Final Report Summary - FROMBCTOBA (Evolution and ecology of Bacillus anthracis: the transition from a soil organism to a mammalian pathogen) Executive Summary:In this project we have used a unique isolate of Bacillus cereus (strain G9241) to understand the evolution of the mammalian pathogen Bacillus anthracis, the causative agent of anthrax. The plasmid pXO1 of B. anthracis encodes a transcriptional regulator called AtxA that regulates the expression at 37ºC of the two major virulence factors, the tripartite toxin and the antiphagocytic capsule. The temperature-dependence of AtxA expression could suggest that it has evolved to control virulence in warm blooded animals rather than insects and would explain why anthrax is primarily a disease of domestic herbivores, while other species of the B. cereus group, remain invertebrate associated. In the B. cereus group, PlcR is a pleiotropic regulator which play an important role in the virulence in insects. This regulator is functionally inactive in B. anthracis due to a point mutation.It has been proposed that in a B. cereus ancestor strain, the horizontal acquisition of AtxA on pXO1 created a dynamic tension between the mammalian virulence regulon and the established insect virulence chromosomal PlcR regulon that resulted in the functional inactivation of the latter.To investigate the evolution between B. cereus and B. anthracis, we have the exceptional B. cereus G9241 strain, which caused an illness resembling inhalational anthrax, yet possess potentially functional versions of both the atxA and plcR regulator genes. This strain contains plasmids pBCXO1 and pBC218, which are seen to be homologues of the pXO1 and pXO2 B. anthracis plasmids.OBJECTIVES1. The initial goal was to investigate if B. cereus G9241 strain is really “half way between” B. cereus and B. anthracis, containing the functional PlcR and also the mammalian virulence factors and high temperature responsive atxA regulon. We attempted to differentiate between two hypothesis; (1) that B. cereus G9241represents an early stage isolate recently derived from an insect pathogenic ancestor, in which the competition of the atxA and plcR regulons has yet to be resolved; or alternatively (2) that the G9241 strain has evolved a mechanism to accommodate both regulators.2. Since the interaction between strain G9241 and insects had not been investigated, the second objective was to evaluate the ability of this strain to infect insects. 3. Once we had confirmed the hypothesis (2) and the defined the infectivity of this strain to insects we added value to the project by defining further objectives not described in the initial application. These objectives focused on the behavior of this pathogen during host interactions and are listed below. 3.a. Testing the toxicity of G9241 strain to insect immune cells (hemocytes) and mammalian cells, including human and sheep erythrocytes, and human immune cells such as lymphocytes, macrophages and neutrophils.3.b. Testing the surface adhesion ability and the cell mobility of G9241 strain under static growth conditions.4. Genomic analysis of the genes involved in virulence and regulation of virulence factors in G9241 strain.SUMMARY OF PROGRESS TOWARD OBJECTIVES AND MAIN RESULTSThe evaluation of the virulence of B. cereus G9241 against insect was carried out with the insect model Manduca sexta. Infections of M. sexta with B. cereus G9241 strain were performed at the rearing temperature for this insect (25ºC), showed a similar level of virulence to a B. cereus type strain. The ability of Bc G9241 to perform in these infections suggested that the mammalian virulence regulator atxA did not overtly effect the activity of the invertebrate virulence regulator PlcR and that, at least in this strain, they can co-exist. To further investigate if the AtxA regulator could exert any effect upon PlcR regulation, an atxA mutant of Bc G9241 was generated by curing Bc G9241 of the pBCXO1 plasmid, upon which AtxA is encoded. This plasmid cured strain (designated Bc G9241-atxA) showed a growth pattern similar to its otherwise isogenic parent strain and it could no longer produce the protective anthrax toxin encoded on pBCXO1. Insect infection experiments with Bc G9241-atxA actually showed a reduction in virulence compared to the parent strain Bc G9241. This not only indicated that AtxA is not influencing or interfering with PlcR, but that the cured pBCXO1 plasmid actually encodes any factor relevant to insect virulence.To study in depth the virulence mechanisms of Bc G9241, we performed experiments to analyse the secretion of lytic enzymes (likely regulated by PlcR) and their toxicity against insect and mammalian cells. Secreted toxicity against M. sexta haemocytes was demonstrated for both the Bc G9241 and Bc G9241-atxA strains when grown at 25ºC. However, this toxicity was lost when these strains were cultured at 37ºC. We also tested the ability of supernatants to lyse sheep and human erythrocytes from strains cultured at these two temperatures. Both sheep and human erythrocytes were lysed by supernatants from Bc G9241 and the Bc G9241-atxA mutant when grown at 25ºC, although the lysis was slightly lower than that of the B. cereus type strain. Interestingly, as was the case with the negative control strain Bt-plcR, no haemolysis of erythrocytes occurred when incubated with supernatants from any of the Bc G9241 strains grown at 37ºC, With the knowledge that Bc G9241 strain was able to infect insects and mammals, and given the findings from the haemolysis tests, we decided to analyse the effect of any secreted virulence factors, against cells of the human immune system. The cytotoxicity of supernatants from 37ºC cultures was tested against three types of human immune cells, macrophages, lymphocytes and neutrophils. Whereas supernatants from the B. cereus type strain produced lysis of the three types of immune cells, the Bc G9241 and Bc G9241-atxA strains did not show any cytotoxic effect. Interestingly, supernatants from the attenuated B. anthracis Sterne strain tested in parallel in these assays also did not result in lytic activity either. The surface adhesion ability and the motility of Bc G9241 and Bc G9241-atxA strains strains were tested and compared with the B. cereus type strain and the Bt-plcR non-virulent strain. The Bc G9241 strains and Bt-plcR strain showed a good ability to adhere to surfaces, as well as less motility in static growth assays compared to the B. cereus type strain. The formation of an unusual novel developmental form of long intertwined cell structures resembling a matrix of ropes was also observed in the Bc G9241 strains.Our work has allowed us to discard the hypothesis that active PlcR and AtxA responsive regulons within the same cell are incompatible. Nevertheless, it should be noted that in all strains of B. anthracis, the plcR regulator gene has been made inactive by mutation. We therefore propose that, rather than incompatibility with AtxA, the simpler hypothesis that an active PlcR regulon may confer some disadvantage in mammal infections. We began the investigation of our new hypothesis by performing a comparative analysis of the genome sequences of the B. cereus type strain, the Bc G9241 strain and that of B. anthracis. Known PlcR-regulated genes involved in virulence were compared. This analysis revealed that numerous B. cereus virulence genes were either mutated or absent in strain Bc G9241. Pertinent examples being: the hblCDBA hemolytic enterotoxin operon, the nheABC non-hemolytic enterotoxin operon, haemolysin II, phosphatidylinositol-specific phospholipase C, microbial collagenases, enhancing and neutral proteases. Interestingly, even though some of these virulence genes remain intact in B. anthracis, it should be noted that they cannot be expressed because of a lack of a functional PlcR global regulator.These observations suggest that, in the absence of a mutation in the plcR gene, that mutations in these multiple virulence genes may have conferred an evolutionary advantage to Bc G9241 in its transition from insect to mammalian pathogen. The overall hypothesis that arises from our work is that B. cereus G9241 as well as B. anthracis have adopted a specific virulence strategy that allows them to infect mammals, which is distinct from that used by their insect associated B. cereus ancestor. This involves the production of the plasmid encoded tripartite toxin, which disables immune cell activity and a capsule that protects the bacteria from a range of immune responses. Furthermore it is believed that spores of B. anthracis must first be taken up by lung macrophages by phagocytosis before they can germinate. It could be anticipated that the production of non-specific lytic enzymes, which appears effective in insects, would negate this otherwise stealth-like approach to establishing a mammalian infection. Moreover, the lack of cytotoxicity of the supernatants from 37ºC strain Bc G9241 cultures suggests that, either all the relevant virulence genes are mutated or that a temperature-regulated system exists which is able to suppress expression of any intact virulence genes that may interfere with mammalian infection. One such mechanism could involve a temperature dependent suppression or failure of the PlcR-PapR quorum-sensing regulator system at 37ºC.
