Coxiella burnetii Type 4B Secretion System-dependent manipulation of endolysosomal maturation is required for bacterial growth

Upon host cell infection, the obligate intracellular bacterium Coxiella burnetii resides and multiplies within the Coxiella–Containing Vacuole (CCV). The nascent CCV progresses through the endosomal maturation pathway into a phagolysosome, acquiring endosomal and lysosomal markers, as well as acidic pH and active proteases and hydrolases. Approximately 24–48 hours post infection, heterotypic fusion between the CCV and host endosomes/lysosomes leads to CCV expansion and bacterial replication in the mature CCV. Initial CCV acidification is required to activate C. burnetii metabolism and the Type 4B Secretion System (T4BSS), which secretes effector proteins required for CCV maturation. However, we found that the mature CCV is less acidic (pH~5.2) than lysosomes (pH~4.8). Further, inducing CCV acidification to pH~4.8 causes C. burnetii lysis, suggesting C. burnetii actively regulates pH of the mature CCV. Because heterotypic fusion with host endosomes/lysosomes may influence CCV pH, we investigated endosomal maturation in cells infected with wildtype (WT) or T4BSS mutant (ΔdotA) C. burnetii. In WT-infected cells, we observed a significant decrease in proteolytically active, LAMP1-positive endolysosomal vesicles, compared to mock or ΔdotA-infected cells. Using a ratiometric assay to measure endosomal pH, we determined that the average pH of terminal endosomes in WT-infected cells was pH~5.8, compared to pH~4.75 in mock and ΔdotA-infected cells. While endosomes progressively acidified from the periphery (pH~5.5) to the perinuclear area (pH~4.7) in both mock and ΔdotA-infected cells, endosomes did not acidify beyond pH~5.2 in WT-infected cells. Finally, increasing lysosomal biogenesis by overexpressing the transcription factor EB resulted in smaller, more proteolytically active CCVs and a significant decrease in C. burnetii growth, indicating host lysosomes are detrimental to C. burnetii. Overall, our data suggest that C. burnetii inhibits endosomal maturation to reduce the number of proteolytically active lysosomes available for heterotypic fusion with the CCV, possibly as a mechanism to regulate CCV pH.


67
Coxiella burnetii is a gram-negative obligate intracellular bacterium which causes human  132 We previously observed the CCV pH in both HeLa cells and cholesterol-free mouse 133 embryonic fibroblasts (MEF) to be pH~5.2, with increased CCV acidification to pH~4.8 leading to 134 C. burnetii degradation [24]. Given this apparent narrow pH tolerance inside the host cell, we 135 hypothesized that C. burnetii regulates the CCV pH at a less acidic pH than host lysosomes. To 136 test this hypothesis, we compared the CCV pH from wild type (WT) C. burnetii-infected HeLa cells 137 at 3 days post infection (dpi) to mature endosomal/lysosomal pH of mock-infected cells using a 138 ratiometric microscopy-based method using pH-sensitive Oregon Green 488 dextran and pH-

139
stable Alexa fluor 647 dextran [23]. Dextran is internalized through fluid-phase endocytosis and 140 delivered to the CCV lumen through CCV-endosome fusion ( Fig 1A). HeLa cells were pulsed with 141 both dextrans for 4 h followed by a 1 h chase to allow for endosomal maturation, as newly formed 142 endosomes mature and fuse with lysosomes in about 40 min [32]. The mean endosomal pH of 143 mock-infected cells was determined from the entire cell area (Fig 1B). We found CCVs to be 144 significantly less acidic (pH~5.2) compared to mature endosomes/lysosomes of mock-infected 145 cells (pH~4.8) ( Fig 1C). Further, the CCV pH was stable at pH~5.2 during a 6-day infection (Fig   146  1D), starting at 24 hour post infection (hpi). Taken together, these data suggest that C. burnetii 147 regulates the CCV pH and maintains it in a less acidic range relative to host lysosomes.

148
As the CCV is highly fusogenic and acquires lysosomal characteristics at least in part by 153 endosomes in infected cells, indicating that the mature endosomes are less acidic in the infected 154 cells (Fig 2A). Indeed, pH measurement revealed that endosomes in infected cells were 155 significantly less acidic, with a pH~5.8 compared to those in mock-infected cells (pH~4.8) (Fig   156  2B). To determine if this is a bacterial-driven process, we measured mature endosomal pH in 157 mock, WT, and ∆dotA (T4BSS mutant) C. burnetii-infected HeLa cells at 1, 2, and 3 dpi. While 158 mature endosomes in mock and ∆dotA-infected cells maintained a pH ≤ 5.0, those in WT-infected 159 cells were significantly less acidic (pH~5.8) starting 2 dpi (Fig 2C), suggesting C. burnetii T4BSS 160 actively manipulates endosomal pH.

C. burnetii T4BSS reduces lysosomal content in infected cells
163 Endosomes are progressively acidified as they mature from early endosomes (pH 6.1-6.8)

165
infected cells indicates incomplete endosomal maturation, leading us to hypothesize that the less-166 acidic vesicles in the WT C. burnetii-infected cells are not lysosomes. To test this hypothesis, we 167 quantitated the endosomal content of mock, WT, and ∆dotA C. burnetii-infected HeLa cell using 168 the early endosomal marker EEA1 (early endosome antigen 1) and the lysosomal marker LAMP1.

169
A late endosomal marker such as Rab7 or CD63 was not included, as these proteins are also 170 found on lysosomes, making it difficult to distinguish late endosomes from lysosomes [35,36].

171
EEA1 and LAMP1 fluorescence intensities from fixed-cell microscopy images were normalized to 172 cell area. While the entire cell was measured for mock-infected cells, the CCV was excluded for

173
WT and ∆dotA-infected cells. LAMP1 intensity was reduced by 57% in WT-infected cells 174 compared to both mock and ∆dotA-infected cells beginning at 3 dpi, while EEA1 intensity was 175 unaffected at all time points (Fig 3A, B). At 6 dpi, LAMP1 was also reduced by 56% in WT-infected 176 cells compared to mock (Fig 3A, B). However, due to the reduced viability of T4BSS mutant at 177 late time points, the ∆dotA mutant was not included at 6 dpi [9]. These data suggest that the C.

178
burnetii T4BSS reduces lysosomal content in infected host cells starting 3 dpi, and that the C.

179
burnetii T4BSS inhibits endosomal maturation. As discussed earlier, the CCV is highly fusogenic and undergoes homotypic and 181 heterotypic fusion with endosomes and lysosomes during expansion between 2 and 3 dpi [37].

182
Thus, it is possible that the decrease in lysosomal content between 2 and 3 dpi is due to increasing 183 CCV-lysosome fusion. To address this possibility, we measured CCV fusogenicity using a 184 quantitative dextran trafficking assay [38]. HeLa cells infected with WT mCherry-C. burnetii were 185 pulsed with fluorescent Alexa fluor 488 dextran for 10 min, followed by imaging every 4 min for 186 28 minutes using live-cell confocal microscopy ( Fig 4A). The dextran accumulation in the CCV with anti-C. burnetii antibody, or processed for CFU assay. Immunofluorescence showed fewer 258 and smaller CCVs in TFEB-GFP cells ( Fig 7C). Growth assay revealed a 50% and 70% reduction 259 in C. burnetii growth at 4 and 6 dpi, respectively, in TFEB-GFP cells compared to parental cells 260 (Fig 7D), indicating that TFEB overexpression attenuates C. burnetii growth. These data suggest 261 that TFEB-induced lysosome biogenesis negatively affects C. burnetii growth. 3) and a core (C). In C. burnetii-infected cells, the CCV was subtracted from each of these areas.    Central PMCID: PMCPMC3200357.

895
Ratiometric pH measurement revealed the pH of "mature" endosomes in WT C. burnetii-infected