A novel gluconeogenic route enables efficient use of erythritol in zoonotic Brucella.

Brucellosis is a worldwide extended zoonosis caused by pathogens of the genus Brucella. While most B. abortus, B. melitensis, and B. suis biovars grow slowly in complex media, they multiply intensely in livestock genitals and placenta indicating high metabolic capacities. Mutant analyses in vitro and in infection models emphasize that erythritol (abundant in placenta and genitals) is a preferred substrate of brucellae, and suggest hexoses, pentoses, and gluconeogenic substrates use in host cells. While Brucella sugar and erythritol catabolic pathways are known, growth on 3-4 carbon substrates persists in Fbp- and GlpX-deleted mutants, the canonical gluconeogenic fructose 1,6-bisphosphate (F1,6bP) bisphosphatases. Exploiting the prototrophic and fast-growing properties of B. suis biovar 5, we show that gluconeogenesis requires fructose-bisphosphate aldolase (Fba); the existence of a novel broad substrate bisphosphatase (Bbp) active on sedoheptulose 1,7-bisphosphate (S1,7bP), F1,6bP, and other phosphorylated substrates; that Brucella Fbp unexpectedly acts on S1,7bP and F1,6bP; and that, while active in B. abortus and B. melitensis, GlpX is disabled in B. suis biovar 5. Thus, two Fba-dependent reactions (dihydroxyacetone-phosphate + glyceraldehyde 3-phosphate ⇌ F1,6bP; and dihydroxyacetone-phosphate + erythrose 4-phosphate ⇌ S1,7bP) can, respectively, yield fructose 6-phosphate and sedoheptulose 7-phosphate for classical gluconeogenesis and the Pentose Phosphate Shunt (PPS), the latter reaction opening a new gluconeogenic route. Since erythritol generates the PPS-intermediate erythrose 4-phosphate, and the Fba/Fbp-Bbp route predicts sedoheptulose 7-phosphate generation from erythrose 4-phosphate, we re-examined the erythritol connections with PPS. Growth on erythritol required transaldolase or the Fba/Fbp-Bbp pathway, strongly suggesting that Fba/Fbp-Bbp works as a PPS entry for both erythritol and gluconeogenic substrates in Brucella. We propose that, by increasing erythritol channeling into PPS through these peculiar routes, brucellae proliferate in livestock genitals and placenta in the high numbers that cause abortion and infertility, and make brucellosis highly contagious. These findings could be the basis for developing attenuated brucellosis vaccines safer in pregnant animals.

Cell and Tissue Tropism of Brucella spp.

Brucella spp. are facultatively intracellular bacteria that can infect, survive, and multiply in various host cell types in vivo and/or in vitro. The genus Brucella has markedly expanded in recent years with the identification of novel species and hosts, which has revealed additional information about the cell and tissue tropism of these pathogens. Classically, Brucella spp. are considered to have tropism for organs that contain large populations of phagocytes such as lymph nodes, spleen, and liver, as well as for organs of the genital system, including the uterus, epididymis, testis, and placenta. However, experimental infections of several different cultured cell types indicate that Brucella may actually have a broader cell tropism than previously thought. Indeed, recent studies indicate that certain Brucella species in particular hosts may display a pantropic distribution in vivo. This review discusses the available knowledge on cell and tissue tropism of Brucella spp. in natural infections of various host species, as well as in experimental animal models and cultured cells.

The Phospholipid N-Methyltransferase and Phosphatidylcholine Synthase Pathways and the ChoXWV Choline Uptake System Involved in Phosphatidylcholine Synthesis Are Widely Conserved in Most, but Not All Brucella Species.

The brucellae are facultative intracellular bacteria with a cell envelope rich in phosphatidylcholine (PC). PC is abundant in eukaryotes but rare in prokaryotes, and it has been proposed that Brucella uses PC to mimic eukaryotic-like features and avoid innate immune responses in the host. Two PC synthesis pathways are known in prokaryotes: the PmtA-catalyzed trimethylation of phosphatidylethanolamine and the direct linkage of choline to CDP-diacylglycerol catalyzed by the PC synthase Pcs. Previous studies have reported that B. abortus and B. melitensis possess non-functional PmtAs and that PC is synthesized exclusively via Pcs in these strains. A putative choline transporter ChoXWV has also been linked to PC synthesis in B. abortus. Here, we report that Pcs and Pmt pathways are active in B. suis biovar 2 and that a bioinformatics analysis of Brucella genomes suggests that PmtA is only inactivated in B. abortus and B. melitensis strains. We also show that ChoXWV is active in B. suis biovar 2 and conserved in all brucellae except B. canis and B. inopinata. Unexpectedly, the experimentally verified ChoXWV dysfunction in B. canis did not abrogate PC synthesis in a PmtA-deficient mutant, which suggests the presence of an unknown mechanism for obtaining choline for the Pcs pathway in Brucella. We also found that ChoXWV dysfunction did not cause attenuation in B. suis biovar 2. The results of these studies are discussed with respect to the proposed role of PC in Brucella virulence and how differential use of the Pmt and Pcs pathways may influence the interactions of these bacteria with their mammalian hosts.

Development of attenuated live vaccine candidates against swine brucellosis in a non-zoonotic B. suis biovar 2 background.

Brucella is a genus of gram-negative bacteria that cause brucellosis. B. abortus and B. melitensis infect domestic ruminants while B. suis (biovars 1-3) infect swine, and all these bacteria but B. suis biovar 2 are zoonotic. Live attenuated B. abortus S19 and B. melitensis Rev1 are effective vaccines in domestic ruminants, though both can infect humans. However, there is no swine brucellosis vaccine. Here, we investigated the potential use as vaccines of B. suis biovar 2 rough (R) lipopolysaccharide (LPS) mutants totally lacking O-chain (Bs2ΔwbkF) or only producing internal O-chain precursors (Bs2Δwzm) and mutants with a smooth (S) LPS defective in the core lateral branch (Bs2ΔwadB and Bs2ΔwadD). We also investigated mutants in the pyruvate phosphate dikinase (Bs2ΔppdK) and phosphoenolpyruvate carboxykinase (Bs2ΔpckA) genes encoding enzymes bridging phosphoenolpyruvate and the tricarboxylic acid cycle. When tested in the OIE mouse model at the recommended R or S vaccine doses (108 and 105 CFU, respectively), CFU/spleen of all LPS mutants were reduced with respect to the wild type and decreased faster for the R than for the S mutants. At those doses, protection against B. suis was similar for Bs2ΔwbkF, Bs2Δwzm, Bs2ΔwadB and the Rev1 control (105 CFU). As described before for B. abortus, B. suis biovar 2 carried a disabled pckA so that a double mutant Bs2ΔppdKΔpckA had the same metabolic phenotype as Bs2ΔppdK and ppdK mutation was enough to generate attenuation. At 105 CFU, Bs2ΔppdK also conferred the same protection as Rev1. As compared to other B. suis vaccine candidates described before, the mutants described here simultaneously carry irreversible deletions easy to identify as vaccine markers, lack antibiotic-resistance markers and were obtained in a non-zoonotic background. Since R vaccines should not elicit antibodies to the S-LPS and wzm mutants carry immunogenic O-chain precursors and did not improve Bs2ΔwbkF, the latter seems a better R vaccine candidate than Bs2Δwzm. However, taking into account that all R vaccines interfere in ELISA and other widely used assays, whether Bs2ΔwbkF is advantageous over Bs2ΔwadB or Bs2ΔppdK requires experiments in the natural host.

Glucose Oxidation to Pyruvate Is Not Essential for Brucella suis Biovar 5 Virulence in the Mouse Model.

Brucella species cause brucellosis, a worldwide extended zoonosis. The brucellae are related to free-living and plant-associated α2-Proteobacteria and, since they multiply within host cells, their metabolism probably reflects this adaptation. To investigate this, we used the rodent-associated Brucella suis biovar 5, which in contrast to the ruminant-associated Brucella abortus and Brucella melitensis and other B. suis biovars, is fast-growing and conserves the ancestral Entner-Doudoroff pathway (EDP) present in the plant-associated relatives. We constructed mutants in Edd (glucose-6-phosphate dehydratase; first EDP step), PpdK (pyruvate phosphate dikinase; phosphoenolpyruvate ⇌ pyruvate), and Pyk (pyruvate kinase; phosphoenolpyruvate → pyruvate). In a chemically defined medium with glucose as the only C source, the Edd mutant showed reduced growth rates and the triple Edd-PpdK-Pyk mutant did not grow. Moreover, the triple mutant was also unable to grow on ribose or xylose. Therefore, B. suis biovar 5 sugar catabolism proceeds through both the Pentose Phosphate shunt and EDP, and EDP absence and exclusive use of the shunt could explain at least in part the comparatively reduced growth rates of B. melitensis and B. abortus. The triple Edd-PpdK-Pyk mutant was not attenuated in mice. Thus, although an anabolic use is likely, this suggests that hexose/pentose catabolism to pyruvate is not essential for B. suis biovar 5 multiplication within host cells, a hypothesis consistent with the lack of classical glycolysis in all Brucella species and of EDP in B. melitensis and B. abortus. These results and those of previous works suggest that within cells, the brucellae use mostly 3 and 4 C substrates fed into anaplerotic pathways and only a limited supply of 5 and 6 C sugars, thus favoring the EDP loss observed in some species.

The Fast-Growing Brucella suis Biovar 5 Depends on Phosphoenolpyruvate Carboxykinase and Pyruvate Phosphate Dikinase but Not on Fbp and GlpX Fructose-1,6-Bisphosphatases or Isocitrate Lyase for Full Virulence in Laboratory Models.

Bacteria of the genus Brucella infect a range of vertebrates causing a worldwide extended zoonosis. The best-characterized brucellae infect domestic livestock, behaving as stealthy facultative intracellular parasites. This stealthiness depends on envelope molecules with reduced pathogen-associated molecular patterns, as revealed by the low lethality and ability to persist in mice of these bacteria. Infected cells are often engorged with brucellae without signs of distress, suggesting that stealthiness could also reflect an adaptation of the parasite metabolism to use local nutrients without harming the cell. To investigate this, we compared key metabolic abilities of Brucella abortus 2308 Wisconsin (2308W), a cattle biovar 1 virulent strain, and B. suis 513, the reference strain of the ancestral biovar 5 found in wild rodents. B. suis 513 used a larger number of C substrates and showed faster growth rates in vitro, two features similar to those of B. microti, a species phylogenomically close to B. suis biovar 5 that infects voles. However, whereas B. microti shows enhanced lethality and reduced persistence in mice, B. suis 513 was similar to B. abortus 2308W in this regard. Mutant analyses showed that B. suis 513 and B. abortus 2308W were similar in that both depend on phosphoenolpyruvate synthesis for virulence but not on the classical gluconeogenic fructose-1,6-bisphosphatases Fbp-GlpX or on isocitrate lyase (AceA). However, B. suis 513 used pyruvate phosphate dikinase (PpdK) and phosphoenolpyruvate carboxykinase (PckA) for phosphoenolpyruvate synthesis in vitro while B. abortus 2308W used only PpdK. Moreover, whereas PpdK dysfunction causes attenuation of B. abortus 2308W in mice, in B. suis, 513 attenuation occurred only in the double PckA-PpdK mutant. Also contrary to what occurs in B. abortus 2308, a B. suis 513 malic enzyme (Mae) mutant was not attenuated, and this independence of Mae and the role of PpdK was confirmed by the lack of attenuation of a double Mae-PckA mutant. Altogether, these results decouple fast growth rates from enhanced mouse lethality in the brucellae and suggest that an Fbp-GlpX-independent gluconeogenic mechanism is ancestral in this group and show differences in central C metabolic steps that may reflect a progressive adaptation to intracellular growth.