brown; the pale colored ORFs (2080, 2079, 2078, 2071, 2068 and 2067) in M. leprae and gene 10 in R. fascians represent pseudogenes; triangles on the R. fascians sequence indicate the location of mutations.
An important feature of the lifestyle of inter- and intracellular pathogens is the acquisition of nutrients from their host. For intracellular pathogens of mammals, a two-carbon diet through the glyoxylate cycle is crucial for virulence and persistence. The persistence of the bacterium Mycobacterium tuberculosis and the yeast Candida albicans inside macrophages depends on the activity of the glyoxylate shunt of the tricarboxylic acid (TCA) cycle to allow growth in an environment where fatty acids are probably the main carbon source [1–3]. Here, we show that the phytopathogenic bacterium Rhodococcus fascians has similarly adopted the glyoxylate shunt as a persistence pathway in infected plant tissues. Persistence of pathogens of mammals
The glyoxylate shunt is a bypass of the TCA cycle that permits gluconeogenesis starting from acetyl-coenzyme A (acetyl-CoA), which is generated following fatty acid catabolism [4]. The shunt is composed of isocitrate lyase, which cleaves isocitrate to succinate and glyoxylate, and malate synthase, which condenses glyoxylate and acetyl-CoA to malate. As such, it circumvents the CO2-generating steps of the TCA cycle and converts one molecule each of acetyl-CoA and isocitrate to two C4 compounds that can be fed into biosynthetic processes. Mycobacterial persistence occurs after the emergence of the adaptive immune response. The intracellular environment of activated macrophages becomes more hostile and bacterial growth is http://tim.trends.com
restrained [3]. The pathogen adapts to the nutrient deprivation by shifting its metabolism towards the degradation of fatty acids. Expression of the isocitrate lyase gene is induced after phagocytosis in activated macrophages and its deletion results in reduced virulence, coinciding with a loss of persistence during lung infection of mice [1]. The reliance on a fatty acid diet is further illustrated by the relative abundance of M. tuberculosis genes encoding proteins involved in fatty acid oxidation [5]. Using whole-genome microarray analysis, the genes of the glyoxylate cycle have been shown to be induced in phagocytosed Saccharomyces cerevisiae cells [2]. In the related fungal pathogen C. albicans, both the malate synthase and isocitrate lyase genes are induced upon contact with macrophages [2]. When the isocitrate lyase gene is deleted, the virulence of C. albicans is reduced, suggesting that the interior of a macrophage is a glucose-deficient environment. These observations show that for these bacterial and fungal pathogens of mammals persistence and virulence rely on the ability to switch diet within host cells. Furthermore, they suggest that an active glyoxylate cycle has widespread significance in the strategies used by these pathogens. Persistence of the plant pathogen Rhodococcus fascians
The genus Rhodococcus is closely related to Mycobacterium and includes the species Rhodococcus equi, a facultative
intracellular pathogen of macrophages in different animals [6]. The phytopathogenic species R. fascians infects a wide range of plants, provoking the formation of leafy galls consisting of masses of shoot buds that are suppressed for further growth [7]. Pathogenesis relies on a linear plasmid carrying virulence genes involved in synthesizing signal compounds that initiate cell division in plant cortical tissues, leading to the formation of shoot meristems [8,9]. Upon epiphytic colonization, endophytic forms are found in the intercellular spaces of gall tissues and, sometimes, inside plant cells [10]. The persistent presence and activity of this bacterial subpopulation is essential for the growth and maintenance of the leafy gall [11,12]. We have characterized a chromosomal locus (vic) that contributes to virulence: an insertion mutant was strongly affected in gall formation and the few buds formed were weakly suppressed for further growth. The insertion mutation maps to a gene homologous to malate synthase and although the vic mutant grew equally well in rich medium and defined glycerol medium, it failed to grow in a medium with acetate as sole carbon source. The mutant also lacked traceable malate synthase activity, and accumulated glyoxylate in acetate medium. The expression of the malate synthase gene was induced by extracts of plant and gall tissues. Growth of the mutant ceased rapidly when confronted with gall extracts, but not with extracts from uninfected plants. In symptomatic plant tissues the number of vic mutant bacteria was significantly lower than that of wild-type bacteria. These results suggest that there is a shift in the diet of R. fascians during plant infection that requires specific metabolic reactions involving vic-encoded biochemistry. The absence of an active malate synthase in the vic mutant results in the accumulation of glyoxylate when bacteria are grown on gall extracts. Glyoxylate accumulation interferes with the metabolism of the bacteria and, ultimately, their viability, with a reduced virulence phenotype as a consequence [12]. Syntenic regions comprise ‘persistence’ loci
Interestingly, the malate synthase gene is part of a locus carrying 14 genes that
Research Update
TRENDS in Microbiology Vol.10 No.11 November 2002
Table 1. Syntenic genes in Mycobacterium tuberculosis, M. leprae and Rhodococcus fascians M. tuberculosis (bp, aa) M. leprae (bp, aa) Rv1821 (2424, 808)
b
R. fascians (bp, aa)
ML2082 (2337, 778)
Rv1822 (627, 209)
ML2081 (621, 206)
Gene 1 (606, 202)
Rv1823 (921, 307) Rv1824 (363, 121) Rv1825 (876, 292) Rv1826 (402, 134)
ML2080 (927, PG) ML2079 (325, PG) ML2078 (692, PG) ML2077 (399, 132)
Gene 2 (855, 285) Gene 3 (441, 147) Gene 4 (732, 244) Gene 5 (423, 141)
Rv1827 (486, 162) Rv1828 (741, 247) Rv1829 (492, 164) Rv1830 (675, 225) Rv1831 (255, 85) Rv1832 (2823, 941)
ML2076 (489, 162) ML2075 (756, 251) ML2074 (495, 164) ML2073 (696, 231)
Gene 6 (498, 166) Gene 7 (738, 245) Gene 8 (474, 158) Gene 9 (678, 226) Gene 10 (315, PG) Gene 11 (2853, 951) Gene 12 (1137, 379)
Rv1833c (858, 286) Rv1834 (864, 288) Rv1835c (1884, 628) Rv1836c (2031, 677) Rv1837c (2223, 741) Rv1838c (393, 131) Rv1839c (261, 87) Rv1840c (1545, 515)
a
c
Function
secA2
Gene 0 (627, 209)
ML2072 (2859, 952)
Gene
487
ML2071 (803, PG)
ML2070 (2202, 733) ML2069 (2196, 731)
ORF5 (1715, 571) vicA (2174, 724)
Rv1841c (1035, 345)
ML2068 (948, PG)
ORF3 (915, 305) ORF2 (1144, 381)
Rv1842c (1365, 455)
ML2067 (1303, PG)
ORF1 (1350, 450)
Rv1843c (1437, 479)
ML2066 (1437, 478)
Preprotein translocase subunit (Class III.D: protein and peptide secretion) mutT2 Homologue of MutT/NudiX-family protein of Brucella melitensis (Q8YII2) pgsA2 CDP-diacylglycerol-glycerol-3-P-3-phosphatidyltransferase (Class I.H.3: lipid biosynthesis) Conserved hypothetical protein (Class V) Membrane protein (Class II.C.5: cell envelope) Unknown hypothetical protein (Class VI) gcvH Glycine cleavage system protein H (Class I.C.1: general central intermediary metabolism) Conserved hypothetical protein (Class V) Conserved hypothetical protein (Class V) Conserved hypothetical protein (Class V) Conserved hypothetical protein (Class V) Unknown hypothetical protein (Class VI) gcvB Glycine cleavage system protein P (Class I.C.1:general central intermediary metabolism) Homologue of putative integral membrane protein of Streptomyces coelicolor (AL353872) Class IV.I: miscellaneous phosphatases, lyases and hydrolases Conserved hypothetical protein (Class V) Conserved hypothetical protein (Class V) Conserved hypothetical protein (Class V) glcB Malate synthase (Class I.B.4: glyoxylate bypass) Conserved hypothetical protein (Class V) Conserved hypothetical protein (Class V) PE_PGRS Class IV.C.1.b: PE family, PGRS subfamily Unknown hypothetical protein Putative integral membrane protein (Class II.C.5: cell envelope) Putative integral membrane protein (Class II.C.5: cell envelope) guaB1 Inosine-5'-monophosphate dehydrogenase (Class I.F.1: purine ribonucleotide biosynthesis)
a
Abbreviations: aa, amino acid(s); bp, base pair(s); PG, pseudogene. The sequence of the region of the R. fascians chromosome is submitted under accession number AJ301559. c The protein classes are according to the functional classification of the gene products of the Mycobacterium genomes (http://www.sanger.ac.uk/Projects/M_tuberculosis/ and http://www.sanger.ac.uk/Projects/M_leprae/). b
are highly conserved both in sequence and organization with the corresponding genomic regions of M. tuberculosis and Mycobacterium leprae (Fig. 1). Also comparable to M. tuberculosis, the malate synthase gene of R. fascians is not linked to the previously cloned isocitrate lyase gene; its gene product has high sequence identity (55%) with malate synthase G, which is involved in glycolate utilization, and a lower identity (20%) with malate synthase A of Escherichia coli [5,12–14]. In addition to genes encoding hypothetical and membrane-targeted proteins, the syntenic region also carries the genes comprising the glycine cleavage system (Table 1), which is responsible for the oxidative decarboxylation of glycine, generating http://tim.trends.com
CO2 and NADH. This pathway is thought to be involved (as glycine synthase in the reverse reaction) in the maintenance of the NAD pool under O2-limiting conditions in M. tuberculosis [15] and is required for chronic persistence of the intracellular pathogen Brucella abortus [16]. Mutagenesis of the R. fascians gene encoding the glycine dehydrogenase P homologue results in a marked decrease in virulence. This gene is not transcribed in a medium containing glycine as the sole nitrogen source, as has been found for the genes of E. coli and S. cerevisiae [17, 18], but, similar to the malate synthase gene, expression is strongly induced by extracts of infected plant tissues (D. Vereecke et al., unpublished).
What are the nutrient sources for R. fascians in leafy galls?
In congruence with Mycobacterium, the vic locus together with isocitrate lyase can be involved in the specific metabolism of fatty acids in leafy gall tissues. However, the physiology of the niche that is occupied by R. fascians – arial plant parts that are active in photosynthesis – could suggest other gall-derived nutrients. From our data, it is conceivable that the vic locus functions in the metabolism of glycine, glycolate and/or glyoxylate, in a pathway that does not require isocitrate lyase. Interestingly, these compounds are shuttled between several cellular organelles of C3 plants during a process known as photorespiration [19,20]. Assuming that photorespiration is high in leafy galls, these intermediates could be
488
Research Update
produced in sufficiently large amounts to be scavenged by R. fascians and metabolized by the vic-derived enzymes. Future research
Several questions remain to be answered to demonstrate further the analogy between the strategies used by R. fascians and M. tuberculosis and their respective hosts. What is the relationship between the metabolic adaptation and the intra- and intercellular forms of R. fascians? Are there leafy gall-abundant fatty acids that can sustain growth of R. fascians? Is the photorespiration level higher in leafy galls? Are the photorespiration intermediates used by R. fascians? If so, how are they obtained by the intercellular population? What is the role of isocitrate lyase in these processes, and hence in virulence? In conclusion, it is apparent that R. fascians and its close relative M. tuberculosis have adopted similar metabolic pathways that are linked to persistence in animal hosts or particular plant tissues. The integration of the corresponding biochemistry into the host’s metabolism might differ between pathogens of plants and animals, but relies on common genetic grounds, thus illustrating the versatility of these persistence pathways in adaptation to different host environments. Acknowledgements
We thank Wilson Ardiles for sequencing the vic locus, Martine De Cock for help with the manuscript, and Rebecca Verbanck for artwork. W.T. and K.C. are indebted to the Vlaams Instituut voor de
TRENDS in Microbiology Vol.10 No.11 November 2002
Bevordering van het WetenschappelijkTechnologisch Onderzoek in de Industrie and the ‘von Humbolt Gesellschaft’ for a postdoctoral fellowship, respectively. References 1 McKinney, J.D. et al. (2000) Persistence of Mycobacterium tuberculosis in macrophages and mice requires the glyoxylate shunt enzyme isocitrate lyase. Nature 406, 735–738 2 Lorenz, M.C. and Fink, G.R. (2001) The glyoxylate cycle is required for fungal virulence. Nature 412, 83–86 3 Höner zu Bentrup, K. and Russell, D.G. (2001) Mycobacterial persistence: adaptation to a changing environment. Trends Microbiol. 9, 597–605 4 Kornberg, H.L. (1966) The role and control of the glyoxylate cycle in Escherichia coli. Biochem. J. 99, 1–11 5 Cole, S.T. et al. (1998) Deciphering the biology of Mycobacterium tuberculosis from the complete genome sequence. Nature 393, 537–544 6 Hondalus, M.K. (1997) Pathogenesis and virulence of Rhodococcus equi. Vet. Microbiol. 16, 257–268 7 Goethals, K. et al. (2001) Leafy gall formation by Rhodococcus fascians. Annu. Rev. Phytopathol. 39, 27–52 8 Crespi, M. et al. (1992) Fasciation induction by the phytopathogen Rhodococcus fascians depends upon a linear plasmid encoding a cytokinin synthase gene. EMBO J. 11, 795–804 9 de O. Manes, C-L. et al. (2001) De novo cortical cell division triggered by the phytopathogen Rhodococcus fascians in tobacco. Mol. Plant–Microbe Interact. 14, 189–195 10 Cornelis, K. et al. (2001) The plant pathogen Rhodococcus fascians colonizes the exterior and interior of the aerial parts of plants. Mol. Plant–Microbe Interact. 14, 599–608 11 Vereecke, D. et al. (2000) The Rhodococcus fascians–plant interaction: morphological traits and biotechnological applications. Planta 210, 241–251 12 Vereecke, D. et al. (2002) Chromosomal locus that affects the pathogenicity of Rhodococcus fascians. J. Bacteriol. 184, 1112–1120 13 Vereecke, D. et al. (1994) Cloning and sequence analysis of the gene encoding isocitrate
14
15
16
17
18
19
20
lyase from Rhodococcus fascians. Gene 145, 109–114 Höner zu Bentrup, K. et al. (1999) Characterization of activity and expression of isocitrate lyase in Mycobacterium avium and Mycobacterium tuberculosis. J. Bacteriol. 181, 7161–7167 Wayne L.G. and Lin K-Y. (1982) Glyoxylate metabolism and adaptation of Mycobacterium tuberculosis to survival under anaerobic conditions. Infect. Immun. 37, 1042–1049 Hong P.C. et al. (2000) Identification of genes required for chronic persistence of Brucella abortus in mice. Infect. Immun. 68, 4102–4107 Stauffer, L.T. and Stauffer G.V. (1994) Characterization of the gcv control region from Escherichia coli. J. Bacteriol. 176, 6159–6164 Sinclair, D.A. et al. (1996) Specific induction by glycine of the gene for the P-subunit of glycine decarboxylase from Saccharomyces cerevisiae. Mol. Microbiol. 19, 611–623 Husic, D.W. et al. (1987) The oxidative photosynthetic carbon cycle or C2 cycle. CRC Crit. Rev. Plant Sci. 5, 45–100 Kozaki, A. and Takeba, G. (1996) Photorespiration protects C3 plants from photooxidation. Nature 384, 557–560
Danny Vereecke Wim Temmerman Marcelle Holsters* Koen Goethals Dept of Plant Systems Biology, Flanders Interuniversity Institute for Biotechnology, Ghent University, K.L. Ledeganckstraat 35, B-9000 Gent, Belgium. *e-mail: [email protected] Karen Cornelis Present address: Entwicklungsgenetik, Zentrum für Molekularbiologie der Pflanzen, Universität Tübingen, Auf der Morgenstelle 3, D-72076 Tübingen, Germany.
Letters to Trends in Microbiology Would you like to respond to any of the issues raised in this month’s Trends in Microbiology? Letters to the Editor can have a maximum length of 750 words and can include up to ten references and one figure or table. If you are interested in contributing a letter, please contact the Editor ([email protected]) with a brief outline. Letters commenting on Trends in Microbiology articles are always sent to the appropriate authors for review and to give them the opportunity to reply.
Please note: submission does not guarantee publication.
http://tim.trends.com