The role of autophagy in tick-endosymbiont interactions: insights from Ixodes scapularis and Rickettsia buchneri

ABSTRACT Approximately 80 identified tick species are recognized as vectors and impact global public health by transmitting a wide range of pathogens; however, little is known about the interactions of ticks with the microbiome that they harbor, let alone their bacterial symbionts. In this study, we used the black-legged tick, Ixodes scapularis Say (Acari: Ixodidae), a vector of seven human pathogens in the United States, and utilized Rickettsia buchneri sensu stricto ISO7T (Rickettsiales: Rickettsiaceae), a spotted fever group (SFG) rickettsia that is an endosymbiont of I. scapularis, to investigate the role of autophagy in symbiont and tick interactions. We found that the expression profile of most autophagy family member proteins (ATGs) is down-regulated after R. buchneri infection in tick cell cultures. The autophagic process was observed by assessment of autophagosome formation and maturation in vitro (tick cell culture) and in vivo (tick ovary) in the presence of R. buchneri, whereas apoptosis was not induced. We further showed that R. buchneri infection triggered the accumulation of plasma membranes within cells. Suppressing autophagy via Atg8 siRNA interference inhibited intracellular rickettsial replication. This research indicates that autophagy regulation is important for the maintenance of R. buchneri in its I. scapularis tick host and provides more clues to solve the tick-symbiont interaction puzzle. IMPORTANCE Ticks are second only to mosquitoes in their importance as vectors of disease agents; however, tick-borne diseases (TBDs) account for the majority of all vector-borne disease cases in the United States (approximately 76.5%), according to Centers for Disease Control and Prevention reports. Newly discovered tick species and their associated disease-causing pathogens, and anthropogenic and demographic factors also contribute to the emergence and re-emergence of TBDs. Thus, incorporating different tick control approaches based on a thorough knowledge of tick biology has great potential to prevent and eliminate TBDs in the future. Here we demonstrate that replication of a transovarially transmitted rickettsial endosymbiont depends on the tick’s autophagy machinery but not on apoptosis. Our findings improve our understanding of the role of symbionts in tick biology and the potential to discover tick control approaches to prevent or manage TBDs.

mammals; yet, this does not include ticks, in which these processes are poorly under stood (4)(5)(6)(7).A series of cellular morphological and biochemical features can be used as markers for autophagy and apoptosis, and these are also applicable to studies in ticks.For instance, visualization of the autophagosome (large cytosolic double-membrane vesicle) is a hallmark of the initiation of the autophagy process, which is controlled by a series of autophagy-related proteins (ATG) beginning with the initial sequestering of the phagophore (a cup-shaped membrane structure).Subsequently, fusion with the lysosome forms autolysosomes in which cytosol and organelles are degraded to recycle the macromolecules (8).Cell shrinkage, chromatin condensation, nuclear and chromoso mal DNA fragmentation, and caspase cleavage, can be used to measure apoptosis, being the major characteristics of the apoptotic process (9).Current studies are focused on two main aspects of tick autophagy and apoptosis: (i) contribution to tick development, such as initiation of salivary gland degeneration through apoptosis following completion of a blood meal, embryonic patterning via autophagy during development, and autophagy induction in response to amino acid starvation; (ii) evolution of interactions between ticks and their transmitted pathogens, such as supporting successful infection/replica tion (in vitro studies) and colonization/persistence (in vivo studies) of some bacteria, or regulating pathogen activity as part of the innate surveillance mechanism (10)(11)(12)(13)(14)(15)(16)(17)(18).Our understanding of all the pathways in which autophagy plays a role is still far from complete, in part due to the broad range of pathogens transmitted by tick species.Moreover, only certain tick species and the pathogens they harbor have been studied with a multi-omics approach, which limits advances in the field in general.Interestingly, although non-pathogenic microbes make up the majority of the complex microbial communities within ticks, their impact on tick development and TBDs has received less attention, let alone how they might interfere with autophagy (19).
Thus, we utilized a medically important tick species, the black-legged tick, Ixodes scapularis Say (Acari: Ixodidae), to investigate the effect of non-pathogenic bacteria on tick autophagy.As a three-host tick with four life stages, I. scapularis feeds on a large variety of host animals and is an important vector for seven human pathogens in the United States, including Borrelia burgdorferi and Borrelia mayonii (the causative agents of Lyme disease), Borrelia miyamotoi (the causative agent of B. miyamotoi disease), Anaplasma phagocytophilum (the causative agent of human anaplasmosis), Babesia microti (a causative agent of human babesiosis), Ehrlichia muris eauclairensis (a causative agent of ehrlichiosis), and Powassan virus (the causative agent of Powassan encephalitis) (20)(21)(22)(23)(24)(25).Rickettsia buchneri (formerly rickettsial endosymbiont of I. scapularis, or REIS) have been shown to be present in I. scapularis throughout the tick's life cycle and to be highly prevalent in tick populations (26,27), consistent with a role as primary symbiotic bacteria.The species status of R. buchneri sensu stricto using the ISO-7 T isolate was recently evaluated using digital DNA: DNA hybridization and a genome-togenome distance calculator program (28).The analysis suggested that R. buchneri may be a subspecies of Rickettsia tamurae (AT-1T) (29), but this should be confirmed using additional studies.R. buchneri sensu stricto are transovarially transmitted endosymbionts that are mainly found in the ovaries and surrounding the nucleus of the developing oocytes of female ticks (30)(31)(32).The R. buchneri genome encodes synthesis pathways to produce essential nutrients (biotin and folate) as well as for the production of anti-bac terial that could act to defend its tick host against invading bacteria (27).In fact, R. buchneri was shown to block tick cell superinfection with rickettsial pathogens in vitro (33), supporting this notion.However, their significance and potential impact on the fitness of I. scapularis are still largely unknown (34,35).In this study, we investigated the interactions between the tick's autophagy process and R. buchneri.We found that the expression profile of most autophagy family member proteins was down-regulated after R. buchneri infection.Autophagy was observed by autophagosome/autolysosome accumulation in vitro (tick cell culture) and in vivo (female tick ovary) in the presence of R. buchneri, while apoptosis was not induced.Inhibiting autophagy by siRNA interference could hamper intracellular rickettsial replication.This research on how R. buchneri affects its I. scapularis tick host will provide more clues to solve the tick-symbiont interaction puzzle and advance knowledge of tick biology.

Autophagy activation in vitro after R. buchneri infection
Our previous study identified 14 ATG in I. scapularis and showed these family mem bers are highly conserved in ticks.An analysis of protein motif compositions indicated that the ATGs were evolutionarily closely related to their homologs in D. melanogaster where they are central to this pathway (Fig. 1A) (17).For instance, the ATG1 complex is recognized as the most upstream factor early in autophagy after starvation (36).We also found autophagy was activated after amino acid starvation in tick cells, confirming that the ATG1 complex has critical roles in initiating tick autophagy (17).To determine whether R. buchneri infection can trigger autophagy in vitro, we used a cell line from the European sheep tick, Ixodes ricinus, which has been shown to be superior to other cell lines for the growth and maintenance of R. buchneri (31).First, we examined the expression profiles of ATG genes in R. buchneri-infected IRE11 cells compared to uninfected controls.The proportion of R. buchneri-infected cells in culture was deter mined as previously reported (33).Considering the slow growth rate of R. buchneri (10-14 days population doubling time) in comparison to pathogens such as Salmonella typhimurium that triggered autophagy by damaging membranes at the early infection phase (hours), we used cultures in which 10% and 30% of the cells, respectively, were infected for this research (31,37).Real-time PCR results showed a significant increase in Atg8A and Atg8C gene expression in cultures with 10% R. buchneri infected cells, whereas the expression of other IsAtg genes was reduced relative to uninfected control cells.In contrast, the expression of all IsAtg genes except Atg8A and Atg8C decreased significantly in the more heavily infected cultures (Fig. 1B; Fig. S1).Identification of autophagosomes by electron microscopy (EM) relies on the detection of the charac teristic spherical structure enclosed by a double-layered membrane, and ATG8-family proteins facilitate membrane elongation during autophagosome formation (38).To verify that autophagy was induced by R. buchneri infection, we monitored the localization of ATG8-family proteins by immunofluorescence assays.Relative to uninfected IRE11 cells, R. buchneri-infected cells exhibited a remarkably increased localization of ATG8 to punctate structures representing autophagosomes (Fig. 1C and D).EM images also indicated a significant accumulation of autophagosomes, including initial autophagic vacuoles (AVi) with double-layered membrane and late/degradative autophagic vacuoles (AVd) with multiple membrane-enclosed structures (white arrows) relative to controls.We also found degraded R. buchneri (red arrow) surrounded by a double-layered membrane and cytoplasmic components (fragments of endoplasmic reticulum and Golgi) in AVi and AVd (Fig. 1E through G).This observation was further supported by the visualized autophagosomes, which were consistently present in most other cell samples (Fig. S3), confirming that autophagy was activated by R. buchneri infection.In addition, we found the presence of R. buchneri associated with autophagosomes by employing ATG8 (in red) and R. buchneri expressing GFPuv from a shuttle plasmid (R. buchneri-GFPuv; in green) (Fig. 1H and I).A similar overall effect was shown in R. buchneri-infected ISE6 cells (Fig. S2), i.e., Atg8A gene expression was significantly increased and more autophagosomes were observed compared to uninfected ISE6 cells.Thus, these data strongly suggest that R. buchneri infection (10%) triggered autophagosome formation in IRE11 cells.The later stage of autophagy involves the maturation of autophagosomes, which requires lysosome fusion, and this is where autophagosomes degrade and recycle cellular contents to supply nutrients (39).To further verify that autophagy was induced by R. buchneri infection, we examined lysosomal associated membrane protein 1 (LAMP-1, a lysosome marker) distribution after R. buchneri-GFPuv infection (Fig. 1J through L; Fig. S4).The degree of co-localization of R. buchneri-GFPuv (in green) and lysosome (in red) was analyzed by calculating Manders' overlap Coefficient (MOC) (40), indicating the degree of co-localization of both fluorescence signals was positively correlated.Together, these results confirmed that autophagy induction is strongly related to R. buchneri infection in vitro.

Altered plasma membrane structure in R. buchneri infected cells
Although multiple-layer membrane structures are a common feature of IRE11 cells (white arrows, Fig. 2A), R. buchneri infection induced much greater layer accumulation of these membrane structures, as seen using EM (Fig. 2B).To verify that R. buchneri infection induced this greatly amplified membrane accumulation, and to investigate the potential source of the membrane, we monitored the appearance of the plasma membrane in live cells by fluorescence microscopy.Compared with uninfected cells, R. buchneri-infected IRE11 (10% infection) displayed a greatly increased accumulation of plasma membrane, indicated by CellMask Deep Red Plasma membrane dye (red; Fig. 2C and D; Fig. S5).Additionally, R. buchneri-GFPuv (green) infected cells confirmed that accumulation of plasma membrane was associated with the presence of R. buchneri (Fig. 2E).Thus, these data strongly suggest that R. buchneri infection of IRE11 cells boosts the accumulation of multiple-membrane structures, which appear to be derived from the plasma membrane.

No apoptosis induction after R. buchneri infection
Martins and colleagues demonstrated that Rickettsia rickettsii inhibited apoptosis in tick cells (41).In contrast, our previous study showed that the cleavage of caspase-3 promoted DNA fragmentation and led to apoptosis in Rickettsia parkeri-infected tick cells, suggesting species-specific interactions of rickettsiae with tick cells.With that research, we confirmed that Caspase3/7 activity can be measured with Magic Red dye, and that terminal deoxynucleotidyl transferase-mediated dUTP nick-end labeling (TUNEL) accurately identified DNA fragmentation in tick cells (42).When we measured the intensity of red fluorescence representing caspase activity (Fig. 3A; excitation wavelength of 592 nm), the plate reader showed no difference between control IRE11 and R. buchneri-infected IRE11 cells (10% infection level).Likewise, live cells stained with Magic Red dye showed no difference in red intensity representing caspase activity (Fig. 3B and  C).To investigate whether DNA fragmentation was induced by R. buchneri infection, cells with different treatments were fixed and labeled with TUNEL, and images showed no difference in the percentage of apoptotic cells between uninfected and infected cells (Fig. 3D through F).These results confirm that there was no apoptosis induction after R. buchneri infection.

siRNA cocktail targeting Atg8A and Atg8C inhibits R. buchneri replication
Gene silencing in tick cell lines using siRNA to investigate the function of tick genes has been successful in some cell lines, including IRE11 (43).Given that ATG8 proteins are key players in regulating autophagosome formation and fusion with lysosomes, we target two I. scapularis ATG8 proteins via the cocktail of Atg8A-specific and Atg8C-specific siRNA, while the cocktail of Atg8A-scrambled and Atg8C-scrambled siRNA (ssiRNA) served as control.As shown in Fig. 4, significant knockdown of Atg8A gene expression in R. buchneri-infected IRE11 was achieved after 6 days of incubation with siRNA relative to ssiRNA, while the siRNA8C silencing experiment resulted in a slight reduction in gene expression without statistical significance.However, the cocktail of siRNAs resulted in an effective reduction of R. buchneri genomic DNA abundance (Fig. 4C).These results indicated that knockdown of the Atg8A and Atg8C gene inhibited R. buchneri intracellular replication in tick cells.

R. buchneri association with autophagosomes and autolysosomes in vivo
We previously showed that R. buchneri resided in the ovaries and surrounded the nucleus of the developing oocytes of female ticks (31).Therefore, to determine whether R. buchneri infection can trigger autophagy in vivo, we used unfed female tick ovaries to observe autophagosome formation and fusion with lysosome by EM.As shown in Fig. 5A, partially developed ovaries displayed oocytes with prominent nuclei.Due to yolk granule accumulation, oocytes appeared dark when viewed using a confocal microscope and bright field illumination.EM images showed that R. buchneri mainly accumulated in the developing oocytes (Fig. S6A and B), and individual bacteria also disseminated to the ovary duct (indicated by *; Fig. 5B).Under high magnification, autophagosomes containing R. buchneri (black arrows) surrounded by double membranes (red arrows) (Fig. 5C, panel 1, Fig. S6C and D), and AVd/autolysosomes containing degraded R. buchneri (black arrows) with a multi-membrane structure (red arrows) were observed around the nucleus of an interstitial cell (Fig. 5C, panel 2).We also observed a typical autopha gosome with the phagophore containing fragments of cytoplasmic organelles, Golgi apparatus, and degraded R. buchneri (Fig. 5D).In addition, an autophagosome containing Golgi apparatus and R. buchneri merged with a lysosome was seen in a developing oocyte (Fig. 5E).Autophagosomes were frequently observed in cells/tissues occupied by R. buchneri.These images provide evidence that R. buchneri may induce autophagosome and autolysosome formation in vivo, suggesting R. buchneri triggers autophagy but not xenophagy (a selective autophagy that specifically targets intracellular microbes to lysosomes).

DISCUSSION
Although the significance of autophagy for the interactions of pathogens and their vectors is generally acknowledged, little is yet known regarding non-pathogenic microbes, let alone whether and how a vector's autophagic pathway determines its fate.In this study, we show that infection of I. scapularis tick cell cultures by a symbiotic rickettsia induces autophagy and that the inhibition of autophagy impedes rickettsial intracellular growth, and this pathway is likely independent of apoptosis.We further identified two types of autophagy-related organelles (autophagosome/autolysosome) prevalent in female tick ovaries occupied by this rickettsia.
Autophagy-related genes and proteins are central to the autophagic process, and have been well-documented in eukaryotes, and we have also thoroughly characterized them in I. scapularis ticks (17).Although not all homologs encoding core ATG proteins in D. melanogaster have been found in I. scapularis, certain ATG proteins that contribute to autophagosome biogenesis were identified (Fig. 1A).For example, the initial step for autophagosome formation is the nucleation of the phagophore, which is regulated via the ULK1/2 complex (44), and ULK1/2 (or ATG1), ATG13, FIP200 (or ATG11), and ATG101 are recognized as critical components for this complex in mammals.It is not surprising that ticks lack ULK1/2 but can still induce autophagosome formation (Fig. 1) because some ATG proteins can compensate for or replace others in the recruitment of ATG components (45,46).Additionally, EM showed autophagosome (with R. buchneri inside) fusion and cargo degradation, suggesting this represented the later phase of autophagy in 10% R. buchneri-infected cultures.An examination of the expression profile of 14 Atg genes showed that transcripts from eight of them were significantly reduced, but not for two Atg8 subfamily members (8A and 8C) (Fig. 1), as well as Atg3.However, higher infection (30%) with R. buchneri decreased the transcription of most Atg genes (Fig. S1 and S2).We suspect that (i) R. buchneri induced the later phase of autophagy and decreased Atg gene transcription, thereby balancing the tick autophagic process for the benefit of the bacteria, and (ii) different infection levels led to different outcomes of the autophagy response.This hypothesis is supported by the finding that Listeria monocytogenes activate autophagy at the early stage of infection in order to evade containment in autophagosomes at the later infection stage (47,48).Similarly, R. parkeri leads to apoptosis inhibition at early stages of infection but activates apoptosis during late infection (42).Unlike the well-known systems of pathogenic bacteria and their hosts, examining tick in vivo autophagy in response to R. buchneri is very challenging because of strict intracellular growth requirements and slower growth rates compared to other Rickettsia spp.(49).Although R. buchneri infection level could be a factor that initiates different responses in vitro (33, this manuscript), future studies using infection by host-cell free R. buchneri instead of infected cells (49), may help to determine the specific requirements (such as a multiplicity of infection and cell types) that regulate tick autophagy at different phases.
Finding the balance between avoidance of autophagy and successful replication is necessary for the survival of intracellular bacteria, given that autophagy is an innate immune response to target bacteria for degradation and that bacteria are evolving strategies to subvert autophagy for their own needs (50).Although much progress has been achieved in elucidating the mammalian host autophagy response to cer tain tick-borne intracellular bacterial pathogens, such as A. phagocytophilum, Ehrlichia chaffeensis, and R. parkeri, this is still completely unknown in their tick vectors (51)(52)(53).For example, A. phagocytophilum used the effector protein Ats-1 to employ autophago somes to promote their intracellular growth and block the fusion of the autophagosome with the lysosome.By contrast, R. buchneri infection led to autophagosome formation and fusion with lysosomes.The co-localization of R. buchneri and lysosomal marker also indicated a different mechanism by which non-pathogenic bacteria interfere with tick autophagy (Fig. 1).This is also confirmed by spotted fever group (SFG) rickettsia behavior in human macrophage-like cells, in which symbiotic or non-pathogenic rickettsiae colocalize with lysosomal markers, whereas pathogenic rickettsiae evade lysosomal destruction (54,55).This suggests that endosymbiotic Rickettsia species are cleared by autophagy as part of the mammalian immune response.It is unclear whether the induction of late-stage autophagy and lysosomal degradation of R. buchneri in ticks is an immune response to control the symbiont to numbers that are tolerable to the tick, or whether autophagy is induced by R. buchneri to aid tick survival, for example, as a mechanism to obtain nutrients.R. buchneri-induced autophagy appeared to be a non-selective process, with R. buchneri-containing autophagosomes also engulfing cellular cytoplasmic contents for lysosomal degradation, including Golgi and mitochon dria (Fig. 5).This differs from xenophagy, a selective form of autophagy which specifically targets intracellular pathogens to lysosomes and mediates their destruction, and would be expected to occur if R. buchneri-induced autophagy were a targeted immune response.It is possible that in heavily infected cells that are packed with R. buchneri (as can be seen in Fig. 1 and 5), activating non-selective autophagy in tick cells leads to the unavoidable engulfment and destruction of some R. buchneri whilst the remain ing multitude of bacteria receives the benefits accompanying increased autophagy.It should be noted that most of the results presented here were obtained in tick cells, which could have different immune responses compared to the symbiont's usual home in the tick ovaries.However, our findings suggest that R. buchneri infection in both tick cells and ovaries triggered similar responses, in terms of inducing autophagosome formation.Therefore, autophagy may serve as a strategic option for R. buchneri, whether it's activated or not, depending on the advantages it provides to the bacteria, espe cially during different phases such as during or after a bloodmeal.We are uncertain of how R. buchneri induces autophagy and whether this involves interactions between bacterial factors (such as effectors) and tick ATG proteins.Some R. buchneri proteins are recognized as potential effectors that may be secreted via the type IV secretion system (34).We used RNA interference (RNAi) as a powerful tool to demonstrate that ATG8 protein expressed during autophagy promotes rickettsial intracellular replication (Fig. 4).Although the reduction in R. buchneri replication upon the cocktail of Atg8A/8C siRNA treatment was relatively small, it was significant, and the effects on R. buchneri growth over time would be even greater considering its slow doubling rate of approx imately 10 days (31,33).It should be noted that the siRNA8C silencing experiment resulted in a slight reduction in gene expression without statistical significance, we speculated that ATG8C protein may have a different expression pattern after R. buchneri infection.Furthermore, the functional redundancy of the three I. scapularis ATG8 proteins and whether they act sequentially or in concert during autophagosome formation are unknown; thus, it is possible that upon knockdown of Atg8A, the other two ATG8 proteins may have compensated for the loss of Atg8A activity, resulting in a smaller effect on R. buchneri replication than if all Atg8 genes had been targeted.This is common in mammals where the ATG8 family comprises six members, including microtubule-asso ciated protein 1 light chain 3 (LC3) and GABARAP subfamilies.LC3 isoforms partially compensate for LC3 loss, and GABARAP isoforms are recognized as redundant proteins that restore autophagy in vitro (56).Further study to define the function of tick ATG proteins will be necessary to identify the basic autophagy machinery in ticks.This will provide the basis for exploring R. buchneri effectors that may interfere with tick autophagy and determine whether they function in an SFG rickettsia-specific manner or trigger generally conserved responses in their tick vectors.
We also observed some unique features during R. buchneri-induced autophagy processes, such as accumulation of plasma membrane (indicated by white arrow) and an increased appearance of multi-membrane structures (Fig. 2), compared to uninfected tick cells (shown here in IRE11 cells).This multiple-membrane structure might be a potential source for the autophagosomal membrane in ticks (57), as plasma membrane, as well as the endoplasmic reticulum, the Golgi, and the mitochondria, can supply membrane lipids for autophagosome formation; however, the relationship between the plasma membrane and this structure needs further investigation.
Considering that autophagy and apoptosis are integrated and have been implicated in microbial infections, we monitored caspase activity and DNA fragmentation and found no notable variation, suggesting R. buchneri employs distinct strategies to activate tick autophagy and apoptosis (Fig. 3).This is probably because endosymbionts should preferentially induce the cell survival pathway (autophagy) to benefit their intracellular growth, instead of the death pathway (apoptosis), which would be detrimental to them.Our lab previously demonstrated that R. parkeri, a pathogenic SFG rickettsia, induced tick apoptosis in vitro (42).Apparently, the species-specific behavior of rickettsiae can drive different outcomes of apoptosis, even though both R. buchneri and R. parkeri belong to the SFG.Additionally, it is not known whether there is cross-talk between autophagy and apoptosis.This cross-talk could govern the cells' fate in response to pathogen stimuli, thus affecting pathogen destiny, i.e., their ability to successfully complete their life cycle through infection, replication, and transmission.For example, Coxiella burnetii can generate a persistent bacterial infection by preventing apoptosis and inducing autophagy through interaction with Bcl-2 and Beclin-1 (58,59).The knowledge gained from the communication between autophagy and apoptosis will significantly enhance our understanding of the defense responses that regulate pathogen infection and burden in vector ticks.Here, we revealed that a symbiotic rickettsia induces autophagy in vitro and in vivo through morphological and genetic methods and demonstrated that autophagy restricted intracellular endosymbiont replication.This finding advances our knowledge of the biology of symbiotic rickettsiae and their impact on arthropod hosts, and sheds light on the basic autophagy machinery of ticks and other arthropod vectors.

RNA extraction and real-time quantitative reverse transcription PCR (qRT-PCR)
Total RNA was extracted from cells using TRI Reagent (Sigma, St. Louis, MO, USA), and purified by an RNA Clean & Concentrator kit (Zymo Research, Irvine, USA).The quantity and quality of RNA were assessed via a DS-11 Series Spectrophotometer/Fluorometer (DeNovix, Wilmington, USA).Genomic DNA contamination was eliminated and cDNA was synthesized using PrimeScript RT Reagent Kit with gDNA Eraser (Takara, Otsu, Shiga, Japan).The expression profile of tick ATG (17) at different infection rates was assessed using qPCR on the Mx3005P Real-Time system (Stratagene, La Jolla, USA) with SYBR green detection (Agilent Technologies, Santa Clara, USA).All protocols were according to the manufacturer's instructions.Primers used in this study are listed in Table S1.The analysis of the Atg gene expression profile included a one-way analysis of variance for each gene, with subsequent Bonferroni correction to address multiple comparisons, and the glyceraldehyde 3-phosphate dehydrogenase (gapdh) gene from the I. scapularis tick genome as a reference control.Relative gene expression levels of individual Atg genes were calculated using the comparative cycle threshold (CT) method (2− ΔΔ CT).Statistical analysis was performed using the data obtained from three separate cDNA sets from three independent biological samples.

R. buchneri genomic DNA isolation and qPCR
Total genomic DNA was isolated from WT-R.buchneri infected cells using the Qiagen Puregene Core A kit, following the manufacturer's gram-negative bacteria instructions.The quantity and quality of genomic DNA were assessed via a DS-11 Series Spectropho tometer/Fluorometer.The single-copy rickettsial citrate synthase (gltA) gene was used to determine the copies of WT-R.buchneri in different treatments using qPCR as previously described (61).Primers used in this study are listed in Table S1.

Immunofluorescence
WT-R. buchneri or R. buchneri-GFPuv infected cells (10% infection) and uninfected control cells were immobilized onto slides and fixed in 4% paraformaldehyde for 1 h at room temperature.They were then permeabilized with 0.1% Tween 20 (in 1× PBS) for 1 h, blocked in 5% BSA for 2 h and subsequently incubated with primary antibody overnight at 4°C.
The slides were then washed twice in TBST (10 mM Tris-HCl, 150 mM NaCl, 0.05% Tween-20, pH 7.5) and incubated with secondary antibody for 2 h at room temperature.They were then mounted in Fluoroshield mounting medium with DAPI (VECTASHIELD, Vector Laboratories), and imaged under a Nikon A1si Spectral Confocal Microscope using a 60X objective or an Olympus BX61 Disk Scanning Unit confocal microscope fitted with a 60X objective.The anti-LC3B antibody (purchased from Abcam, Cambridge, UK) was used as the primary antibody at 1:2,000 dilution for staining ATG8 protein in cells.The anti-LAMP1 antibody (purchased from Cell Signaling Technology, Danvers, MA, USA) was the primary antibody at 1:2,000 dilution for staining lysosome/autolysosome in cells.A goat anti-rabbit IgG (Alexa Fluor 594) was used as the secondary antibody at a 1:3,000 dilution (purchased from Abcam).The fluorescence properties were observed using different wavelength filters (4, 6-diamidino-2-phenylindole (DAPI): excitation at 365 nm and emission at 480 nm; FITC: excitation at 495 nm and emission at 519 nm; TRITC: excitation at 560 nm and emission at 590 nm; Cy5: excitation at 647 nm and emission at 665 nm) depending on the fluorophores used for secondary antibodies.All treatments were replicated three times.Images were processed using the program NIS-Elements View 4.50 (University Imaging Centers at the University of Minnesota, Twin Cities).The colocalization of R. buchneri and lysosomes was assessed by determining the degree of the area of two signals overlapping, according to Pearson's coefficient and overlap coefficient according to Manders' calculations by ImageJ Fiji (JaCoP plugin and Colocalization Threshold plugin) (62).

Transmission electron microscopy
WT-R. buchneri-infected tick cells (10% infection in IRE11) and uninfected control cells were pre-fixed with 2.5% glutaraldehyde at 4°C over three nights.They were rinsed in 1× PBS three times for 10 min and post-fixed in 2% OsO 4 for 2 h.Samples were rinsed in 1× PBS three times for 10 min and dehydrated in a graded ethanol series.For 25% and 50%, the samples were rinsed once for 10 min.For 75%, 95%, and 100%, the samples were rinsed twice for 10 min.Subsequently, the samples were rinsed twice in 100% acetone for 10 min and infiltrated in 50% Embed 812 resin/acetone solutions overnight.Then the solvent was replaced with 100% Embed 812 resin/BDMA (N-Benzyl-N, N-Dimethylamine C 6 H 5 CH 2 N(CH 3 ) 2 ) twice for 12 h at room temperature.Next, the samples were polymer ized at 60°C for 48 h.Finally, the pieces were cut into semi-thin sections (toluidine blue stain for 2-5 min) and thin sections (uranyl acetate stain for 15 min and lead citrate stain for 5 min) and examined using a JEM-1400Plus TEM.
To examine tick ovaries, the dorsal cuticle was dissected away from three unfed I. scapularis female ticks (from a laboratory colony maintained at the University of Minnesota) and immersed in Ito's modified fixative (63) for pre-fixation.They were then rinsed in 1× PBS three times for 10 min and post-fixed in 1% OsO 4 in 0.4M sodium cacodylate buffer for 1 h.Secondary fixation used 2% uranyl acetate, en bloc for 1 h.First, the samples were rinsed in 1× PBS three times for 10 min and dehydrated in a graded ethanol series (25%, 50%, 75%, and 95%) for 10 min.Then the samples were rinsed twice in 100% ethanol for 10 min and infiltrated in 25%, 50%, and 75% Spurr's low-viscosity resin/ethanol in each solution for 24 h.Then the solvent was replaced with 100% Spurr's low viscosity resin for 24 h at room temperature.Next, the samples were polymerized at 70°C for 8-12 h.Finally, the pieces were cut into semi-thin sections (toluidine blue stain for 3-4 min) and thin sections (uranyl acetate stain for 20 min and modified Sato's lead solution stain for 5 min) examined using Philips CM12 TEM.

Plasma membrane staining
Tick cell plasma membranes were stained using the CellMask Plasma Membrane Stains Kit (Thermo Fisher Scientific), following the manufacturer's instructions.WT-R.buchneri or R. buchneri-GFPuv infected cells (10% infection in IRE11) and uninfected control cells were incubated with Deep Red plasma membrane stain substrate and NucBlue Live ReadyProbes Reagent (Hoechst 33342) (Thermo Fisher Scientific), and then immobilized onto slides.Specimens were mounted in a Fluoroshield mounting medium and imaged under a Nikon A1si Spectral Confocal Microscope using a 60X objective.Dual fluorescence properties were observed using filters (DAPI and Cy5) as above.All treatments were replicated three times.Images were processed using the program NIS-Elements View 4.50 (University Imaging Centers at the University of Minnesota, Twin Cities).The mean of fluorescence intensity and a correlation analysis between the intensity of red and green staining were performed using ImageJ Fiji.

Monitoring apoptosis in response to R. buchneri infection in tick cells
Caspase enzyme activity assay (Magic Red Caspase3/7 Assay Kit, Immunochemistry), and DNA fragmentation via TUNEL (in situ Cell Death Detection Kit, Roche) were applied to monitor apoptosis in WT-R.buchneri-infected cells (10% infection in IRE11), and uninfected cells served as controls, following the protocols previously reported (42).For each treatment, TUNEL-positive cells and DAPI-positive cells in a total of three slides were quantified.From each slide, three fields were randomly selected, and the percentage of apoptotic cells was calculated as the number of TUNEL-positive cells divided by the number of DAPI-positive cells in these chosen fields.The student's two-tailed t-test was applied for the analysis of percentages of apoptotic cells.All treatments were replicated three times.

RNA interference
The I. scapularis Atg8A and Atg8C sequence (gene ID: ISCW000710-RA, gene ID: ISCW017654-RA) are accessible at VectorBase (https://www.vectorbase.org/organisms/ixodes-scapularis). siRNA primers targeting I. scapularis Atg8A and Atg8C were designed by BLOCK-iT RNAi Designer (Thermo Fisher Scientific) and are listed in Table S1.Primers were searched against the I. scapularis genome database using BLASTN to evaluate target-specificity.Scrambled and Atg8A or Atg8C-specific siRNA were synthe sized using the Silencer siRNA Construction Kit (Thermo Fisher Scientific), following the manufacturer's protocols.Lipofectamine 3,000 Transfection Reagent was used for the transfection of siRNA into tick cells (Thermo Fisher Scientific).Uninfected and R. buchneri infected cells were transfected with 160 nM siRNA (or 160 nM ssiRNA as control) and harvested for mRNA isolation 6 days after initiation of transfection.The cocktail of Atg8A-specific and Atg8C-specific siRNA (ssiRNA as control) were transfected into R. buchneri infected cells and harvested for mRNA isolation 6 days after initiation of transfection.Atg8A and Atg8C gene relative expression were detected by qRT-PCR and normalized to gapdh.R. buchneri genomic DNA was measured by qPCR and normalized to citrate synthase-encoding gene (gltA) after siRNA treatments as previously described (42).The student's two-tailed t-test was applied to analyze the quantities of R. buchneri DNA and the relative gene expression.

FIG 1
FIG 1 Autophagy activation in vitro after R. buchneri infection.(A) ATG in I. scapularis (highlighted in yellow), compared to D. melanogaster.(B) Expression profiles of IsAtg genes in IRE11 cells and R. buchneri-infected IRE11 cells (10% and 30% infection levels).The left clustering showed the gene expression patterns, and the heatmap represents different treatments (each column) and IsAtg expression relative to gapdh (each row).The color bars show the scale of relative expression (Continued on next page)

FIG 3
FIG 3 No apoptosis induction after R. buchneri infection.(A) Caspase3/7 enzyme activity in IRE11 and R. buchneri-infected IRE11 (10% infection) detected by the fluorescence plate reader.(B and C) R. buchneri-infected IRE11 (10% infection) cells showed no difference in red intensity representing caspase activity, relative to uninfected IRE11 cells.Blue DAPI staining corresponds to the nuclei.(D and E) IRE11 cells and R. buchneri-infected IRE11 cells (10% infection) were fixed and labeled with TUNEL (green).Blue DAPI staining corresponds to the nuclei.(F) Percentage of apoptotic cells (number of TUNEL-positive cells/number of DAPI-positive cells) in different treatments.In panels A and F, data are mean ± SD, and different letters above the columns indicate significant differences, P = 0.9496 (panel A), P = 0.1813 (panel B) (Student's two-tailed t-tests).