Rickettsia rickettsii RoaM negatively regulates expression of a limited number of rickettsial genes

Abstract

The recently described rickettsial protein RoaM (regulator of actin-based motility) negatively regulates the production of actin tails, and its abrogation induces hyper-spreading behavior in many laboratory-adapted strains of Rickettsia rickettsii. RoaM is not surface exposed; thus, its mechanism of regulating actin-based motility is unclear. Using R. rickettsii strains derived from the virulent Sheila Smith strain that express varying levels of roaM, an RNA-seq experiment was performed. We found that roaM-overexpressing strains downregulate expression of at least six genes which may link the regulatory effects of RoaM to the phenotypic effect on motility. Genes regulated by RoaM were confirmed by RT-qPCR. Among the genes regulated is the secreted effector RarP2, which disrupts the trans-Golgi network. Two of the hypothetical proteins were shown to be secreted via fusion to a glycogen synthase kinase tag, which when phosphorylated reveals exposure to the host-cell cytosol. Taken together, these data support the hypothesis that RoaM affects transcription, downregulating rickettsial genes important for pathogenicity in the mammalian host but which are perhaps otherwise detrimental within the tick vector. To determine how RoaM activity may itself be regulated, we investigated a role of temperature in roaM transcription. RoaM expression itself is not temperature dependent, but many other rickettsial genes are, including some also regulated by RoaM. This suggests that rickettsiae utilize multiple mechanisms to control gene expression in response to environmental signals.

Importance: RoaM was previously shown to repress the production of actin tails by unknown mechanisms. The roaM gene is negatively selected for in cell culture resulting in hyper-spreading mutants. This work reveals that rather than specifically regulating motility in Rickettsia rickettsii, a set of rickettsial genes is downregulated that includes the type IV secreted effector, rarP2, as well as two other secreted, putative effectors. Relatively few secreted effectors have been identified in Rickettsia. RoaM appears to be part of a larger biological program encompassing active spreading in mammalian cells and may be a critical component for R. rickettsii to transition from arthropod to mammalian host.

Keywords: Rickettsia; environmental signaling; gene expression; transcriptomics.

Fig 1

Volcano plots of RoaM overexpression and deletion from R. rickettsii. (A) Overexpressed vs knockout. Down (blue) indicates decreased expression in the RoaM overexpressing strain (SS_S-pROAM-FL) relative to roaM knockout (SS_S-D roaM) (6). (B) Knockout (SS_S-D roaM) vs wildtype. Up (red) indicates increased expression in the RoaM deletion strain relative to wild-type Sheila Smith (SS_S). Note that RoaM (A1G_6520) is overexpressed or deleted, thus, shows exceptionally large changes in abundance.

Fig 2

Expression of six genes is negatively regulated by RoaM. (A) Six genes are downregulated greater than twofold with a significance less than P < 0.05. Wild type = SS_S; Truncated = SS_L (RoaM truncated at aa211); Knockout = SS_S D roaM; and overexpresser = SS_S-pROAM-FL as described in reference . Color scale represents the Z-score of the rlog transformed counts. (B) Downregulation of expression confirmed by qRT-PCR. An adjacent gene to A1G_01170 (A1G_01175) was also analyzed as was the entire region of A1G_02955 – A1G_02960, which together comprise an apparent pseudogene. Sca2 was included as a gene not regulated by RoaM (6).

Fig 3

Secretion of RoaM-regulated genes. All significantly regulated genes were expressed as GSK fusions and probed with anti-phospho-GSK specific antibodies. In addition to RarP2, which was previously shown to be secreted (13), A1G_01170 and A1G_01175 were also phosphorylated and, therefore, exposed to the host cytosol. Asterisks identify the recombinant products detected by the anti-phospho-GSK antibody.

Fig 4

Occurrence of RoaM-regulated genes in Rickettsia species.

Fig 5

Effect of temperature on R. rickettsii gene expression. R. rickettsii wild type and a RoaM deletion mutant were analyzed for changes in gene expression at 22°C vs 34°C. Volcano plots of (A) R. rickettsii Sheila Smith (SS_S) and (B) a RoaM deletion mutant of Sheila Smith (SS_S-D roaM) (6) are shown. Up (red) indicates increased expression at 34°C relative to 22°C. Down (blue) indicates decreased expression at 34°C relative to 22°C.

Fig 6

Heat map of the 30 most up- or downregulated genes showing a fold change of >2 and padj <0.05 in response to temperature. SS_22 and SS_34 represent the Sheila Smith strain (SS_S) at 22°C and 34°C, respectively. DROAM_22 and DROAM_34 represent the Sheila Smith strain (SS_S-D roaM) at 22°C and 34°C, respectively. A full heat map of all temperature-regulated genes is shown in Fig. S1.

Fig 7

Confirmation of changes in rickettsial gene expression in Vero cells due to temperature shift by RT-qPCR. RT-qPCR was performed on RNA extracted from Vero cells infected with (A) R. rickettsii Sheila Smith WT (SS_S) or (B) R. rickettsii Sheila Smith DroaM (B) at an MOI of 0.1. Infected cells were first grown at 34°C for 24 h, then shifted to 22°C for the times indicated. Experiments were performed in biological triplicate, with three technical replicates each, using rpoD as a reference gene and normalized to the 34°C condition with error bars representing standard deviation. Data were analyzed using standard two-way ANOVA followed by Šídák’s multiple comparisons test. ns, Not significant; **P < 0.01; ***P < 0.001; ****P = < 0.0001.

Fig 8

Expression of temperature regulated genes after temperature downshift and after subsequent upshift. RT-qPCR was performed on RNA extracted from Vero cells infected with R. rickettsii Sheila Smith WT at an MOI of 0.1. Infected cells were first grown at 34°C for 24 h, a subset of samples was shifted to 22°C for 2 h, then a subset of those shifted back to 34°C for an additional 2 h. Where present, arrowheads indicate temperatures at which final samples were collected and processed. Experiments were performed in biological triplicate, with three technical replicates each, using rpoD as a reference gene and normalized to the 34°C condition with error bars representing standard deviation. Data were analyzed using standard two-way ANOVA followed by Šídák’s multiple comparisons test. Significance was assessed by two-way ANOVA with Sidaks multiple comparison test. ns, Not significant; **P < 0.01; ***P < 0.001; ****P < 0.0001.