Autophagy facilitates intracellular survival of pathogenic rickettsiae in macrophages via evasion of autophagosomal maturation and reduction of microbicidal pro-inflammatory IL-1 cytokine responses

ABSTRACT The genus Rickettsia is comprised of obligate intracellular bacterial parasites of a wide range of arthropod and vertebrate hosts. Some Rickettsia species (spp.) are responsible for serious human diseases globally. One interesting feature of these stealthy group of pathogens is their ability to exploit host cytosolic defense responses to their benefits. However, the precise mechanism by which pathogenic Rickettsia spp. elude host immune defense responses remains to be determined. Here, we observed that pathogenic Rickettsia typhi and Rickettsia rickettsii (Sheila Smith [SS]), but not non-pathogenic Rickettsia montanensis, become ubiquitinated and induce autophagy upon entry into bone marrow-derived macrophages. Moreover, unlike R. montanensis, both R. typhi and R. rickettsii (SS) colocalized with LC3B and not with Lamp2 upon host cell entry. Finally, we observed that both R. typhi and R. rickettsii, but not R. montanensis, reduce pro-inflammatory interleukin-1 (IL-1) cytokine responses, likely via an autophagy-mediated mechanism. In summary, we identified a previously unappreciated pathway by which both pathogenic R. typhi and R. rickettsii (SS), but not the non-pathogenic R. montanensis, become ubiquitinated, induce autophagy, avoid autolysosomal destruction, and reduce microbicidal IL-1 cytokine responses to establish an intracytosolic niche in macrophages. IMPORTANCE Rickettsia spp. are intracellular bacterial parasites of a wide range of arthropod and vertebrate hosts. Some rickettsiae are responsible for several severe human diseases globally. One interesting feature of these pathogens is their ability to exploit host cytosolic defense responses to their benefits. However, the precise mechanism by which pathogenic Rickettsia spp. elude host defense responses remains unclear. Here, we observed that pathogenic Rickettsia typhi and Rickettsia rickettsii (Sheila Smith [SS]), but not non-pathogenic Rickettsia montanensis, become ubiquitinated and induce autophagy upon entry into macrophages. Moreover, unlike R. montanensis, R. typhi and R. rickettsii (SS) colocalized with LC3B but not with Lamp2 upon host cell entry. Finally, we observed that both R. typhi and R. rickettsii (SS), but not R. montanensis, reduce pro-inflammatory interleukin-1 (IL-1) responses, likely via an autophagy-mediated mechanism. In summary, we identified a previously unappreciated pathway by which both pathogenic R. typhi and R. rickettsii (SS) become ubiquitinated, induce autophagy, avoid autolysosomal destruction, and reduce microbicidal IL-1 cytokine responses to establish an intracytosolic niche in macrophages.

prompted us to test the hypothesis that survival of both pathogenic R. typhi and R. rickettsii (SS), but not the non-pathogenic Rickettsia montanensis, involves evasion of autophagosomal maturation and reduction of microbicidal pro-inflammatory interleu kin-1 (IL-1) cytokine responses to establish a replication niche in phagocytic host immune defense cells, like macrophages.

Pathogenic rickettsiae are ubiquitinated and induce autophagy upon entry into macrophages
Preceding findings suggest that intracellular pathogens, like rickettsiae, not only encounter inflammasome-dependent defense mechanisms but are also confronted by another cytosolic defense pathway, autophagy (22-25, 30, 32-35).Both responses not only are key to mount the appropriate host defense responses, but are also functionally interconnected (15,16).In one of our preceding reports, we showed that R. typhi is ubiquitinated upon host entry and induces autophagy, but escapes autophagosomal maturation for intracellular colonization in non-phagocytic cells (30).More recently, we demonstrated that unlike R. montanensis (a non-pathogenic SFG member), R. rickettsii (SS) and R. typhi, two highly pathogenic Rickettsia spp., preferentially targeted the non-canonical inflammasome-IL-1α signaling axis in macrophages to support their replication (32).Given these reports by others and our recent findings, we first evaluated the ubiquitination status of R. typhi, R. rickettsii (SS), or R. montanensis during invasion of bone marrow-derived macrophages isolated from C57BL/6J WT mice.Similar to infection studies using R. australis (23), we observed that both R. rickettsii (SS) and R. typhi spp.were ubiquitinated during the course of invasion of WT BMDMΦ (Fig. 1A and B).In contrast, but similar to R. parkeri infections, R. montanensis was not ubiquitinated in WT BMDMΦ (Fig. 1A and B).These data suggest that the life cycle of R. montanensis is restricted by phagolysosomal fusion allowing for only a small number of bacteria to escape into cytosol of macrophages and support the findings from another laboratory using a human monocytic Tohoku Hospital Pediatrics-1 (THP-1) cell line (36,37).To test that hypothesis, we evaluated the bacterial burdens of R. typhi, R. rickettsii (SS), or R. montanensis during invasion of WT BMDMΦ and demonstrated that, unlike R. montanen sis, both R. rickettsii (SS) and R. typhi replicated in macrophages (Fig. 1C), which is in agreement with our previous published work (32).Furthermore, we evaluated the status of autophagy markers p62 (also known as sequestosome-1 [SQSTM1]) and autophagic vesicle formation marker LC3B (38) during infection of WT BMDMΦ by Western blot analyses.Our data revealed that, unlike R. montanensis, infection with R. rickettsii (SS) and R. typhi induced autophagy, as evidenced by an enhanced induction of LC3B and the simultaneous down-regulation of p62 expression (Fig. 1D and E).Given these findings, we performed similar infection studies in the presence of bafilomycin A1 (Baf A1), a late-stage autophagy inhibitor, and found that treatment with Baf A1 resulted in a further increase in LC3B levels providing additional evidence for a direct modulation of autophagy during the invasion of pathogenic rickettsiae (Fig. S1).

Pathogenic rickettsiae avoid autolysosomal destruction to establish a replicative niche in macrophages
Autophagy is an intracellular process that delivers autophagosomes to the lysosomes for degradation and is considered as one of the host defense pathways to combat bacterial infection, including other cellular functions (19,20,39).Given our above presented data that pathogenic, but not non-pathogenic, Rickettsia spp.are ubiquitinated upon host cell entry (Fig. 1A and B), we next evaluated the status of autophagy marker, LC3B, and lysosomal marker, Lamp2, during infection of WT BMDMΦ by immunofluorescence assay (IFA).We observed that, unlike R. montanensis, R. typhi and R. rickettsii (SS) colocalized with autophagy marker, LC3B, over the course of infection (Fig. 2A through D).As we previously demonstrated that R. typhi was able to avoid autolysosomal destruction in non-phagocytic cells (30), we next assessed the colocalization pattern of all three Rickettsia spp. with Lamp2.We observed that, unlike R. montanensis, both R. typhi and R. rickettsii (SS) spp.did not colocalize with Lamp2 in WT BMDMΦ (Fig. 2A through E), suggesting that both pathogenic Rickettsia spp.induce autophagy and escape autolyso somal destruction to facilitate their intracytosolic survival.

Intracytosolic survival of pathogenic rickettsiae requires induction of autophagy and reduction of microbicidal pro-inflammatory IL-1 cytokine responses
Based on recent findings from other laboratories and ours (22-25, 30, 32-35, 40), it has become evident that Rickettsia spp.exhibit variable degree of pathogenicity, indicat ing species-specific strategies to respond to mechanism of host defense surveillance, including autophagy and inflammasome responses.To address the role of autophagy in modulating the intracellular survival of Rickettsia spp., we utilized BMDMΦ from mice genetically lacking the ATG5 gene (ATG5 fl/fl -LysM-Cre) in myeloid cells, a well-accepted mutant model defective in the autophagy pathway (41).Next, we determined the colocalization status of R. montanensis, R. typhi, and R. rickettsii (SS) with LC3B during infection of ATG5 fl/fl or ATG5 fl/fl -LysM-Cre BMDMΦ by IFA.We found that, unlike R. montanensis, both R. typhi and R. rickettsii (SS) strains colocalized with LC3B in ATG5 fl/fl BMDMΦ, and not in ATG5 fl/fl -LysM-Cre BMDMΦ (Fig. 3A through D; Fig. S2).Moreover, we observed that, unlike R. typhi and R. rickettsii (SS), R. montanensis colocalized with Lamp2 in ATG5 fl/fl BMDMΦ (Fig. 3A through D; Fig. S2).However, all three Rickettsia strains colocalized with Lamp2 in ATG5 fl/fl -LysM-Cre BMDMΦ (Fig. 3A through D; Fig. S2).As it has been reported that ATG5 can modulate other host events independently of autoph agy (42), we employed BMDMΦ from mice genetically deprived in the Beclin1 gene function (Beclin1 fl/fl -LysM-Cre) in myeloid cells, an alternative mutant model defective in the autophagy pathway (41).Accordingly, we determined the colocalization status of R. montanensis, R. typhi, and R. rickettsii (SS) with LC3B and Lamp2 during infection of Beclin1 fl/fl -LysM-Cre or Beclin1 fl/fl (corresponding WT control) BMDMΦ by IFA.In line with our infection studies using ATG5 fl/fl and ATG5 fl/fl -LysM-Cre BMDMΦ, we found that, unlike R. montanensis, both R. typhi and R. rickettsii (SS) strains colocalized with LC3B in Beclin1 fl/fl BMDMΦ, and not in Beclin1 fl/fl -LysM-Cre BMDMΦ (Fig. S3).Furthermore, we observed that, unlike R. typhi and R. rickettsii (SS), R. montanensis colocalized with Lamp2 in Beclin1 fl/fl BMDMΦ (Fig. S3), while all three Rickettsia spp.colocalized with Lamp2 in Beclin1 fl/fl -LysM-Cre BMDMΦ (Fig. S3), suggesting that the intracytosolic survival of both pathogenic R. typhi and R. rickettsii (SS) strains in macrophages requires the activation of autophagy in an ATG5-/Beclin1-dependent manner and escape from autolysosomal destruction.
Our recent findings identified a previously unappreciated mechanism by which R. typhi and R. rickettsii (SS), unlike R. montanensis, benefited from a reduced IL-1 cytokine response, specifically IL-1α, to support their replication within the host (32).Intriguingly, R. australis benefited from ATG5-dependent autophagy induction and reduction of proinflammatory cytokine responses (22,23); however, a successful host colonization of R. parkeri involved the evasion of autophagy and inflammasome-mediated host cell death (24,25), leaving the precise mechanism inconclusive.To gain further mechanistic insight, we first evaluated the IL-1α and IL-1β cytokine responses in cultured supernatants of R.  montanensis-, R. typhi-, and R. rickettsii (SS)-infected ATG5 fl/fl BMDMΦ, as well as ATG5 fl/fl -LysM-Cre BMDMΦ.We observed that infection of ATG5 fl/fl BMDMΦ with R. typhi or R. rickettsii (SS) produced significantly lower levels of IL-1β or IL-1α cytokine, as compared to that of R. montanensis (Fig. 4A).Infection of ATG5 fl/fl -LysM-Cre BMDMΦ resulted in an increase of IL-1β or IL-1α cytokine levels for both R. typhi and R. rickettsii (SS) (Fig. 4A).Furthermore, we observed that, unlike R. montanensis, both R. typhi and R. rickettsii (SS) replicated in infected ATG5 fl/fl BMDMΦ; however, the replication of both bacteria was significantly impaired in ATG5 fl/fl -LysM-Cre BMDMΦ (Fig. 4B).Next, we validated our IL-1β or IL-1α cytokine and bacterial burden data in R. montanensis-, R. typhi-, or R. rickettsii (SS)-infected Beclin1 fl/fl BMDMΦ and Beclin1 fl/fl -LysM-Cre BMDMΦ.Our assays showed that infection of Beclin1 fl/fl BMDMΦ with R. typhi or R. rickettsii (SS) produced lower IL-1β or IL-1α cytokine levels, as compared to that of R. montanensis (Fig. S4A).Infection of Beclin1 fl/fl -LysM-Cre BMDMΦ with Rickettsia spp.resulted in an increase of IL-1β or IL-1α cytokine levels for both R. typhi and R. rickettsii (SS) (Fig. S4A).Furthermore, we found that, unlike R. montanensis, both R. typhi and R. rickettsii (SS) replicated in infected Beclin1 fl/fl BMDMΦ; however, the replication of both bacteria was impaired in Beclin1 fl/fl -LysM-Cre BMDMΦ (Fig. S4B).Given these data, we sought to determine the biological importance of autophagy induction and IL-1β or IL-1α cytokine responses in modulating the survival of all three Rickettsia spp. in macrophages.In this effort, we neutralized the activity of IL-1β or IL-1α cytokines using anti-IL-1β or anti-IL-1α antibodies and assessed the effect on the bacterial burdens for R. montanensis-, R. typhi-, and R. rickettsii (SS)infected ATG5 fl/fl BMDMΦ, as well as ATG5 fl/fl -LysM-Cre BMDMΦ.Neutralization of IL-1α and IL-1β (with a lower effectiveness) resulted in an increase in bacterial loads of R. typhi-, R. rickettsii (SS)-, and R. montanensis-infected ATG5 fl/fl BMDMΦ (Fig. 4C).Importantly, antibody neutralization of IL-1α and IL-1β (with a lower effectiveness), increased the bacterial burdens of R. typhi-, and R. rickettsii (SS)-infected ATG5 fl/fl -LysM-Cre BMDMΦ, reaching levels observed in infected ATG5 fl/fl BMDMΦ (Fig. 4C).Of note, concurrent anti-IL-1α and anti-IL-1β antibody treatments in R. typhiand R. rickettsii (SS)-infected ATG5 fl/fl BMDMΦ or ATG5 fl/fl -LysM-Cre BMDMΦ did not result in a synergistic effect (Fig. 4C).
Neutralization of either IL-1α or IL-1β (with a lower effectiveness) cytokine resulted in an increase in the bacterial loads of R. montanensis-infected ATG5 fl/fl BMDMΦ or ATG5 fl/fl -LysM-Cre BMDMΦ (Fig. 4C), while concurrent treatment with both antibodies resulted in synergistic increase in bacteria loads (Fig. 4C).The efficiency of antibody-mediated blocking was validated by comparing the levels of IL-1β and IL-1α cytokines in the supernatants of IgG isotype control, IL-1β and IL-1α antibody-treated Rickettsia-infected ATG5 fl/fl BMDMΦ, as well as ATG5 fl/fl -LysM-Cre BMDMΦ (Fig. S5).
Collectively, the data presented here are in agreement with prior reports from our laboratory and others (22,23,30,32), further confirming that the intracytosolic survival of R. typhi and R. rickettsii (SS), but not R. montanensis, in macrophages requires the activation of autophagy in an ATG5-/Beclin1-dependent manner, escape from autolyso somal destruction, and inhibition of IL-1α and IL-1β cytokines responses (Fig. 4D).

DISCUSSION
Various intracellular bacterial pathogens employ sophisticated mechanisms to hijack host cellular processes to facilitate their host survival.Such strategies entail reprogramming host phosphoinositide metabolism, which can facilitate uptake into host cells, modify phagosomes, undercut apoptosis, and interfere with other cellular defense mechanisms, including inflammasomes and autophagy.However, in the case of rickett siae, the mechanisms by which these intracytosolic pathogens modulate both inflammasome and autophagy responses to facilitate their replication in endothelial cells and immune cells, like macrophages, are only now emerging (22-25, 30, 32-35, 40, 43).In fact, recent findings from others and our laboratory have provided some intriguing findings that may suggest that Rickettsia spp.exhibit species-specific inflammasomemediated immune responses to establish an intracellular niche (22-24, 32, 34).However, the precise role of autophagy in regulating host colonization by Rickettsia spp.remains to be determined (22)(23)(24)(25).For instance, R. australis benefited from autophagy induction and reduction of pro-inflammatory cytokine responses (22,23), while R. parkeri evaded autophagic responses to colonize the host (24,25).Our recent reporting on R. typhi showed that ubiquitination followed by autophagy induction, and the escape from autolysosomal destruction, were crucial steps for R. typhi to colonize non-phagocytic cells (30).Based on these findings from others and our laboratory, we tested the hypothesis that both pathogenic R. typhi and R. rickettsii (SS), but not the non-pathogenic R. montanensis, induce autophagy, evade autolysosomal destruction, and reduce microbicidal pro-inflammatory IL-1 responses to establish an intracytosolic replication niche in host immune defense cells, like macrophages.Our data revealed that, using partially purified rickettsiae, R. montanensis is not ubiquitinated, while both R. typhi and R. rickettsii (SS) are ubiquitinated, during infection of macrophages.Ubiquitination of R. typhi and R. rickettsii (SS) is in agreement with infection studies using the Renografin-purified R. australis spp.(23).In contrast, no ubiquitination of partially purified R. montanensis was observed in our assays, which is in agreement with reports using the mild-pathogenic R. parkeri spp., where a MD-76R (overlaid or step gradient)-purified bacterium was used for ubiquitination studies (24,25).Collectively, these findings by our laboratory and others suggest that variations in the intracellular behavior of Rickettsia spp.are likely due to species-specific strategies employed by the bacteria and not because of differences in the purification methodol ogies.However, the precise mechanism of these phenotypic differences on rickettsial ubiquitination requires further investigation.In addition, our data demonstrated that, unlike R. montanensis, both R. typhi and R. rickettsii (SS) spp.induced autophagy and evaded lysosomal destruction during macrophage invasion.Importantly, our presented data using ATG5-/Beclin1-deficient macrophages further strengthen our hypothesis that R. typhi and R. rickettsii (SS) utilize autophagy for establishing an intracytosolic replication niche in macrophages (Fig. 4D).
Our findings also raised another intriguing question as to why both R. typhi and R. rickettsii (SS) spp.seem to possess the ability to actively induce autophagy, in contrast to R. montanensis.Furthermore, our data suggest that both R. typhi and R. rickettsii (SS) spp.can efficiently escape from phagolysosomal destruction, become ubiquitinated, and induce autophagy.However, R. montanensis localized with Lamp2 + -lysosomes and failed to evade lysosomal destruction, a phenotype previously observed in THP-1 cells (36,37), suggesting that the phagolysosomal maturation likely contributes to the lack of ubiquitination and autophagic recognition of R. montanensis.Also, as Rickettsia-specific immunodominant outer membrane proteins (e.g., Scas) are predicted to be expressed on either pathogenic or non-pathogenic Rickettsia spp.(44), another factor to consider is the presence of an effector repertoire that differs depending on the rickettsial virulence.In fact, we recently reported that R. typhi induces autophagy upon infection of non-phag ocytic cells, while subsequently avoiding autolysosomal destruction, via the function of secreted effectors, including phosphatidylinositol 3-kinase Risk1 and phospholipases Pat1 and Pat2 (30,44).This mechanism of host invasion seems consistent with rickettsiae close relatives, such as Anaplasma phagocytophilum and Ehrlichia chaffeensis (45)(46)(47)(48), or other intracellular pathogens such as Shigella (49,50).So, it is conceivable that effectors (e.g., Risk1 and Pat1/2 [30,44]), either alone or in combination with other currently yet to be identified effectors, could account for the observed phenotypic differences in host colonization among Rickettsia spp.(R. montanensis, R. typhi, and R. rickettsii [SS]).The precise mechanism and composition of the effector repertoire for each Rickettsia spp.are currently a matter of active research and still remain to be determined.
Our preceding report suggest that IL-1 signaling responses played a role in limiting rickettsial infection in vivo and in vitro (32).Specifically, we observed that both R. typhi and R. rickettsii (SS), but not R. montanensis, blocked cell death and reduced non-canon ical inflammasome-dependent IL-1α responses, in order to establish an intracytosolic replication niche in macrophages.In line with these findings, we now report that the R. montanensis, but not R. typhi and R. rickettsii (SS), is efficiently cleared by macrophages, a mechanism supported by previously published findings with various SFG Rickettsia spp.using THP-1 cells (36,37).Another intriguing observation is that the survival of both R. typhi and R. rickettsii (SS) was restricted in ATG5-or Beclin1-deficient macrophages, as evident by their localization with destructive Lamp2 + -lysosomes and the increase in pro-inflammatory IL-1 cytokine response.These findings suggests that activation of autophagy is critical for the intracellular survival of both R. typhi and R. rickettsii (SS), likely via an autophagy-mediated proteasomal degradation of pro-inflammatory IL-1 cytokines and/or signaling components/molecules involved in their induction (17)(18)(19), a mecha nism supporting the report on R. australis (23).However, the precise mechanism on how the intracellular survival of R. typhi and R. rickettsii (SS) benefits from the activation of an ATG5-/Beclin1-dependent autophagy pathway to limit pro-inflammatory IL-1 cytokine responses remains unclear and requires further investigation.Given the data presented in this manuscript, our recent report (30,32), and published reports by others (13,14,17,19,22,23), we now present a working model of host invasion by which both pathogenic R. typhi and R. rickettsii (SS) become ubiquitinated, induce autophagy, avoid autolysosomal destruction, and reduce microbicidal IL-1 cytokine responses to establish an intracytosolic niche in macrophages (Fig. 4D).

Bacterial strains, cell culture, and infection
Vero76 cells (African green monkey kidney, American Type Culture Collection [ATCC], RL-1587) were maintained in minimal Dulbecco's modified Eagle's medium (DMEM) supplemented with 10% heat-inactivated fetal bovine serum (FBS) at 37°C with 5% CO 2 .R. montanensis (M5/6) and R. rickettsii (SS) strains were obtained from Dr. Ted Hackstadt (Rocky Mountain Laboratories, NIH, MT, USA), while the R. typhi strain (Wilmington) was obtained from CDC.All Rickettsia strains were propagated in Vero76 cells grown in DMEM supplemented with 5% FBS at 34°C with 5% CO 2 .All Rickettsia strains were partially purified as described previously (27,51).Partially purified R. montanensis, R. rickettsii (SS), and R. typhi were used for infection of BMDMΦ at 34°C.At early stages of infection (before the doubling time [8 to 10 h]) of rickettsiae), a higher multiplicity of infection (MOI) of 20 (for 0.5 and 2 h post-infection [hpi]) was used to ensure the presence of sufficient number of bacteria, as compared to MOI of 5 at later time points (24 and 48 hpi), to determine the biological functions of the bacteria during host infection (26-28, 30, 32).

Measurement of cytokines and chemokines
IL-1 cytokine concentrations in supernatants from cultured BMDMΦ were assessed using the Legendplex mouse inflammation kit (BioLegend) following the manufacturer's instructions as described previously (32,52).

RNA isolation and quantitative real-time PCR
To determine viable bacterial number during the course of host infection, we performed Real time-quantitative polymerase chain reaction (RT-qPCR) assay on isolated RNA (26,53,54).In this effort, BMDMΦ samples were collected at 0.5, 2, 24, and 48 h post-infec tion.RNA was extracted from 1 × 10 6 BMDMΦ using the Quick-RNA miniprep kit (Zymo Research).The iScript Reverse Transcription Supermix kit (Bio-Rad; 1708841) was used to synthesize cDNAs from 200 ng of RNA according to the manufacturer's instructions.Quantitative real-time PCR was performed using SYBR Green (Thermo Fisher Scientific), 2 µL cDNA, and 1 µM each of the following oligonucleotides for rickettsial (housekeep ing) citrate synthase gene (GltA), and host (housekeeping) GAPDH gene.Relative copy number (RCN) of GltA expression was normalized by the expression of the GAPDH and calculated with the equation RCN = E −ΔCt , where E = efficiency of PCR and Ct = Ct target − Ct GAPDH.Melting curve analyses were performed at the end of each run to ensure that only one product was amplified as described previously (32,55).

Immunofluorescence
Eight-well chamber slides were seeded with BMDMΦ (30-50 × 10 4 cells per well) and infected with partially purified pathogenic and non-pathogenic Rickettsia spp.(MOI = 20 [0.5 and 2 h] and MOI = 5 [24 and 48 h]) as described previously (26-28, 30, 32).Briefly, partially purified rickettsiae were added to BMDMΦ and incubated for various lengths of time at 34°C with 5% CO 2 .Following incubation, cells were washed three times with 1× phosphate-buffered saline (PBS) and fixed with 4% PFA for 20 min at room temperature.Cells were than permeabilized in blocking buffer (0.3% saponin and 0.5% normal goat serum in 1× PBS) for 30 min and incubated for 1 h with the following primary antibodies diluted in antibody dilution buffer (0.3% saponin in 1× PBS): anti-Rickettsia (1:100 or 1:500), anti-Ub (1:100), anti-LC3B (1:100), and anti-Lamp2 (1:100).Cells were then washed with 1× PBS and incubated for 1 h with anti-Alexa Fluor 488 or anti-Alexa Fluor 594 secondary antibodies diluted 1:1,500 in antibody dilution buffer.Next, cells were washed with 1× PBS and mounted with ProLong Gold antifade mounting medium containing DAPI.Images were acquired using the Nikon W-1 spinning disk confocal microscope (University of Maryland Baltimore, Confocal Core Facility) and degree of colocalization (yellow + -stained bacteria) between Rickettsia and ubiquitin; LC3B or Lamp2 was analyzed using Fiji software as described previously (30).Quantification of the mean percentage of cellular marker positive bacteria was calculated from approximately 100 host cells for each experiment, while each experiment was repeated five times.The percentage of internalized bacteria (approximately 100-200 bacteria were counted per strain and time point) was calculated by dividing the number of extracellular bacteria by the total number of bacteria, multiplying by 100, and then subtracting this number from 100% to get the percentage of intracellular bacteria.

Extract preparation and Western blot analysis
Rickettsia-infected BMDMΦ cells were lysed for 2 h at 4°C in ice-cold lysis buffer (50 mM HEPES [pH 7.4], 137 mM NaCl, 10% glycerol, 1 mM EDTA, 0.5% NP-40, and supplemented with protease and phosphatase inhibitory cocktails) as descri bed previously (30,32).Equal amounts of protein were loaded for SDS-PAGE and membranes were probed with anti-p62, anti-LC3B, anti-GAPDH, and anti-EF-Ts antibodies, followed by enhanced chemiluminescence with secondary antibodies conjugated to horseradish peroxidase.

Statistical analysis
The statistical significance was assessed using analysis of variance with Bonferroni's procedure and Student's t-test (GraphPad Prism Software, version 8).Data are presented as mean ± standard error of the mean (SEM), unless stated otherwise.Alpha level was set to 0.05.contributed to the overall project administration and supervision; O.H.V., M.S.R., and A.F.A. wrote the manuscript; and all authors participated in editing the manuscript.
The funding sources had no role in the design of the study, in the collection, analyses, or interpretation of data, in the writing of the manuscript, or in the decision to publish the results.

FIG 1 R
FIG 1 R. rickettsii (SS) and R. typhi, but not R. montanensis, are ubiquitinated and induced autophagy in macrophages.(A) BMDMΦ from WT mice were infected with R. montanensis, R. rickettsii (SS), or R. typhi at amultiplicity of infection (MOI) of 20 (for 0.5 and 2 h) and 5 (for 24 and 48 h).Samples were fixed with 4% paraformaldehyde and Rickettsia spp.were detected using specific Alexa Fluor 488-conjugated Rickettsia (SFG or TG) antibodies, while Ub status was assessed using Alexa Fluor 594-conjuagted anti-Ub antibody.Images represent Rickettsia-infected macrophages after 2 h post-infection.DNA was stained using 4´,6-diamidino-2-phenylindole (blue).Colocalization between Rickettsia and Ub was analyzed using Coloc 2 plugin Fiji software.Bars in panel A, 10 µm.(B) Graph shows the percentage of Ub positive stained Rickettsia at indicated time points (0.5, 2, 24, and 48 h) of infection.Approximately 100-200 bacteria were counted (Continued on next page)

FIG 4 (
FIG 4 (Continued) (BioLegend) followed by flow cytometry.(B) Bacterial burdens in infected BMDMΦ were evaluated 0.5, 2, 24, and 48 h post-infection by GltA reverse transcription quantitative real-time PCR (RT-qPCR).Expression of the host housekeeping gene GAPDH was used for normalization.(C) Bacterial burdens in Rickettsia-infected ATG5 fl/fl or ATG5 fl/fl -LysM-Cre treated with anti-IL-1α, IL-1β, both anti-IL-1α and anti-IL-1β antibodies, or IgG isotype controls.Samples were evaluated at 24 h post-infection by GltA RT-qPCR.Expression of the host housekeeping gene GAPDH was used for normalization.Error bars (A-C) represent means ± standard error of the mean (SEM) from five independent experiments; NS, non-significant; *P ≤ 0.05; **P ≤ 0.01; ***P ≤ 0.005.(D) Proposed working model on how pathogenic Rickettsia spp.initiate autophagy, evade autolysosomal destruction, and suppress IL-1 cytokine responses to establish an intracytosolic replication niche in macrophages.Of note, a sizable amount of non-pathogenic Rickettsia is likely destroyed by phagolysosomal fusion, while a subpopulation may escape lysosomal fusion, ultimately resulting in the induction of IL-1 responses.