Bartonella taylorii: A Model Organism for Studying Bartonella Infection in vitro and in vivo

Abstract

Bartonella spp. are Gram-negative facultative intracellular pathogens that infect diverse mammals and cause a long-lasting intra-erythrocytic bacteremia in their natural host. These bacteria translocate Bartonella effector proteins (Beps) into host cells via their VirB/VirD4 type 4 secretion system (T4SS) in order to subvert host cellular functions, thereby leading to the downregulation of innate immune responses. Most studies on the functional analysis of the VirB/VirD4 T4SS and the Beps were performed with the major zoonotic pathogen Bartonella henselae for which efficient in vitro infection protocols have been established. However, its natural host, the cat, is unsuitable as an experimental infection model. In vivo studies were mostly confined to rodent models using rodent-specific Bartonella species, while the in vitro infection protocols devised for B. henselae are not transferable for those pathogens. The disparities of in vitro and in vivo studies in different species have hampered progress in our understanding of Bartonella pathogenesis. Here we describe the murine-specific strain Bartonella taylorii IBS296 as a new model organism facilitating the study of bacterial pathogenesis both in vitro in cell cultures and in vivo in laboratory mice. We implemented the split NanoLuc luciferase-based translocation assay to study BepD translocation through the VirB/VirD4 T4SS. We found increased effector-translocation into host cells if the bacteria were grown on tryptic soy agar (TSA) plates and experienced a temperature shift immediately before infection. The improved infectivity in vitro was correlating to an upregulation of the VirB/VirD4 T4SS. Using our adapted infection protocols, we showed BepD-dependent immunomodulatory phenotypes in vitro. In mice, the implemented growth conditions enabled infection by a massively reduced inoculum without having an impact on the course of the intra-erythrocytic bacteremia. The established model opens new avenues to study the role of the VirB/VirD4 T4SS and the translocated Bep effectors in vitro and in vivo.

Keywords: Bartonella; NanoLuc; STAT3; effector proteins; luciferase; type IV secretion system.

Figure 1

Bartonella taylorii grown on TSA trigger a stronger STAT3 activation compared to bacteria grown on CBA. JAWS II cells were infected at MOI 50. At the depicted time points, cells were harvested and analyzed by immunoblot with specific antibodies against p-STAT3 (Y705), STAT3, and actin. Phosphorylated STAT3 was quantified over the total amount of STAT3. (A) Shown is an immunoblot of JAWS II cell lysates infected with Bartonella henselae wild-type or ΔvirD4, B. taylorii wild-type or ΔvirD4 grown on CBA (orange). Cells were infected for 24 h with ΔvirD4 mutants. (B) Quantification of three independent immunoblots as shown in (A). (C) Shown is an immunoblot of JAWS II cell lysates infected with B. henselae wild-type or ΔvirD4, B. taylorii wild-type or ΔvirD4 grown on TSA (blue). Cells were infected for 24 h with ΔvirD4 mutants. (D) Quantification of three independent immunoblots as shown in (C). (E) JAWS II dendritic cells were infected as described in (A). During the last 2 h of infection, cells were treated with 100 ng/ml LPS. After 24 hpi TNF-α concentration in the cell culture supernatant was assessed by ELISA. (F) Cells in (E) were analyzed by Western Blot for phosphorylated STAT3 (Y705). Actin was used as loading control. Data for immunoblots were acquired by pooling three technical replicates. Data from one representative experiment (n = 3) are presented. Data were analyzed using one-way ANOVA with multiple comparisons (Tukey’s multiple comparison test), *p < 0.05 and ****p < 0.0001.

Figure 2

Temperature shift increases effector translocation (A) Bacteria were cultured in M199 + 10% FCS at 28°C or 35°C for 24 h prior to infection. RAW 264.7 macrophages were infected for 3 h at MOI 50 with B. taylorii wild-type, the BepD-deficient mutant ΔbepD or the translocation-deficient mutant ΔvirD4. Cell lysates were analyzed by Western Blot for phosphorylated STAT3 (Y705), STAT3, and actin. Shown is an immunoblot of RAW cell lysates infected for 3 h with bacteria grown at 28°C or 35°C. (B) Quantification of pSTAT3 signal over STAT3 control of three independent immunoblots as shown in (A). Bacteria were grown at 28°C (light grey) or 35°C (dark grey). (C) Schematic overview showing the split NLuc assay principle. Bacteria were allowed to infect RAW LgBiT macrophages for 24 h. HiBiT-BepD was translocated inside host cells via the VirB/VirD4 T4SS. The substrate Furimazine was added and luminescence measured. (D) Bacteria were cultured in M199 + 10% FCS for 24 h at 28°C or 35°C prior to infection. RAW LgBiT macrophages were infected at MOI 50 for 24 h with B. henselae ΔbepA-G or ΔvirD4 containing pHiBiT-bepD_Bhe (blue) or pHiBiT (grey). Luminescence of the complemented split NLuc was measured. (E) RAW LgBiT were infected for 24 h with B. taylorii ΔbepA-I or ΔvirD4 containing pHiBiT-bepD_Bta (blue) or the control pHiBiT-FLAG (grey) either cultured in M199 + 10% FCS at 28°C or 35°C for 24 h prior to infection. Luminescence of the complemented split NLuc was measured. (F,G) Bartonella henselae or B. taylorii expressing GFP under the corresponding virB2 promoter on a plasmid were grown on CBA (orange) or TSA (blue) plates and cultured for 24 h in M199 + 10% FCS at 28°C or 35°C. GFP expression was analyzed by FACS measurement. Bacteria containing the empty plasmid (pCD366, grey) were used as control. Data for immunoblots were acquired by pooling three technical replicates. All experiments were performed in three independent biological replicates. Data were analyzed using one-way ANOVA with multiple comparisons (Tukey’s multiple comparison test), ****p < 0.0001.

Figure 3

Bartonella taylorii efficiently downregulates the innate immune response using the novel in vitro infection protocol. (A) Scheme of B. taylorii culture conditions used for infection. (B–G) RAW 264.7 macrophages were infected at MOI 50 with B. taylorii wild-type, the ΔbepD or ΔbepA-I mutants or the translocation-deficient mutant ΔvirD4. Secreted TNF-α was quantified by ELISA. Cells were harvested, lysed, and analyzed by immunoblot using specific antibodies against p-STAT3 (Y705), STAT3, and actin. (B) Immunoblot of cellular lysates after 6h infection. (C) Quantification of pSTAT3 signal over STAT3 control of three independent immunoblots as shown in (B). (D) TNF-α secreted by cells in (B) was quantified by ELISA. (E) Immunoblot of RAW macrophages infected for 20 h. (F) Quantification of pSTAT3 signal over STAT3 control of three independent immunoblots as shown in (E). (G) TNF-α secreted by cells in (E) was quantified by ELISA. Data for immunoblots were acquired by pooling three technical replicates. Data representative for three independent biological replicates. Data were analyzed using one-way ANOVA with multiple comparisons (Tukey’s multiple comparison test), ns, not significant, ****p < 0.0001.

Figure 4

Growth on TSA primes B. taylorii for high infectivity in mice. (A) C57BL/6 mice were infected i.d. with 10², 10³, or 10⁵ cfu of B. taylorii wild-type either grown on CBA (orange) or TSA (blue) plates. At several time points after infection, blood was drawn from the tail vein and plated on CBA plates to assess the amount of bacteria inside the blood. Table represents number of mice developing bacteremia vs. mice remaining abacteremic after infection with B. taylorii for nine animals per condition. Bacteremia kinetics (cfu/ml blood) in infected mice shown for (B) 10⁵ (CBA 3 of 3 infected mice bacteremic; TSA 3 of 3 infected mice bacteremic), (C) 10³ (CBA 3 of 3 infected mice bacteremic; TSA 3 of 3 infected mice bacteremic) and (D) 10² (CBA 1 of 3 infected mice bacteremic; TSA 3 of 3 infected mice bacteremic). Symbols represent the mean ± SD. (A) Shows pooled data from three independent experiments and (B–D) show representative data from three independent experiments.