Intracellular Trafficking and Persistence of Acinetobacter baumannii Requires Transcription Factor EB

Intracellular Trafficking and Persistence of Acinetobacter baumannii Requires Transcription Factor EB Raquel Parra-Millán, a David Guerrero-Gómez, b Rafael Ayerbe-Algaba, a Maria Eugenia Pachón-Ibáñez, a Antonio Miranda-Vizuete, b Jerónimo Pachón, a Younes Smani a a Clinic Unit of Infectious Diseases, Microbiology and Preventive Medicine, Institute of Biomedicine of Seville, IBiS, University Hospital Virgen del Rocío/CSIC/University of Seville, Seville, Spain b Redox Homeostasis Group, Institute of Biomedicine of Seville, IBiS, University Hospital Virgen del Rocío/CSIC/ University of Seville, Seville, Spain ABSTRACT Acinetobacter baumannii is a significant human pathogen associated with hospital-acquired infections. While adhesion, an initial and important step in A. baumannii infection, is well characterized, the intracellular trafficking of this pathogen inside host cells remains poorly studied. Here, we demonstrate that transcription factor EB (TFEB) is activated after A. baumannii infection of human lung epithelial cells (A549). We also show that TFEB is required for the invasion and persistence inside A549 cells. Consequently, lysosomal biogenesis and autophagy activation were observed after TFEB activation which could increase the death of A549 cells. In addition, using the Caenorhabditis elegans infection model by A. baumannii, the TFEB orthologue HLH-30 was required for survival of the nematode to infection, although nuclear translocation of HLH-30 was not required. These results identify TFEB as a conserved key factor in the pathogenesis of A. baumannii. IMPORTANCE Adhesion is an initial and important step in Acinetobacter baumannii infections. However, the mechanism of entrance and persistence inside host cells is unclear and remains to be understood. In this study, we report that, in addition to its known role in host defense against Gram-positive bacterial infection, TFEB also plays an important role in the intracellular trafficking of A. baumannii in host cells. TFEB was activated shortly after A. baumannii infection and is required for its persistence within host cells. Additionally, using the C. elegans infection model by A. baumannii, the TFEB orthologue HLH-30 was required for survival of the nematode to infection, although nuclear translocation of HLH-30 was not required. KEYWORDS Acinetobacter baumannii,Caenorhabditis elegans, HLH-30, TFEB, bacterial invasion Acinetobacter baumannii is an important cause of severe hospital-acquired infections, such as ventilator-associated pneumonia, bacteremia, skin and soft tissue infections, surgical site infections, urinary tract infections, sepsis, and meningitis in humans (1, 2). The virulence of A. baumannii is based on multiple secreted and surface-associated components. An important group of virulence factors are the outer membrane proteins (OMPs). Among the OMPs, the outer membrane protein A (OmpA) interacts with biotic and abiotic surfaces (3–5). Several other surface and intracellular proteins have been identified, and their role on the virulence of A. baumannii has been characterized (6). For many of these proteins like OmpA, Omp33, phosphorylcholine, lipopolysaccharide, K1 capsular polysaccharide, penicillin-binding protein, and phospholipase D, isogenic mutants are less virulent in vitro and in vivo (4, 5, 7–12). Adhesion is an initial and important step in A. baumannii infections. Following Received 25 February 2018 Accepted 5 March 2018 Published 28 March 2018 Citation Parra-Millán R, Guerrero-Gómez D, Ayerbe-Algaba R, Pachón-Ibáñez ME, MirandaVizuete A, Pachón J, Smani Y. 2018. Intracellular trafficking and persistence of Acinetobacter baumannii requires transcription factor EB. mSphere 3:e00106-18. https://doi.org/10.1128/ mSphere.00106-18. Editor Sarah E. F. D'Orazio, University of Kentucky Copyright © 2018 Parra-Millán et al. This is an open-access article distributed under the terms of the Creative Commons Attribution 4.0 International license. Address correspondence to Jerónimo Pachón, [email protected], or Younes Smani, [email protected]. RESEARCH ARTICLE Host-Microbe Biology crossm March/April 2018 Volume 3 Issue 2 e00106-18 msphere.asm.org 1 on May 25, 2018 by guesthttp://msphere.asm.org/Downloaded from adhesion, A. baumannii can invade host cells such as human lung, laryngeal, and cervical epithelial cells (4, 5). A. baumannii enters epithelial cells by way of a microfilamentand microtubule-dependent, zipper-like mechanism, and upon internalization, it localizes to membrane-bound vacuoles (4). Clathrin and ␤ -arrestins are also engaged during the uptake into human lung epithelial cells (8). Interestingly, A. baumannii can persist within host cells; however, no intracellular replication has been reported. In vitro and in vivo data from our group have demonstrated that A. baumannii induces host cell death and disseminates in tissues and bloodstream (7, 8). Although the dissemination to deeper tissues leads to invasive diseases, their intracellular trafficking is unclear and remains to be understood. A better understanding of the host factors involved in the A. baumannii intracellular trafficking will help to elucidate the role of this process during infection. Given that A. baumannii is localized to membrane-bound vacuoles (4), we hypothesize that endosomes and lysosomes are involved in the passage of A. baumannii to access the basal side of host cells, through which it can transit across to the underlying tissue. In mammalian host cells, transcription factor EB (TFEB) is known to control the transcription of autophagy and lysosomal biogenesis genes in response to nutritional stress (13). TFEB is regulated by the kinases mammalian target of rapamycin complex 1 (mTORC1) and extracellular signal-regulated kinase 2 (ERK2) (14, 15). During bacterial infection, TFEB and its Caenorhabditis elegans orthologue HLH-30 are regulated by the phospholipase C-protein kinase D (PLC-PKD) pathway (16) and play an important and evolutionary role in the host defense against Gram-positive bacterial infection (16, 17). In this study, we report that, in addition to its known role in host defense against Gram-positive bacterial infection, TFEB also plays an important role in the intracellular trafficking of a Gram-negative bacillus (GNB), such as A. baumannii, in host cells. TFEB was activated shortly after A. baumannii infection and is required for its persistence within host cells. In addition, HLH-30 is required for C. elegans survival to A. baumannii infection. RESULTS TFEB expression in A549 cells infected by A. baumannii.A. baumannii cellular infection increased the expression of TFEB in a time-dependent manner after6hby ⬇83% (Fig. 1A). TFEB immunostaining showed that in noninfected control cells, most of TFEB staining is found in the cytoplasm, while 2 h after infection, TFEB is robustly translocated into the nucleus (Fig. 1B). Role of TFEB in A. baumannii internalization by A549 cells. We first tested the ability of TFEB small interfering RNA (siRNA) to deplete the TFEB levels in A549 cells. TFEB siRNA transfection reduced the expression of targeted TFEB by ⬇55% compared with either nonsilenced or control siRNA-transfected A549 cells (Fig. 2A). TFEB siRNAtransfected A549 cells were found to be effective in decreasing A. baumannii invasion to 36.32% ⫾17.6%. However, total cell-adhered bacteria did not differ between control and TFEB siRNA-transfected A549 cells, indicating that invasion did not decrease due to inefficient binding of A. baumannii to A549 cells (Fig. 2B). In contrast, control siRNAtransfected A549 cells did not show significant blocking of A. baumannii adherence and invasion in A549 cells (Fig. 2B). It is important to note that A549 cells transfected with scrambled or TFEB siRNA had no effect on the viability of A549 cells during 24 h of incubation (see Fig. S1 in the supplemental material). In addition, we evaluated the effect of TFEB overexpression in A549 cells on A. baumannii adherence and invasion. We first tested the ability of pEGFP-N1-TFEB (EGFP stands for enhanced green fluorescent protein) to increase the TFEB levels in A549 cells. pEGFP-N1-TFEB transfection increased the expression of targeted TFEB by ⬇80% (Fig. 2C). The pEGFP-N1-TFEB-transfected A549 cells were found to be effective in increasing A. baumannii invasion to 195.69% ⫾50.43%. However, total cell-adhered bacteria did not differ between control and pEGFP-N1-TFEB-transfected A549 cells, indicating that this increase was not due to high binding of A. baumannii to A549 cells (Fig. 2D). Interestingly, TFEB siRNAand pEGFP-N1-TFEB-transfected A549 cells proParra-Millán et al. March/April 2018 Volume 3 Issue 2 e00106-18 msphere.asm.org 2 on May 25, 2018 by guesthttp://msphere.asm.org/Downloaded from longed the significant reduction and increase of the A. baumannii persistence inside A549 cells during8hofbacterial infection by 2.95 and 1.17 log CFU/ml compared with control cells, respectively (Fig. 2E). Taken together, these results demonstrate that TFEB is involved in the A. baumannii invasion of human lung epithelial cells. Implication of the autophagosome-lysosome system in A. baumannii intracellular trafficking. To evaluate the role of the autophagosome-lysosome system in A. baumannii intracellular trafficking, we studied the activation of lysosomes upon bacterial infection. We showed that incubation of A549 cells with A. baumannii for2h increased the numbers of lysosomes by ⬇50%. In contrast, heat-killed A. baumannii did not increase the number of lysosomes significantly (Fig. 3A). In addition, live bacteria persist inside A549 cells for at least 8 h even if the lysosomes were more abundant (Fig. 3B), unlike heat-killed A. baumannii, which activates few lysosomes. Analysis of lysosome membrane damage showed that cathepsin D, an enzyme present inside lysosomes, was released by ⬇50% in cytosol after bacterial infection (Fig. 3C). With this precedent in mind and given that lysosomal acidification is crucial to the antimicrobial function of host cells (18), we sought to determine the impact of acidic and neutral conditions on bacterial intracellular viability by testing the effect of NH 4 Cl (19) and KCl (20), respectively, on intracellular persistence of A. baumannii. In A549 cells treated with NH 4 Cl (40 mM, 30 min) before bacterial infection, survival of A. baumannii inside these cells was reduced by 2.42 log CFU/ml (P⬍0.05) at8h( Fig. 3D). In contrast, treatment of A549 cells with KCl (0.2 mM, 30 min) before bacterial infection increased the persistence of A. baumannii inside these cells by 1.24 log CFU/ml (P⬍0.05) at 8 h (Fig. 3D). It is important to note that treating A549 cells with NH 4 Cl and KCl had no effect on the viability of A549 cells during 24 h of incubation (data not shown), which suggests that persistence of A. baumannii inside A549 cells was not due to the death of these cells. Thus, the observed persistence of A. baumannii inside A549 cells may be due to less lysosome acidification. It is noteworthy that A. baumannii growth under acidic and neutral pH in LB broth at 24 h was unchanged, although their growth dynamic is different in the first hours of bacterial growth (Fig. 4A). In acidic conditions, the pH value of LB medium during A. baumannii growth shifted from 4.72 to 6.17 FIG 1 Expression of TFEB in A549 cells by A. baumannii. (A) Western blot analysis of TFEB in A549 cells after incubation with A. baumannii ATCC 17978 for 0.5, 2, and 6 h. The solid black lines in the blots separate the spliced portions of the blots between 2 and 6 h. Values shown in the bar graph are the percentage of TFEB expression in control (CTL) and infected A549 cells. (B) TFEB in A549 cells after incubation with A. baumannii ATCC 17978 for 0.5 and 2 h, immunostaining, and imaging by immunofluorescence microscopy. TFEB detected with rabbit anti-TFEB antibodies and labeled with Alexa Fluor 594-tagged secondary antibodies (red). Blue staining with DAPI shows the location of nuclei of A549 cells. The percentage of TFEB expression in the nuclei of A549 cells was calculated as follows: (number of A549 cells that expressed TFEB in the nuclei of A549 cells/total number of A549 cells) ⫻100. Results are from three independent experiments, and data are means plus standard errors of the means (SEM) (error bars). Values that are significantly different (P⬍0.05) between untreated (control [CTL]) and treated groups are indicated by an asterisk. Acinetobacter baumannii Invasion Mediated by TFEB March/April 2018 Volume 3 Issue 2 e00106-18 msphere.asm.org 3 on May 25, 2018 by guesthttp://msphere.asm.org/Downloaded from between 0 and 24 h (Fig. 4B), while the pH of the LB medium after 24 h of A. baumannii growth in neutral pH was unchanged and shifted only from 6.96 to 6.81 (Fig. 4B). These data confirm that A. baumannii is acid resistant, as are other GNB (21), and suggest that in A549 cells, A. baumannii is able to resist lysosome acidification. Lu et al. showed that Streptococcus group A induces low lysosome acidification without sufficient activation of autophagy in endothelial cells (18). To test whether A. baumannii may activate autophagy, we determined the expression profiles of 84 autophagic genes by real-time PCR in A. baumannii-infected A549 cells compared with FIG 2 Role of TFEB in A. baumannii internalization by A549 cells. (A and C) Immunoblot analysis of A549 cells transfected with scrambled (SC) and TFEB siRNA and pEGFP-N1-TFEB for 48 and 24 h, respectively. Values in the bar graphs are the percentages of TFEB level in control (CTL) and transfected A549 cells. (B, D, and E) A549 cells were transfected with SC and TFEB siRNA and pEGFP-N1-TFEB and infected with 10 8 CFU/ml A. baumannii ATCC 17978 for 2, 4, or 8 h. An assay of adherence and invasion of A. baumannii ATCC 17978 into A549 cells was performed as described in Materials and Methods. The effect of TFEB siRNA and pEGFP-N1-TFEB mediated TFEB depletion and overexpression, respectively, on adherence or invasion of A. baumannii ATCC 17978. The percentages of total nontransfected A549 cells and A549 cells incubated with A. baumannii ATCC 17978 are shown for both adhesion and invasion. Results are from three independent experiments, and data are the means plus SEM (error bars). Values for untransfected and transfected groups in panels B and E that are significantly different (P⬍0.05) are indicated by an asterisk. Values in panel E that are significantly different (P⬍0.05) are indicated by bars and asterisks as follows: **, ATCC 17978 cells transfected with siRNA TFEB and ATCC 17978 cells or ATCC 17978 cells transfected with pEGFP-N1-TFEB; ***, ATCC 17978 cells and ATCC 17978 cells transfected with pEGFP-N1-TFEB. Parra-Millán et al. March/April 2018 Volume 3 Issue 2 e00106-18 msphere.asm.org 4 on May 25, 2018 by guesthttp://msphere.asm.org/Downloaded from noninfected control cells. After2hofbacterial infection, 79 genes were upregulated (Fig. 5A), including MAP1LC3B encoding LC3B, the most studied gene in autophagy (22). Western blot analysis of LC3BII showed that LC3BII protein levels were increased by ⬇50% (Fig. 5B) and confirmed the result obtained by the autophagic gene expression profiling. Next, we addressed whether the autophagosome-lysosome system might also participate in the intracellular trafficking of A. baumannii inside A549 cells. In this regard, pharmacological inhibition by pepstatin, an inhibitor of lysosomal degradation (23), bafilomycin, an inhibitor of the fusion between autophagosomes and lysosomes FIG 3 Evaluation of the role of the autophagosome-lysosome system in A. baumannii intracellular trafficking. (A) The lysosomes in A549 cells were incubated with A. baumannii ATCC 17978 for 2 h, immunostained, and imaged by immunofluorescence microscopy. Acidic organelles were detected with LysoTracker red (75 nM), and mitochondria were detected with MitoTracker green (250 nM). The values for labeling of lysosomes in infected A549 cells in the bar graph are percentages compared to the value for noninfected cells. Values that are significantly different (P⬍ 0.05) are indicated by bars and asterisks as follows: *, ATCC 17978 and control (CTL) cells; **, ATCC 17978 and heat-killed (HK) ATCC 17978 cells. (B) A. baumannii ATCC 17978 and ATCC 17978 HK invasion into A549 cells for up to8hofinfection. (C) Western blot analysis of cathepsin D in A549 cells infected with A. baumannii ATCC 17978 for 2 h. Blots were part of the same internally controlled experiment in Fig. 5B. Values are expressed as the percentage of cathepsin D expression level in control and infected A549 cells. Values that are significantly different (P⬍0.05) are indicated by an asterisk. (D) A. baumannii ATCC 17978 invasion into A549 cells pretreated for 30 min with NH 4 Cl or KCl for various lengths of time up to8hofinfection. Values that are significantly different (P⬍0.05) are indicated by asterisks as follows: *, ATCC 17978 cells and ATCC 17978 cells treated with NH 4 Cl; **, ATCC 17978 and ATCC 17978 treated with KCl. Acinetobacter baumannii Invasion Mediated by TFEB March/April 2018 Volume 3 Issue 2 e00106-18 msphere.asm.org 5 on May 25, 2018 by guesthttp://msphere.asm.org/Downloaded from (24), or wortmannin, an inhibitor of the class III phosphatidylinositol 3-kinase (PI3K) activity (24), reduced A. baumannii invasion inside A549 cells to 49.14% ⫾17.11%, 41.02% ⫾8.97%, or 40.79% ⫾17.9%, respectively. However, the total numbers of cell-adhered bacteria did not differ in untreated and pepstatin-, bafilomycin-, or wortmannin-treated A549 cells, indicating that the inhibition was not due to inefficient binding of A. baumannii to A549 cells (Fig. 6A). Interestingly, the pretreatment of A549 cells by pepstatin, bafilomycin, or wortmannin together with the inhibition of TFEB by TFEB siRNA amplified the reduction of A. baumannii invasion of A549 cells to 8.25% ⫾ 5.13%, 16.05% ⫾7.73%, or 26.6% ⫾11.89%, respectively, compared to treatment with pepstatin, bafilomycin, wortmannin, or TFEB siRNA alone (Fig. 6B). Conversely, the pretreatment of A549 cells by pepstatin, bafilomycin, or wortmannin together with the TFEB overexpression by pEGFP-N1-TFEB reduced A. baumannii invasion of A549 cells to 63.34% ⫾5.89%, 46.5% ⫾14.17%, or 66.29% ⫾11.89%, respectively, reducing invasion less than treatment with pepstatin, bafilomycin, or wortmannin alone (Fig. 6B). It is important to note that treating A549 cells with pepstatin, bafilomycin, and wortmannin had no effect on the viability of A549 cells during 24 h of incubation (Fig. S2). Together, these results support a key role for autophagosome-lysosome system in the intracellular persistence of A. baumannii. Collectively, these data suggest a hypothetical model whereby infection triggers TFEB activation, which could induce the autophagosome-lysosome system to promote the intracellular trafficking and persistence of A. baumannii within cells. HLH-30 is necessary for C. elegans survival, but not for A. baumannii infection. To address the role of TFEB in the context of an infective process in a complete FIG 4 Bacterial acid resistance. (A) Bacterial growth in LB medium during 24 h under acidic or neutral conditions. (B) pH determination during 24 h of LB medium in the presence of A. baumannii ATCC 17978. Parra-Millán et al. March/April 2018 Volume 3 Issue 2 e00106-18 msphere.asm.org 6 on May 25, 2018 by guesthttp://msphere.asm.org/Downloaded from organism and because TFEB and its C. elegans orthologue HLH-30 are both regulated by infection (16, 17), we hypothesized that HLH-30 might also be required for C. elegans to cope with A. baumannii infection. To test this idea, we first performed a longevity assay of wild-type and hlh30(tm1978) mutant worms growing in A. baumannii culture. Interestingly, hlh-30 mutant worms have a strong reduction in the mean life span compared with wild-type control worms (11 versus 16 days). In addition, a dramatic decrease occurred in the maximum life span of hlh-30 mutants compared to the wild-type controls (29 versus 15 days) (Fig. 7A). Note that the life span of hlh-30 mutant worms was barely affected when grown on Escherichia coli OP50 (17). Similarly, while the brood size of hlh-30 mutant worms was not different from that of wild-type controls when grown on E. coli OP50, the brood sizes of both wild-type and hlh-30 mutant worms were affected significantly FIG 5 A. baumannii stimulates the autophagy. (A) Expression of human autophagy genes after A. baumannii infection. Total RNA was isolated from A549 cells infected with A. baumannii ATCC 17978 and uninfected cells. cDNAs were synthesized by reverse transcription of the total RNA. A real-time PCR analysis was performed by using the Stratagene Mx3005p system. Samples were normalized to beta-2-microglobulin. Human autophagy gene expression after infection is represented by the heat map. Results are representative of two independent experiments. (B) Western blot analysis of LC3B in A549 cells infected with A. baumannii ATCC 17978 for 2 h. The blots were part of the same internally controlled experiment in Fig. 3C. Results are representative of three independent experiments. The solid white line separates the spliced portions between control and infected cells. Acinetobacter baumannii Invasion Mediated by TFEB March/April 2018 Volume 3 Issue 2 e00106-18 msphere.asm.org 7 on May 25, 2018 by guesthttp://msphere.asm.org/Downloaded from when raised on A. baumannii, although to different extents. Thus, wild-type control worms had a brood size reduction of 42%, while hlh-30 mutant worms displayed a dramatic reduction of 93% when grown in A. baumannii compared to E. coli OP50 (Fig. 7B), suggesting that A. baumannii causes defects in germline function in C. elegans. Visual inspection by differential interference contrast (DIC) microscopy of hlh-30 mutant and wild-type worms growing on A. baumannii confirmed this hypothesis and identified severe germline phenotypes like oocytes with abnormal size (enlarged and small), binucleated oocytes, enlarged and deformed embryos, blisters and blebs in the head, vulva, and tail and also in some cases extruded intestinal/uterine contents (Fig. 7C to H). Together, these data suggest that HLH-30 is a key factor for C. elegans to survive A. baumannii infection. A fluorescent HLH-30::GFP reporter was previously shown to translocate to the nucleus upon Staphylococcus aureus infection (16, 17). We then tested whether A. baumannii infection would induce a similar response. Unexpectedly, and in sharp contrast to S. aureus infection, A. baumannii induced very weak HLH-30 nuclear translocation in C. elegans intestinal cells after2hofinfection (Fig. 7I) or 12 to 24 h (data not shown). FIG 6 A. baumannii activates the autophagosome-lysosome system. (A and B) Additive effect of autophagy and TFEB on A. baumannii internalization by A549 cells. A549 cells were transfected with scrambled (SC) or TFEB siRNA or pEGFP-N1-TFEB and treated with pepstatin (20 ␮ g/ml), bafilomycin (0.8 ␮ M), or wortmannin (1 ␮ M), and infected with 10 8 CFU/ml A. baumannii ATCC 17978 for2htostudy bacterial adherence and invasion to host cells. Results are representative of three independent experiments, and data are the means plus SEM. Values for treated and untreated groups that are significantly different (P⬍0.05): are indicated by an asterisk. CTL, control. Parra-Millán et al. March/April 2018 Volume 3 Issue 2 e00106-18 msphere.asm.org 8 on May 25, 2018 by guesthttp://msphere.asm.org/Downloaded from FIG 7 HLH-30 is required for C.elegans survival against A.baumannii infection. (A) Longevity assay of hlh-30(tm1978) mutant worms compared to wild-type control worms growing on A. baumannii at 20°C. Survival of wild-type (n⫽140) and hlh-30(tm1978) mutants (n⫽134) of C. elegans when growing at the same temperature in the nonpathogenic E. coli OP50 or in the pathogenic A. baumannii ATCC 17978. Two independent experiments were performed, and the data for both experiments are shown (***,P⬍0.001). (B) Brood size quantification of wild-type and hlh-30(tm1978) mutants growing from eggs on E. coli OP50 and A. baumannii. Values are means plus SEM for 30 individuals (***,P⬍0.001; ns, not significantly different). (C to H) Differential interference contrast (DIC) micrographs of wild-type and hlh-30(tm1978) mutants growing from eggs on E. coli OP50 and A. baumannii showing alterations in worm germline like enlarged oocytes and embryos (white asterisks in panels F, G, and H), binucleated oocytes (white arrowhead in panel F), vulva extrusion (white arrowhead in panel G), and extensive blebbing in the intestine, vulva, and head regions (H). The solid white line in panel D separates the spliced portions. (I) Fluorescence micrographs (left) and quantification (right) of HLH-30::GFP nuclear translocation in intestinal cells of transgenic worms expressing the integrated array sqIs17 [Phlh-30::hlh-30::GFP;rol-6(su1006)] when grown in S. aureus 29213 or A. baumannii ATCC 17978. Bars, 100 ␮ m (C, D, E, and I) and 50 ␮ m (F, G, and H). D, dorsal; V, ventral; A, anterior; P, posterior. Acinetobacter baumannii Invasion Mediated by TFEB March/April 2018 Volume 3 Issue 2 e00106-18 msphere.asm.org 9 on May 25, 2018 by guesthttp://msphere.asm.org/Downloaded from