Coxiella burnetii Nine Mile Phase II (NMII) (Clone 4, RSA 439) and mCherry-expressing C. burnetii NMII (Beare et al., 2009) were grown and stored as previously described (Samanta et al., 2019). Human cervical epithelial cells (HeLa, ATCC CCL-2) were maintained in Roswell Park Memorial Institute (RPMI) 1640 medium with glutamine (Cat. 10–040-CV, Corning, New York, NY) and 10% fetal bovine serum (FBS; Cat. S11150, R&D Systems, Inc., Minneapolis, MN). The genome equivalents (GE) for the Coxiella stocks were determined by qPCR using primers for the Coxiella burnetii dotA gene (Coleman et al., 2004). The multiplicities of infection (MOI) of the bacteria stocks were optimized for HeLa cells and each type of culture vessels to ~1 internalized bacterium per cell at 37 °C and 5% CO₂.

The plasmids encoding EGFP-tagged Rab proteins were generous gifts from Dr. Mary Weber, University of Iowa (Faris et al., 2019). The plasmids were first transformed into competent E. coli DH5α (Cat. C2988J, New England Biolabs, Ipswich, MA) and stored at -80˚C. A single colony of the transformed E. coli was grown in 5 mL lysogeny broth (LB) containing 50 µg/mL kanamycin or 100 µg/mL carbenicillin overnight, which was then used to inoculate a 150 mL LB culture and incubated overnight. The GeneJET Endo-free Plasmid Maxiprep Kit (Cat. K0861, Thermo Fisher Scientific Baltics UAB, Lithuania) was used to isolate and purify plasmid DNA from bacterial cultures according to the manufacturer’s protocol. Plasmids were then concentrated using the MilliporeSigma Amicon Ultra-0.5 Centrifugal Filter Units (Cat. UFC503096, Fisher Scientific, Waltham, MA) and stored at -20˚C until used for cell transfection.

HeLa cells, grown in RPMI with 10% FBS (hereafter referred to as 10% RPMI), were plated in a 24-well plate (2x10⁴ cells per well) and simultaneously transfected with 0.4 µg of respective EGFP-Rab plasmids using FuGENE 6 transfection reagent (Cat. E269A, Promega, Madison, WI) per the manufacturer’s protocol. At 24 hours post-transfection, the cells were infected with mCherry-Coxiella in 250 uL 10% RPMI for 1 h, washed extensively with sterile phosphate-buffered saline (PBS; HyClone, Cat. SH30256.01, Cytiva, Marlborough, MA), and incubated in 10% RPMI at 37˚C. On the day before the indicated time points, cells were trypsinized, counted, and diluted to 1x10⁵ cells/mL. Diluted cells were then plated onto coverslips placed in a separate 24-well plate (5x10⁴ cells per coverslip) and allowed to adhere overnight. The next day, cells were fixed in 2.5% paraformaldehyde (PFA; Cat. 15710, Electron Microscopy Sciences, Hatfield, PA) for 15 min, washed in PBS, and blocked/permeabilized for 20 min in 1% bovine serum albumin (BSA; Cat. BP9700–100, Fisher BioReagents, Pittsburgh, PA) and 0.1% saponin (Cat. S0019, TCI America, Portland, OR) in PBS. Cells were then incubated in rabbit anti-LAMP1 (1:1000; Cat. PA1–654A, Thermo Fisher Scientific, Waltham, MA) primary antibody for 1h followed by Alexa Fluor secondary antibody (1:1000; Life Technologies) for 1 h to label the CCV. Following washing with PBS, coverslips were mounted using ProLong Gold Antifade Mountant with 4’, 6’-diamidino-2 phenylindole (DAPI; Cat. P36941, Invitrogen, Carlsbad, CA) and visualized on a Leica Stellaris confocal microscope using a 63x oil immersion objective. Images were processed and analyzed in ImageJ FIJI (Schindelin et al., 2012).

HeLa cells were plated at 5x10⁴ cells/well in a 6-well plate and incubated overnight. On the following day, the cells in each well were infected with mCherry-Coxiella in 500 µL of 10% RPMI for 1 hour at 37°C, washed with PBS, and incubated in 10% RPMI. On the day before the indicated time points, the infected cells were trypsinized and harvested from the 6-well plate, diluted to 1x10⁵ cells/mL, and plated on coverslips in a 24-well plate (5x10⁴ cells per coverslip) and incubated overnight. The cells were fixed the following day using 2.5% PFA for 15 min at room temperature. Immunofluorescent assay (IFA) was performed by blocking and permeabilizing the cells for 20 min at room temperature using 1% BSA-PBS with 0.1% saponin. Subsequently, the coverslips were incubated in rabbit anti-Rab11a for 72 h at 4°C (1:100; Cat. 715300, Invitrogen) or rabbit anti-Rab35 for 72 h at 4°C (1:500; Cat. ab152138, Abcam, Waltham, MA). Coverslips were then washed with PBS and incubated in Alexa Fluor 488 goat anti-rabbit (1:1000; Cat. A11034, Invitrogen) for 1 h at room temperature. The coverslips were washed with PBS once more before mounting in ProLong Gold Antifade Mountant with DAPI. The cells were visualized with a Leica Stellaris confocal microscope using a 63x oil immersion objective. For both 3 dpi and 6 dpi, 50–60 CCVs were inspected for Rab11a or Rab35 localization, and the percentage of Rab11a- and Rab35-positive CCVs was determined. The statistical difference in the percent of Rab11a- and Rab35-positive CCVs at each time-point was determined by Welch’s t-test from three independent experiments (N=3).

HeLa cells were either mock-infected or infected with mCherry-Coxiella in 24-well plates (2.5x10⁴ cells per well; two wells per condition) for 1 h, washed extensively with PBS, and incubated in 10% RPMI. For 3-day experiments, the cells were trypsinized at 2 dpi, diluted to 1x10⁵ cells/mL, and replated onto coverslips in a new 24-well plate (5x10⁴ cells per coverslip). At 3 dpi, cells were fixed with 2.5% PFA for 15 min and washed thoroughly with PBS. For 6-day experiments, cells were trypsinized, diluted, and replated first at 3 dpi and again at 5 dpi (5x10⁴ cells per coverslip). Next day, cells were fixed with 2.5% PFA and washed thoroughly with PBS. For both 3-day and 6-day experiments, following fixation, the cells were blocked/permeabilized with 1% BSA in PBS with 0.1% saponin for 20 min. Cells were then incubated overnight at 4°C in rabbit anti-Rab35 (1:100; Cat. 11329–2-AP, Proteintech, Rosemont, IL) or for 1 h at room temperature (RT) in rabbit anti-Rab11a (1:500; Cat. 20229–1-AP, Proteintech) primary antibody. The coverslips were washed with PBS and then incubated in Alexa Fluor 488 goat anti-rabbit (1:1000; Cat. A11034, Invitrogen) secondary antibody for 1 h at RT. Upon washing with PBS, the coverslips were mounted using ProLong Gold with DAPI. Cells were visualized with a Leica Stellaris confocal microscope using a 63x oil immersion objective. To quantify Rab content from mock and WT-infected cells, identically captured and processed images were imported into ImageJ FIJI and the total fluorescent intensity of Rab11a or Rab35 was measured and normalized to cell area. At least 25 cells were analyzed per condition in each of three independent experiments (N=3). The statistical difference in Rab11a and Rab35 intensities between mock and Coxiella-infected cells was measured by Welch’s t-test.

HeLa cells (10⁵ cells per well, 6-well plate) were reverse transfected with 50 nM small-interfering RNA (siRNA) SMARTpools specific for human Rab11a (ON-TARGETplus, Cat. L-004726–00-0005, Dharmacon Inc., Horizon Discovery Ltd., Lafayette, CO), Rab35 (ON-TARGETplus, Cat. L-009781–00-0005, Dharmacon Inc.), or a non-targeting (NT) control (Cat. D-001810–10-05, Dharmacon Inc.) using DharmaFECT-1 transfection reagent (Cat. T-2001–01, Dharmacon Inc.) in RPMI with 5% FBS according to the manufacturer’s protocol. At 48 h post-transfection, cells were infected with mCherry-Coxiella for 1 h at 37°C. Following washing with PBS, cells were harvested by trypsinization and subjected to a second round of siRNA transfection in a 24-well plate (2.5x10⁴ cells per well). On the day before the indicated time points, cells were trypsinized, resuspended to 1x10⁵ cells/mL, and either replated onto coverslips (2.5x10⁴ cells per coverslip) or into another 24-well plate without coverslips. The next day, the cells on coverslips were fixed with 2.5% PFA and subjected to immunofluorescent assay with rabbit anti-LAMP1 primary and Alexa Fluor 488 secondary antibodies to label the CCVs. Concurrently, mock-infected cells from the second 24-well plate were lysed with RIPA lysis buffer (Cat. 89900, Thermo Fisher) supplemented with HALT protease inhibitor (Cat. 87786, Thermo Fisher) on ice. The soluble fraction from the cell lysates were collected and protein concentrations were determined using the Pierce Rapid Gold BCA kit (Cat. A53226, Thermo Fisher). To confirm Rab11a and Rab35 silencing, 10 µg of total protein for both NT and Rab siRNA-transfected samples were resolved in 4–20% gradient SDS-PAGE (Mini-PROTEAN Gel, Cat. 4561094, Bio-Rad Laboratories, Hercules, CA) and subsequently transferred to a PVDF membrane for immunoblotting (Immun-Blot PVDF Membrane, Cat. 1620174, Bio-Rad). After blocking in Bio-Rad EveryBlot buffer (Cat. 12010020, Bio-Rad) for 5 min, the membranes were then probed separately using rabbit anti-Rab11a (Cat. 20229–1-AP, Proteintech), rabbit anti-Rab35 (Cat. 11329–2-AP, Proteintech), and anti-β-tubulin (BT7R) (HRP-conjugated loading control; Cat. MA5–16308-HRP, Invitrogen) antibodies in 1% BSA in PBS. After washing, the membranes were incubated with horseradish peroxidase-conjugated goat anti-rabbit (1:1000; Cat. 31460, Thermo Fisher) secondary antibody in EveryBlot blocking buffer, and developed using enhanced chemiluminescence (ECL) reagent (Cat. 170–5060, Bio-Rad).

HeLa cells were reverse-transfected with NT control, Rab11a, or Rab35-specific siRNAs as described above, and at 48 h post-transfection, were infected with mCherry-Coxiella for 1 h. Upon washing with PBS, cells were subjected to a second round of siRNA transfection as described above. On the day before the indicated time points, cells were trypsinized, resuspended to 1x10⁵ cells/mL, and replated onto coverslips (2.5x10⁴ cells/coverslip). The next day, the coverslips were fixed with 2.5% PFA and cells were immunostained with rabbit anti-LAMP1 primary antibody and Alexa Fluor secondary antibody to label the CCVs. Mounted coverslips were visualized using a Keyence BZ-X710 fluorescent microscope. Images of individual CCVs were captured and processed identically between conditions, and the CCV areas from exported images were measured in ImageJ FIJI. A 20 µm scale bar placed on a representative image for each sample was used for setting the scale in ImageJ. CCV areas were expressed as pixels and plotted using GraphPad Prism. At least 25 CCVs were measured per condition for each of the three independent experiments (N=3). The statistical differences in CCV areas between non-targeting and Rab11a/Rab35-specific siRNA-transfected cells were measured by Welch’s t-test.

To quantitate Coxiella growth in Rab11a and Rab35 knockdown cells we performed an agarose-ACCM-2 media-based colony forming unit (CFU) assay as described previously (Samanta et al., 2019) HeLa cells were reverse-transfected with NT control, Rab11a, or Rab35-specific siRNAs as described above, and at 48 h post-transfection, were infected with mCherry-Coxiella for 1 h. Upon washing with PBS, cells were scraped into 2 mL fresh 10% RPMI. Infected cells were then subjected to a second round of siRNA transfection in a new 24-well plate where 2.5x10⁴ cells/well were plated for 3 dpi, and 1.25x10⁴ cells/well were plated for 6 dpi time-points. To determine viable Coxiella at 0 dpi (day of infection), 500 µL (5x10⁴) of infected cells were lysed in sterile water for 5 min. The released bacteria were diluted 1:5 in ACCM-2 and plated in 2-fold serial dilutions onto 0.25% ACCM-2 agarose plates. For day 3 and day 6 time points, first, the spent media from the culture wells were collected and centrifuged at 20,000xg for 10 min to pellet any previously released bacteria, and supernatants were discarded (tube 1). The cell monolayers were then lysed in sterile water for 5 min and the released bacteria were diluted 1:5 in ACCM-2 (tube 2). The contents of tubes 1 and 2 were then mixed and spotted in 10-fold serial dilutions onto 0.25% ACCM-2 agarose plates. The plates were incubated for 6–7 days at 37 °C in 2.5% O₂ and 5% CO₂, and the number of colonies counted to measure bacterial viability. Two independent experiments were performed, each with biological duplicates for each condition, and the bacteria were spotted in triplicate. Bacterial growth was expressed as fold change in CFU over 0 dpi and the statistical difference in fold change between NT, siRab11a, and siRab35 conditions were measured using one-way ANOVA with multiple comparisons for each time point.

Fluorescent and confocal images were processed and analyzed in ImageJ FIJI software (Schindelin et al., 2012). All statistical analyses were performed using unpaired t-test with Welch’s correction (Welch’s t-test) using GraphPad Prism (GraphPad, La Jolla, CA). A p-value < 0.05 was considered significant. Data are plotted as mean ± standard error of mean (SEM).

We previously observed that Coxiella actively regulates CCV acidity by inhibiting host endosomal maturation in a T4BSS-dependent manner (Samanta et al., 2019). However, by impeding endosomal maturation, Coxiella gives rise to population of ‘immature’ endosomes with a higher mean pH compared to mature endosomes. As the immature endosomes may be impaired in cargo trafficking and degradation, they may be detrimental to the host cells. However, whether Coxiella manipulates the trafficking of the ‘immature’ endosomes in the infected cells has not been clear. Since Rab proteins are the master regulators of endocytic trafficking, we hypothesized that Coxiella manipulates one or more Rab proteins to modulate endosomal trafficking in the infected cells. To test this, we screened 12 major Rab proteins in Coxiella-infected cells for their interaction with the CCV. Plasmids encoding EGFP-tagged Rab proteins ( Table 1 ; a gift from Dr. Mary Weber, University of Iowa) were transfected into HeLa cells followed by infection with mCherry-Coxiella. At 3 dpi, the cells were immunostained with LAMP1 primary antibody to label the CCVs. During image analysis, CCVs were considered positive for Rab trafficking based on the qualitative similarity of Rab proteins and LAMP1 localization at the CCV membrane. EGFP tagged-Rab7 was included as a positive control in the screen as previous reports suggest that Rab7 localizes to the CCV during infection (Beron et al., 2002; Romano et al., 2007). As expected, all CCVs were positive for Rab7 and LAMP1. In addition to that, we interestingly observed that some CCVs were also positive for Rab11a and Rab35 as evidenced by their colocalization with LAMP1 at the CCV membrane ( Figure 1 ). Because ectopic expression may affect protein localization, we confirmed Rab11a and Rab35 localization to CCV using respective antibodies ( Figure 2A ). Both Rab11a and Rab35 are associated with host recycling endosomes, where they regulate the ‘slow’ and ‘fast’ endosomal recycling pathways, respectively (Ullrich et al., 1996; Grant and Donaldson, 2009; Chaineau et al., 2013). Therefore, these data suggest that, in addition to interacting with late endosomes and lysosomes, at least a subset of CCVs also interacts with the host recycling endosomes in both the ‘slow’ and the ‘fast’ recycling pathways. A CCV area comparison between the Rab11a/Rab35-positive and the negative CCVs showed no significant difference in areas between the positive and the negative CCVs (data not shown). Moreover, Coxiella colonies appeared unaffected by the presence or absence of Rab11a and Rab35 at the CCV. This study is the first report suggesting that the CCV interacts with host recycling endosomes and acquires recycling endosomal markers during maturation.

In our Rab CCV localization screen, we interestingly observed that only a subset of the CCVs were positive for Rab11a or Rab35 at a given time point. This led us to hypothesize that the interaction between the CCV and Rab11a/Rab35 are dynamic in nature. To test this, we quantified the percentage of Rab11a- and Rab35-positive CCVs in a population of infected HeLa cells at 3 and 6 dpi using HeLa cells immunostained with anti-Rab11a and anti-Rab35 antibodies. Our data revealed that at 3 dpi, ~47% of CCVs were positive for Rab11a, whereas only ~28% of CCVs were positive for Rab35 ( Figures 2B, C ). Interestingly this pattern reversed at 6 dpi, when only ~19% of CCVs were positive for Rab11 and ~57% of CCVs were positive for Rab35 ( Figures 2B, C ). Together, these data suggest that, while CCVs preferentially interact with the Rab11a-positive endosomes at 3 dpi, at 6 dpi, the preferential interaction shifts towards the Rab35-positive endosomes indicating that Rab11a and Rab35 interactions with the CCV are dynamic and vary depending on the stage of infection.

Our observation that Coxiella interacts with two different populations of recycling endosomes in the host cells led to the hypothesis that Coxiella manipulates host recycling endosomes during infection. To test this hypothesis, we quantitated the total cellular Rab11a and Rab35 content in mock and Coxiella-infected HeLa cells by immunofluorescent assay. Mock and Coxiella-infected cells were immunostained using anti-Rab11a or anti-Rab35 antibodies at 3 and 6 dpi and identically captured images were analyzed in ImageJ FIJI. The total fluorescent intensity of Rab11 or Rab35 was measured from mock and Coxiella-infected cells normalized to cell area and were considered indicative of total recycling endosome content in each cell. Our data revealed a ~19% increase in Rab11a fluorescent intensity (p < 0.0001; Figures 3A, B ) and a ~27% increase in Rab35 fluorescent intensity in infected cells compared to mock at 3 dpi (p < 0.0001; Figures 4A, B ). However, interestingly, we did not observe any difference in Rab11a or Rab35 fluorescent intensities between mock and Coxiella-infected cells at 6 dpi ( Figures 3C , 4C ). These data suggest that Coxiella increases the recycling endosome content in infected host cells at 3 dpi, but not at 6 dpi.

Based on our observation that the CCV dynamically interacts with recycling endosome-associated Rab proteins, and that Coxiella increases the recycling endosome content of infected cells, we hypothesized that Rab11a and Rab35 positively affects CCV maturation and Coxiella intracellular growth. To test this, we first quantified CCV areas in NT control, Rab11a, and Rab35 knockdown HeLa cells using our dual-hit siRNA knockdown protocol ( Figure 5A ) (Samanta et al., 2019) followed by immunofluorescent assay. Since CCV interaction with Rab11a is higher at 3 dpi, and that with Rab35 is higher at 6 dpi, we measured the CCV areas in Rab11a and Rab35 knockdown cells at 3 and 6 dpi, respectively, to detect maximum effect of the respective knockdowns. Immunoblotting with cell lysates from NT, siRab11a, and siRab35-transfected cells using anti-Rab or anti-β-tubulin antibodies confirmed depletion of both Rab11a ( Figure 5B ) and Rab35 ( Figure 5C ) at 3 and 6 dpi, respectively. The CCV area measurement data showed a ~45% reduction in CCV area in Rab11a knockdown cells compared to NT (p<0.0001) at 3 dpi ( Figure 6A ), and a ~28% reduction in CCV area in Rab35 knockdown cells (p<0.0001) at 6 dpi ( Figure 6B ) suggesting that both Rab11a and Rab35 positively affect CCV expansion at 3 and 6 dpi, respectively. Next, Coxiella growth was quantitated in NT, Rab11a, and Rab35 knockdown HeLa cells by agarose-based CFU assay. Compared to NT, Coxiella growth was significantly reduced in both Rab11a and Rab35 knockdown cells at both 3 and 6 dpi ( Figure 6C ; Table 2 ). Therefore, these data strongly suggest that host recycling endosomes are essential for CCV expansion and Coxiella intracellular growth.