First Steps Towards a Vaccine against Acinetobacter baumannii

Depósito de investigación de la Universidad de Sevilla https://idus.us.es/ Esta es la versión aceptada del artículo publicado en: This is a accepted manuscript of a paper published in: Current Pharmaceutical Biotechnology (2013): Volume 14, Issue 10, 2013 DOI: https://doi.org/10.2174/1389201014666131226123511 El acceso a la versión publicada del artículo puede requerir la suscripción de la revista. Access to the published version may require subscription. El manuscrito publicado está disponible en EurekaSelect a través de https://www.eurekaselect.com/openurl/content.php?genre =artículo&doi=10.2174/1389201014666131226123511. En caso de cualquier forma de archivo, se debe dar reconocimiento a la fuente original de publicación y se debe insertar un enlace al artículo publicado en el sitio web de la revista/editor. https://www.eurekaselect.com/article/58291. First steps towards a vaccine against Acinetobacter baumannii 1 2 Meritxell García-Quintanilla*, Marina R. Pulido* and Michael J. McConnell 3 4 Unit of Infectious Diseases, Microbiology, and Preventive Medicine and Biomedical Institute of Seville (IBiS), 5 University Hospital Virgen del Rocío/CSIC/University of Sevilla, 41013, Sevilla, Spain. 6 7 8 *These authors contributed equally to this work. 9 10 Author to whom correspondence should be addressed: 11 Michael J. McConnell 12 Unit of Infectious Disease, Microbiology, and Preventive Medicine 13 Hospital Universitario Virgen del Rocío/Instituto de Biomedicina de Sevilla 14 Avenida Manuel Siurot s/n, 41013 Sevilla, Spain 15 e-mail: [email protected] 16 Phone: +34 955923104 17 Fax: +34 955013292 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 Abstract 38 39 Acinetobacter baumannii has become an important cause of human infections, most notably in 40 the hospital setting. In addition, the global dissemination of multidrug resistant strains has 41 complicated effective antibiotic therapy of infections produced by this pathogen, necessitating 42 the development of novel treatment and prevention strategies. Active and passive immunization 43 approaches have begun to be explored in experimental animal models as potential alternative 44 therapies for A. baumannii. In the present review, we discuss the advantages and disadvantages 45 of each therapeutic strategy with respect to A. baumannii infections, and summarize the recent 46 studies that have explored these approaches. The single antigen candidates that have been tested 47 include, the outer membrane protein OmpA, the membrane transporter Ata, the biofilm48 associated protein Bap, the K1 capsular polysaccharide and the membrane associated 49 polysaccharide poly-N-acetyl-β -(1-6)-glucosamine. Strategies employing multicomponent 50 antigens include inactivated whole cells, outer membrane complexes and outer membrane 51 vesicles. The strengths and limitations of each approach are discussed and the challenges that 52 remain to be addressed for successful A. baumannii vaccine development are highlighted. 53 54 55 Keywords: Acinetobacter baumannii, Vaccine, Passive immunization. 56 INTRODUCTION 57 Acinetobacter baumannii is a Gram-negative coccobacillus that has been associated with a number of 58 human infections including pneumonias, bloodstream infections, meningitis, urinary tract infections, 59 and skin and soft tissue infections. The majority of infections produced by A. baumannii are hospital 60 acquired, although community acquired infections have been reported [1]. Within the hospital 61 setting, 62 A. baumannii has the ability to persist on surfaces within the patient environment for long periods of 63 time, and is thought to be transmitted from patient to patient by contaminated equipment and contact 64 with hospital personnel [2]. Infections caused by this organism have been especially problematic in 65 critically-ill patients in intensive care units, and are associated with increased patient morbidity and 66 mortality [1]. The prevalence of infections caused by A. baumannii can vary widely between 67 institutions and geographic locations. However, a recent report from the European Centre for Disease 68 Control and Prevention which included data from over 900 hospitals from 29 countries indicated that A. 69 baumannii accounted for 3.6% of all hospital acquired infections, 8.7% of all pneumonias/lower 70 respiratory tract infections, and 4.1% of all bloodstream infections [3]. 71 The clinical management of infections caused by 72 A. baumannii has become increasingly difficult due to the high prevalence of infections caused by 73 multidrug resistant strains. In many cases, these strains are resistant to the majority of clinically-available 74 antibiotics, leaving clinicians with few options that still retain adequate antimicrobial activity against 75 the infecting organism. As an example, the carbapenem class of antibiotics has traditionally been a 76 mainstay of treatment for infections caused by A. baumannii. However, the global emergence of strains 77 with resistance to carbapenems over the proceeding two decades has severely compromised the efficacy 78 of this antibiotic class for the treatment of A. baumannii infections. This is evidenced by a recent 79 European point prevalence study that included data from 29 countries in which 81.2% of isolates for 80 which susceptibility data were available demonstrated non-susceptibility to carbapenems [3]. This 81 worrisome trend has prompted the increasing use of novel therapeutic approaches, one of which is the 82 reintroduction of colistin for the treatment of infections caused by multidrug resistant gram negative 83 infections. Colistin is an “old” peptide antibiotic that was introduced half a century ago, although its use 84 has not been widespread in recent decades due to concerns regarding nephrotoxicity [4]. More recently, 85 however, colistin has been used increasingly in the treatment of infections for which local epidemi86 ology indicates that traditional therapies, such as carbapenems, will not have adequate activity. 87 Unfortunately, strains that acquire resistance in response to treatment with colistin have been described 88 [5-7], making the emergence of pandrug resistant strains, with resistance to all clinically-used antibi89 otics, a reality that has already been reported in sporadic cases [8, 9]. 90 In this current context of antibiotic resistant A. baumannii, the development of novel treatment and 91 prevention strategies is of interest. While antibiotic stewardship practices and hospital hygiene-based 92 approaches will undoubtedly play a role in combating the appearance of resistant strains, new 93 therapeutics with the ability to prevent and/ortreat infections caused by highly resistant strains of A. 94 baumannii could provide an important alternative to existing treatments, which are increasingly 95 ineffective [10]. Vaccination represents a therapeutic approach that has the potential to reduce patient 96 morbidity and mortality, and at the same time help to prevent the emergence of resistance by 97 decreasing clinicians’ dependence on antibiotics for treating infections caused by A. baumannii. A 98 handful of preclinical studies have begun to characterize different vaccine candidates in animal models 99 of infection. In general, these vaccines can be divided into two broad categories, vaccines based on 100 single purified bacterial antigens, and vaccines that contain multiple antigens. In this review, we 101 summarize the results that have been obtained from these initial studies, and comment on the 102 remaining challenges that still must be addressed for successful development of a vaccine against A. 103 baumannii. 104 105 106 107 ACTIVE IMMUNIZATION VS. PASSIVE IMMUNIZATION/ANTIBODY-BASED THERAPY 108 Active immunization, which involves the administration of an antigen that stimulates a protective 109 immune response, and passive immunization/antibody-based therapy, which relies on the 110 administration of antibodies with antibacterial activity, both represent potential novel therapeutic ap111 proaches which could be effective for prevention and/or treatment of infections caused by A. 112 baumannii. Advantages of the active immunization approach are that it is a costeffective method with 113 a proven track record regarding safety and efficacy. The major drawback of active immunization for 114 preventing infections caused by A. baumannii is that patients at risk for infection must be identified 115 with sufficient lead time to allow for the host to mount a protective immune response after 116 immunization. Due to the fact that A. baumannii produces predominately nosocomial infections, it may 117 not be possible to vaccinate acute patients admitted urgently with sufficient lead time for achieving 118 protective immunity. In contrast, antibody-based passive immunization has the potential to provide 119 instantaneous protective immunity, thus avoiding the problem of lead time that could occur with active 120 immunization. In addition, antibodies with antibacterial activity could potentially be used for treating 121 established A. baumannii infections, either alone or in combination with traditional antibiotic therapy. 122 The drawbacks associated with antibody-based approaches are that they are largely unproven in the 123 clinical setting for bacterial infections, and that they are typically significantly more costly than active 124 immunization approaches. Regardless of the immunization approach that is employed, active 125 vaccination and passive immunization, in many cases, both rely on the identification of an antigen(s) 126 that can induce protective immunity. In the sections below, the antigens that have been tested in 127 preclinical models are presented and their advantages and disadvantages discussed. 128 129 SINGLE ANTIGEN VACCINES 130 Vaccines based on a single bacterial antigen are attractive because high levels of antigen purity can be 131 obtained, thus avoiding the presence of contaminating bacterial components such as lipopolysaccharide 132 (LPS). In addition, manufacturing and regulatory issues may be less cumbersome for vaccines based on 133 a single, well-defined bacterial component (Table 1 ). A potential disadvantage of a vaccine based on a 134 single antigen includes the possibility that not all circulating strains express the chosen antigen, likely 135 resulting in the ineffectiveness of the vaccine against such a strain. It is also possible that bacteria may 136 adapt to the selective pressure produced by an immune response generated by immunization through 137 downregulation of the antigen targeted by the vaccine. In this scenario, the downregulation of a single 138 antigen may be more readily achieved than the downregulation of several targeted antigens. Similarly, 139 the targeting of a single antigen may result in the selection and expansion of a variant that was a 140 minority member of the bacterial population and does not possess the epitope to which the vaccine141 induced antibodies are directed. The single antigen vaccines that have been studied for A. baumannii 142 include three outer membrane proteins, outer membrane protein A (OmpA), the biofilm associated 143 protein (Bap) and a surface autotransporter (Ata), and the surface polysaccharide poly-N-acetyl-β -(1144 6)- glucosamine (PNAG) and the capsular polysaccharide. 145 Acinetobacter baumannii OmpA is a transmembrane protein that shares sequence and structural 146 homology with the OmpA superfamily of proteins, which are highly conserved among Gram-negative 147 bacteria [11]. OmpA has been implicated in a number of virulence traits expressed by A. baumannii 148 including induction of host cell apoptosis, adherence, biofilm formation and surface motility [12-14]. 149 OmpA may also play a role in bacterial survival and/or dissemination during infection given data 150 showing that blood bacterial loads in mice with A. baumannii pneumonia were lower in mice 151 infected with an ompA mutant compared to the wild type strain [13]. Lou et al. identified OmpA as a 152 target of the humoral immune response produced in mice during sublethal intravenous infection with A. 153 baumannii [15]. Based on these findings, purified recombinant OmpA was combined with an 154 aluminium hydroxide adjuvant and used for immunization. Vaccination stimulated high levels of 155 OmpA-specific antibodies and provided partial protection from intravenous infection in a diabetic 156 mouse model. Vaccination also resulted in a 10-fold reduction in tissue bacterial loads compared to 157 control mice. Importantly, anti-OmpA antibodies were shown to enhance opsonophagocytic killing of 158 A. baumannii in vitro, and were able to mediate passive protection in mice. OmpA is known to be a 159 predominant component of the bacterial outer membrane, and sequence analysis has shown that it is 160 highly conserved between A. baumannii strains [15], supporting the use of this antigen for vaccine 161 development. 162 Bap is an 854 kDa surface exposed protein that is involved in biofilm formation [16]. Fattahian et al. 163 evaluated a 371 amino acid region of Bap that had previously been predicted to be a conserved 164 functional motif in the native protein as a vaccine antigen [17]. Three administrations of the Bap 165 subunit in combination with Freund’s complete and incomplete adjuvants produced antigen-specific 166 antibodies that were able to recognize intact bacterial cells in a whole cell ELISA. Immunized mice had 167 lower bacterial loads in spleen and liver 18 hours after intraperitoneal infection, and showed increased 168 survival compared to unimmunized controls. An important consideration for a vaccine based on Bap is 169 the expression of this protein in circulating clinical strains of A. baumannii, since it has been shown that 170 not all strains produce biofilm [18], and it is unknown if strains express Bap independent of whether or 171 not biofilm is produced. 172 The A. baumannii Ata protein is a surface-exposed, trimeric autotransporter that has been shown to 173 participate in biofilm formation and the adhesion of A. baumannii cells to host extracellular and 174 basement membrane proteins such as collagen type IV [19]. Ata also appears to play a role in 175 pathogenesis as mice infected with wild type A. baumannii showed decreased survival compared to 176 mice infected with an isogenic mutant strain lacking Ata expression in an intraperitoneal model of 177 infection. Rabbit antisera to Ata were able to block the binding of A. baumannii to immobilized 178 collagen type IV and promote the opsonophagocytic killing and complement-dependent bactericidal 179 killing of A. baumannii strains in vitro, including multidrug resistant clinical isolates [20]. Passive 180 immunization of mice with Ata antisera resulted in reduced lung bacterial loads after intranasal infec181 tion compared to control mice receiving non-immune serum, demonstrating that antibodies against Ata 182 have antibacterial activity in vivo. It remains to be seen if the antibacterial activity mediated by Ata 183 antibodies is sufficient for reducing post-infection mortality in animal models that evaluate survival. As 184 with other antigens, the presence and expression of Ata in circulating strains of A. baumannii is a 185 critical issue. Bentancor et al. used PCR to demonstrate that the ata gene was present in 44/75 (58.6%) 186 of isolates from different geographic locations, and that the levels of surface expression of Ata in PCR187 positive isolates showed considerable variation, indicating that the presence and expression of Ata 188 varies between clinical strains. 189 The surface-associated polysaccharide PNAG is produced by a variety of Gram-positive and Gram190 negative bacterial pathogens and consists of linked subunits of N-acetylD-glucosamine [21, 22]. In A. 191 baumannii, PNAG is synthesized by the products of the pgaABCD genes, as a mutant lacking pgaABC 192 genes did not produce PNAG [23]. PNAG has been shown to play a role in biofilm formation in A. 193 baumannii. Bentancor et al. raised rabbit antisera against a synthetic nonameric oligonucleotide to 194 evaluate the ability of PNAG-specific antibodies to mediate bacterial killing and protection from 195 infection [24]. The antisera were able to promote the opsonophagocytic killing of a PNAG-producing 196 strain, but not its isogenic mutant lacking the pgaABC genes. In addition, the antisera were able to 197 promote opsonophagocytolysis of four multidrug resistant clinical isolates. In animal models of 198 pneumonia and bacteremia, passive immunization with the anti-PNAG sera resulted in lower bacterial 199 loads in lung and blood, respectively. Protection studies evaluating survival after administration of 200 PNAG antisera were not reported. The presence of the pgaABCD locus in a collection of 30 multidrug 201 resistant clinical isolates was evaluated by PCR, and showed that all 30 produced amplification 202 products corresponding to the predicted size [23]. An immunoblot assay demonstrated that of these 30 203 strains, 14 showed high levels of PNAG expression, 14 showed low levels of expression, and 2 204 demonstrated no PNAG expression. The use of PNAG as a vaccine antigen is supported by studies with 205 other PNAG-expressing bacterial species that have shown encouraging results in preclinical models 206 [25, 26]. 207 The A. baumannii capsular polysaccharide has been shown to play a role in bacterial growth and 208 survival in human ascites fluid, human serum and a rat soft tissue model of infection [27]. Based on 209 these findings, and the fact that the capsular polysaccharide is surface exposed and that bacterial 210 capsules have formed the basis for vaccines against a number of pathogens [28-30], Russo et al. 211 characterized a monoclonal antibody directed against the K1 capsular polysaccharide of A. baumannii 212 in passive immunization studies [31]. Their results showed that the monoclonal antibody could 213 promote neutrophil-mediated bactericidal activity in vitro and reduce post-infection bacterial loads in a 214 rat soft tissue model of infection. Using an immunoassay, the monoclonal antibody reacted with 13 of 215 100 (13%) strains of A. baumannii from different geographic locations and isolated from different body 216 sites or environmental sources. Clearly in the case of the capsular polysaccharide more work is 217 necessary in order to characterize the different capsular serotypes in circulating strains in order to 218 facilitate broad coverage of a potential vaccine or antibody-based therapy. 219 220 VACCINES CONTAINING MULTIPLE BACTERIAL ANTIGENS 221 Vaccines containing multiple bacterial components, such as whole cells or membrane complexes, have 222 the advantage of potentially producing an immune response against multiple antigens. This may result 223 in increased vaccine coverage of strains within a bacterial species compared to vaccines based on 224 single antigens given that antigen expression can vary widely between strains. Potential disadvantages 225 of these types of vaccines include difficulties associated with achieving consistent levels of all vaccine 226 components between production lots, and the presence of bacterial components that could produce 227 unwanted side effects such as LPS. Multi-antigen vaccines against A. baumannii that have been re228 ported to date include a vaccine based on outer membrane complexes, an inactivated whole cell 229 vaccine and a vaccine consisting of outer membrane vesicles (OMVs). 230 Outer membrane complexes are prepared by isolating whole bacterial membranes and then solubilizing 231 the inner membrane component using a detergent before subsequent removal. Proteomic analysis of 232 outer membrane complexes prepared from the A. baumannii ATCC 19606 strain grown in laboratory 233 media identified 61 protein components, 41 of which were predicted to be located on the cell surface 234 [32]. Immunization of mice with the outer membrane complexes induced antibodies against multiple 235 bacterial outer membrane proteins that were able to recognize surface proteins from multiple clinical 236 isolates. Immunized mice infected with A. baumannii using an intraperitoneal sepsis model 237 demonstrated dramatically reduced post-infection tissue bacterial loads and lower serum levels of the 238 pro-inflammatory cytokines IL-6, IL-1β and TNF-a compared to unimmunized controls. Vaccinated 239 mice also showed increased survival after infection, including after infection with a pandrug resis tant 240 clinical isolate. Importantly, treatment of previously infected mice with antisera raised against outer 241 membrane complexes 1 hour after infection was able to therapeutically rescue mice from infection. 242 These results indicate that a mix ture of outer membrane components can induce a potent immune 243 response. However, the use of outer membrane complexes as a vaccine is limited by the difficult nature 244 of standardizing the levels of the multiple components that are present in the vaccine preparation. 245 Vaccines based on whole bacterial cells have the advantage of potentially inducing a response against 246 multiple surface antigens in their native conformation. An inactivated whole cell vaccine prepared by 247 formalin inactivation of the ATCC 19606 strain was highly immunogenic and produced antibodies 248 against multiple bacterial outer membrane proteins [33]. Immunized mice showed reduced bacterial 249 loads and serum cytokine levels compared to control mice, and increased survival after infection with 250 the ATCC 19606 strain and two clinical isolates. While these results are promising, concerns about 251 using whole cells due to the high levels of LPS present a crucial limitation. 252 OMVs are vesicles produced from the bacterial outer membrane that are actively produced by 253 numerous Gramnegative bacteria [34]. They are typically 20-200 nm in size and consist of multiple 254 bacterial components including periplasmic and outer membrane proteins, as well as LPS. A. baumannii 255 OMVs have been shown to contain a number of potential virulence factors and immune modulating 256 proteins [35]. In vaccination studies, A. baumannii OMVs were highly immunogenic, producing 257 antibodies against multiple bacterial outer membrane proteins [36]. Mice vaccinated with OMVs had 258 reduced post-infection tissue bacterial loads and increased survival after intraperitoneal infection com259 pared to control mice. Similar to whole cell vaccines, vaccines based on OMVs have the advantage of 260 presenting multiple antigens in their native form. Unfortunately, OMVs, like whole cells, have the 261 limitation of containing high levels of LPS. OMVs that have been detergent extracted to remove LPS, or 262 OMVs isolated from strains genetically modified to lack LPS have been employed as vaccines for other 263 bacterial species [37], raising the possibility that this approach may hold promise for A. baumannii 264 OMVs as well. 265 266 REMAINING CHALLENGES 267 The studies summarized above demonstrate the proof of concept that protective immunity can be 268 achieved against A. baumannii through both active and passive immunization. However there is clearly 269 a great deal of further work that must be performed before these therapies can begin to be considered 270 as viable alternative therapies for treating and preventing infections caused by this pathogen. An ideal 271 antigen would be present on the cell surface, highly conserved between strains within the species, and 272 be highly expressed during infection. Although some of the candidates described above hold promise, 273 the continued identification of antigens that meet such criteria is warranted. An additional aspect that 274 must be clarified is the definition of the correlates of protective immunity for A. baumannii. While 275 passive immunization experiments indicate that antibodies alone are sufficient for providing protection 276 against infection, the role of the cell-mediated immune response has not been characterized. The 277 findings that the cell-mediated immune response may play a role in controlling infections caused by 278 Pseudomonas aeruginosa, a bacterium that is phylogenetically related to A. baumannii and produces 279 similar types of infections, supports the idea that this aspect of the immune response may have 280 importance [38, 39]. Finally, although a number of animal models of infection by A. baumannii have 281 been developed [1], in some cases these models are not ideal for characterizing the protective capacity 282 of vaccines. Due to the low virulence of A. baumannii in mice, some models employ neutropenic mice 283 or virulence enhancing agents (e.g. porcine mucin) in order to achieve mortality, although a recent 284 study has identified a strain the produces lethal infection without manipulation [40]. These limitations 285 demonstrate the importance of the further development of additional models that more accurately 286 reflect human infection with A. baumannii (i.e. models that permit bacterial growth and do not 287 provoke a cytokine “storm” in order to produce mortality). Although further studies are required, the 288 continued development of active and passive immunization strategies for A. baumannii is of interest 289 given their potential for reducing the morbidity and mortality produced by this drug resistant pathogen. 290 CONFLICT OF INTEREST 291 MJM owns stock in and act as scientific advisors for Vaxdyn, S.L., a biotechnology company 292 developing vaccines for multidrug resistant bacteria, including A. baumannii. The other authors declare 293 no potential conflicts of interest. 294 295 ACKNOWLEDGEMENTS 296 This work was funded by a grant from the European Community’s 7 th Programme Framework 297 (MagicBullet; Grant Agreement Number: 278232) and the Ministerio de Economía y Competitividad, 298 Instituto de Salud Carlos III - co-financed by European’s Development Regional Fund "A way 299 toachieve Europe" ERDF, Spanish Network for the Research in Infectious Diseases (REIPI 300 RD06/0008/0000). MJM is supported by the Subprograma Miguel Servet from the Minis terio de 301 Economía y Competitividad of Spain (CP11/00314) and a grant from the Consejería de Salud y 302 Bienestar de la Junta de Andalucía (PI-0046-2011). 303 304 305 LIST OF ABBREVIATIONS 306 307 Ata = Acinetobacter trimeric autotransporter Bap = Biofilm-associated protein ELISA = Enzyme -linked immunosorbant assay LPS = Lipopolysaccharide OmpA = Outer membrane protein A OMV = Outer membrane vesicles PNAG = poly-N-acetyl-β-(1-6)- glucosamine 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 Table 1. Advantages and disadvantage of single and multiple antigen vaccine strategies. 326 327 Vaccine Type Advantages Disadvantages Single antigen vaccines Well - defined composition, low levels of reactogenic impurities, existence of standardized methods for industrial production Concerns regarding expression of the antigen in all strains, adaptation to immune pressure via antigen down - regulation more feasible, purification process can alter native antigen conformation Multicomponent vaccines Higher stain coverage due to targeting of multiple antigens, reduced risk of adaptation due to immune pressure, antigens can be maintained in their native conformation Difficult to standardize all vaccine components between production lots, presence of impurities that could produce side effects (e.g. LPS) 328 329 330 REFERENCES 331 [1] McConnell, M.J.; Actis, L.; Pachón, J. Acinetobacter baumannii: human infections, factors 332 contributing to pathogenesis and animal models. FEMS Microbiol. Rev., 2013, 37(2), 130-155. 333 [2] Dijkshoorn, L.; Nemec, A.; Seifert, H. An increasing threat in hospitals: Multidrug-resistant 334 acinetobacter baumannii. Nat. Rev. Microbiol., 2007, 5(12), 939-951. 335 [3] European Centers for Disease Control and Prevention.. ECDC, 336 2013.. 337 [4] Biswas, S.; Brunel, J.M.; Dubus, J.C.; Reynaud-Gaubert, M.; Rolain, J.M. Colistin: An update on the 338 antibiotic of the 21st century. Expert. Rev. Anti Infect. Ther., 2012, 10(8), 917-934. 339 [5] Adams, M.D.; Nickel, G.C.; Bajaksouzian, S.; Lavender, H.; Murthy, A.R.; Jacobs, M.R.; Bonomo, 340 R.A. Resistance to colistin in Acinetobacter baumannii associated with mutations in the PmrAB two-341 component system. Antimicrob. Agents Chemother., 2009, 53(9), 3628-3634. 342 [6] López-Rojas, R.; Jiménez-Mejias, M.E.; Lepe, J.A.; Pachón, J. acinetobacter baumannii resistant to 343 colistin alters its antibiotic resistance profile: A case report from Spain. J. Infect. Dis., 2011, 204(7), 1147-344 1148. 345 [7] Rolain, J. M.; Roch, A.; Castanier, M.; Papazian, L.; Raoult, D. Acinetobacter baumannii resistant to 346 colistin with impaired virulence: A case report from France. J. Infect. Dis., 2011, 204(7), 1146-1147. 347 [8] Taccone, F.S.; Rodríguez-Villalobos, H.; De Backer, D.; De Moor, V.; Deviere, J.; Vincent, J.L.; 348 Jacobs, F. Successful treatment of septic shock due to pan-resistant acinetobacter baumannii using combined 349 antimicrobial therapy including tigecycline. Eur. J. Clin. Microbiol. Infect. Dis., 2006, 25(4), 257-260. 350 [9] Valencia, R.; Arroyo, L.A.; Conde, M.; Aldana, J.M.; Torres, M.J.; Fernandez-Cuenca, F.; Garnacho-351 Montero, J.; Cisneros, J.M.; Ortiz, C.; Pachón, J.; Aznar, J. Nosocomial outbreak of infection with pan-drug-352 resistant acinetobacter. Baumannii in a tertiary care university hospital. Infect. Control Hosp. Epidemiol., 353 2009, 30(3), 257-263. 354 [10] García-Quintanilla, M.; Pulido, M.R.; López-Rojas, R.; Pachón, J.; McConnell, M.J. Emerging 355 therapies for multidrug resistant acinetobacter baumannii. Trends Microbiol., 2013, 21(3), 157-163. 356 [11] Smith, S.G.; Mahon, V.; Lambert, M.A.; Fagan, R.P. A molecular Swiss army knife: OmpA 357 structure, function and expression. FEMS Microbiol. Lett., 2007, 273(1), 1-11. 358 [12] Choi, C.H.; Lee, E.Y.; Lee, Y.C.; Park, T.I.; Kim, H.J.; Hyun, S.H.; Kim, S.A.; Lee, S.K.; Lee, J.C. 359 Outer membrane protein 38 of acinetobacter baumannii localizes to the mitochondria and induces apoptosis 360 of epithelial cells. Cell Microbiol., 2005, 7(8), 11271138. 361