MIAMI UNIVERSITY the Graduate School

Home , Formimidoylglutamase, T7 RNA polymerase

MIAMI UNIVERSITY The Graduate School

Certificate for Approving the Dissertation

We hereby approve the Dissertation

Cecily Renate-Maria Wood

Candidate for the Degree

Doctor of Philosophy

______Luis A. Actis, Director

______Mitchell F. Balish, Reader

______Rachael M. Morgan-Kiss, Reader

______Donald J. Ferguson, Reader

______Carole Dabney-Smith, Graduate School Representative

ABSTRACT

LIGHT SENSING IN A HUMAN PATHOGEN: GENETIC, BIOCHEMICAL, FUNCTIONAL AND PROTEOMICS ANALYSES OF BLUE LIGHT REGULATION IN ACINETOBACTER BAUMANNII by

Cecily Renate-Maria Wood

Acinetobacter baumannii is a prevalent human pathogen commonly associated with severe nosocomial infections such as ventilator-associated pneumonia and wound infections in immunocompromised individuals. This pathogen is a global concern due to its ability to acquire resistance to antimicrobials and persist within the hospital environment for weeks under different pressures such as nutrient limitation and desiccation. The capacity of A. baumannii to survive in different niches, including in and on human hosts or abiotic environments within clinical settings suggests that this pathogen senses and responds to its surroundings to modulate its physiology to survive. Recently, it has been reported that A. baumannii senses and responds to blue light through the photoreceptor protein BlsA, with both factors playing significant roles in the regulation of surface motility, biofilm formation, metabolism and antibiotic resistance responses. The work presented here shows that the interaction of the BlsA N-terminal region with flavin chromophores determines light sensory functions and protein stability, while the C-terminal region could play critical photocycling and downstream regulatory functions. This work also demonstrates that the ability of A. baumannii to differentially display surface motility and biofilm formation on plastic at 24°C depends on the active expression of the PrpABCD type I pilus assembly system. Unexpectedly, analysis of a PrpA deficient mutant resulted in the detection of light regulated motility responses by bacteria cultured at 37°C, a condition that impairs BlsA production and its sensory functions. These unexpected observations suggest that A. baumannii senses and responds to illumination when incubated at 24°C and 37°C using different light sensing and regulatory systems. The application of a shotgun proteomics approach not only confirmed the predicted light-mediated BlsA-dependent differential expression of proteins at 24°C, but also showed that light regulates protein expression by bacterial cells cultured at 37°C by uncharacterized sensory and regulatory mechanisms. Taken together, the observations made during this work demonstrate that light is a ubiquitous environmental signal that allows A. baumannii to sense and adapt to conditions this pathogen would encounter while interacting with the human host and the surrounding nosocomial environment where it is exposed to different illumination and temperature conditions.

LIGHT SENSING IN A HUMAN PATHOGEN: GENETIC, BIOCHEMICAL, FUNCTIONAL AND PROTEOMICS ANALYSES OF BLUE LIGHT REGULATION IN ACINETOBACTER BAUMANNII

A DISSERTATION

Presented to the Faculty of

Miami University in partial

fulfillment of the requirements

for the degree of

Doctor of Philosophy

Department of Microbiology

Cecily Renate-Maria Wood

The Graduate School Miami University Oxford, Ohio

2019

Dissertation Director: Luis A. Actis

Cecily Renate-Maria Wood

TABLE OF CONTENTS LIST OF TABLES ...... VI LIST OF FIGURES ...... VII CHAPTER 1 ...... 1

GENERAL INTRODUCTION ...... 1 PROJECT STATEMENT ...... 19

CHAPTER 2 ...... 20

A LIGHT-REGULATED TYPE I PILUS CONTRIBUTES TO ACINETOBACTER BAUMANNII BIOFILM, MOTILITY AND VIRULENCE FUNCTIONS ...... 20 ABSTRACT ...... 21 INTRODUCTION ...... 22 MATERIALS AND METHODS ...... 24 RESULTS ...... 32 DISCUSSION ...... 60 ACKNOWLEDGEMENTS ...... 64

CHAPTER 3 ...... 65

STRUCTURAL AND FUNCTIONAL ANALYSIS OF THE ACINETOBACTER BAUMANNII BLSA PHOTORECEPTOR AND REGULATORY PROTEIN ...... 65 ABSTRACT ...... 66 INTRODUCTION ...... 68 MATERIALS AND METHODS ...... 70 RESULTS AND DISCUSSION ...... 80 CONCLUDING REMARKS ...... 115 ACKNOWLEDGEMENTS ...... 116 AUTHOR CONTRIBUTIONS...... 116

CHAPTER 4 ...... 117

ANALYSIS OF THE HUMAN PATHOGEN ACINETOBACTER BAUMANNII PROTEOME IN RESPONSE TO BLUE LIGHT EXPOSURE ...... 117 ABSTRACT ...... 118 INTRODUCTION ...... 119 MATERIALS AND METHODS ...... 121 RESULTS AND DISCUSSION ...... 123 CONCLUDING REMARKS ...... 140 ACKNOWLEDGEMENTS ...... 141

iv CHAPTER 5 ...... 142

SUMMARY ...... 142

REFERENCES ...... 152

v List of Tables Table 1 Bacterial strains and plasmids used in this study ...... 29 Table 2 Primers used in this study ...... 30 Table 3 RNA-Seq data showing the downregulation of the A1S_2091- A1S_2088 operon when bacterial cells were cultured in SB at 24°C in the presence of light ...... 31 Table 4 Bacterial strains and plasmids used in this work ...... 76 Table 5 Primers used in this study ...... 79 Table 6 Kinetic values for light-to-dark recovery of His-BlsA and derivatives generated by site-directed point and deletion mutations ...... 102 Table 7 Proteins with higher abundance in light compared to dark at 24°C ...... 128 Table 8 Proteins with decreased abundance in the light compared to dark at 24°C ...... 133 Table 9 Proteins with increased abundance in light compared to dark at 37°C ...... 138 Table 10 Proteins with decreased abundance in the light compared to dark at 37°C ...... 139

List of Figures Figure 1. Basic architectures of the two classes of BLUF domain-containing proteins...... 9 Figure 2. Light-induced signaling state formation of a BLUF-domain containing photoreceptor...... 13 Figure 3. Predicted structure of the BlsA BLUF domain...... 16 Figure 4. Genetic organization and expression analysis of the prpABCD operon...... 34 Figure 5. Transcriptional analyses of the prpABCD operon...... 36 Figure 6. Analysis of the 17978.prpA insertion derivative...... 39 Figure 7. Role of PrpA in surface motility at 24°C...... 43 Figure 8. TEM analysis of samples collected from the surface of SA plates...... 48 Figure 9. Role of PrpA in pellicle and biofilm formation in response to illumination at 24°C...... 50 Figure 10. SEM analysis of biofilms formed on plastic...... 52 Figure 11. Role of PrpA in surface motility and biofilm formation in response to illumination at 37°C...... 54 Figure 12. Role of PrpA in A549-bacteria interactions...... 56 Figure 13. Role of PrpA in virulence...... 58 Figure 14. Comparative analysis of short BLUF-containing photoreceptors...... 82 Figure 15. Predicted tertiary structure of BlsA...... 84 Figure 16. Light-regulated surface motility of 17978 cells producing native BlsA or derivatives generated by site-directed mutagenesis...... 87 Figure 17. Growth curves of the 17978 parental strain and isogenic complemented derivatives...... 90 Figure 18. Overexpression of His-tagged BlsA and derivatives in E. coli BL21...... 96 Figure 19. Analyses of His-tagged BlsA...... 98 Figure 20. Figure 20. BlsA responses to low and high light intensities...... 100 Figure 21. Spectral analysis of purified His-tagged BlsA derivatives generated by site- directed deletion and point mutagenesis...... 106 Figure 22. CD spectra of dBlsA, lBlsA and related derivatives...... 109

vii Figure 23. HPLC of flavin standards...... 111 Figure 24. HPLC of heat-denatured supernatants of purified His-tagged BlsA recombinant derivatives generated by site-directed deletions (BlsA.152-156 and BlsA.143-156) or point mutations (BlsA.K144E and BlsA.K145E)...... 113 Figure 25. Effects of light and temperature on the expression of csuAB in 17978 cells. .... 146 Figure 26. CsuAB detection in 17978 wild type and prpA mutant cells...... 148

viii Chapter 1

General Introduction Taxonomical and biological traits of Acinetobacter species Historically, Acinetobacters were classified simply as Gram-negative non-motile microorganisms lacking pigmentation (1). Today, the classification system for Acinetobacter species is much more complex and the genus now comprises over 30 species that have been isolated from a variety of water, soil and clinical environments (2, 3). Acinetobacter spp. have been traditionally considered non-motile (1). However, various studies have shown that many members of the Acinetobacter genus are capable of motility on different surfaces under different conditions (4, 5), even though the Acinetobacter genus name is a Greek derivation from “akineto” or non-motile (6, 7). Notably, the core genomes of many Acinetobacter species include both type I and type IV pilus-encoding genes (3) that may contribute to various forms of motility. Most members of the Acinetobacter genus are non-pathogenic (8), and due to the non- fastidious nature of these microorganisms, they can also persist on the human skin of non- infected healthy adults and in the clinical setting for extended periods of time (1). The most medically relevant Acinetobacters constitute the Acinetobacter calcoaceticus – Acinetobacter baumannii (Acb) complex, which is associated most often with human disease (1). A. baumannii is almost always isolated from the nosocomial environment and is of the highest concern due to its ability to survive in clinical settings and rapidly acquire resistances to antibiotics (9). The resistance to various antimicrobials are typically carried on plasmids and mobile genetic elements, which can undergo rearrangements into the chromosome (10), contributing to the multidrug resistant (MDR) strains emerging worldwide in healthcare settings today.

Clinical relevance of Acinetobacter baumannii The World Health Organization (WHO) has recently ranked A. baumannii as the highest priority among common pathogenic bacteria for which antibiotics are critically needed and pose the greatest threat to human health, including Pseudomonas aeruginosa and Enterobacteriaceae (WHO, 2017). Although A. baumannii is most often isolated from nosocomial environments, it is also a source of community-acquired infections in immunocompromised individuals (11).

1 Outside the hospital, most A. baumannii outbreaks occur during hot and humid months or in tropical regions (12), and primarily affect individuals with underlying immune disorders and comorbidities, especially alcoholism and diabetes (13-15). Although Acinetobacter spp. can survive well under humid conditions, these organisms have a remarkable ability to survive from weeks to months in hospital environments under desiccation stress and on abiotic surfaces, including bed rails, linens and even on the hands of hospital workers (1, 16). The ability of A. baumannii to survive pressures imposed by the nosocomial environment and acquire resistances to a wide variety of antimicrobials make this opportunistic pathogen a great concern to human health. In the hospital, A. baumannii is a common cause of ventilator-associated pneumonia, pulmonary infections, urinary tract infections and septicemia, with mortality rates ranging from 5% - 54% (10). These infections are primarily caused by colonization of ventilators and indwelling surgical and catheter devices and occur most often in compromised patients in intensive care units (17). The prevalence of these infections can be attributed to the ability of A. baumannii to adhere to various abiotic and biotic surfaces and form biofilms. These are multicellular structures surrounded by an extracellular matrix composed of protein, DNA and/or carbohydrates. One of the best characterized mechanisms by which A. baumannii forms biofilms is through the products of the chaperone-usher CsuA/BABCDE pilus assembly system. The production of pili, such as the CsuA/B pilin subunit, plays a role in adherence and biofilm formation to plastic surfaces in the ATCC 19606T A. baumannii isolate (18). Other factors that may contribute to its ability to survive for extended time periods in the hospital include its genome plasticity and ability to acquire resistance to antibiotics and antimicrobials and its nutrient acquisition mechanisms. In 2009, the first fatal cases of monomicrobial necrotizing fasciitis caused by A. baumannii were reported (19). Although the severity of the infections and increased mortality reported for the two patients is unusual for A. baumannii, and considering that the necrotizing fasciitis was caused solely by A. baumannii, the cases are evidence of increased virulence due to MDR phenotypes and the possible expression of alternative virulence functions that are yet to be characterized. The blood isolates from both patients displayed resistance to all antimicrobials tested, including the aminoglycoside amikacin, various carbapenems, fluoroquinolones and cephalosporins (19). There are few to no therapeutic options for treatment of MDR or pan-drug resistant (PDR) A. baumannii isolates. Underlying immune disorders of the patients, combined

2 with A. baumannii’s ability to acquire resistance to current antimicrobials, increase the lethality of infections. Although the mechanisms contributing to the MDR and PDR phenotypes of A. baumannii are being uncovered, including the contribution of efflux pumps and target modifications (17), little is known about the biological mechanisms underlying A. baumannii’s ability to survive in various intra- or extra-host environments. The ability of A. baumannii to sense and respond to the variety of stressors, including desiccation, light or ethanol, and adapt its physiology to each signal accordingly (4, 20, 21) may contribute to its success as a facultative pathogen.

Environmental sensing in Acinetobacter baumannii The prevalence of A. baumannii has increased worldwide due inefficient hospital disinfection procedures and misuse of antibiotics (22). The capacity of A. baumannii to thrive in nutrient-limited medical settings or overcome challenges imposed by the human host can be attributed to the diverse and complex sensory and regulatory mechanisms expressed by this opportunistic pathogen. To become a successful facultative pathogen, it must sense its surroundings and adapt its physiology to the various host-associated or environmental signals it encounters. To colonize the host and establish infection, the pathogen must overcome the host’s immune response and nutrient limitation, including iron and zinc sequestration (23). However, outside of the host, A. baumannii must survive desiccation and adapt to temperature fluctuations. The ability of A. baumannii to sense light may also play a role in its establishment of an environmental niche. Sensing a light signal may seem surprising, considering that pathogens such as A. baumannii do not encode light-harvesting proteins or perform photosynthesis, although light is crucial for all living organisms. Light serves as an energy source for phototrophs, as a warning signal to organisms to avoid cellular damage or as a beacon for some pathogens like Brucella abortus, alerting them to upregulate virulence factors to prepare them to enter a host (24, 25). For A. baumannii, it senses its surroundings and adjusts its lifestyle to survive as a host-associated facultative pathogen or in the environment by regulating its physiology through complex mechanisms, including the differential expression antibiotic resistance, particular metabolic pathways and high-affinity iron acquisition systems (26, 27). One of the first obstacles towards establishing infection faced by A. baumannii in a human host is the limitation of critical nutrients including essential metals. The human host has

3 found ways to sequester iron and other transition metals such as zinc and magnesium from invading pathogens through metal-binding proteins, including hemoglobin, lactoferrin, transferrin and calprotectin (28). Limiting nutrient metals keeps invading pathogens from using them as cofactors in enzymes or in primary metabolic processes (28, 29). To combat iron limitation in the host, many pathogens use siderophore-based high-affinity iron acquisition systems. These siderophores strip iron from the host’s iron-binding proteins and bring this essential metal into the bacterial cell. In most Gram-negative bacteria including A. baumannii, the ferric uptake regulator, Fur, regulates iron acquisition genes such as those involved in siderophore biosynthesis and transport (30). A. baumannii adapts its physiology to survive in free iron-deprived conditions by increasing the expression of genes involved in acinetobactin siderophore biosynthesis (bas) and utilization (bau) pathways (31, 32). However, when the extracellular iron concentration is high, expression of Fur-regulated genes is repressed. Outside the host environment, A. baumannii encounters different types of signals and stressors, including fluctuations in nutrients, temperature and light – an important ubiquitous environmental cue. Light plays a significant role not only for photosynthetic organisms, but also for chemotrophic, non-photosynthetic organisms, including the pathogens B. abortus, Klebsiella pneumoniae, and Listeria monocytogenes (24, 33, 34). The ability to sense and respond to light is important for all organisms and especially for A. baumannii, considering its ability to survive in a variety of natural environmental sources including water, plants, soil, and even insects (35). This pathogen has been frequently isolated from a variety of wound infections on patients and even from the skin and clothing of the workers in the hospital (1, 16). Whereas A. baumannii is not exposed to light inside the host where it experiences a relatively stable temperature of 37°C, it experiences large temperature and light fluctuations while in the nosocomial environment or a wound bed. Wound temperatures can range from 23°C to 37°C, depending on the room temperature, the type of dressing used, the location of the wound or if the wound is left uncovered (36, 37). These findings indicate that temperatures lower than 37°C are relevant to the pathogenicity of A. baumannii and how this pathogen senses and interacts with its surroundings to thrive and cause disease. Notably, recent work showed that there is a functional connection between Fur, temperature and light responses mediated by the BlsA photoreceptor and transcriptional regulators in A. baumannii ATCC 19606T (38). At 23°C, the A. baumannii BlsA photoreceptor

4 antagonizes Fur, allowing the constitutive expression of genes coding for the acinetobactin biosynthesis and transport functions. In contrast, at higher temperatures, such as 37°C, when BlsA production is negligible and the protein is non-functional (4, 39), Fur is no longer bound by BlsA and the repressor is free to regulate the transcription of genes coding for iron acquisition and metabolism in response to changes in extracellular iron concentration.

Physiological role of photoreceptors in bacteria Photosensory proteins were first discovered in plants approximately six decades ago (40). Since their discovery, genes coding for light sensing proteins, or photoreceptors, have been discovered in genomes of organisms from all domains of life and are especially prevalent in bacteria (41). Photoreceptors, based on structure and photochemistry, are classified into six families, which include the retinal-containing rhodopsins, bilin-binding phytochromes, coumaric acid-containing xanthopsins, the flavin-binding cryptochromes, phototropins, and proteins harboring the blue light using FAD (BLUF) domain (42). All photoreceptors covalently or non- covalently bind a non-protein chromophore, or a molecule that absorbs light and is responsible for a shift in conformation that causes photoactivation of the protein to which it is bound (43). Rhodopsins are distinct from the other five members of the photosensory proteins in that they are transmembrane rather than cytoplasmic proteins and lack the common architecture of other photosensors with modular domains. They also have the unusual feature of their N-termini protruding into the periplasm side, rather than the cytoplasm, defying the “positive-inside” rule (43, 44). The photochemistry of rhodopsins involves the photoisomerization of a retinal molecule after absorbing a photon of blue light. After light absorption, downstream cellular processes, including light-gated proton pumps and ion channels, are signaled. Light-gated channels are only some of the characterized components involved in downstream signaling in the rhodopsin blue light sensory protein family. Phytochromes, on the other hand, bind a bilin chromophore that undergoes isomerization upon sensing red light. This red light-sensing protein family is responsible for regulating photomorphogenesis in plants and phototaxis in some microorganisms (25). Like rhodopsins, light stimulates a large conformational change to the protein through isomerization of the chromophore. Bacteriophytochomes, the phytochrome-like proteins found in prokaryotes, are the most prevalent photosensory proteins in all sequenced microbial genomes to date (41). The

5 discovery of the first bacteriophytochromes in the human pathogen Pseudomonas aeruginosa and the extremophile Deinococcus radiodurans provided useful information for the field of photoreceptors and revealed that non-phototrophic bacteria have the ability to perceive light (25). Photoactive yellow proteins (PYP), which are members of the xanthopsin family, undergo isomerization of their covalently bound 4-hydroxycinnamyl chromophore upon blue light illumination (43). Like rhodopsins and phytochromes, a light-dependent reversible photocycle occurs after absorption of a blue light photon to induce a large conformational change. However, the model PYP in Halorhodospira halophila is a special case as it is a small, soluble cytoplasmic protein without other linked domains as in different PYP proteins, which are typically linked to a PAS (Per-Arnt-Sim) domain in other prokaryotic proteins (42, 43). To date, there are no identified homologs of PYP in plants, and xanthopsins with linked PAS domains are the only ones that have been identified in prokaryotes. The PAS protein superfamily encompasses both PYP proteins, which bind coumaric acid, and sensors of light, oxygen and voltage (LOV) proteins, which bind flavin chromophores (25). LOV domain photoreceptors, cryptochrome/photolyase proteins and BLUF-containing photoreceptors use a flavin cofactor to sense and respond to light (43, 45, 46). Flavins are isoalloxazine molecules that have diverse functions as non-protein cofactors. Flavin adenine dinucleotide (FAD) and flavin adenine mononucleotide (FMN) act as primary electron acceptors during reduction-oxidation reactions. Due to their versatility of having different redox states, flavins function in cell repair, metabolism and more recently have been identified as chromophores in photoreceptors (47, 48). LOV proteins use a covalently-bound oxidized FMN chromophore to sense blue light (43, 45). The primary photochemistry of this family involves cysteinyl adduct formation between the flavin and the LOV protein domain, leading to downstream signal transfer to other domains or proteins. The signaling states of LOV proteins have been probed by spectroscopic and structural techniques, including NMR studies (42, 49). The cryptochrome family, first described in Arabidopsis thaliana (50), has subsequently been identified in mammals, insects, algae and prokaryotes including cyanobacteria (51). These proteins function in the regulation of seed germination in plants and regulation of circadian clocks in mammals (42, 43). Photolyases, the bacterial homologs of cryptochromes, function in light-dependent DNA repair using FAD and either a folate- or deazaflavin-based cofactor/chromophore (52). The photochemistry of the cryptochrome/photolyase family involves

6 electron transfer between FAD, a conserved tryptophan triad and a second chromophore, which lead to more subtle changes in protein conformation rather than large isomerization-based structural changes as seen in the other photoreceptor families (43, 53). Cryptochromes, unlike their photolyase homologs, have no function in the repair of DNA, but are thought to have evolved from the photolyase family because of their high sequence homology (48). Large conformational changes in protein structure are not always necessary to stimulate signal transduction from the light sensing domain of the protein to its regulatory or output domain. In the BLUF photoreceptors, subtle rearrangements involving the flavin chromophore and a β-strand in the C-terminus triggers signal transduction leading to the light or excited state of the protein (54). The subtle rearrangement and subsequent signal transduction are due to the protein’s ability to sense the excited state of the flavin and undergo a unique chemistry involving proton-coupled electron transfer (PCET) specific to the BLUF proteins (42, 55). Analyses of the N-terminal BLUF domains of six different BLUF photoreceptors with solved crystal structures, including both “long” and “short” proteins, show that they share only about 20% sequence identity (56). However, their domain architecture is similar and includes a conserved tyrosine and glutamine pair that binds the flavin chromophore. This flavin chromophore, either FAD or FMN, is non-covalently bound between the two N-terminal α-helices and lies parallel or perpendicular to a five-stranded β-sheet (56, 57). The class I or “long” BLUF photoreceptors include the model protein AppA of Rhodobacter sphaeroides and BlrP1 of Klebsiella pneumoniae. Class I BLUFs are linked to an effector domain or protein that executes gene regulation in response to illumination. AppA, for example, senses light through its BLUF domain and the redox state of the cell through its cysteine-rich carboxy-terminal region to modulate photosystem gene expression through its interaction with the redox repressor, PpsR (48, 58). In contrast, “short” or class II BLUF photoreceptors have a stand-alone BLUF domain with an additional 30-50 amino acid C-terminal extension, which show no significant similarity to other known bacterial proteins (43, 57) (Fig. 1). Due to the lack of identified effector domains and no known DNA-binding activity, these short proteins are hypothesized to interact with other proteins to transmit the light signal to regulate cellular processes including phototaxis, biofilm formation and even virulence (59). Some of these short BLUF proteins have been crystallized and their structures solved, including the Tll0078 protein (57), which forms oligomeric structures to interact with other proteins and control cellular functions such as phototaxis (60). Others have

7 been shown to regulate levels of the secondary messenger cyclic-di-guanosine monophosphate (c-di-GMP) in a light-dependent manner through an interaction with an EAL domain-containing regulator, such as the PapBBLUF-PapA complex in the purple bacterium Rhodopseudomonas palustris (Fig. 1).

8 Figure 1. Basic architectures of the two classes of BLUF domain-containing proteins. A) “Long” BLUF domains can be fused to known effectors such as the EAL domain as seen in the Klebsiella BlrP1 or Escherichia coli YcgF proteins, respectively, or a cysteine-rich (CxCCCxC) redox sensing domain as identified in AppA. B) “Short” BLUF proteins are defined as BLUF domains with no link to effector domains as described for the A. baumannii BlsA photoreceptor. Notably, BlsA has no identified sequences potentially involved in oligomerization or intermolecular interactions (middle diagram) in contrast to the characterized PapBBLUF (left) or Tll0078 (right) sensor proteins. In the case of Tll0078, two pentamers oligomerize to form a decamer structure (not shown) involved in light-dependent responses including phototaxis, whereas the PapB protein interacts with its cognate EAL-containing regulatory protein, PapA, to regulate biofilm formation via c-di-GMP metabolism.

10 The unique photochemistry of BLUF domain containing photoreceptors Upon blue light excitation, the flavin in the active site of the BLUF domain undergoes a red shift in the absorbance spectrum consistent with the characteristic BLUF photocycle (58). This photocycle, or blue light-induced reaction series of photoactivation and dark state reversion, is dependent on conserved amino acid residues of the N-termini (NT) of BLUF proteins. BLUF proteins undergo PCET that depends on the highly conserved tyrosine and glutamine residues near the flavin chromophore. After blue light illumination, the protein achieves the red-shifted or light-adapted state which is measured through UV-visible spectroscopy (Fig. 2). This is the photoactive signaling state of the protein. The timing of photocycling of BLUF proteins varies, with light-adapted states relaxing back to their dark-adapted states within 10 seconds to 30 minutes depending on the protein (58, 61). Although it is not completely understood how the tyrosine, glutamine and other residues interact with the flavin chromophore and transmit the light signal, various studies have shown that without these key residues, photoreceptors lose their ability to perceive and respond to light, leading to a protein that is not normally active in the dark, otherwise known as a “pseudo-lit” state (56). Although some BLUF protein domain crystal structures have been solved and this photoreceptor family is an attractive model for optogenetics, or the manipulation of cellular processes through light-sensitive proteins, conflicting models regarding the photoactivation and photocycle are under debate (56, 62, 63). The signaling mechanism of BLUF photoreceptors begins with the oxidation of the flavin chromophore after blue light illumination, which occurs on a picosecond timescale (64). The early photoactivation dynamics are still debated and are different for each BLUF-containing protein. For some proteins, including Tll0078, the details leading to the change around the flavin environment and the tyrosine, glutamine and other important residues including methionine are better characterized than others (57, 65). Two models have been proposed to account for the protein-FAD dynamics. One model suggests that glutamine rotation leads to a H-bond rearrangement process after the early PCET step. Another model proposes that glutamine tautomerization after PCET is the driving force for the H-bond rearrangement of the BLUF domain. For other proteins, such as AppA and BlsA, no accumulation of flavin radical intermediates have been identified by spectroscopic studies after this early photoactivation step and the initial signaling process is proposed to rely more on the glutamine residue changing its conformation from the keto into its enol form (63, 66). It is also proposed that a methionine or

11 tryptophan residue just before the β5 strand of the BLUF domain may have a significant role in signaling to downstream effectors (54). The steps that occur after the proton and electron transfer from the tyrosine are under debate and whether the methionine and/or tryptophan residues are important for all BLUF proteins is still unknown. These observations are significant and indicate the diverse roles and functions of the BLUF-containing family of photoreceptors in different organisms.

12 Figure 2. Light-induced signaling state formation of a BLUF-domain containing photoreceptor. Excitation of the dark-adapted BlsA protein (black solid line) by blue light at 100 μmol m−2 s−1 for 3 min at 22°C leads to the light-adapted state of the protein (blue line). The blue arrow is indicative of the 10-nm red-shifted signaling state.

14 The human pathogen A. baumannii senses light The observation that A. baumannii responds to light via the BLUF-domain containing photoreceptor protein BlsA has provided valuable insight into its physiology and cellular responses that are modulated not only by blue light, but also by temperature (4). Surface motility on semisolid media, one of the cellular responses regulated by light, was decreased when bacterial cells were incubated at 24°C in blue light but not in darkness (4). This finding was particularly fascinating considering that members of the Acinetobacter genus were typically defined as non-motile (67). Blue light also decreased biofilm formation on abiotic surfaces and pellicle formation. These cellular responses depended on the BlsA photoreceptor, as a blsA::aph isogenic insertion mutant formed biofilms and was motile in blue light at 24°C (4). Further analyses of the A. baumannii ATCC 17978 clinical strain under blue light illumination showed that virulence mechanisms were also regulated by light at 24°C, a finding also observed for the intracellular pathogen B. abortus (24). Killing of Candida albicans filaments and the expression of genes coding for a type VI secretion system were increased when A. baumannii cells were cultured in the presence of blue light (4, 27). Blue light and BlsA, the only BLUF photoreceptor identified in the genome of A. baumannii to date, also have roles in regulating metabolic pathways and antibiotic resistance responses by this pathogen (26, 27). Like other BLUF photosensory proteins, BlsA is predicted to contain two N-terminal α- helices with a flavin chromophore non-covalently bound by conserved and semi-conserved amino acids that make up an intricate hydrogen-bonding network of the 95-amino acid BLUF domain (4). BlsA’s NT BLUF domain, similar to other short BLUF-containing photoreceptors, has a typical β1α1β2β3α2β4β5 architecture (Fig. 3). The conserved amino acids of all BLUF domain photoreceptors include the tyrosine-glutamine pair responsible for the unique PCET photochemistry of this protein. The remaining BlsA region is composed of a 61-amino acid divergent C-terminal (CT) sequence with no known homology to other BLUF-containing photoreceptors. The C-terminal regions of other photosensory proteins with BLUF domains, including Slr1694 and Tll0078, contain an extension of at least two α-helices, α3α4, following β5 (57). Secondary structural analysis of BlsA shows the presence of at least three helical structures

C-terminal to β5 that may be involved in binding other proteins and allow BlsA to exert its regulatory functions. The presence of a fifth α-helix (α5), encompassing residues 146-150, might be unique to BlsA.

15 Figure 3. Predicted structure of the BlsA BLUF domain.

Labeled in dark blue is β1, one of the 5-stranded β sheet components, followed by α1 (light blue) and α2 (green), which make up a typical BLUF domain. The model was created using Phyre2 and the predicted BlsA amino acid sequence. The model only includes residues 1-137. The labeled secondary structural components indicate the proposed flavin binding region.

17 Photoreceptor Applications The diversity of photoreceptors has given rise to an interest in optogenetics. The first application of photoreceptors was demonstrated using a retinal-binding rhodopsin photoreceptor coupled to other proteins from Drosophila melanogaster to make neurons photosensitive (68). Initial studies from D. melanogaster led to the use of the Chlamydomonas reinhardtii channelrhodopsin-2 protein as a light-gated cation channel due to its size, kinetics and conductance properties for neuronal stimulation (69). These observations make some photosensory proteins, like the opsins, an appealing target for optogenetic applications not only for neuronal excitement, but also for neuronal inhibition and biochemical control over neurotransmitters or hormones because of their function as G protein-coupled receptors (70). The flavin-binding photoreceptors are of particular interest to those in the field of optogenetics because of their uses as light-activated regulatory circuitry. Flavin-binding photoreceptors are also attractive models due to the low cytotoxicity of blue light needed to penetrate tissue for their activation and the relatively small sizes of their flavin-binding domains (71, 72). Because flavin intermediates are unstable in solution, the use of BLUF-containing proteins, in particular, allow the study of electron transfer from conserved aromatic residues, such as tyrosine and tryptophan, and create an ideal environment to study their photophysical characteristics (62, 73). Beyond controlling gene expression or oligomerization state by light, the BLUF-containing proteins and other photoreceptors also allow the study of charge transfer and protein conformational changes, even in non-photoresponsive proteins. However, the conflicting photocycle dynamics must be resolved before optogenetics could be used as a successful biological tool. Outside of the field of optogenetics, blue light and photoreceptors also have an impact in medicine. The use of photodynamic therapy on localized wound infections caused by A. baumannii has already been successfully demonstrated in 2009 using mouse models (74). These early studies had significant impact on the future of therapeutics and combatting A. baumannii infections considering how desperately new treatments are needed for the rise of the growing number of MDR and PDR strains.

18 Project Statement Acinetobacter baumannii infections have been increasing over the last decade due to the prevalence of multidrug-resistant clinical isolates, but research has yielded little information contributing to the treatment of these infections. The life cycle and physiology of A. baumannii are also poorly understood. The lack of new treatment options, combined with only a few characterized virulence factors, have made A. baumannii a serious threat to human health worldwide. For these reasons, there is a fundamental need to examine the interactions of this pathogen and its environment that may affect its pathogenicity. In this work, we aimed to understand how A. baumannii perceives light through the BlsA photoreceptor and regulatory protein to control the expression of various cellular processes, including surface motility and biofilm responses. These cellular responses are significant because they may directly contribute to its ability to survive in different ecological niches and cause disease. The goals of the studies in this dissertation were to characterize a light-regulated type I pilus system that was identified through RNA-Seq experiments. Through our work, we showed that the photo-regulated pilus subunit (prpA) contributes to motility, biofilm and virulence functions in a BlsA- and light-dependent manner. Furthermore, we determined that this regulation was lost at 37°C; however, we found that a BlsA-independent light-regulated surface motility function was present in the prpA mutant derivative. This finding led us to perform a proteomics study on A. baumannii bacterial cells cultured under blue light or dark conditions at 24°C or 37°C. We also performed a structure-function analysis of the N- and C-terminal regions of the BlsA photoreceptor protein to examine how BlsA conducts its regulatory functions as a flavin-binding photosensor. The findings presented here provide insight about the relationship between light, temperature and how a facultative human pathogen senses and responds to its environment. These responses are important to understand how A. baumannii regulates its lifestyle to live as a nosocomial- or host-associated pathogen.

19 Chapter 2

A light-regulated type I pilus contributes to Acinetobacter baumannii biofilm, motility and virulence functions

Cecily R. Wood, Emily J. Ohneck, Richard E. Edelmann, and Luis A. Actis Infection and Immunity. 2018. 86: e00442-18.

20 ABSTRACT Transcriptional analyses of A. baumannii ATCC 17978 showed that the expression of A1S_2091 was enhanced in cells cultured in darkness at 24°C through a process that depended on the BlsA photoreceptor. Disruption of A1S_2091, a component of the A1S_2088-A1S_2091 polycistronic operon predicted to code for a type I chaperone/usher pilus assembly system, abolished surface motility and pellicle formation but significantly enhanced biofilm formation on plastic by bacteria cultured under darkness. Based on these observations, the A1S_2088-A1S_2091 operon was named the photo-regulated pilus ABCD (prpABCD) operon, with A1S_2091 coding for the PrpA pilin subunit. Unexpectedly, the comparative analyses of ATCC 17978 and prpA isogenic mutant cells cultured at 37°C showed the expression of light-regulated biofilm biogenesis and motility functions under a temperature condition that drastically affects BlsA production and its light sensing activity. These assays also suggest that 17978 cells produce alternative light- regulated adhesins and/or pili systems that enhance bacterial adhesion and biofilm formation both at 24°C and 37°C on plastic as well as on the surface of polarized A549 alveolar epithelial cells, where formation of bacterial filaments and cell chains was significantly enhanced. The inactivation of prpA also resulted in a significant reduction in virulence when tested using the Galleria mellonella virulence model. All these observations provide strong evidence showing the capacity of A. baumannii to sense light and interact with biotic and abiotic surfaces using undetermined alternative sensing and regulatory systems as well as alternative adherence and motility cellular functions that allow this pathogen to persist in different ecological niches.

21 INTRODUCTION Acinetobacter are Gram-negative aerobic, non-spore-forming bacteria that do not require complex media for growth (67). Members of this genus, which comprises more than 30 species (3), are present in the environment (2) and frequently found in different food sources, even under refrigerated conditions or after irradiation (75). They are also normal inhabitants of the human skin and are widely distributed in hospital environments (1, 76). Not surprisingly, these microorganisms have been implicated in a variety of food spoilage processes and human infections (7, 8). These diseases involve a group of isolates represented by members of the “A. calcoaceticus – A. baumannii (Acb) complex”, which are relevant to human health since they are associated with intra- and extra-nosocomial infections worldwide (77). A. baumannii is the most medically relevant member of the Acb complex since it has been associated with hospital infections as well as community-acquired and injury infections, particularly those caused after natural disasters or wound infections in military personnel deployed to Iraq and Afghanistan (35). Furthermore, A. baumannii has been isolated from a wide range of environmental sources and samples including water and aquaculture environments (78), soil (79), different food sources (80), animals (81, 82) and insects (83), all of which could be reservoirs for this bacterium outside the hospital environment (35). As a facultative pathogen that lives in different ecological niches, A. baumannii uses a wide range of extracellular signals to adapt to either an environmental or a host lifestyle, with each style playing a critical role in the limited understood life cycle and ecology of this bacterium. Accordingly, our work and that of others showed that A. baumannii senses and responds to extracellular signals including iron limitation, changes in salt concentrations, desiccation stress and the presence of antibiotics and disinfectants (17). These signals affect A. baumannii’s ability to interact with the host, form biofilms on abiotic surfaces and display motility on semi-solid media (18, 21, 84, 85). However, these are only some of the numerous extracellular signals that modulate the physiology of A. baumannii and its interaction with environments in which this microorganism is normally found. Thus, while we were studying the effect of factors such as temperature and media composition on the expression of some of the aforementioned phenotypes, we unexpectedly observed that A. baumannii senses and responds to light by differentially affecting biofilm formation, motility and the interaction with Candida albicans, a response that is widespread among different members of the Acinetobacter genus (4,

22 86, 87). The ability of A. baumannii to sense and respond to light when cultured at 24°C depends on the expression of blsA, a gene that codes for a “short” photoreceptor that harbors a N-terminal blue-light sensing using flavin (BLUF) domain, which binds the chromophore flavin adenine dinucleotide (FAD), but lacks identifiable regulatory output motifs (4, 41). Albeit limited, there is some understanding of the molecular interactions that play a role in the photosensing functions exerted by BlsA (54). In contrast, almost nothing is known regarding the mechanisms by which this protein works as a regulator and the cellular targets it affects in response to illumination. In this report, we show that the light-regulated motility, biofilm and pellicle formation responses displayed by A. baumannii ATCC 17978 (17978) depend on the active expression of the photo-regulated pilus assembly system PrpABCD in a BlsA-mediated process when bacteria are cultured at 24°C. PrpABCD also proved to be critical for the interaction of polarized A549 human alveolar epithelial cells with 17978 and the virulence of this strain to Galleria mellonella when tested at 37°C, a temperature at which BlsA-mediated light regulation is lost because of a drastic reduction in blsA expression and an alteration of the BlsA reversible photocycle at temperatures higher than 30°C (4, 39). However, the ability of a 17978 prpA mutant to display light-regulated surface motility and biofilm responses at 37°C indicate that undetermined additional 17978 factors involved in these responses are controlled by light in a BlsA- independent regulatory process that remains to be identified and characterized.

23 MATERIALS AND METHODS Bacterial strains, plasmids, and culture conditions All bacterial strains and plasmids used in this work are listed in Table 1. Bacterial strains were routinely maintained and cultured on Luria-Bertani (LB) agar or broth (88) unless otherwise indicated and supplemented with appropriate antibiotics. All cultures were grown at 37°C for 16-18 h (overnight) or at 24°C under dark or illuminated conditions as described before (4).

General DNA procedures A commercial kit (Qiagen) and a phenol-based method adapted from Barcak et al. (89) were used to isolate plasmid and total DNA, respectively. DNA was amplified with DNA polymerases and digested with restriction enzymes as indicated by the supplier (New England Biolabs). Automated DNA sequencing using BigDye-based chemistry (Life Technologies) was performed prior to subcloning reactions and electroporation using primers supplied with cloning vector kits (Life Technologies) or custom-designed primers (Integrated DNA Technologies) listed in Table 2.

Construction and complementation of a 17978 prpA::ermAM isogenic insertion derivative A 2.4-kb genomic fragment containing the prpA coding region and 1-kb up- and downstream flanking regions were PCR amplified using Taq DNA polymerase and primers 4151 and 4152 (Fig. 4A and Table 2). This amplicon was ligated into pCR8-TOPO, generating pMU1135 (Table 1), which was further confirmed by automated DNA sequencing. Inverse PCR of pMU1135 with Phusion DNA polymerase and primers 4153 and 4154 (Table 2), which anneal within the prpA coding region and results in a 61-bp intragenic deletion, were used to generate an amplicon that was ligated to the Lactococcus lactis pIL252 ermAM cassette, which codes for erythromycin resistance (Emr) (Table 1). Phusion DNA polymerase and primers 4100 and 4101 (Table 2) were used to PCR amplify the ermAM cassette from pIL252. The resulting pMU1142 derivative harboring the prpA::ermAM fragment (Table 1) was PCR-amplified using primers 4151 and 4152 (Fig. 4A and Table 2). The amplicon was ligated into pEX100T, which was previously digested with SmaI, generating pMU1143 (Table 1). This plasmid was then electroporated into electrocompetent 17978 cells as described before (90). Gene disruption

24 mutants were selected for based on their erythromycin and sucrose resistance phenotypes. PCR using primers 4151 and 4152 (Fig. 4A and Table 2) was performed to verify the mutation by allelic exchange in the 17978.prpA isogenic derivative (Fig. 4A and Table 1). The nature of this mutant was further confirmed by automated DNA sequencing. For complementation studies, a 662-nt chromosomal region encompassing the prpA wild type allele and its promoter region was PCR-amplified from 17978 total genomic DNA with Q5 high-fidelity DNA polymerase (New England Biolabs) using primers 4413 and 4465 (Fig. 4A and Table 2), each flanked with BamHI restriction sites, and cloned into the cognate site of the A. baumannii-E. coli shuttle vector pWH1266 (91) to generate the pMU1269 complementing plasmid (Table 1). This plasmid, which was checked by automated DNA sequencing, was electroporated into 17978.prpA electrocompetent cells and the resulting 17978.prpA.CV transformants (Table 1) were selected on LB agar plates containing 500 μg/ml Amp and 20 μg/ml Em. The 17978.prpA.EV derivative harboring empty pWH1266 shuttle vector (Table 1) was used as a negative control.

RNA isolation and transcriptional analysis of prpABCD Three independent biological replicates of the 17978 parental strain and the blsA (17978.OR) mutant were grown in 25 ml of swimming broth (SB, 10 g/l tryptone, 5g/l NaCl) at

24°C or 37°C under illumination or darkness until an OD600 of 0.8 was reached. Bacterial cells were collected by centrifugation and total RNA was isolated using the Direct-Zol Miniprep Plus kit (Zymo Research) according to the manufacturer’s protocol. Following treatment with DNaseI (Invitrogen), RNA quality was assessed using the Bioanalyzer 2100 and the RNA Pico 6000 kit (Agilent Technologies). cDNA used for RT-PCR and qRT-PCR assays was synthesized using the iScript cDNA synthesis kit (Bio-Rad) following manufacturer’s protocol. Briefly, 100 ng of DNA-free total RNA isolated from 17978 cells cultured as described in the previous paragraph were used as template with random hexamer primers and the following cycling conditions: priming for 5 min at 25°C, reverse transcription for 20 min at 46°C, and inactivation for 1 min at 95°C. RT-PCR was performed using the cDNA samples described above as template and the primer pairs 4470-4615 and 4471-4618, which link the prpA-prpC and prpC-prpD coding regions, respectively (Fig. 4, Table 2), to determine the predicted polycistronic nature of the prp

25 operon and the potential polar effect of the insertion of the ermAM cassette using the following conditions: initial denaturation at 95°C for 3 min, denaturation at 95°C for 30 s, annealing for 30 s, and extension for 1 min at 72°C. The denaturation, annealing, and extension steps were repeated for 29 more cycles followed by a final extension of 5 min at 72°C. The production of the predicted amplicons was determined by ethidium bromide agarose gel electrophoresis (88). Amplification of total RNA without reverse transcription was used as a negative control. The differential transcription of prpA was measured by qRT-PCR using cDNA synthesized from DNA-free total RNA isolated from cells of the 17978 parental strain, the 7978.OR blsA mutant and the 17978.prpA.CV derivative cultured under illumination or darkness as described above and the primers 4419 and 4420 (Fig 4A and Table 2) using the experimental condition described previously (92). Primers 4488 and 4489 (Table 2) were used to detect the transcription of an internal fragment of the recA housekeeping gene, which was used as a constitutively expressed gene. Triplicates of three independent biological samples were used to statistically validate the data collected for each tested strain.

Bacterial growth analysis Fresh bacterial cultures of 17978 and the prpA insertion derivative were passaged from LB agar plates to LB broth and grown overnight at 37°C in a shaking incubator set at 200 rpm. Bacterial cultures were then diluted 100 times into fresh SB and transferred to a 96-well microtiter plate. OD600 measurements were recorded hourly for 24 h at 37°C in a shaking plate reader (BioTek Instruments, Inc.) using two independent biological samples in six technical replicates.

Motility Assays Fresh bacterial cultures grown on LB agar plates supplemented with antibiotics when necessary were used to inoculate the surface of swimming agar (SA; SB containing 0.3% agarose) plates and incubated overnight at 24°C or 16-18 h at 37°C as previously described (4). Motility assays were performed in triplicate using three independent biological samples each time. The surface motility area of each replicate was measured using the ImageJ image- processing program (National Institutes of Health).

26 Transmission electron microscopy (TEM) Bacterial cells were inoculated on the surface of SA plates that were incubated overnight at 24°C in darkness. Freshly prepared carbon coated, nitrocellulose substrated TEM grids were placed substrate side down at the edge or immediately in front of the motility zone to analyze the 17978 strain. Grids were placed at the edge of the colony formed by the non-motile 17978.prpA mutant. The grids were carefully removed and negative stained with 5 µl of 1.5% (wt./vol.) ammonium molybdate for 5 min. The grids were then wicked dry with filter paper and allowed to air dry. Images were captured at 120kv with a JEOL 1200-EX II TEM.

Biofilm and pellicle Assays Fresh bacterial cultures of 17978 and the prpA mutant were grown in SB overnight with shaking at 37°C. To assess pellicle formation, overnight cultures were diluted 1:100 into fresh SB and 1 ml of each diluted culture was inoculated into glass tubes, which were incubated statically at 24°C or 37°C for 96 h or 48 h, respectively, as reported before (4). The formation of pellicles on the surface of liquid cultures was assessed visually. To assess biofilm formation, bacterial cultures were prepared as described above and 200 µl of each strain were inoculated into 96-well microtiter plates, which were either incubated in light or dark conditions at 24°C for 96 h or at 37°C for 48 h without shaking. Biofilms were quantified by crystal violet staining as previously described (18). Biofilm experiments were performed in duplicate using fresh biological samples for each experiment. To examine biofilm structures formed on plastic surfaces, three independent fresh SB cultures of 17978 or the isogenic prpA derivative grown overnight at 37°C were diluted 1:100 and 5 ml of each diluted culture were inoculated into 50-ml conical tubes. A sterile plastic coverslip was placed semi-submerged into each tube as described previously (93). All cultures were incubated statically at either 24°C for 96 h or at 37°C for 48 h in light or darkness. Coverslips were recovered from each 50-ml conical tube, washed, processed and examined by scanning electron microscopy (SEM) on a Zeiss Supra 35 VP microscope as described before (93).

Infection of polarized A549 human alveolar epithelial cells

27 A549 cells were polarized and infected with bacteria cultured in LB broth as described 6 before (94), using 10 bacteria/ml and incubated for 48 h at 37°C in the presence of 5% CO2. All polarized cell samples were fixed after 48 h of infection and analyzed by SEM as previously described (93). Polarized cell assays were performed in duplicate for each condition using different biological samples each time. Galleria mellonella virulence assays Fresh cultures of 17978 or the prpA mutant were passaged from LB agar plates and grown overnight in LB broth without antibiotics. Cells were harvested by centrifugation, washed twice in sterile phosphate-buffered saline (PBS) solution and then resuspended in 1 ml PBS. 8 Bacterial cells were standardized to an OD600 of 0.5, which corresponded to 10 bacterial cells/ml. Healthy G. mellonella larvae weighing between 250-300 mg, which were previously sorted into groups of 10 based on yellow coloration and high activity, were either injected with 5 µl of the 17978 parental strain or the prpA mutant. Control groups included larvae inoculated with 5 µl sterile PBS to account for physical trauma caused by the injection process or larvae that were not injected to confirm larvae viability throughout the experiment. Infected and control larvae were stored for 5 days at 37°C in a humidifying chamber and assessed for death at 24-h intervals. Death events were based on lack of movement and melanisation as previously described (84). If more than two deaths occurred in the control groups, the experiment was discarded and repeated. Kaplan-Meier survival curves were plotted using Prism Graphpad 7 and log-rank tests were performed using the statistical analyses tools included with the Prism Graphpad software. A total of four replicas per experimental condition (n = 40) were used to validate the experimental data. P values ≤ 0.05 were considered statistically significant for the log rank test of survival curves (SAS Institute Inc., Cary, NC).

Statistics The GraphPad 7 InStat software package (GraphPad Software, Inc.) was used to analyze the statistical significance of data sets by the student’s t-test or analysis of variance (ANOVA) as appropriate for each experiment. P values ≤ 0.05 were considered statistically significant for all aforementioned assays. Error bars represent the standard error for each data set shown in the figures.

28 Table 1 Bacterial strains and plasmids used in this study Strain/plasmid Relevant characteristic(s)a Source/reference Strains A. baumannii ATCC 17978 Clinical isolate ATCC 17978.OR blsA::aph derivative of 17978; KmR (4) 17978.prpA prpA::ermAM derivative of 17978; EmR This work 17978.prpA.EV prpA mutant harboring pWH1266; EmR, TetR This work 17978.prpA.CV prpA mutant harboring pMU1269; AmpR, EmR, TetS This work

E. coli DH5α DNA recombinant methods Gibco-BRL Top10 DNA recombinant methods Life Technologies

Lactococcus lactis Strain MG1363 harboring pIL252; EmR Smidt and Kruse subsp. cremoris

Plasmids pCR8-TOPO Gateway cloning vector; SpR Life Technologies pEX100T Suicide vector for allelic exchange; AmpR ATCC pIL252 Cloning vector; source of ermAM encoding EmR Smidt and Kruse pWH1266 A. baumannii-E. coli shuttle vector; AmpR TetR (91) pMU1135 pCR8-TOPO harboring prpA; SpR This work pMU1142 pCR8-TOPO harboring prpA::ermAM; SpR, EmR This work pMU1143 pEX100T harboring prpA::ermAM; AmpR, EmR This work pMU1269 pWH1266 harboring prpA; AmpR, TetS This work aAmpR, ampicillin resistance; EmR, erythromycin resistance; KmR, kanamycin resistance; SpR, spectinomycin resistance; TetR, tetracycline resistance; TetS, tetracycline sensitivity.

29 Table 2 Primers used in this study Primer number Nucleotide Sequencea 4100 5’-GCAAACTTAAGAGTGTGTTG-3’ 4101 5’-CCTTTAGTAACGTGTAACTTTC-3’ 4151 5’-CGAAGCATTGTCTTTAGACC-3’ 4152 5’-GGCTACTAAATCCTCAGAAGG-3’ 4153 5’-ATTGGAGTTGCAACAACTGC-3’ 4154 5’-CCACCAATCATACGACGTTG-3’ 4413 5’-ATGGTAGGATCCGTATGATGCCAAAAATAGAG-3’ 4419 5’-GTCCACCATCAAATGACAAAGTCC-3’ 4420 5’-CTGTGTCCTGAATACCTCAGC-3’ 4465 5’-TGATGTGGATCCTTAATAAGTTACTGTGACTGTCAC-3’ 4470 5’-CACATTCGTATTGTCCTGTAACTG-3’ 4471 5’- ACTTGTAGTCGATCTTCTTGATC-3’ 4488 5’-GCCCAGAAACTACCACTGG-3’ 4489 5’-GCTTCTTTAAACGGAGGAGCC-3’ 4615 5’-GGTAATGTCCAGACAGTTAACG-3’ 4618 5’-TGCATCACTATTAATATTTGCTGC-3’ aUnderlined nucleotides identify BamHI restriction sites.

30 Table 3 RNA-Seq data showing the downregulation of the A1S_2091- A1S_2088 operon when bacterial cells were cultured in SB at 24°C in the presence of light Gene Identifiera Fold Change P value Predicted function A1S_2088 -2.75 0.0008 Hypothetical A1S_2089 -4.41 0.0001 Fimbrial usher protein A1S_2090 -3.08 0.0014 P pilus assembly protein A1S_2091 -5.67 0.0001 Hypothetical aThe A1S_2091-A1S_2088 operon is described as the prpABCD operon in this report.

31 RESULTS Identification of a light-regulated locus coding for a predicted type I pilus Our initial analysis of RNA-Seq data collected from 17978 cells cultured in SB in the presence or absence of light (B.A. Arivett, S.E. Fiester, C.R. Wood, and L.A. Actis, unpublished data) led to the identification of the A1S_2088-A1S_2091 4-gene locus (Fig. 4A), in which the expression of all of its components was significantly repressed when bacteria were cultured at 24°C under illumination (Table 3). The initial annotation of the 17978 genome describes A1S_2089 and A1S_2090 as genes coding for putative fimbrial usher and P pilus assembly proteins, respectively (95). Our in silico analysis using several bioinformatics tools, including Phyre2, predicted that: 1) A1S_2091 is a 147-amino acid protein significantly related to type I fimbrial proteins, 2) A1S_2090 is a 234-amino acid protein significantly related to fimbrial chaperone proteins, 3) A1S_2089 is an 819-amino acid protein significantly related to type I fimbrial usher proteins, and 4) A1S_2088 is a 339-amino acid protein significantly related to fimbrial tip adhesins. Thus, we named this 4-gene cluster as the photo-regulated pilus (prp) ABCD operon, which has the potential for assembling a type I pili in A. baumannii differentially produced in response to illumination. Interestingly, BLAST searches proved that the prpABCD operon is present in the genome of a large number of A. baumannii strains and analysis of the strain MAR002 showed that this biofilm hyperproducing isolate harbors a prpABCD homolog, which was annotated as the LH9211085-LH9211080-LH9211075-LH9211070 operon (96). The prpABCD was also detected in the genome of different members of the Acinetobacter genus including the clinical species gyllenbergii, pitti and nosocomialis, and the environmental species calcoaceticus, oleivorans and nectaris. Taken together, these observations, which are in agreement with our reports describing light-regulated responses in A. baumannii and different members of the Acinetobacter genus (4, 87), support the potential importance of PrpABCD in the biology and physiology of a bacterium that has a remarkable ability to persist under different environmental conditions.

Expression analysis of the prp locus The prp locus is preceded by a 638-nucletide intergenic region that separates prpA from A1S_2092, which is predicted to code for a PepN homolog, and followed by a 367-nucleotide intergenic region that separates prpD from A1S_2087, a gene coding a putative glutathione S-

32 transferase (Fig. 4A). This genomic structure together with the fact that prpA-prpB, prpB-prpC and prpC-prpD are separated by 72-, 9- and 5-nucleotide intergenic regions, respectively, suggest that prpABCD is a polycistronic operon. This possibility was tested by RT-PCR of total RNA isolated from 17978 wild type bacteria cultured in darkness at 24°C. Figure 5A shows that amplification of 17978 cDNA with primers 4470 and 4615 resulted in the predicted 975- nucleotide amplicon encompassing the 3’-end of prpA and the 5’end of prpC. The size of this amplicon matches the size of the amplicon obtained using total DNA as a template. Similarly, PCR amplification of 17978 cDNA using primers 4471 and 4618 produced a 288-nucleotide amplicon linking the 3’-end of prpC with the 5’-end of prpD, which matches the amplicon obtained with total DNA (Fig. 5C). In contrast, no amplicons were produced when the same primer sets were used with total RNA samples that were not reverse transcribed. The same results were obtained when total RNA isolated from bacteria cultured in the presence of light was used as a template (data not shown). The initial RNA-Seq data showing that the prpABCD locus is differentially expressed in response to light was further confirmed by qRT-PCR using recA as a constitutively expressed internal control. Our RNA-Seq data show that there are not significant differences in the transcription of recA in 17978 cells cultured under darkness or illumination (Arivett et al., manuscript in preparation). This approach showed that the transcription of prpA is increased about 2.2-fold in cells cultured in darkness when compared to transcript levels detected in bacteria cultured under illumination at 24°C (Fig. 4B). This observation together with the polycistronic nature of this pilus-coding operon are in agreement with our RNA-Seq data showing a significant reduction in the transcription of all four prp coding regions in cells cultured in the presence of light as compared to data collected from cells cultured in darkness (Table 3). Since the BlsA photoreceptor protein plays a critical role in 17978 light responses (4), the transcription of prpA was further measured in the 17978 parental strain and the 17978.OR blsA::aph isogenic derivative. Figure 4B shows that the insertion inactivation of blsA results in comparable prpA transcriptional levels in bacteria cultured at 24°C in the presence or absence of light. Taken together, these results indicate that prpABCD is a polycistronic operon that is differentially transcribed in response to light via a BlsA-mediated regulatory process when bacteria were cultured at 24°C by a mechanism that remains to be determined.

33 Figure 4. Genetic organization and expression analysis of the prpABCD operon. (A) In silico analysis of the A1S_2089-A1S_2091 coding region. The horizontal arrows represent predicted coding regions and their direction of transcription. Coding regions are identified by their original genomic annotation, the names assigned in this work and their predicted protein products. Numbers on the top of the vertical bars represent size in bp. The location of the insertion of the EmR DNA cassette within prpA is indicated by the black triangle. The connected vertical bars indicate primer locations and the length of cognate amplicons used to clone and mutagenize prpA, construct the complementing vector pMU1269 (Table 1), test the expression of prpA by qRT-PCR, and determine the polycistronic nature of the prpABCD operon. Numbers underneath the vertical bars identify primers listed in Table 2. (B) qRT-PCR of prpA using RNA isolated from 17978 and 17978.OR cells cultured in SB at 24°C in darkness or illumination. (C) qRT-PCR of prpA using RNA isolated from 17978 cells cultured in SB at 37°C in darkness or illumination. recA was used as a constitutively expressed control gene. Horizontal bars identify statistically different values (P ≤ 0.0001, ) and error bars represent the standard error of each data set.

35 Figure 5. Transcriptional analyses of the prpABCD operon. RT-PCR with the intergenic primer pairs shown in Fig. 4A using as templates DNA, RNA and cDNA samples obtained from cells of the 17978 parental strain (A and C) or the 17978.prpA mutant (B and D) cultured in SB at 24°C in darkness. Numbers represent primer pairs used to amplify the sequences located between coding regions as displayed in Fig. 4A. λ HindIII- digested DNA and a 100-bp ladder were used as molecular weight markers in panels A/B and C/D, respectively. (E) qRT-PCR analysis of RNA isolated from 17978 and 17978.prpA.CV cells cultured in SB at 24°C in darkness until they reached on OD600 of 0.8. recA was used as a constitutively expressed control gene. Horizontal bars identify statistically different values (P ≤ 0.05, ) and error bars represent the standard error of each data set.

37 Functional analysis of the prp locus in bacteria cultured at 24°C The biological role of the prpABCD locus was tested using the isogenic 17978.prpA mutant (Table 1), which harbors the insertion of a DNA cassette coding for erythromycin resistance (EmR) within a 61-bp prpA intragenic deletion (Fig. 4A) that resulted in the predicted 800-nucleotide size increase when compared with the 2.4-kbp parental amplicon (Fig. 6A). As previously reported (4), the 17978 parental strain displayed surface motility when inoculated on the surface of a SA plate incubated at 24°C in darkness (Fig. 2A). In contrast, only growth at the inoculation site without detectable surface motility was observed when the 17978.prpA derivative was tested under the same experimental conditions (Fig. 7A). The effect of the prpA mutation was further examined by transmission electron microscopy of cells lifted from SA plates incubated in darkness at 24°C. Figure 3A shows the presence of long and thin pili associated with bacteria located at the edge of the motility zone. These pili, some of which formed bundles (white arrow in Fig. 8B), were clearly detected in the area located immediately in front of the motility zone. In contrast, none of the pili structures produced by the parental strain could be detected in 17978.prpA samples collected and analyzed under the same experimental conditions (Fig. 8C). Electroporation of the complementing vector pMU1269, a derivative of the pWH1266 shuttle cloning vector that harbors the prpA parental allele expressed from its native promoter (Table 1), into 17978.prpA mutant cells (17978.prpA.CV) significantly increased surface motility (P ≤ 0.05) albeit not to the wild type levels (Fig. 7A). This response by 17978.prpA.CV cells could be due to a potential polar effect on the expression of downstream coding regions because of the insertion of the EmR cassette within prpA. RT-PCR of RNA isolated from 17978.prpA cells with primers 4470-4615 and 4471-4618 (Fig. 4A), which were used to determine the polycistronic nature of the prpABCD operon (panels A and C of Fig. 5), produced prpA-prpC and prpC-prpD intergenic amplicons of the same size as those detected using 17978 RNA as a template (compare panels A and B, and C and D of Fig. 5, respectively). These results demonstrate that the insertion of the EmR DNA cassette within prpA did not cause polar effects on the prpBCD downstream coding regions. Alternatively, the reduced motility response displayed by 17978.prpA.CV cells could be due to a gene-dosage effect because of the genetic complementation of this mutant with a plasmid copy of the prpA parental allele. The qRT-PCR

38 Figure 6. Analysis of the 17978.prpA insertion derivative. (A) Agarose gel electrophoresis of the amplicons generated using primers 4151 and 4152 (Fig. 4A and Table 2) and total DNA isolated from 17978 (lane 2) or 17978.prpA cells (lane 3). Lane 1, λ HindIII-digested DNA. (B) Growth curves of 17978 and 17978.prpA cells cultured in LB broth at 37°C in a shaking plate reader set at 200 rpm. Error bars represent the standard error of each data set.

40 data shown in Fig. 5E proves that the electroporation of pMU1269 indeed results in a 2.2-fold increase (P = 0.008) in prpA transcript levels when compared with the parental 17978 strain. Thus, the higher copy number of prpA in the complemented strain most likely affects the stoichiometry of the PrpABCD protein components needed for proper motility functions. As expected from our previous observations (4), 17978 cells displayed no motility when SA plates were incubated at 24°C under illumination, a response that was also observed with 17978.prpA and 17978.prpA.CV bacteria (Fig. 7B). The motility phenotype of the 17978.prpA.EV strain, which was electroporated with the empty vector pWH1266, was comparable to that displayed by the 17978.prpA mutant when tested under the same experimental conditions (data not shown). Taken together, these results show that the PrpABCD pilus assembly system is required for 17978 bacterial cells to display motility on a semi-solid surface in the absence of light. Growth analysis of the 17978 parental strain and the prpA derivative proved that the insertion inactivation of this coding region did not result in significant growth defects when the strains were cultured in SB (Fig. 6B) or LB broth (data not shown) without selective pressure. Biofilm assays showed that 17978 cells were able to form readily detectable pellicles on the surface of SB cultures statically incubated at 24°C in darkness (Fig. 9A). In contrast, 17978.prpA cells produced no detectable pellicles when incubated under the same experimental conditions. As expected from our previous observations (4), illumination resulted in no detectable pellicle formation by cells of the 17978 parental strain as well as by cells of the 17978.prpA mutant when the SB cultures were statically incubated in the presence of light (data not shown). The biofilm assays showed that the inactivation of prpA significantly affected biofilm biogenesis on plastic, with 2.5-fold and 3.0-fold increases under illumination and darkness, respectively, when the 17978 and the 17978.prpA strains were compared side-by-side under the same experimental conditions (Fig. 9B). Genetic complementation of the 17978.prpA derivative with a plasmid-copy of the prpA parental allele reduced biofilm formation to levels comparable to those produced by 17978 under illumination but lower than those produced by the parental strain under darkness. The latter response could also be due to a prpA gene-dosage effect that affects the proper stoichiometry and function of the PrpABCD pilus assembly system.

41 SEM analysis of plastic coverslips semi-submerged in SB inoculated with 17978 bacteria showed that this strain produced denser and more structured biofilms under darkness than under illuminated conditions, particularly at the liquid-air interface (Fig. 10A). A similar outcome was

42 Figure 7. Role of PrpA in surface motility at 24°C. Cells of the parental strain (17978), the prpA insertion mutant (prpA) and its complemented derivative (prpA.CV) were surface-inoculated on SA plates and incubated at 24°C in darkness (A) or illumination (B). Horizontal bars in A represent statistically different values (P ≤ 0.0001, ; P ≤ 0.001, ) between cognate samples and error bars represent the standard error of each data set.

44 detected with the 17978.prp strain, which covered the surface of the coverslips with a more uniform and denser layer of cells at the liquid-air interface when incubated in darkness (Fig. 10A). This analysis also showed that overall, the 17978.prp insertion derivative produced more biofilm than the 17978 strain both under darkness and illumination, an observation that matches the biofilm data collected using standard crystal violet assays shown in Fig. 9B. Taken together, these observations underscore the positive role the prpABCD locus and the products it codes for play in the capacity of 17978 cells to differentially display surface motility and produce pellicles at the surface of static broth cultures incubated at 24°C. On the other hand, the biofilm results indicate that cellular products other than those coded for by prpABCD could be responsible for the light-mediated responses shown in Fig. 9B.

Functional analysis of the prp locus in bacteria incubated at 37°C Considering that the interaction of A. baumannii with the human host mostly occurs at 37°C, we determined whether the prpABCD operon plays a role when bacterial cells are incubated at this temperature. Surface motility assays showed that 17978 cells displayed comparable high motility responses on SA plates incubated 37°C in the presence or absence of light (Fig. 11A). These responses are in accordance with the finding that the transcriptional levels of prpA were not significantly different in cells cultured in darkness or illumination (Fig. 4C). Altogether, these results reinforce the correlation between light-regulated expression of the prpABCD operon and the consequent differential display of surface motility on a semi-solid surface, which was apparent when bacteria were cultured at 24°C but not at 37°C. In contrast to the parental strain, 17978.prpA cells showed no motility and about half of the motility displayed by 17978 parental cells when the SA plates were incubated in the presence and absence of light, respectively (Fig. 11A). Electroporation of 17978.prpA cells with pMU1269, a pWH1266 derivative harboring a copy of the prpA parental allele, fully restored motility in the absence of light and significantly increased the motility response to illumination when compared to the prpA mutant, although not to the levels displayed by the 17978 parental cells (Fig. 11A). Crystal violet assays of biofilms formed on plastic surfaces using cells cultured in SB at 37°C showed that the 17978 parental strain and the prpA mutant formed more biofilm in darkness compared to illuminated conditions; although the amount of biofilm produced by these

45 two strains under illumination and darkness was not significantly different from each other (Fig. 11B). Genetically complemented 17978.prpA.CV cells displayed the same response as 17978 parental and 17978.prpA cells under illumination and significantly reduced biofilm formation when cultured in darkness (Fig. 11B). Further analysis of biofilms using SEM supports the data collected with crystal violet assays. The 17978 strain formed more complex biofilm structures under darkness than in the presence of light, mainly at the liquid-air interface (Fig. 10B). Interestingly, 17978.prpA cells formed more biofilms under darkness not only at the liquid-air interface, but also above and below this interface when compared to samples of this strain cultured under illumination. Overall, the SEM results support the data collected with standard crystal violet assays shown in Fig. 11B. Although not shown, incubation of the 17978 parental strain and the 17978.prpA derivative in SB at 37°C resulted in non-detectable formation of pellicles on the surface of SB either in the presence or absence of light. Taken together, these results not only reveal the importance of the PrpABCD pilus system in surface motility and biofilm biogenesis, but also suggest the expression of alternative light-regulated motility and biofilm mechanisms when 17978.prpA cells are cultured at 37°C.

Contribution of the prp locus to virulence The virulence role of the 17978 PrpABCD pilus assembly system was tested using the ex vivo polarized A549 human alveolar epithelial cell system and the in vivo G. mellonella experimental infection model we used before to examine the role of A. baumannii factors and products in its pathophysiology (84, 96). Infection of A549 polarized cells with 17978 bacteria resulted in a remarkable damage to the cells and the disappearance of the surfactant layer that covers their surfaces, which was readily detectable on polarized cells cultured under sterile conditions (compare panel A with panels B, C and D of Fig. 12). Furthermore, 17978 bacteria formed clusters, filaments and cell chains attached to the surface of damaged A549 cells (aqua, yellow and red arrows in panels B and C of Fig. 12, respectively). Examination at higher magnification showed that some of the bacterial filaments were formed by incomplete or lack of cell division, while chains were formed by bacterial cells attached by their ends (yellow and red arrows in Fig. 12C, respectively). Infection with the 17978.prpA insertion derivative resulted in apparent A549 damage with bacterial clusters attached to them as well as the formation of areas covered by a dense layer of bacterial filaments (pink and red squares in Fig. 12D, respectively).

46 Examination at higher magnification showed that, as shown for the 17978 strain, cell chains were formed by bacteria attached by their ends, while filaments were formed by incomplete or lack of cell division, with the latter being the predominant structures (red and yellow arrows in panels E and F of Fig, 12, respectively). Infection of G. mellonella larvae, which were incubated for 5 days at 37°C after injection with cells of the 17978 strain, resulted in a survival rate of 47.5%. This rate is significantly lower (P < 0.05) than the rates displayed by non-injected larvae or larvae injected with the same volume of sterile PBS (Fig. 13). This figure also shows that the infection of caterpillars with 17978.prpA bacteria resulted in an 82.5% survival rate, which is significantly higher (P ≤ 0.05) than the rates scored with larvae infected with cells of the 17978 parental strain but comparable to the survival rates displayed by the negative controls. Taken together, these observations indicate that the PrpABCD pilus assembly system plays a role in the interaction of host cells with 17978 and its virulence when tested with two experimental models already used to study A. baumannii virulence (17).

47 Figure 8. TEM analysis of samples collected from the surface of SA plates. TEM grids were placed substrate-side down on 17978 cells at the edge of the motility zone (A) or immediately in front of this zone where cells were not visible (B). (C) Micrograph of cells lifted from the edge of the colony formed by the non-motile 17978.prpA mutant. The white arrows in panels A and B identify pili attached to bacterial cells as well as pili present in front of the of the motility zone. Plates were incubated overnight at 24°C under darkness. Images are representative of at least three fields of view of each sample at 50,000X magnification. Scale bars, 0.5 µm.

49 Figure 9. Role of PrpA in pellicle and biofilm formation in response to illumination at 24°C. (A) Pellicle formation by the parental strain (17978) and the prpA insertion mutant (prpA) on the surface of SB cultures statically incubated at 24°C in darkness. (B) Biofilm formation by the same two strains and the complemented prpA derivative (prpA.CV) on microtiter plate wells statically incubated at 24°C in darkness or illumination. Horizontal bars identify statistically different values (P ≤ 0.01, ; P ≤ 0.001, ) and error bars represent the standard error of each data set.

Figure 10. SEM analysis of biofilms formed on plastic. Plastic coverslips were semi-submerged in SB inoculated with 17978 or prpA bacteria and statically incubated at 24°C (A) or 37°C (B) under illumination (L) or darkness (D) for 96 h or 48 h, respectively. Samples were fixed, gold coated and examined by SEM. Images taken above, at and below the air-liquid interface are representative of at least three fields of view of each sample at 5,000X magnification. Scale bars, 2 µm.

53 Figure 11. Role of PrpA in surface motility and biofilm formation in response to illumination at 37°C. (A) Cells of the parental strain (17978), the prpA insertion mutant (prpA) and its complemented derivative (prpA.CV) were surface-inoculated on SA plates and incubated at 37°C in the presence or absence of light. (B) Biofilm formation by the same three strains on microtiter plate wells statically incubated at 37°C in darkness or illumination. Horizontal bars identify statistically different values (P ≤ 0.001, ; P ≤ 0.001, ) and error bars represent the standard error of each data set.

55 Figure 12. Role of PrpA in A549-bacteria interactions. SEM micrographs of uninfected (sterile) polarized A549 human alveolar epithelial cells (A) or polarized cells infected with bacteria of the 17978 strain (B and C) or the prpA insertion derivative (D-F) and incubated for 48 h at 37°C in the presence of 5% CO2. The green and white boxes in panel A identify the surface of a healthy A549 cell and the mucin layer covering them when cultured under sterile conditions, respectively. The red and pink boxes in panel D identify areas covered by a layer of bacterial filaments and areas where the mucin layer was degraded upon bacterial infection, respectively. The aqua arrows in panels B and C identify bacterial clusters attached to damaged epithelial cells. The yellow and red arrows in panels B, C, E and F identify bacterial filaments and bacterial chains attached to A549 cell surfaces, respectively. Scale bars, 2 μm and 10 μm. Micrographs were taken at a magnification of 1,000x, 5,000x and 10,000x as shown in the micrographs.

57 Figure 13. Role of PrpA in virulence. G. mellonella larvae (n = 40 per experimental condition) were injected with 105 bacterial cells of the 17978 parental strain or the prpA isogenic insertion derivative. Larvae injected with sterile PBS and non-injected larvae were used as negative controls. Larvae were incubated at 37°C and assessed for death at 24-h intervals over 5 days with removal of dead larva at times of inspection. The asterisk below the prpA line data indicates P ≤ 0.05 when compared with the data representing the virulence response of 17978.

59 DISCUSSION Our original report describing the unexpected ability of A. baumannii ATCC 17978 to respond to light showed that this ubiquitous environmental signal controls surface motility, biofilm biogenesis and virulence through the function of BlsA, a “short” BLUF-containing photoreceptor (4). The fact that the molecular and cellular factors involved in these light- regulated responses are mostly unknown prompted us to identify and characterize some of these factors and associated responses in more detail using RNA-Seq data as a guide. This approach (Table 3) together with qRT-PCR analyses (Fig. 4B) resulted in the identification of prpABCD, a polycistronic operon coding for a predicted type I pilus assembly system, the expression of which was significantly increased under darkness by an uncharacterized regulatory mechanism that depended on the production of active BlsA when cells were cultured at 24°C in SB (Fig. 4C). Such a response strongly indicates that the effect of light on surface motility could be associated with the expression of this operon when cells are incubated under the aforementioned conditions. This hypothesis is fully supported by the failure of the 17978.prpA derivative, which harbors a deletion-insertion that removed an internal region of this gene, to display surface motility on SA plates incubated at 24°C in darkness (Fig. 7). Furthermore, the comparative analysis of 17978 and 17978.prpA cells lifted from SA plates incubated under the aforementioned conditions using TEM showed that prpA inactivation abolishes the production of long and thin pili structures detected only in the 17978 samples (Fig. 8). Interestingly, the prpA mutation also resulted in a significant reduction in the presence of vesicle-like structures that were readily apparent in the 17978 samples. Currently it is not known whether these structures play a role in surface motility and/or any aspect of the A. baumannii pathophysiology. The lack of motility of the parental strain, the 17978.prpA mutant and the 17978.prpA.CV derivative when cultured at 24°C under illumination is not surprising considering our initial observation that light drastically reduces this cellular response (4). Altogether, these results not only demonstrate the effect of light on the differential expression of the prpABCD operon, but also provide novel insights into the cellular factors and the mechanism by which a microorganism is able to move on the surface of a semi-solid medium in spite of a genus name that derives from the Greek term “akineto”, which denotes lack of motility. The pellicle and biofilm biogenesis responses suggest different relationships between these two cellular responses and motility (Fig. 9). In the case of pellicle formation, it is apparent

60 that this relationship is direct since inactivation of prpA drastically reduced this cellular response, a behavior that most likely depends on the ability of bacterial cells to migrate toward the liquid- air interface by a mechanism that so far has not been described in any member of the Acinetobacter genus. The role of PrpA, and predictably the entire PrpABCD pilus assembly system, in pellicle formation is further supported by the identification of PrpA as a component of the proteins embedded in the matrix of pellicles formed by 17978 cells statically cultured at 25°C in Mueller Hinton Broth or T broth (97). Taken together, our surface motility and pellicle formation results provide strong support for the potential involvement of the PrpABCD pili in A. baumannii motility functions, which remain to be understood at the functional and cellular levels. In contrast to pellicle formation at 24°C, the relationship between motility and biofilm formation on an abiotic surface is inverse; non-motile 17978.prpA bacteria formed more biofilms on plastic, both under illumination and darkness, when tested using crystal violet assays (Fig. 9B). This observation is supported by SEM analysis, which showed that 17978. prpA form more uniform and denser biofilms at the liquid-air interface under illumination and darkness, respectively, when compared to 17978 (Fig. 10A). This outcome could be due to a decrease in motility and the consequent increased sessility that results in enhanced bacterial adherence and biofilm development as reported for other bacteria (98-100). Alternatively, the absence of prpABCD-mediated pili could facilitate or trigger the interaction of 17978 bacteria with solid substrates through additional or alternative light-regulated adhesins. The annotation of additional 17978 putative pilus assembly systems, including the A1S_0690-A1S_0695 and A1S_1507- A1S_1510 operons, and predicted pili proteins, such as A1S_3167 (PilY1) and A1S_3177 (fimbrial protein), lend support to this possibility that remains to be tested experimentally. The studies conducted at 37°C resulted in particularly interesting and novel information since they showed that light still controls cellular functions under conditions that have a negative impact on BlsA’s production and function. Our original report describing light-regulated responses in A. baumannii ATCC 17978 showed that the incubation of bacteria at 37°C resulted in the constitutive expression of surface motility and biofilm biogenesis on plastic (4). This response could be attributed to the significantly reduced constitutive transcription of blsA in cells cultured at this temperature (4). The negative effect of 37°C on blsA transcription was recently confirmed by work published by Abatedaga et al. (39), which also showed the predicted

61 reduction in BlsA production as well as the observation that temperatures higher than 30°C significantly impair the photocycle of purified BlsA protein. Thus, the ability of the 17978 parental strain incubated at 37°C to produce comparable surface motility responses in the presence or absence of light under the conditions described in this report (Fig. 11A) is in accordance with all the aforementioned observations regarding the effect of temperature on BlsA production and activity, and the consequent constitutive expression of the PrpABCD pilus assembly system (Fig. 4C). Conversely, the ability of 17978.prpA cells to display light-regulated motility at 37°C was an unexpected response that has relevant functional and regulatory implications. First, the expression of light-regulated motility at 37°C by 17978.prpA (Fig. 11A) strongly indicates that this cellular response involves additional motility functions that are not operational at 24°C since this mutant is non-motile at that temperature under darkness or light (Fig. 7A). Second, considering the observation that the amount of biofilm produced by the 17978 parental strain under darkness and illumination are similar to those produced by the 17978.prpA mutant, it is possible to speculate that the same cellular component and/or process is responsible for the biofilm responses shown in Fig. 11B. Third, the detection of light-regulated responses by the 17978 parental strain and the isogenic 17978.prpA under temperature conditions that negatively affect the production of BlsA and its photoreceptor activity (Fig. 11) strongly indicate that an alternative and uncharacterized photodetection and regulatory system must be operational when A. baumannii cells persist and/or grow at 37°C. Taking into account all these observations, it is tempting to speculate that the BlsA-mediated system is one of the mechanisms A. baumannii uses to persist outside the host in nosocomial environments at temperatures lower than 37°C, while at least one different light-dependent regulatory system, which remains to be identified and characterized, is responsible for the capacity of this pathogen to adapt to and persist under higher temperature conditions imposed by the human host. Our data also suggest that A. baumannii is capable of producing different adhesins and pili assembly systems in response to changes in environmental conditions. This possibility is supported by our report showing that an A. baumannii ATCC 19606T derivative deficient in the expression of the CsuAB-A-B-C-D-E chaperone-usher pili assembly system does not produce pili and does not to attach to and develop biofilms on plastic when cultured in LB broth at 37°C (18, 101). However, the inactivation of this system enhanced rather than abolished the adherence of the ATCC 19606T Csu-deficient mutant to A549 monolayers and sheep erythrocytes when

62 compared to the parental strain (102). Furthermore, the Csu mutant produced an increased amount of thin and short pili when compared with the ATCC 19606T parental strain, whose surface contained thin and short as well as long pili-like structures (102). Thus, several uncharacterized sensory and regulatory functions, in addition to those mediated by BlsA, should be involved in these adaptive responses, which allow facultative pathogens, such as A. baumannii, to respond to different extracellular cues they encounter throughout their life cycles. The A. baumannii MAR002 biofilm hyper-producer clinical isolate is among the strains that harbor a homolog of the prpABCD operon, which was annotated as the LH92_11070- LH92_11085 4-gene cluster (96). This report (96) showed that LH92_11085 and its 17978 prpA homolog were significantly overexpressed in biofilm cells when compared to planktonic bacteria of the cognate strains cultured at 37°C. This observation indicates that the expression of the prpABCD operon is controlled not only by light, but also the bacterial lifestyle. It is interesting to note that the analysis of the MAR002 11085 derivative, which harbors a LH92_11085 partial deletion, resulted in functional outcomes that were similar as well as different from those we collected with the 17978.prpA derivative. The MAR002 11085 derivative produced less rather than more biofilms on plastic as well as on the surface of A549 polarized cells when compared with the parental strain, without detectable formation of filamentous bacteria as we observed with the 17978 strain and particularly the 17978.prpA derivative (Fig. 12). In contrast, the A549 polarized culture model showed that inactivation of LH92_11085 reduced the virulence of MAR002 against host cells (96), a response that parallels our data collected with the G. mellonella experimental infection model. This virulence model showed that the survival rates of caterpillars infected with the 17978.prpA mutant were not only significantly higher than those displayed by larvae infected with the 17978 parental strain, but also comparable with the negative controls (Fig. 13). Overall, these observations indicate that the 17978 PrpABCD system and its MAR002 LH92_11070-LH92_11085 homolog play a role in the physiology and virulence of A. baumannii. However, some of the biological roles associated with these pilus assembly systems, such as adherence and biofilm biogenesis, seem to be strain specific, a phenomenon that has been described for other A. baumannii traits, including the different effects of mutations in genes coding for the same antimicrobial resistance functions in different clinical isolates (103-105). Furthermore, it is important to note that A. baumannii MAR002 was classified as a ST271 genotype isolate that does not relate to any of the high-biofilm producers

63 reported and does not map into any International Clonal linage (96). Whether these taxonomical differences or the potential ability to differentially express distinct adhesins and biofilm functions under different conditions are responsible for the distinct MAR002 phenotypes when compared to 17978 are possibilities that remain to be explored. ACKNOWLEDGEMENTS This project was supported by National Institutes of Health Public Health grant R15GM117478- 01 as well as by Miami University research funds. We thank Hauke Smidt and Thomas Kruse (Wageningen University, The Netherlands) for providing the Lactococcus lactis subsp. cremoris strain that was used as a source of plasmid pIL252.

64 Chapter 3

Structural and functional analysis of the Acinetobacter baumannii BlsA photoreceptor and regulatory protein

Cecily R. Wood, Mariah S. Squire, Natosha L. Finley and Luis A. Actis

To be submitted

65 ABSTRACT The Acinetobacter baumannii BlsA photoreceptor has an N-terminal (NT) BLUF domain and a C-terminal (CT) amino acid sequence with no significant homology to characterized bacterial proteins. In this study, we tested the biological role of specific residues located in these BlsA regions. Site-directed mutagenesis, surface motility assays at 24°C and protein overexpression showed that residues Y7, Q51 and W92 are essential for not only light-regulated motility, but also BlsA’s solubility when overproduced in a heterologous host. In contrast, residues A29, W78 and F32, the latter representing a significant difference when compared to other BLUF- containing photoreceptors, do not play a major role in BlsA’s biological functions. Analysis of the CT region showed that the deletion of the last five BlsA residues has no significant effect on the protein’s light-sensing and motility regulatory functions, but the deletion of the last 14 residues as well as K144E and K145E substitutions significantly alter light-regulated motility responses. In contrast to the NT mutants, these CT derivatives were overexpressed and purified to homogeneity to demonstrate that although these mutations do not significantly affect flavin binding and photocycling, they do affect BlsA’s photodynamic properties. Notably, these mutations map within a potential fifth α-helical component that could play a role in predicted interactions between regulatory partners and BlsA, which could function as a monomer according to gel filtration data. All these observations indicate that although BlsA shares common structural and functional properties with unrelated photoreceptors, it also exhibits unique features that make it a distinct BLUF photoreceptor.

66 Importance As a facultative pathogen that thrives in different ecological niches, including the human host and the medical environment, Acinetobacter baumannii senses and responds to different extracellular signals. Among these signals is light, which regulates the expression of a wide range of cellular functions via the BlsA photoreceptor and transcriptional regulator. Our work has established significant unreported N- and C-terminal structure-function relationships that determine BlsA’s photosensing and regulatory functions involved in the differential expression of surface motility upon illumination as well as protein stability. Our findings provide critical insights needed to understand the molecular mechanisms by which BlsA controls a broad range of light-regulated cellular responses.

67 INTRODUCTION Acinetobacter baumannii is a Gram-negative opportunistic pathogen and the causative agent of both nosocomial- and community-acquired infections (106). Although A. baumannii is commonly associated with hospital infections, it has been isolated from a wide range of environmental sources and samples including water and aquaculture environments (78), soil (79), different food sources (80), animals (81, 82) and insects (83), all of which could be reservoirs for this bacterium outside the hospital environment (35). A. baumannii’s prevalence in the environment and hospital settings and its ability to spread throughout the community can be attributed to its capacity to sense and respond accordingly to different environmental cues, including its ability to sense light through the blue-light-sensing protein A (BlsA) (4). BlsA is a blue light using flavin (BLUF) domain-containing photoreceptor involved in controlling biofilm, motility, and virulence functions in response to light at temperatures lower than 30°C (4, 39). BlsA shares a similar architecture with other BLUF photoreceptors: the flavin chromophore is inserted between the first two N-terminal (NT) α-helices of the BLUF domain, which also includes a 5-stranded β-sheet component (107). BlsA’s NT BLUF domain, which is predicted to noncovalently bind the flavin chromophore through conserved and semiconserved amino acids including a tyrosine-glutamine pair and a tryptophan residue, is hypothesized to act as a light sensor for this microorganism (24, 46). The light signal, which is generated by proton-coupled electron transfer (PCET), is initiated by the conserved tyrosine upon blue light illumination (108) and then transmitted to the C-terminus (CT) of BlsA through rearrangements of an intricate hydrogen bonding network involving the side chains of the conserved glutamine and other semiconserved amino acids of the BLUF domain around the N5 of the isoalloxazine ring of the flavin chromophore. Hydrogen bond switching between the side chain glutamine, the flavin, and other residues in the BLUF domain are proposed to trigger a signal through the β5 strand of the protein, particularly through the side chain of the tryptophan residue, which leads to conformational changes in the BlsA structure and ultimately, a potential complex formation with unknown target proteins (54). UV-visible spectroscopic studies can reveal the signaling state of BLUF domains characterized by a 10-nm redshift, which is indicative of photoexcitation (109). The light state is reversible and depending on the BLUF protein, relaxation back to the dark- adapted state occurs in as little as 10 seconds (t1/2, 5 s), as in the case of PixD (Slr1694), or can take up to 30 minutes (t1/2, 15 min), as with the large BLUF protein AppA (58, 61). BlsA is

68 intermediate between these proteins and relaxes back to its dark state after illumination with a half-life of approximately 8 minutes (54). Although the basic photochemistry and architecture of the BLUF photoreceptors are similar, the proteins are further classified into one of two groups: “long” multidomain BLUF- containing proteins or “short” BLUF proteins. Unlike the first and one of the best-characterized BLUF domain-containing photoreceptors, AppA, which is a BLUF multidomain blue light sensing protein (55), BlsA is considered a short BLUF-containing photoreceptor protein. Like other short BLUF proteins, such as Tll0078 (57), the first 95 amino acids of BlsA are predicted to be dedicated to light sensing, whereas the remaining 61 CT residues have no known function. Because their CT regions show no sequence similarity to known DNA-binding motifs or other proteins involved in regulation of gene expression, these short photoreceptor proteins are predicted instead to participate in protein-protein interactions, such as those used for oligomerization (55, 57). Structural and photochemical characterizations of BLUF domain- containing photoreceptors have been the focus of several studies since the identification of this family of light sensing proteins. However, the biological activity, especially of the divergent CT regions, as it pertains to the organism remains to be elucidated. In this work we further characterize the biological function and biochemical properties of BlsA by targeting specific amino acid residues of the N- and C-termini. This approach showed that NT amino acid residues predicted to interact with the flavin adenine dinucleotide (FAD) chromophore are critical not only for its light-dependent regulatory functions, but also for protein stability. The analysis of BlsA’s CT region showed that although the last five amino acid residues are dispensable for its photosensing and regulatory roles, the 7-residue region located between amino acids 144 and 150 could represent a functionally significant fifth α-helix component. Surface motility assays showed that deletion or site-directed mutagenesis of residues mapped within this region affect the ability of isogenic derivatives to display differential motility responses at 24°C upon illumination. Additionally, size exclusion chromatography resulted only in the isolation of BlsA monomers without any indication of dimerization or polymerization of this protein under the native conditions used in this work. Taken together, our results provide critical insights that contribute to understanding the structure-function relationships of a photosensory and regulatory protein that plays a key role in A. baumannii’s pathophysiology.

69 MATERIALS AND METHODS Bacterial strains and culture conditions All strains and plasmids used in this study are listed in Table 4. Escherichia coli and A. baumannii strains were routinely cultured in Luria Bertani (LB) broth or on agar plates and supplemented with appropriate antibiotics when necessary (88). Swimming broth (SB; 10 g.L-1 tryptone; 5 g.L-1 NaCl) and swimming agar (SA; SB containing 0.3% agarose) were used for growth and motility assays, respectively. A. baumannii was routinely grown at 37°C for 16-18 h (overnight).

DNA procedures Plasmid DNA was isolated from E. coli cells using a commercial kit (Qiagen). A. baumannii total DNA was isolated using a phenol-based method described previously (89). DNA polymerases and restriction enzymes were used according to the manufacturer’s protocols (New England Biolabs). Custom-designed primers (Integrated DNA Technologies) or kit-supplied M13 or T7 primers (Life Technologies) were used to perform all sequencing reactions using BigDye-based chemistry (Applied Biosystems) prior to subcloning reactions and DNA transformations. All primers used in this work are listed in Table 5. blsA mutagenesis The 760-bp chromosomal region encompassing the blsA promoter and coding sequences was amplified with Q5 DNA polymerase and primers 4379 and 4380, each flanked with BamHI restriction sites (Table 5), digested with BamHI and then ligated into the BamHI site of pMU368, a pMAC A. baumannii-E. coli shuttle vector derivative (110), generating pMU1202 (Table 4). Conserved and non-conserved amino acids of the BlsA NT region were targeted by site-directed mutagenesis with the QuikChange Lightning Multi (QCLM) kit (Agilent) using pMU1202 as template and primers 4388, 4392, 4394, 4396, 4412 and 4426 (Table 5) to generate the six corresponding derivatives: A29S (pMU1222), F32N (pMU1213), Q51A (pMU1226), Y7A (pMU1229), W78A (pMU1240) and W92A (pMU1270), respectively (Table 4). Primer 4394 and pMU1229 (Table 5) were used to generate a Y7A/Q51A double point mutant (pMU1243) (Table 4). BlsA’s CT point mutations were generated following the same protocol as described for the NT mutagenesis using pMU1202 as template and primers 4425 and 4478 (Table 5) to

70 generate two cognate derivatives: K145E (pMU1250) and K144E (pMU1276) (Table 4). Two CT deletion mutations were generated by inverse PCR using Q5 DNA polymerase and pMU1202 as template with primer pairs 4383/4385 and 4384/4385 (Table 5) to produce pMU1232 (Δ152-156) and pMU1235 (Δ143-156) harboring deletions of the 5- and 14-most CT residues, respectively (Fig. 14). Primers 4386 and 4387 (Table 5) and plasmid pMU1202 were used to generate pMU1236 (Δ121-135) (Table 4) harboring a deletion of residues located in

BlsA’s predicted α4 component (Fig. 14). Primers 4421/4423 (Table 5) and pMU1202 were used to generate pMU1251 (Table 4) harboring the deletion of BlsA’s CT residues 135-147 (Δ135- 147) shown in Fig. 14.

Complementation and motility assays The biological effect of the BlsA mutations described above was tested by comparing surface motility of the 17978 wild type strain and the 17978.OR blsA insertion derivative electroporated, as described before (90), with the pMU368 derivatives listed in Table 4. Transformants were selected on LB agar containing 100 µg.mL-1 zeocin (Zeo). Fresh bacterial cultures of each strain were touch-inoculated onto the surface of SA plates with blunt-ended sterile wooden sticks and incubated at 24°C for 24 h under darkness or light using a blue light LED array as described previously (4). Motility assays were performed with three independent fresh bacterial samples in triplicate for each tested sample (n = 9). The ImageJ image processing program (National Institutes of Health) was used to measure the surface motility area of each replicate. The hot Triton plasmid isolation method (28) and agarose gel electrophoresis (88) were used to confirm the presence of complementing plasmids in 17978 derivatives tested for motility.

Overexpression and purification of BlsA and BlsA derivatives Q5 DNA polymerase and primers 4427 and 4428 (Table 5) were used to PCR-amplify the parental blsA coding region from 17978 genomic DNA. The 480-bp amplicon, flanked with NdeI and BamHI restriction sites, was ligated into the cognate restriction sites of the pET-15b expression vector (Novagen), generating pMU1254 (Table 4), which codes for a His-tagged derivative of the parental BlsA protein. Clones harboring blsA mutations were PCR-amplified from pMU1213, pMU1226, pMU1229, pMU1243 and pMU1270 (Table 4) using primers 4427 and 4428 (Table 5), as described above. The resulting amplicons were cloned into pET-15b to

71 generate the cognate His-tagged BlsA derivatives harboring NT amino acid mutations. Corresponding plasmids are listed in Table 4 as pMU1299, pMU1255, pMU1273, pMU1271 and pMU1298, respectively. BlsA derivatives with CT deletions were generated by inverse PCR using pMU1254 as a template and primer pairs 4439/4440 and 4439/4441 (Table 5) to generate two pET-15b derivatives: pMU1258 and pMU1264 harboring deletions of residues 143-156 and 152-156, respectively (Table 4). Inverse PCR using pMU1254 as a template and primer pairs 4386/4387 and 4421/4423 (Table 5) were used to generate two pET-15b derivatives: pMU1256 and pMU1257 (Table 4) harboring in-frame deletions of residues 121-135 and 135-147, respectively. The BlsA CT point-mutation derivatives in pET-15b were constructed by subcloning the blsA derivatives from pMU1276 and pMU1250 (Table 4) into pET-15b, as described for the construction of pMU1254, to generate the derivatives pMU1282 and pMU1259, which code for the BlsA K144E and K145E derivatives, respectively (Table 4). Proper plasmid construction was verified by automated DNA sequencing using T7 primers. E. coli BL21(DE3) were transformed with each pET-15b derivative described above and selected for on LB agar containing 150 µg.mL-1 ampicillin (LB-Amp). A single colony was used to inoculate LB-Amp broth, which was grown with shaking overnight at 37°C. Overnight cultures were diluted 100 times and used to inoculate 500 mL of LB-Amp. Cultures were grown with shaking at 30°C until an OD600 of 0.7-0.8 was reached. The temperature was then lowered to 18°C before the addition of IPTG to a final concentration of 0.8 mM. Bacterial cells were grown at 18°C, with shaking, for an additional 15 h in the dark. Cells were collected by centrifugation and pellets were stored at -80°C until further use. Bacterial cell pellets were resuspended in lysis buffer (20 mM Tris-HCl, pH 8.0, 500 mM NaCl; 20 mM imidazole; 1 mM 2-mercaptoethanol [2-ME] and 20 mg.L-1 DNase and RNase) containing cOmplete EDTA-free protease inhibitor cocktail tablets (Sigma) and lysed with a French Press. Cell debris was removed by ultracentrifugation at 100,000 x g for 90 min at 4°C. Debris-free lysates were loaded onto a 5-mL HisTrap column (General Electric Healthcare, GE) and washed with 5 vol of lysis buffer without 2-ME. His-BlsA and derivatives were eluted using an imidazole step gradient ranging from 20 to 500 mM. His-tagged BlsA and derivatives dissolved in 20 mM Tris-HCl, pH 8.0, 500 mM NaCl were further purified and characterized by size-exclusion chromatography (SEC) on an ÄKTA instrument using a Superose 6 Increase 10/300 GL column (GE), at a flow rate of 0.5 mL.min-1.

72 Protein concentration and purity were determined by Bradford assays (111) and SDS-PAGE analysis (112), respectively.

Spectroscopy analyses Pure protein samples in 20 mM Tris-HCl (pH 8.0) containing 500 mM NaCl were analyzed with an EPOCH/2 microplate reader/spectrophotometer (BioTek) or a Perkin Elmer Lambda 35 UV/Vis spectrometer. Protein samples were kept in the dark at 4°C for dark-adapted BlsA samples (dBlsA) before blue light illumination, with light intensities ranging between 20 and 200 μmol m−2 s−1, for 3 min at 22°C and read using a spectral scan of 1 nm. The spectral absorbance properties of all derivatives were recorded at least twice from two different purified protein samples. For dark state recovery kinetics, purified samples were exposed to blue light for 3 min at 22°C and spectral scans were recorded at 505 nm for 20 min.

Circular dichroism (CD) Protein samples dissolved to a final concentration of 3 µM in 10 mM phosphate buffer, pH 8.0 containing 20 mM NaCl were analyzed with an Aviv model 435 CD spectrophotometer at 20°C using a 0.1 cm quartz cuvette. Scans of light-adapted (lBlsA) and dBlsA were collected at a speed of 500 nm.min-1 between 190 nm and 260 nm. The data are representative of two protein samples run at least twice under the described experimental conditions.

Chromophore detection A protocol based on the method described before (113) was used to detect flavin bound to BlsA. Briefly, a BlsA sample containing 40 µM of protein dissolved in 20 mM Tris-HCl, pH 8.0, 500 mM NaCl was incubated at 90°C for 10 min followed by centrifugation at 20,000 x g for 15 min at 4°C. The supernatant was analyzed by HPLC with an Agilent 1100 LC instrument and a Waters Symmetry C18 reversed-phase column (100Å, 5 µm, 4.6 mm X 150 mm) at room temperature. The column was equilibrated with 50 mM sodium acetate (pH 5.0) containing 5% methanol, and a linear gradient of 5-70% methanol at a flow rate of 0.5 mL.min-1 was used to elute the flavins, which were detected at a wavelength of 450 nm. Standards were made using 100 µM solutions of FAD, flavin mononucleotide (FMN) and riboflavin (Ribo).

73 Bacterial growth analysis Fresh bacterial cultures of the 17978 parental strain and selected derivatives were grown on LB agar plates, inoculated into LB broth supplemented with appropriate antibiotics and cultured overnight at 37°C in a shaking incubator. The bacterial strains were diluted 1:100 into 50 mL of fresh SB without antibiotics and incubated under darkness as described above with constant shaking in an incubator set at 24°C. OD600 reads were taken hourly for 12 h and 24 h after inoculation using two independent biological samples.

Detection of BlsA in E. coli BL21 cells and antibody production Whole-cell lysates of overnight cultures of E. coli BL21 harboring pET-15b derivatives coding for parental BlsA or mutant derivatives were prepared and size-fractionated by SDS- PAGE using 4%-20% polyacrylamide gradient gels (Bio-Rad) as described before (114). Proteins were detected by staining with Coomassie Brilliant Blue or western blotting with anti- BlsA antiserum. Protein A labeled with horseradish peroxidase was used to detect the immunocomplexes (31). Polyclonal anti-BlsA antibodies were generated by immunizing a female New Zealand white rabbit with purified protein as described before (114). Polyclonal anti-BlsA antibodies were generated by immunizing a female New Zealand white rabbit with purified protein as described before (114). The polyclonal serum was pre-adsorbed with E. coli and A. baumannii 17978.OR cell extracts to reduce non-specific reactions as described previously (115). The immunization protocol was carried out as approved by the Miami University Institutional Animal Care and Use Committee.

Bioinformatic methods Amino acid sequences were compared and aligned using EMBOSS Needle and MUSCLE, respectively. The secondary structure of BlsA was initially determined using the SABLE server. The primary amino acid sequences of short BLUF proteins used for alignments and modeling were downloaded from the Protein Data Bank (PDB). Additional modeling and graphics analysis was performed using the Chimera interface to Modeller (116) and the full- chain protein structure prediction server Robetta (117). The Modeller9v8 was also used to generate a 3D model of BlsA (118, 119). The stereochemical quality of the 3D model was assessed using a Ramachandran plot.

74 Statistics The statistical significance of collected experimental data was analyzed by the student’s t-test or analysis of variance (ANOVA) using the GraphPad Prism version 7.0d for Max OS X (GraphPad Software). P values ≤ 0.05 were considered statistically significant for all aforementioned assays. Error bars represent the standard error for each data set shown in the figures.

75 Table 4 Bacterial strains and plasmids used in this work Strains/plasmids Relevant characteristic(s)a Source/reference Strains A. baumannii ATCC 17978 Clinical isolate ATCC 17978.OR blsA::aph derivative of 17978; KmR (4) 17978.OR.V blsA::aph harboring pMU368; KmR; ZeoR This work 17978.OR.W blsA::aph harboring pMU1202; KmR; ZeoR This work 17978.OR.Y7A 17978.OR harboring pMU1229; KmR; ZeoR This work 17978.OR.A29S 17978.OR harboring pMU1222; KmR; ZeoR This work 17978.OR.F32N 17978.OR harboring pMU1213; KmR; ZeoR This work 17978.OR.Q51A 17978.OR harboring pMU1226; KmR; ZeoR This work 17978.OR.Y7A/Q51A 17978.OR harboring pMU1243; KmR; ZeoR This work 17978.OR.W78A 17978.OR harboring pMU1240; KmR; ZeoR This work 17978.OR.W92A 17978.OR harboring pMU1270; KmR; ZeoR This work 17978.OR.K144E 17978.OR harboring pMU1276; KmR; ZeoR This work 17978.OR.K145E 17978.OR harboring pMU1250; KmR; ZeoR This work 17978.OR.Δ121-135 17978.OR harboring pMU1236; KmR; ZeoR This work 17978.OR.Δ135-147 17978.OR harboring pMU1251; KmR; ZeoR This work 17978.OR.Δ143-156 17978.OR harboring pMU1235; KmR; ZeoR This work 17978.OR.Δ152-156 17978.OR harboring pMU1232; KmR; ZeoR This work

E. coli DH5α DNA recombinant methods Life Technologies Top10 DNA recombinant methods Life Technologies BL21(DE3) λDE3, T7 RNA polymerase Life Technologies XL10-Gold DNA recombinant methods Life Technologies

Plasmids Cloning/overexpression pET-15b Overexpression vector; AmpR Novagen

76 pMU1254 pET-15b harboring blsA; AmpR This work pMU1255 pET-15b harboring blsA:Q51A; AmpR This work pMU1256 pET-15b harboring blsA:Δ121-135; AmpR This work pMU1257 pET-15b harboring blsA:Δ135-147; AmpR This work pMU1258 pET-15b harboring blsA:Δ143-156; AmpR This work pMU1259 pET-15b harboring blsA:K145E; AmpR This study pMU1271 pET-15b harboring blsA:Y7A/Q51A; AmpR This work pMU1264 pET-15b harboring bls:AΔ152-156; AmpR This work pMU1273 pET-15b harboring blsA:Y7A; AmpR This work pMU1282 pET-15b harboring blsA:K144E; AmpR This work pMU1298 pET-15b harboring blsA:W92A; AmpR This work pMU1299 pET-15b harboring blsA:F32N; AmpR This work

Complementationb pMU368 A. baumannii-E. coli shuttle vector; KmR; (110) ZeoR pMU1202 pMU368 harboring blsA from 17978; KmR; This work ZeoR pMU1213 pMU368 harboring blsA:F32N; KmR; ZeoR This work pMU1222 pMU368 harboring blsA:A29S; KmR; ZeoR This work pMU1226 pMU368 harboring blsA:Q51A; KmR; ZeoR This work pMU1229 pMU368 harboring blsA: Y7A; KmR; ZeoR This work pMU1232 pMU368 harboring blsA:Δ152-156; KmR; This work ZeoR pMU1235 pMU368 harboring blsA:Δ143-156; KmR; This work ZeoR pMU1236 pMU368 harboring blsA:Δ121-135; KmR; This work ZeoR pMU1240 pMU368 harboring blsA:W78A; KmR; ZeoR This work pMU1243 pMU368 harboring blsA:Y7A/Q51A; KmR; This work ZeoR

77 pMU1250 pMU368 harboring blsA:K145E; KmR; ZeoR This work pMU1251 pMU368 harboring blsA:Δ135-147; KmR; This work ZeoR pMU1270 pMU368 harboring blsA:W92A; KmR; ZeoR This work pMU1276 pMU368 harboring blsA:K144E; KmR; ZeoR This work aAmpR, ampicillin resistance; KmR, kanamycin resistance; ZeoR, zeocin resistance.

78 Table 5 Primers used in this study Primer Sequencea number 4379 5’-ATGTGAGGATCCAGTATTACAAATTGAACGTG-3’ 4380 5’-TGATATGGATCCGTGATATGGATGCTGTCG-3’ 4383 5’-CCGATTCACCATTCCAAC-3’ 4384 5’-CGTATTCATCTTAGTTTGCTC-3’ 4385 5’-TAGCTATTAGATGTGATACCG-3’ 4386 5’-CAGCAAAAGCGGGTTAAAG-3’ 4387 5’-GAGCAAACTAAGATGAATACGG-3’ 4388 5’-CGTTTAAATCGTTGAAATCACGACTTTCTGTCAGAATATCACGTAAATCTT-3’ 4392 5’-CACAAATCCCGTTTAAATCGTTGTTATCACGAGCTTCTGTCVAGAATAT-3’ 4394 5’-GTTCACCTTCTAGACATGCAAAAAATGCATTATCGGCATAATAAAGCACCC-3’ 4396 5’-CGTTGGCTGGCAGCACACAGGCGAACGTTCATAAGAC-3’ 4412 5’-GATCAAAGGCATTATAATATTAAAGCGTTATGCACATACTCCATTGAT-3’ 4421 5’-AATTAAGAGTTCATTTAAG-3’ 4423 5’-ATGGTGAATCGGGGA-3’ 4425 5’-CAAACTAAGATGAATACGGTTAAAGAAGTTGGAATGGTGAATCGGG-3’ 4426 5’-ATGAACACTCTTTTCAGCGTGCGTCAATGAAATACGTGCAGC-3’ 4427 5’-TAGTGACATATGAACGTTCGCCTGTGT-3’ 4428 5’-TGATATGGATCCCTAGAACGGGTTTACTCC-3’ 4439 5’-TAGGGATCCGGCTGCTAAC-3’ 4440 5’-CGTATTCATCTTAGTTTGCTCTGC-3’ 4441 5’-CCGATTCACCATTCCAACTT-3’ 4478 5’-GCAAACTAAGATGAATACGGTTGAAAAAGTTGGAATGGTGAATCG-3’ aUnderlined nucleotides identify BamHI and NdeI restriction sites.

79 RESULTS AND DISCUSSION Overall predicted BlsA structure The class of short BLUF-containing photoreceptors includes small proteins, such as Tll0078 from Thermosynechococcus elongatus BP-1 (120), Slr1694 from Synechocystis sp. PCC6803 (61), SnfB from Stenotrophomonas sp. SKA14 (121) and the BlsA protein we identified in A. baumannii ATCC 17978 (4). These proteins each have an NT FAD-binding domain and a CT domain containing 40-50 amino acid residues, which are potentially involved in homooligomer formation and predicted to participate in regulatory functions (57), although they do not show significant similarity to known proteins. Comparative analysis showed that the amino acid sequences of BlsA and Tll0078 are 26% identical and 39.9% similar, whereas BlsA and Slr1694 share 30.7% and 47.6% identity and similarity levels, respectively. The same analysis showed that BlsA and SnfB share the lowest amino acid sequence relationship with 21.7% and 41.6% identity and similarity indices, respectively. The amino acid sequence alignments also revealed that 17 of the 19 residues identical in BlsA, Tll0078, Slr1694 and SnfB are located within the BLUF domain, which encompasses BlsA residues 2-96 (Fig. 14). This particular feature reflects the NT flavin-binding and photosensing functions shared among BLUF containing proteins, while the lower similarity at their C-termini is most likely due to the interaction of each photoreceptor with distinct protein partners involved in signal transduction and regulation of gene expression. The secondary structure of BlsA is also predicted to be comparable to those reported for Tll0078, Slr1694 and SnfB (Fig. 14). Analysis of the crystal structures of Tll0078 (57) and Slr1694 (122) revealed homologous BLUF domains that contain five β-strands and two α-helices (β1α1β2β3α2β4β5) with the remaining two α-helices (α3α4) located in the C-terminal region of these proteins (57). Based on all these similarities, we used Tll0078 and Slr1694, short photoreceptors whose crystal structures have been determined experimentally (57, 122), as templates to construct the tertiary structure of the entire 156-amino acid BlsA sequence, which is displayed in Fig. 15A using Chimera. This model shows the typical

β1α1β2β3α2β4β5α3α4 structure. Further modeling using Robetta produced a BlsA secondary structure model similar to that displayed in Fig. 14 with the exception that it did not include the

β5 component, which encompasses residues 93-97, and included an additional α-helix (α5) formed by the VGMVN residues located between positions 146-150 (Fig. 14). The BlsA tertiary structure predicted by Robetta not only included the β5 component, but also displayed the

80 additional α5 component (Fig. 15B). The prediction of the latter component, which encompasses residues 144-150 according to Robetta’s 3D model, is supported by additional modeling, which shows that the surface topology of the BlsA region defined by residues A135-K145 is similar to that displayed by the Tll0078 residues A126-Y137 that map within the predicted α4 component of this protein (Figs. 14 and 15C). The structural assessment of the 3D model quality revealed that > 94% of these residues in are in the most favored regions as evidenced by the Ramachandran plot. The biological effects of the K144E and K145E point mutations described later in the report lend further support for the proposed model, which includes the NT FAD- binding photoreceptor domain and a predicted CT domain containing an additional α5 that could be involved in signal transduction upon illumination. Taken together, these observations indicate that although the overall BlsA structure is similar to other short BLUF-containing photoreceptors, it may have unique structural properties that would explain some of the distinct observations described in this report.

81 Figure 14. Comparative analysis of short BLUF-containing photoreceptors. The amino acid sequence of the BlsA, Tll0078, Slr1694 and SnfB proteins from A. baumannii ATCC 17978, T. elongatus BP-1, Synechocystis sp. PCC6803 and Stenotrophomonas sp. SKA14., respectively, were compared using MUSCLE. Asterisks identify conserved residues. BlsA secondary structure components are represented by black arrows (β strands) and grey rectangles (α-helices), respectively. BlsA’s secondary structure was predicted using the SABLE server. The black horizontal bar identifies BlsA’s BLUF domain residues. Black triangles and corresponding numbers above them indicate site-directed amino acid changes. The boxed BlsA residues identify an additional α-helix component (α5) predicted using Robetta. The boxed

Tll0078, Slr1694 and SnfB residues designate cognate 4 components of these proteins. The black horizontal rectangles shown at the bottom of the figure represent locations of site-directed deletions.

83 Figure 15. Predicted tertiary structure of BlsA. Structural models were drawn using Chimera and the structure of the Tll0078 and Slr1694 proteins as templates (A), and Robetta (B). (C) Surface model of the Tll0078 α4 (left) compared to surface models of two predicted BlsA CT α helices (middle and right) drawn using Modeller9v8. Yellow (polar), red (negative), blue (positive), grey (nonpolar). Visible residues are labeled according to their position in the amino acid sequence of the cognate proteins.

85 Analysis of the BlsA N terminus Functional studies Based on sequence identity among BLUF domains located in different photosensor proteins, Brust et al. (54) proposed a model highlighting, as a distinct BlsA feature, the presence of an F residue at position 32 rather than the N residue found in different BLUF domains, including those of Tll0078 and Slr1694 (Fig. 14), that H-bonds to the C2=O carbonyl group of the isoalloxazine ring of the FAD. The unusual presence of an aromatic amino acid at a position close to the FAD moiety suggests that the rearrangement of the H-bonding network in the FAD- binding pocket in response to light may be affected. TRIR spectroscopy studies, used to investigate the role of this particular BlsA residue, indicated that changing F32 to H, which is the analogous residue in AppA, does not significantly alter the spectra of lBlsA and dBlsA (54). In contrast, changing F32 to N, the analogous residues in Tll0078 and Slr1694 (Fig. 14), results in spectral differences between lBlsA and dBlsA. These differences indicate that the presence of a N residue at position 32 results in stronger H-bonding interactions between BlsA and FAD (54). These observations prompted us to examine the biological effect, measured as the light-mediated BlsA-dependent surface motility response we described before (4), of the F32N mutation. Figure 16 shows that the motility of the 17978.OR BlsA mutant under illumination at 24°C was significantly increased (P ≤ 0.0001) when compared to the response of the 17978 parental strain under the same experimental condition. Analysis of 17978.OR.W showed that the electroporation of pMU1202, a derivative of the pMU368 shuttle vector harboring a copy of the parental blsA allele expressed from its natural promoter (Table 4), reduced motility under illumination to wild type levels. Such a response was not observed when 17978.OR.V cells, which harbored empty pMU368, were tested under the same experimental conditions (data not shown). These results show that the shuttle vectors pWH1266, which we used before for the same purpose (4), and pMU368, which proved to be a more convenient template to generate site- directed mutations by inverse PCR because of its smaller size, produced comparable outcomes. Furthermore, restriction analysis of plasmids isolated from the complemented strains demonstrated that pMU368-based complementing vectors were stably maintained as independent replicons without noticeable rearrangements throughout the motility assays, which were conducted using SA without the addition of a selecting antibiotic (data not shown).

86 Figure 16. Light-regulated surface motility of 17978 cells producing native BlsA or derivatives generated by site-directed mutagenesis. (A) Motility of 17978.OR BlsA deficient cells transformed with a pMU368 derivative coding for the F32N (OR.F32N), Y7A (OR.Y7A), Q51A (OR.Q51A), Y7A/Q51A (OR.Y7A/Q51A) or W92A (OR.W92A) NT mutant proteins. (B) Motility of 17978.OR BlsA deficient cells transformed with a pMU368 derivative coding for the Δ152-156 (OR.Δ152-156), Δ143-156 (OR.Δ143-156), Δ135-147 (OR.Δ135-147), K144E (OR.K144E) or K145E (OR.K145E) CT mutant proteins. The motility responses of the 17978 parental strain (17978), the BlsA 17978.OR mutant (OR) and the 17978.OR mutant transformed with a pMU368 derivative coding for the wild type protein (OR.W) were used as controls. Surface motility was tested using SA at 24°C under darkness or illumination. Horizontal bars identify statistically different values (P ≤ 0.01, ; P ≤ 0.001, ; P ≤ 0.0001, ) and error bars represent the standard error of each data set.

88 The analysis of 17978.OR.F32N, the OR BlsA mutant complemented with pMU1213 coding for the BlsA F32N mutant (Table 4), showed that this amino acid change did not alter the motility responses under darkness or illumination when compared with 17978 and 17978.OR.W (Fig. 16). This observation indicates that although the TRIR spectra of the F-to-N mutation at position 32 showed changes in H-bonding interactions with FAD (54), this mutation did not significantly affect the biological function of this protein when compared with the parental strain. It is noteworthy that the presence of an F residue at position 32 is not unique to BlsA; the SnfB BLUF-containing protein produced by Stenotrophomonas sp. SKA14 also has this aromatic residue at this position (Fig. 14). The model proposed by Brust et al. (54) predicts that Y7 is H-bonded to Q51 to form the critical FAD-protein association network required for light sensing that was also described for other BLUF proteins including Tll0078 (57). As described for the analogous Tll0078 Y8 and Q50 residues, it is possible to predict that upon light stimulation, BlsA structural modifications allow the Q51 side chain oxygen to form an H-bond with the hydroxyl oxygen of Y7, while the side chain amino group of Q51 forms a tight H-bond with the C4=O of FAD. These predictions of critical roles that Y7 and Q51 could play in BlsA’s biological functions were tested using the motility assays described above. In contrast to F32N, the transformation of 17978.OR with pMU1229, pMU1226, or pMU1243, which code for proteins harboring Y7A, Q51A and Y7A/Q51A BlsA mutations, respectively (Table 4), resulted in derivatives with significantly altered light-regulated motility responses. The OR.Y7A and OR.Y7A/Q51A transformants displayed significantly increased motility under illumination when compared with the response of 17978 (P ≤ 0.0001 and P ≤ 0.001, respectively) as well as OR.W harboring a plasmid copy of the parental allele (Fig. 16). The Y7A and Y7A/Q51A residue changes did not significantly alter the motility of these mutants under darkness. Interestingly, although the motility response of the OR.Q51A derivative under illumination was comparable to that displayed by 17978 and OR.W, the Q51A mutation resulted in a significant reduction (P ≤ 0.0001) of surface motility under darkness (Fig. 16). Notably, the growth curve of the OR.Q51A derivative was not significantly different from that displayed by the parental strain when cultured in SB at 24C under darkness (Fig. 17). This observation indicates that the reduced surface motility of the Q51A mutant cannot be explained by a simple negative effect on bacterial growth in the absence of light.

89 Figure 17. Growth curves of the 17978 parental strain and isogenic complemented derivatives. Cells of the 17978 strain (17978) and the OR derivative transformed with either pMU1226 or pMU1270, which code for the BlsA mutant derivatives Q51A and W92A, respectively, were cultured in SA at 24°C under darkness in a shaking incubator. The OD600 of each culture was determined hourly for 12 h and then at 24 h after inoculation using two independent cultures of each tested strain. Error bars represent the standard error of each data set.

91 The W92 residue, which is located just N-terminal to the predicted β5 strand (Fig. 14), could be crucial for BlsA functions based on the analogous residues in other bacterial BLUF- containing photoreceptors, including the Tll0078, Slr1694 and SnfB proteins (Fig. 14). In the case of Slr1694, the analogous W91 residue moves from a hydrophobic to a hydrophilic environment upon illumination, a process that potentially conducts the signal transduction process associated with the response to illumination (122). All these observations, which indicate that W92 plays a role in BlsA regulatory functions in response to light, prompted us to test the effect of the W92A mutation in light-regulated motility. The significantly enhanced motility phenotype (P ≤ 0.0001) of the OR.W92A derivative when compared with 17978 and OR.W upon illumination (Fig. 16) supports the role of this amino acid residue in BlsA activity. However, as was observed with Q51A, the W92A mutation significantly reduced (P ≤ 0.0001) the dark motility response compared with 17978 and OR.W (Fig. 16), without affecting the overall growth of OR.W92A compared to both the 17978 parental strain and the OR.Q51A mutant cultured at 24C under darkness in SB (Fig. 17). It is apparent that, compared to all tested strains, the motility of the OR.Q51A and OR.W92A mutants under darkness is significantly reduced to the point that it resembles the response of the 17978 parental strain under illumination. The reduced motility response of OR.Q51A under darkness could be explained by the observation that Q50A and Q63E mutations in the analogous sites of Slr1694 and AppA, respectively, resulted in derivatives locked in pseudo-light-excited states that produced spectral responses mimicking illumination when tested under darkness (123, 124). Unfortunately, we could not test this hypothesis because the Q51A and W92A point mutations resulted in the overexpression of insoluble proteins as described later in this report. Interestingly, the Y7A/Q51A double mutation resulted in a motility response under darkness similar to that displayed by all other tested strains with the exception of the OR.W92A, which also seems to represent a lit-state derivative (Fig. 16A). Notably, this observation is in contrast to that produced by the analogous W91A Slr1694 derivative, which displayed wild type photochemistry and signal transduction responses (124). Compared to the sequences aligned in Fig. 14, BlsA’s amino acid sequence is distinct because it has a W residue at position 78 within the β4 strand rather than the I residue of the Tll0078 and Slr1694 proteins. Motility assays showed that OR.W78A cells harboring pMU1240, which codes for the recombinant BlsA W78A derivative (Table 4), displayed a slight but not

92 significant reduction of motility under darkness with no differences under illumination when compared with the response of the 17978 parental strain (data not shown). This observation indicates that this particular residue is not critical for the function of short BLUF-containing proteins, a possibility supported by the fact that an E residue is located at the analogous site in SnfB (Fig. 14).

The BlsA A residue at position 29 within α1 is different from the analogous S residue of Tll0078 and Slr1694. This BlsA residue has previously been reported to play a role in the excited state stabilization of the protein. According to Brust et al. (54), the ground state recovery kinetics of the A29S mutant compared to the wild type BlsA protein increases approximately 2-fold in its lifetime. This phenomenon was only observed for the dark-adapted BlsA. Similar to data obtained for the AppA protein, whose analogous residue is S41, no observable effects were measured for the light-adapted state. To test the biological role of this residue, we generated a BlsA A29S derivative. Motility assays showed that the motility response of OR.A29S cells harboring pMU1222 (Table 4) was not significantly different from that displayed by the 17978 parental strain in either light or dark conditions (data not shown). These responses suggest that BlsA’s A29 residue is not essential for its activity, a possibility that is supported by the presence of an analogous A residue in the SnfB photoreceptor (Fig. 14). Taken together, all the observations presented above highlight the fact that BlsA shares some common structural and functional properties with different short BLUF-containing photoreceptors that are produced by unrelated bacteria, yet also exhibits some unique properties of its own.

Protein studies To further characterize the effect of the NT mutations described above, the blsA parental allele and the isogenic derivatives generated by site-directed mutagenesis were cloned in pET- 15b and overexpressed in E. coli BL21. Analysis of whole lysates of cells harboring pMU1254, which codes for a His-tagged BlsA recombinant product, showed that a protein overproduced upon IPTG induction reacted with anti-BlsA antibodies (Fig. 18, lane 3 in panels A and B). Although a significant amount of His-BlsA was present in the pellet after ultracentrifugation (data not shown), the supernatant contained enough recombinant protein that could be isolated for further analysis. The His-tagged BlsA derivative, which was purified to homogeneity by Ni-

93 affinity chromatography and SEC, displayed the characteristic 10-nm red shift upon blue light irradiation (Fig. 19A and inset) we reported before (4). Comparable spectra were produced by protein samples exposed for 3 min at 20, 100 or 200 μmol m−2 s−1 (Fig. 20). The spectral analysis also showed that purified lBlsA relaxed back to the dark-adapted state with a t1/2 of 10.2 min (612 sec) and a  rate of 14.7 min (881 sec) (Table 6), values that are comparable to those reported previously (54). The SEC experiments not only produced pure photoactive BlsA preparations, but also showed that purified dBlsA and lBlsA eluted as a major protein fraction displaying molecular masses of 17.7 kDa and 17.2 kDa, respectively, which match the predicted mass of the monomeric protein (Fig. 19B and inset). Analysis of different protein preparations produced the same outcome without the detection of dimers or larger BlsA multimers under the experimental conditions used to isolate and analyze overexpressed proteins. In contrast to the analysis of BlsA, the same overexpression and purification approach failed to produce meaningful results for the Y7A, Q51A, Y7A/Q51A and W92A His-tagged BlsA recombinant derivatives. Although SDS-PAGE analysis of whole cell lysates of E. coli BL21 transformants harboring pMU1273 (Y7A), pMU1255 (Q51A), pMU1271 (Y7A/Q51A) or pMU1298 (W92A) (Table 4) showed the presence of a protein reacting with anti-BlsA antibodies upon IPTG induction (Fig. 18, lanes 4-7), most of the recombinant proteins were present in the pellets collected after ultracentrifugation with little or no His-tagged BlsA derivatives recovered after Ni-affinity column chromatography (data not shown). This outcome aligns with the fact that the supernatant of E. coli BL21 cleared lysates prepared from cells producing these four BlsA derivatives did not turn yellow upon IPTG induction and did not produce the yellow band at the top of the Ni-agarose column that was readily visible with E. coli BL21 cell lysates containing His-tagged BlsA (data not shown). Although these results are in line with those observed during the analysis of the R. sphaeroides AppA flavoprotein, whose proper folding and consequent solubility depend on flavin binding (48), they contrast with the outcome of the overexpression of AppA126 (residues 1-126 harboring the AppA BLUF domain) and the AppA126 W104A mutant, both of which were obtained as soluble recombinant products of pTY-derivatives overexpressed in E. coli BL21 (125). Our results with the Q51A mutant also contrast those collected during the analysis of the analogous Tll078 Q50A derivative, which was overexpressed as a soluble N-His-tagged recombinant product in E. coli BL21 cells harboring the cognate pET-28a clone even when incubated at 37°C after IPTG induction (57). Our ability to

94 obtain a soluble His-tagged BlsA derivative using an overexpressing vector different from the pET-TEV plasmid we used in our initial report (4) and reproduce the outcome reported during the spectroscopic analysis of BlsA (54) strongly indicate that our failure to obtain soluble Y7A, Q51A, Y7A/Q51A and W92A His-tagged derivatives is not due to technical and/or procedural problems. Taken together, these observations indicate that the interaction of flavin with the Y7, Q51 and W92 residues is critical not only for light sensing and regulatory functions, but also BlsA stability, which seems to depend on its capacity to bind flavins. This is a potentially distinct property of BlsA when compared to related short BLUF-containing chromophores produced by unrelated bacteria.

95 Figure 18. Overexpression of His-tagged BlsA and derivatives in E. coli BL21. SDS-PAGE and immunoblot analysis of whole cell lysates of uninduced cells, lane 2, and induced cells harboring pMU1254 (parental BlsA), lane 3; pMU1273 (Y7A), lane 4; pMU1255 (Q51A), lane 5; pMU1271 (Y7A/Q51A), lane 6; or pMU1298 (W92A), lane 7. The BlsA amino acid changes coded for by each pET-15b derivative is indicated in parentheses. Lane 1, molecular weight markers. Total proteins were detected by Coomassie Blue staining (A) and BlsA was detected by immunoblotting with anti-BlsA polyclonal antibodies (B).

97 Figure 19. Analyses of His-tagged BlsA. (A) Absorption spectra of lBlsA and dBlsA. The UV-Vis light spectra were recorded using His- tagged BlsA purified by Ni-affinity chromatography and SEC. The purity of the protein sample used for spectral analyses was confirmed by SDS-PAGE using 4%-20% polyacrylamide gradient gels (inset). (B) Elution profile of lBlsA and dBlsA upon chromatography in a Superose 6 Increase 10/300 GL column. The inset shows the calibration curve constructed using the elution volume of the gel filtration standards cytochrome C (12.4 kDa), carbonic anhydrase (29 kDa), bovine albumin (66 kDa), alcohol dehydrogenase (150 kDa) and β-amylase (200 kDa). The black and blue triangles indicate the elution position of dBlsA and lBlsA, respectively. (C) HPLC analysis of heat-denatured BlsA supernatant. The retention times for FAD, FMN and Ribo are indicated in min.

99 Figure 20. Figure 20. BlsA responses to low and high light intensities. Difference of light minus dark spectra of purified WT BlsA protein samples illuminated at three different light intensities. Red line, 20 μmol m−2 s−1; blue line, 100 μmol m−2 s−1, black line, 200 μmol m−2 s−1.

100

101 Table 6 Kinetic values for light-to-dark recovery of His-BlsA and derivatives generated by site-directed point and deletion mutations

Protein Half-time – t1/2 (min/sec) Time constant –  (min/sec) His-BlsA 10.2/612 14.7/881 His-BlsA.K144E 5.2/312 7.4/446 His-BlsA.K145E 7.6/456 11.0/658 His-BlsA.Δ152-156 8.6/516 12.4/742 His-BlsA.Δ143-156 5.2/312 7.4/446

102 Analysis of BlsA C terminus Although BLUF domains of bacterial photoreceptors are highly conserved, their C- terminal amino acid sequences are highly divergent. In the case of BlsA, the primary sequence of this region has no significant similarity to known or characterized proteins that could predict its role in BlsA’s overall function. However, as it was described above, BlsA’s CT region might include not only the α3α4 components found in related bacterial BLUF photosensors, including

Tll0078, Slr1694 and SnfB (Figs. 14 and 15A), but also an additional α5 element (Fig. 14), which encompasses residues 144-150 according to the predicted tertiary structures shown in Fig. 15B. The absence of this additional α element in related bacterial chromophores may reflect potential structural and functional differences between BlsA and aforementioned BLUF-containing chromophores. It has been suggested that BlsA’s CT amino acids are involved in particular signal transduction and regulatory processes (4, 27, 38, 54), but their function has not been tested experimentally. Based on these observations, we decided to determine whether BlsA’s CT residues play a role in its biological functions.

Functional studies We started by testing the effect of deleting the last five CT amino acid residues on light- regulated motility. Figure 16B shows that the 17978.OR.Δ152-156 derivative, which was electroporated with pMU1232 (Table 4), displayed a light-dependent differential surface motility response comparable to that of the 17978 parental strain and the OR.W derivative complemented with the blsA parental allele. In contrast, the complementation of OR with pMU1235, which produced the 17978.OR.Δ143-156 BlsA derivative, resulted in a significant increase (P ≤ 0.0001) in motility under illumination when compared to the response of the OR.W, although to a much lower level than the response by the OR mutant (Fig. 16B). The most drastic change in surface motility responses upon illumination (P ≤ 0.0001), which reached the levels displayed by the OR strain, was detected with the complemented derivative OR.Δ135-147 that codes for the blsA:Δ135-147 in-frame deletion mutation (Fig. 16B). Taken together, these results indicate that the region located between residues A135 and N150, which includes our predicted α5 component (Fig. 14), is critical for BlsA functions. This possibility is further supported by the observation

103 that OR.K144E and particularly OR.K145E displayed a significantly increased surface motility responses (P ≤ 0.0001) under illumination when compared to OR.W (Fig. 16B). Although not shown, the complementation of 17978.OR with pMU1236 (blsA:Δ121- 135), which resulted in the OR.Δ121-135 derivative (Table 4), proved that the in-frame deletion of the predicted BlsA α4 component resulted in a motility phenotype comparable to that displayed by the 17978.OR BlsA mutant under darkness or illumination. Such a response is not surprising considering the predictable major structural changes caused by this deletion mutation.

Protein studies Overexpression of E. coli BL21 harboring pMU1264, pMU1258, pMU1282 or pMU1259 coding for the BlsA.152-156, BlsA.143-156, BlsA.K144E or BlsA.K145E derivatives, respectively (Table 4), resulted in readily visible yellow cell lysates that produced yellow-stained homogenous protein fractions upon Ni-affinity chromatography and SEC, the latter of which yielded data indicative of only protein monomers (data not shown). When analyzed by SDS- PAGE, these samples displayed the predicted molecular sizes (Fig. 21 insets). The top panels of Fig. 21 also show that the deletion of neither the last 5 (BlsA.152-156) nor 14 (BlsA.143-156) CT amino acid residues significantly affected the photo-responses of these BlsA recombinant derivatives upon illumination. Both of these derivatives displayed a red shift similar to that of the His-BlsA derivative upon illumination (compare panel A in Fig. 19 with top panels in Fig. 21). However, it is noteworthy that complementation of the OR strain with blsA:152-156 but not with blsA:143-156 fully restored the parental light-regulated motility response (Fig. 16B). Although motility responses of the OR.K144E and OR.K145E complemented strains, particularly the latter mutant, were significantly impaired when compared with the response of the OR.W strain (Fig. 16B), spectral analyses showed that the K144E and K145E point mutations did not affect the photo-response of these BlsA derivatives upon blue light illumination as reflected by a red shift similar to that displayed by the His-BlsA derivative (compare panel A in Fig. 19 with bottom panels in Fig. 21). In contrast to the spectral observations, analysis of recovery kinetics showed that the four BlsA CT mutations described above impacted this BlsA property when compared with the His- BlsA derivative. Table 6 shows that the K144E and 143-156 point and deletion mutations accelerated light-to-dark reversion kinetics to  values of 446 s, rates that represent an almost 2-

104 fold change when compared with His-BlsA. For the K145E ( = 658 s) and the 152-156 ( = 742 s) mutations, the reversion kinetics rates were accelerated by 1.3- and 1.2-fold, respectively, when compared to the BlsA His-tagged parental derivative. Taken together, these results indicate that CT residues play a role in the photodynamic properties of BlsA, although to a different degree as in the case of residues K144 and K145, which may play differential roles in the functional structure of BlsA including its interaction with flavins and photosensing functions at the NT region. Although not shown, SDS-PAGE analysis of E. coli BL21 lysate of cells harboring pMU1256 or pMU1257, which code for BlsA.121-135 and BlsA.135-147, respectively (Fig. 14), resulted in the overproduction and detection of the predicted protein products upon IPTG induction. However, the ultracentrifugation of the cell lysates of each of these overexpression strains produced samples that were not stained yellow and did not result in the detectable recovery of His-tagged proteins after Ni-affinity chromatography. Furthermore, SDS-PAGE analysis of the pellet collected after ultracentrifugation showed that most if not all of the recombinant proteins were located in the insoluble fraction of the cell lysates (data not shown).

These results indicate that the deletion of the α4 component or the adjacent 12-residue region, which includes four residues of the additional α5 component (Fig. 14A and 15B), results in insoluble BlsA derivatives. This outcome is most likely due to significant changes in BlsA structure that result in insoluble products when overexpressed upon IPTG induction under conditions used to isolate overexpressed His-tagged BlsA. Notably, the insolubility of these BlsA derivatives as well as those affecting the NT region of this protein must be taken into consideration in the analysis of the motility responses of bacteria encoding them.

105 Figure 21. Spectral analysis of purified His-tagged BlsA derivatives generated by site- directed deletion and point mutagenesis. The UV-Vis light spectra were recorded using His-tagged BlsA purified by Ni-affinity chromatography and SEC. The purity of the protein samples used for spectral analyses was confirmed by SDS-PAGE using 4%-20% polyacrylamide gradient gels (insets).

106

107 CD spectroscopy The potential effects of the point and deletion mutations described above on the predicted secondary structure and folding properties of BlsA and its derivatives were assessed by CD spectroscopy. This analysis showed that the spectra of dBlsA and lBlsA His-tagged derivatives were comparable and compatible with a protein harboring the α-helix and β-sheet components predicted to be present in BlsA’s tertiary structure (Fig. 22). Furthermore, the spectra of the dark- and light-adapted BlsA.K144E, BlsA.K145E, BlsA.152-156, and BlsA.143-156 point mutation and deletion derivatives, respectively, were not significantly different from each other (Fig. 22). This outcome is further supported by the observation that none of these mutations resulted in predicted secondary structures significantly different from that shown in Fig. 14 for the parental BlsA protein (data not shown). Taken together, these results indicate that the biological effects of these four BlsA mutations cannot be attributed to significant changes in the structure of the cognate translational products.

Chromophore analysis The capacity of BlsA to bind flavins was determined by HPLC analysis, which showed that the supernatant of heat-denatured His-tagged BlsA contains components with the same retention times displayed by the FAD, FMN and riboflavin standards (Figs. 19C and 23). It is apparent that FAD and FMN are the major flavin components bound to BlsA with riboflavin representing a minor component. The presence of these flavins in purified BlsA could be due to the replacement of FAD with FMN and the hydrolysis of the later flavin during overexpression in E. coli BL21, a phenomenon that was reported for the SnfB photoreceptor (121) and the R. sphaeroides AppA BLUF domain (126). The same HPLC analysis of the BlsA.K144E, BlsA.K145E, BlsA.152-156 and BlsA.143-156 derivatives showed that neither the residue substitutions at positions 144 and 145 nor the deletion of the last 5 and 14 amino acid residues impaired the binding of the flavin components listed above (Fig. S6) by the BlsA NT region. This observation is congruent with the fact that the responses of these proteins to illumination are comparable among themselves (Fig. 21) and with the His-tagged BlsA recombinant protein (Fig. 19).

108 Figure 22. CD spectra of dBlsA, lBlsA and related derivatives. Purified 3 µM samples of each protein dissolved in 10 mM phosphate buffer, pH 8.0 containing 20 mM NaCl were analyzed as described in Materials and methods.

109

110 Figure 23. HPLC of flavin standards. A volume of 10 µl of 100 µM stock solutions of FAD, FMN and Ribo was injected into a Waters Symmetry C18 reversed-phase column and eluted and detected as described in Materials and methods. Retention times (min) for each standard is indicated in the cognate panels.

111

112 Figure 24. HPLC of heat-denatured supernatants of purified His-tagged BlsA recombinant derivatives generated by site-directed deletions (BlsA.152-156 and BlsA.143-156) or point mutations (BlsA.K144E and BlsA.K145E). The retention times for each flavin component is indicated in minutes.

113

114 CONCLUDING REMARKS Our work has revealed unexplored BlsA structure-function relationships that are key for its biological functions. The hydrogen bonding network around the FAD chromophore, including the conserved Y7 and Q51 and the semiconserved W92 residues, is critical for A. baumannii light sensing; differential expression of light-controlled functions, such those involved in the interaction with semisolid surfaces as we previously reported (4, 127) and protein stability when overproduced in an unrelated bacterial host. Replacement of the Y7 and Q51 by A residues should abolish the interactions needed not only for electron- and proton-transferring activities, but also BlsA-flavin interactions required for proper protein folding and stability. The effect of the BlsA W92A mutation could be explained by the role the analogous W104 residue plays in the photochemistry of the AppA BLUF domain; the interaction of this residue with Q63, analogous to the BlsA Q51 residue, is critical for AppA light-sensing functions (128). Notably, Slr1694 activity depends on the interaction of Q50, which is the analogous to the BlsA Q51 residue, with an M residue located at position 93 (60). Taken together, these observations indicate that BlsA is functionally more similar to the BLUF domain of AppA than to the Slr1694 photoreceptor, although their photochemistries involving a flavin chromophore are similar. The analysis of the BlsA CT region also provided critical information. Although the last five amino acid residues do not play any photochemical and gene regulatory functions, the region limited by residues A and R located at position 135 and 151 (Fig. 14), respectively, proved to be critical not only for BlsA photodynamic properties, but also gene regulatory functions. The latter functions could depend on the interaction of BlsA with putative protein partners involved in downstream signal transduction pathways and corresponding regulatory processes that remain to be identified and characterized. This possibility is supported by the apparent effect that replacing the K144 or K145 charged residues has on light-regulated surface motility without significantly affecting flavin binding and photocycling. This hypothesis is further supported by the recent observation that BlsA interacts with the Fur iron-dependent regulator to control the transcriptional expression of A. baumannii genes coding for acinetobactin-mediated iron acquisition functions in a temperature-dependent manner (38). Confirming the proposed BlsA-Fur interactions and testing the iron response of the 17978 OR.K144E and OR.K145E mutants under conditions similar to those described by Tuttobene et al. (38) should provide critical information to understand these

115 interactions and shed some light on the mechanism(s) BlsA uses to regulate a wide range of cellular functions in response to illumination. ACKNOWLEDGEMENTS This work was supported by funds from National Institutes of Health Public Health grant R15GM117478-01 and by Miami University research funds. We are grateful to Drs. Carole Dabney-Smith and Rick Page and Mr. C. Paul New (Miami University Department of Chemistry and Biochemistry) for their help with column chromatography, CD assays and protein modeling. We also thank Dr. Andor Kiss and the Miami University Center for Bioinformatics and Functional Genomics for their assistance with column chromatography instrumentation and methods.

AUTHOR CONTRIBUTIONS CW: performed protein modeling, planned and performed experiments, analyzed and interpreted data; MS: planned and performed experiments, analyzed and interpreted data; NF: performed protein modeling and data interpretation; LAA: planned and performed experiments, analyzed and interpreted data, carried out figure design and preparation. All authors contributed to writing the manuscript.

116 Chapter 4

Analysis of the human pathogen Acinetobacter baumannii proteome in response to blue light exposure

Cecily R. Wood, Xin Wang and Luis A. Actis

117 ABSTRACT Acinetobacter baumannii is a nosocomial pathogen that affects immunocompromised individuals primarily in intensive care units and long-term care facilities. Temperature and light fluctuations within the hospital environment and on the host vary considerably, making these environmental signals important cues for A. baumannii to adjust its physiology to survive. Light is a ubiquitous environmental signal that A. baumannii perceives and responds to by regulating different cellular processes such as biofilm formation, motility and virulence. These sensory and regulatory responses depend on the production of the BlsA BLUF-containing photoreceptor that is produced and biologically active only in cells cultured at temperatures lower than 30°C. However, genetic and functional analyses of a mutant impaired in the expression of a pilus assembly system resulted in the unpredicted identification of light-regulated responses in cells cultured at 37°C, a condition in which BlsA biosynthesis and its light sensory activity are abolished. This observation indicates that A. baumannii senses and responds to light when cultured at 24°C and 37°C by BlsA-dependent and -independent mechanisms, respectively. This possibility was tested using a proteomics approach using an unlabeled LC-MS/MS protocol that not only showed the differential expression of 124 proteins at 24°C in response to illumination, but also demonstrated that light affected differentially the expression of 20 proteins in cells cultured at 37°C. The predicted functions of the differentially expressed proteins suggest that light affects a wide range of biological functions including gene regulation, metabolism, cell morphology and cell wall/membrane biosynthesis and stress responses. These observations strongly indicate that as a facultative pathogen that persists in different ecological niches in the nosocomial environment, A. baumannii senses and responds to light at different temperatures through complex mechanisms that are mostly unknown.

118 INTRODUCTION Acinetobacter baumannii is a Gram-negative pathogen that has recently been categorized as a critical medical priority for which antibiotics are urgently needed. Its extreme resistance to antibiotics makes it one of the greatest health risks among other priority pathogenic bacteria such as Pseudomonas aeruginosa and members of the Enterobacteriaceae, including Klebsiella and Escherichia coli (129). A. baumannii has a remarkable ability to acquire resistance to antibiotics that it encounters in the medical environment and persist for weeks under the selective pressure of disinfectants, desiccation and nutrient limitation (10, 20). Exposure to antimicrobials, combined with the limited release of useful antibiotics and ineffective hospital decontamination procedures (22), has led to the emergence of multi-drug resistant (MDR) and pan-drug resistant (PDR) strains worldwide (130, 131). Despite of the critical impact of MDR or PDR strains on human health, little is known about the host and environmental factors that modulate A. baumannii’s persistence and virulence. As a facultative pathogen that persists on the human host and on abiotic elements normally encountered in medical settings (18, 132), A. baumannii must sense and respond to a wide range of extracellular cues, including the presence of antibiotics and antiseptics such as ethanol (21, 133). Surprisingly, light was also found to be one of the signals to which A. baumannii responds, although it does not use it as an energy source. It was initially reported that the capacity of A. baumannii to form biofilms on plastics, display surface motility on semi-solid surfaces and kill fungal filaments at 24°C is affected by illumination through the light sensing and regulatory functions of the blue light sensing A photoreceptor protein (BlsA) (4). The global regulatory effects on BlsA-dependent responses to illumination together with data collected a using global transcriptional analysis (27, 39) and the identification of the light-regulated expression of the PrpABCD type I pilus assembly system (127) further confirmed the global effect of light on A. baumannii’s pathophysiology and the key role BlsA plays in these responses when bacterial cells are cultured at 24°C. However, the analysis of an A. baumannii ATCC 17978 (17978) prpA isogenic mutant led to the novel observation that this mutant differentially displays surface motility in response to illumination when bacteria are cultured at 37°C (127). Notably, growth of 17978 cells at this temperature not only results in a drastic reduction of blsA transcription and the lack of detection of its translational product, but also affects the photosensory functions needed to modulate the light-dependent responses originally detected

119 using cells cultured at 24°C (4, 39). Taken together, these observations indicate that 17978 cells express uncharacterized BlsA-independent light-regulated responses at 37°C that could play a role in the interaction of this pathogen with the human host. In this study, we aimed to understand the physiological changes of 17978 cells in response to blue light at 24°C or 37°C using an unlabeled global proteomics-based approach using liquid chromatography-tandem mass spectrometry (LC-MS/MS). Using this approach, we identified a total of 124 differentially expressed proteins in response to light at 24°C, whereas only 20 proteins were detected as differentially expressed at 37°C. These results not only validate our hypothesis that A. baumannii senses and responds to light in a temperature-sensitive manner, but also indicates that light sensing and consequent responses are indeed complex processes that are poorly understood.

120 MATERIALS AND METHODS Bacterial strains, culture conditions and chemical reagents A. baumannii ATCC 17978 was cultured on Luria-Bertani (LB) agar plates and passaged into swimming broth (10 g/L tryptone; 5 g/L NaCl) for all experiments. Each experiment was performed with three independent biological replicates using fresh bacterial cultures plated on LB agar from glycerol stocks maintained at -80°C. A. baumannii was grown at 37°C for 16-18 h (overnight) for routine growth and maintenance. All LC-MS/MS-grade chemicals, including

H2O, 0.1% TFA, acetonitrile and formic acid were purchased from Thermo Scientific. Iodoacetic acid was purchased from Sigma.

Sample processing Growth Conditions. Fresh cultures of A. baumannii were used to inoculate 3-mL SB starter cultures, which were then incubated overnight with shaking at 37°C. The following day, the starter cultures were diluted 100 times into 25 mL of fresh SB in 125-mL flasks and cells were grown as previously described (127). Briefly, the flasks were incubated under blue light or dark conditions at 37°C or 24°C for approximately 4 to 6 h until an OD600 of 0.8 was reached. Cells from the 25-mL cultures were pelleted by centrifugation at 4°C for 20 min at 2,000 x g, washed with sterile deionized water and stored at -80°C until further use.

Protein preparation and digestion. Protein samples were prepared and digested following a previously published protocol (134) with some modifications. Briefly, the bacterial pellets were thawed, resuspended in 700 µL of lysis buffer (50 mM Tris-HCl, 0.02% n-dodecyl-β-D- maltoside, DDM; pH 8.0) with Halt protease and phosphatase inhibitor cocktail (Thermo Scientific), and sonicated in an ice bath. Cell debris was pelleted by centrifugation at 4°C for 30 min at 21,000 x g and the supernatants were transferred to low protein binding microcentrifuge tubes. The protein concentration of the whole-cell lysate samples was determined by Bradford assay (111). Protein content was standardized to 100 µg prior to denaturation, reduction and alkylation with 8 M urea, 5 mM dithiothreitol (DTT) and 15 mM iodoacetamide (IAA), respectively. Following alkylation in the dark at room temperature for 30 min, 50 mM Tris-HCl (pH 8.0) was added to each sample to dilute the urea concentration to 2 M before treating the samples with trypsin (1 µg) overnight at 37°C.

121

Peptide Fractionation. Trypsin-treated protein samples were acidified with 0.1% formic acid (Buffer A) and applied to Buffer A-washed Sep-Pak C-18 Plus desalting columns (Waters Corporation). Peptides were eluted from the columns with 80% acetonitrile/0.1% formic acid (Buffer B), dried and dissolved in 0.1% TFA. Peptides were fractionated with reversed-phase high pH spin columns (Pierce) following the manufacturer’s protocol. Each fraction was dried and then resuspended in Buffer A. Samples were loaded into the wells of a 96-well microtiter plate and stored at -20°C until analysis by mass spectrometry.

Liquid chromatography-tandem mass spectrometry (LC-MS/MS) Each protein sample prepared as described above was further fractionated using a nanoLC Easy1000 (Thermo Fisher) instrument with a Thermo Scientific Acclaim PepMap RSLC column (C18, 50 µm x 15 cm, nanoViper, 2 µn, 100 Å) at a flow rate of 300 nL/min using a gradient of water containing 0.1% formic acid (Buffer A) and 80% acetonitrile, 0.1% formic acid (Buffer B) and directly injected into a LTQ Orbitrap XL mass spectrometer. The raw spectra were viewed using XCalibur before import and data processing through the proteomics PatternLab pipeline (PatternLab for proteomics 4.0).

Data processing with PatternLab Raw data were extracted with MakeMS2 (http://proteome.gs.washington.edu/). An A. baumannii sequence database was prepared using the ATCC 17978-mff genome downloaded from the National Center for Biotechnology Information (NCBI). A target decoy was prepared using this genome prior to running Comet, a MS/MS sequence search engine integrated into the PatternLab program. A 40-ppm mass tolerance and a post-translational modification of carbamidomethylation of cysteines were set a priori to the Comet search. Post-processing of the data was performed with SEPro included in PatternLab’s pipeline for data analysis. The search was limited to two missed tryptic digests and proteins with a minimum sequence length of 6 amino acids. For post-processing, only proteins with unique peptides, at least two spectra counts and less than 10 ppm were considered. These SEPro parameters led to final false discovery rates (FDR) of all experiments at the protein level to be less than 1%. A Student t test between the

122 light and dark samples at 24°C or 37°C resulted in the identification of differentially expressed proteins with p values of < 0.05 considered statistically significant.

RESULTS AND DISCUSSION Proteins Differentially Expressed by Light- and Dark-adapted Bacteria at 24°C To compare the proteome profiles of light- and dark-adapted 17978 bacterial cells, total proteins from whole lysates prepared from bacteria cultured under both growth conditions were isolated and analyzed using a label-free shotgun proteomics approach. A total of 124 proteins, which met both the fold-change and statistical criteria, were identified as differentially expressed by bacteria that were cultured in the absence or presence of light. Tables 7 and 8 list the 74 and 50 proteins whose abundance was significantly higher in cells cultured at 24°C in light or darkness, respectively. It is apparent that light regulates the expression of proteins predicted to be involved in a wide range of cellular functions including those reported and discussed here. Notably, these differentially expressed proteins include those predicted to play roles in regulation, metabolism and stress responses. Some of these differentially expressed proteins play a role in the physiology and virulence of many different bacterial pathogens.

Proteins with increased abundance in light compared to dark Regulatory proteins. Six regulators or regulator-type proteins had increased abundance in light-adapted bacteria at 24°C (Table 7), including the YebC/PmpR family DNA-binding transcriptional regulator (WP_000907229.1) and the two-component regulatory system (TCS) response regulator proteins WP_000101096.1 and the OmpR protein (WP_000060753.1). The YebC/PmpR protein belongs to the cl00361superfamily, whose members are implicated in the regulation of quorum sensing, biofilm formation and swarming motility responses in P. aeruginosa (135). The OmpR protein is involved in the regulation of OmpF and OmpC porins and surface structures such as curli, which mediate adherence properties in E. coli (136, 137). In A. baumannii, OmpR has been linked to motility and virulence (138), although little is known about its regulation of porins. The predicted WP_000101096.1 protein shares a receiver domain similar to the CheY, OmpR, NtrC, and PhoB response regulators, suggesting a role in regulation. Another regulator identified with increased abundance in light-adapted bacterial cells is the phosphoenolpyruvate synthase (PEP) regulatory protein (WP_000004354.1), or PsrP (Table 7).

123 This regulatory protein has been described in E. coli (139) and has roles in regulating the conversion of PEP into pyruvate or during gluconeogenesis, the anabolism of pyruvate into PEP. Orthologs of a LysR regulator (WP_001047619.1), which is related to transcriptional regulators involved in the catabolism of nitroaromatic/naphthalene compounds, and a Fis regulator (WP_001086304.1), which activates RNA promoters and controls the Hin-mediated DNA inversions, are also proteins with increased abundance in bacterial cells cultured under illumination. Taken together, these observations suggest that light has a global effect on A. baumannii and could control many pathways related to cell surface modifications, virulence, quorum sensing and other cell functions through the control of TCS and other regulatory proteins. Our observations also indicate that responses to light involve different transcriptional regulatory pathways that allow cells to respond and adapt to this ubiquitous extracellular signal.

Metabolic functions. Many metabolism-associated proteins differentially expressed in both the light- and dark-adapted 17978 bacteria were identified. Table 7 lists proteins that were in found in higher abundance in cells cultured in light including those involved in protein biosynthesis (WP_000048256.1, WP_000085212.1, WP_000937220.1, WP_001205031.1 and WP_000854895.1), lipid/fatty acid metabolism (WP_000195147.1, WP_001136744.1, WP_001273591.1), carbohydrate metabolism and energy production (WP_000737519.1, WP_049594827.1, WP_000542121.1, WP_000080757.1 and WP_000712680.1) and amino acid metabolism (WP_000037737.1, WP_000980437.1, WP_000218885.1, WP_000781240.1, WP_000406996.1, WP_000088559.1, WP_000956445.1, WP_001207983.1, WP_000129946.1. A tryptophan synthase subunit protein (WP_000088559.1) was also identified as upregulated in light-adapted bacterial cells. Tryptophan synthase is a multi-subunit enzyme that catalyzes the final reaction of the tryptophan biosynthetic pathway, producing glyceraldehyde-3-phosphate and tryptophan (140). Tryptophan synthase has been implicated in Chlamydia trachomatis survival during interferon-gamma persistence models (141) and the breakdown product of tryptophan, indole, has broad roles in bacterial cell-to-cell communication, chemotaxis and biofilm formation (142, 143). Interestingly, the production of indole 3-acetic acid is important for A. baumannii stress tolerance, virulence against A549 human epithelial cells and the interaction of this pathogen with plants, which could serve as a natural habitat (144).

124 Enzymes involved in the synthesis of tryptophan, alanine and isoprenoid compounds as well as aromatic compound metabolism were differentially expressed in response to illumination. Of particular interest is the phenylacetic acid (PAA) thioesterase PaaI (WP_001016790.1), which had higher accumulation in the light-adapted bacteria than the dark-adapted 17978. The paaI coding sequence is part of a 15-gene operon in A. baumannii that is involved in catabolism of phenylacetate, virulence and affecting neutrophil chemotaxis (145, 146). A septicemia mouse model demonstrated that a deletion of the gene coding for the early PAA enzyme PaaA did not attenuate the virulence of A. baumannii, but deletion of another early enzyme, PaaE, was attenuated virulence (145). It is hypothesized that the latter product may be responsible for the formation toxic epoxides during the metabolism of aromatic compounds (147). While the traditional metabolic roles of the protein products of the paa gene cluster have been described for E. coli (147), it has been suggested that these proteins are also linked to virulence in other organisms besides A. baumannii, including Mycobacterium abscessus (148) and Burkholderia cenocepacia (149). Taken together, the observation of an accumulation of metabolism-associated proteins suggest that light has a significant impact on A. baumannii metabolic functions, some of which are associated with its virulence. Among the proteins with higher abundance in light-adapted bacteria is an alanine racemase (WP_000769882.1) (Table 7). Two alanine racemase enzymes have been identified to date and are involved in the conversion of L- to D-alanine (Alr) or the catabolism of D-alanine (DadX) (150). Another alanine racemase was also identified in the proteins that were downregulated in the light-adapted bacteria (WP_001137871.1) (Table 8). The WP_000769882.1 racemase identified in light-adapted cells is 430 amino acids in length, while the racemase that was identified as downregulated in the light-adapted cells (WP_001137871.1) is only 367 amino acids in length. Although both proteins have the identical type III pyridoxal 5’-phosphate fold characteristic of alanine racemase enzymes (150), and they are both annotated as alr/dadX-like, further validation is required to determine the functional difference between these racemase enzymes in A. baumannii. Because of their role in bacterial peptidoglycan biosynthesis, these enzymes have gained attraction over the years as possible antibiotic targets, even in A. baumannii, as humans have no protein homologs to the dadX-/alr-encoded proteins (150, 151). It is possible that these enzymes might play two different roles in A. baumannii depending on the light condition, as alanine racemases have been proposed to have either a

125 catabolic role or biosynthetic role for cell wall formation (151) as reported for Salmonella typhimurium and E. coli (152). Light exposure also increased the abundance of proteins involved in LPS (WP_000064877.1) and cell wall (WP_000522216.1, WP_000277821.1 and WP_000557503.1) metabolism, a response that may reflect the differential capacity of 17978 cells to form biofilms on abiotic surfaces (97) and display surface motility (153) upon illumination. The abundance of these proteins could also account for the increased virulence of A. baumannii after blue light exposure (4), considering the role of LPS in the virulence of A. baumannii (17).

Adaptation-related functions. An ortholog of OsmC (WP_001195082.1) was identified as an upregulated protein in light-adapted bacteria (Table 7). In E. coli, this protein is produced in response to changes in osmotic pressure associated with cell growth from two overlapping promoter regions (154). The OsmC family proteins also play a role in oxidative stress resistance in Gram-negative and Gram-positive bacteria (155). Light also induced the accumulation of an ortholog of the osmotically-inducible protein OsmY (WP_000919363.1), which harbors a predicted BON domain of unknown function that has been associated with outer membrane/periplasmic proteins involved in protection against osmotic shock (156). Additionally upregulated in light-adapted cells was an enzyme annotated as a 2-C- methyl-D-erythritol 2,4-cyclodiphosphate synthase (WP_000226512.1). This protein is encoded by the gene annotated as ispF and has a role in isoprenoid synthesis in the 2C-methyl-D- erythritol 4-phosphate (MEP) pathway. The MEP pathway has been proposed to produce antioxidants in some organisms in response to oxidative pressures, including the pathogens Brucella abortus (157) as well as corynebacteria and mycobacteria (158). In some pathogenic prokaryotes, components and products of the MEP pathway have a role in virulence by enhancing bacterial survival inside a host by combatting reactive oxygen species (158, 159). Interestingly, there were proteins related to natural competence/transformation/DNA uptake (WP_001235986.1), plasmid maintenance (WP_000757214.1) and DNA partition (WP_000023426.1) that were accumulated in higher amounts in cells cultured in light than in darkness. Based on this observation and the capacity of A. baumannii to acquire foreign DNA(95, 160, 161), it is tempting to speculate that illumination may contribute to the genomic plasticity of A. baumannii and its capacity to adapt to the different conditions this pathogen

126 encounters in the host and the nosocomial environment by acquiring genetic traits coding for adaptation responses.

127 Table 7 Proteins with higher abundance in light compared to dark at 24°C Protein ID Fold Change p value Description WP_000626048.1 3.6733709 1.00E-05 Hypothetical WP_000863328.1 2.2800854 0.00051358 Hypothetical WP_000037737.1 2.1081568 0.00093436 Glutamine amidotransferase WP_000644343.1 1.9940236 0.01262504 DUF4142 domain-containing protein WP_000106721.1 1.9566537 7.89E-05 Purine NTP pyrophosphatase WP_000102943.1 1.8523942 0.00242327 Oxidoreductase WP_000195147.1 1.8366855 1.00E-05 1-acyl-sn-glycerol-3-phosphate acyltransferase WP_000737519.1 1.8240683 0.00141457 Galactose-1-epimerase WP_000858381.1 1.80425 0.01427467 Tol-Pal system beta propeller repeat protein TolB WP_001235986.1 1.7981074 0.02206706 MBL fold metallo-hydrolase WP_000222569.1 1.746396 0.0019165 Porphobilinogen synthase WP_001247947.1 1.7406726 0.01674103 Hydroxymethylpyrimidine/phosphomethy lpyrimidine kinase WP_000980437.1 1.7396039 0.0006288 Aspartate ammonia-lyase WP_049594827.1 1.7107648 0.00020764 Rieske (2Fe-2S) protein WP_000877608.1 1.6972178 0.00413948 YciK family oxidoreductase WP_000064877.1 1.6581024 0.0002624 acyl-ACP-UDP-N-acetylglucosamine O- acyltransferase WP_000048256.1 1.6316134 0.01265847 50S ribosomal protein L28 WP_000085212.1 1.6027652 0.00349015 Glycine-tRNA ligase subunit alpha WP_001115787.1 1.5862988 0.02891854 Hypothetical WP_000216076.1 1.5621308 0.03931782 Phosphoadenylyl-sulfate reductase WP_000218885.1 1.5526872 0.00334785 Histidinol-phosphate aminotransferase WP_001140455.1 1.5305712 1.00E-05 Short chain dehydrogenase WP_001016790.1 1.5305712 1.00E-05 Phenylacetate-CoA oxygenase subunit PaaI WP_000101096.1 1.5305712 1.00E-05 TCS Response regulator WP_000099436.1 1.5225542 0.00847113 Oligoribonuclease WP_000757214.1 1.5204169 0.03978405 Entericidin, EcnA/B family WP_002009385.1 1.5202248 0.01193147 Peptidylprolyl isomerase A WP_000023426.1 1.5197773 0.01884353 Chromosome partitioning protein WP_001047619.1 1.519434 0.02839089 LysR family transcriptional regulator WP_000769882.1 1.4854566 0.02363639 Alanine racemase WP_000907229.1 1.484414 0.01277394 YebC/PmpR family DNA-binding transcriptional regulator WP_000046156.1 1.4790092 0.02811306 Hydrolase WP_001086059.1 1.4782275 0.02459293 Oxidoreductase WP_000781240.1 1.4722992 0.00649795 Maleylacetoacetate isomerase

128 WP_000206127.1 1.4333501 0.02278365 Hypothetical WP_000156920.1 1.4303091 0.00733511 CBS domain-containing protein WP_001086304.1 1.424703 0.0246632 DNA-binding transcriptional regulator Fis WP_000406996.1 1.4245338 0.026623 Bifunctional phosphoserine phosphatase/ homoserine phosphotransferase ThrH WP_000964178.1 1.4245338 0.026623 NADPH-dependent oxidoreductase; flavoprotein WP_000781338.1 1.4212306 0.00098109 Glutathione S-transferase WP_000567434.1 1.3906706 0.03472193 Chaperone protein WP_000088559.1 1.383837 0.04402478 Tryptophan synthase subunit alpha WP_000522216.1 1.3822017 0.02354936 Peptidoglycan-binding protein LysM WP_000956445.1 1.3779184 0.00076037 Type II asparaginase (L-) WP_000169602.1 1.3775141 1.00E-05 Hypothetical WP_000825868.1 1.3775141 1.00E-05 Hypothetical WP_000798673.1 1.3775141 1.00E-05 Hypothetical WP_001269278.1 1.3775141 1.00E-05 DUF1049 domain-containing protein WP_001207983.1 1.3742986 0.02319295 Ornithine carbamoyltransferase WP_000537113.1 1.3679851 0.03270243 ATP-binding protein WP_000542121.1 1.3612539 0.01584885 NADP-dependent isocitrate dehydrogenase WP_000277821.1 1.3284989 0.0319161 Lytic transglycosylase WP_000975113.1 1.3187246 0.00219486 Hypothetical WP_000712680.1 1.3185836 0.03277265 Gfo/Idh/MocA family oxidoreductase WP_000623102.1 1.3141794 0.00098291 NAD-dependent malic enzyme WP_001143885.1 1.3087104 0.02009696 Hypothetical WP_025467431.1 1.3041441 0.00014051 M3 family peptidase WP_000937220.1 1.3033229 0.00403179 30S ribosomal protein S16 WP_000080757.1 1.2994384 0.00010747 Bifunctional 3-demethylubiquinone 3-O- methyltransferase WP_000060753.1 1.2906875 0.04041091 Two-component system response regulator OmpR WP_001136744.1 1.2897629 0.02394952 Acyl-CoA dehydrogenase WP_000557503.1 1.2872781 0.02127764 UDP-N-acetylmuramoyl-tripeptide-D- alanyl-D-alanine ligase WP_000226512.1 1.2872781 0.02127764 2-C-methyl-D-erythritol 2,4- cyclodiphosphate synthase WP_001273591.1 1.2856798 1.00E-05 Acyl-CoA thioesterase WP_000919363.1 1.2830736 0.00626094 BON domain-containing protein WP_000129946.1 1.277098 0.02744196 Aspartate aminotransferase family protein WP_001043187.1 1.258674 0.01643316 Hypothetical WP_001205031.1 1.258674 0.01643316 50S ribosomal protein L33 WP_000696025.1 1.256842 0.0088266 Hypothetical WP_001195082.1 1.224457 1.00E-05 OsmC family protein

129 WP_000086443.1 1.224457 1.00E-05 Hypothetical WP_000698062.1 1.1479284 1.00E-05 Membrane protein WP_000004354.1 1.1101664 0.00031031 Phosphoenolpyruvate synthase regulatory protein WP_000854895.1 1.0648553 6.77E-05 50S ribosomal protein L13

130 Proteins with decreased abundance in the light compared to dark Regulatory proteins. The carbon storage regulator CsrA (WP_000906487.1), was identified as the only regulatory-like protein that was downregulated in light-adapted bacteria (Table 8), having a 1.45-fold change, compared to the dark-adapted state. The CsrA protein regulates global changes in E. coli through catabolite repression, cell surface modifications that modify biofilm properties and stress survival (162). Changes in the abundance of the A. baumannii CsrA ortholog in response to illumination may reflect the light-dependent biofilm responses we described for this pathogen when cultured at 24°C (11, 14).

Metabolic functions. Illumination also resulted in the reduced abundance of proteins involved in different metabolic process and functions associated with membrane and cell wall biosynthesis, and cell division processes (Table 8), although this effect was limited when compared with the number and type of proteins whose abundance was increased in light-adapted bacteria (Table 7). Table 8 also includes a glutathione peroxidase ortholog (WP_000066032.1), which could be involved in protection against reactive oxygen species (ROS) and may even contribute to virulence as demonstrated for some pathogens, including Streptococcus pyogenes and Candida albicans (163). All these observations indicate that light has a global effect on the expression of proteins that play different roles in cell metabolism and physiology. Interestingly, light reduced the accumulation of the WP_001131392.1 and WP_001121174.1 proteins, which represent a putative bifunctional 3,4-dihydroxy-2-butanone-4- phosphate synthase/GTP cyclohydrolase II (DHBP synthase, RibB) and a GTP cyclohydrolase II (RibA), respectively. These two enzymes are part of the biosynthetic pathway that results in the biosynthesis of riboflavin (164), the flavin precursor used by A. baumannii to sense and respond to light via the BlsA photoreceptor protein (4, 39).

Adaptation-related functions. As described in the previous section of this manuscript, light affects the expression of proteins involved in the MEP pathway, inducing the accumulation of a predicted 2-C-methyl-D-erythritol 2,4-cyclodiphosphate synthase (WP_000226512.1) ortholog coded for by ispF (Table 7). However, light seems to negatively affect this pathway by reducing the abundance of the IspG enzyme (WP_000572095.1) in cells grown in light. Interestingly, this enzyme depends on electrons donated by a ferredoxin protein source (165),

131 which could be represented by a predicted ISC system 2Fe-2S type ferredoxin (WP_001137383.1). Notably, this predicted ortholog, which could play a role in isoprenoid biosynthesis, is also reduced in light-adapted bacteria. Taken together, these observations suggest that the effect of light on the virulence of A. baumannii, an effect reported when we described the identification and analysis of BlsA (4), could be due at least in part to the differential accumulation of isoprenoids by this pathogen in response to illumination, a possibility that remains to be determined experimentally. The expression of a bacterioferritin ortholog (WP_001214863.1) was reduced in the light-adapted bacteria (Table 8). Bacterioferritins are iron storage proteins that were originally classified as cytochrome b1 proteins in E. coli (166). In A. baumannii, bacterioferritin is also involved in iron storage and is induced when the iron concentration is abundant (167). In addition to the iron-storing bacterioferritin, a heme-binding flavoprotein was also identified with decreased abundance in the light (Table 8), which could also contribute to regulating the intracellular iron concentration of A. baumannii. These observations suggest that illumination results in a potential reduction of intracellular iron levels that requires a coordinated functional adaptation to persist under iron-limiting conditions imposed by the presence of light.

132 Table 8 Proteins with decreased abundance in the light compared to dark at 24°C Protein ID Fold Change p value Description WP_001228338.1 3.26675425 1.00E-05 Membrane protein WP_000477056.1 3.05644405 0.00238694 Hypothetical WP_000248340.1 2.18497281 1.00E-05 Hypothetical WP_000615571.1 2.11776382 3.85E-05 N-acetyltransferase WP_001137383.1 2.11084669 0.00050918 ISC system 2Fe-2S type ferredoxin WP_000284347.1 1.96951737 0.00435802 16S rRNA (adenine(1518)- N(6)/adenine(1519)-N(6))- dimethyltransferase RsmA WP_000015937.1 1.8374674 0.01233808 Guanylate kinase WP_000923478.1 1.75130611 0.01599013 Diaminopimelate epimerase WP_000572095.1 1.70567156 0.0235883 4-hydroxy-3-methylbut-2-en-1-yl diphosphate synthase WP_000917698.1 1.63804801 1.00E-05 Serine o-acetyltransferase WP_000680696.1 1.63337712 1.00E-05 3-deoxy-manno-octulosonate cytidylyltransferase WP_000894500.1 1.63337712 1.00E-05 Hypothetical WP_000351255.1 1.63337712 1.00E-05 Hypothetical WP_000627259.1 1.63337712 1.00E-05 Peptidylprolyl isomerase WP_000987637.1 1.63337712 1.00E-05 Cytochrome b WP_001259423.1 1.62244196 0.00157473 DUF541 domain-containing protein WP_000108826.1 1.59092956 0.03924548 Ribonuclease WP_000132118.1 1.57443135 0.01976443 3-oxoacyl-ACP reductase WP_001215920.1 1.50445095 0.02606131 Acetolactate synthase small subunit WP_002027464.1 1.46133474 0.00427248 Hypothetical WP_000240312.1 1.45189078 1.00E-05 DUF2058 domain-containing protein WP_000906487.1 1.45189078 1.00E-05 Carbon storage regulator CsrA WP_000778126.1 1.4505737 0.01253082 Hypothetical WP_000842362.1 1.44696679 0.00566977 Outer membrane lipid asymmetry maintenance protein MlaD WP_001137871.1 1.41303728 0.0133454 Alanine racemase WP_000276693.1 1.40723113 0.00291711 NADH-quinone oxidoreductase subunit I WP_000777195.1 1.40380138 0.01939063 Signal peptide protein WP_000460625.1 1.39755444 0.04109557 Signal recognition particle protein WP_001216787.1 1.3611476 1.00E-05 2-nonaprenyl-3-methyl-6-methoxy-1,4- benzoquinol hydroxylase WP_000003719.1 1.3611476 1.00E-05 Cell division protein ZapA WP_000770719.1 1.3611476 1.00E-05 Heavy metal-associated domain protein WP_000047940.1 1.34875458 0.03892386 Acetyl-CoA C-acyltransferase WP_001183739.1 1.34271196 0.0023238 Fatty acid-CoA ligase WP_001138119.1 1.31574021 0.019539 30S ribosomal protein S19

133 WP_000731729.1 1.31358214 0.02766907 Putative porin WP_000066032.1 1.31358214 0.02766907 Glutathione peroxidase WP_000433617.1 1.30057588 0.00142468 Inositol monophosphatase WP_000213740.1 1.29243861 0.02345881 Argininosuccinate lyase WP_001075972.1 1.27952725 0.02176459 Tyrosine protein kinase WP_001131392.1 1.23306694 0.00847657 Bifunctional 3,4-dihydroxy-2-butanone-4- phosphate synthase/GTP cyclohydrolase II WP_001121174.1 1.23306694 0.00847657 GTP cyclohydrolase II WP_000070912.1 1.22976148 0.00386118 30S ribosomal protein S10 WP_000205997.1 1.21560742 0.00549957 Integration host factor subunit Beta WP_000188888.1 1.09561054 1.49E-05 Flavohemoprotein WP_001094391.1 1.09248641 0.00010474 DUF2171 domain-containing protein WP_001025680.1 1.09248641 0.00010474 N-acetyltransferase WP_001214863.1 1.09203201 9.13E-05 Bacterioferritin WP_000056784.1 1.08891808 1.00E-05 Hypothetical WP_001006517.1 1.08891808 1.00E-05 Hypothetical WP_000564581.1 1.08891808 1.00E-05 Hypothetical

134 Proteins Differentially Expressed in Light- and Dark-adapted Cells at 37°C Considering our previous findings (127) and that the BlsA photoreceptor is not active as a light sensor and its production is negligible at 37°C (4, 39), we speculate that light-mediated regulation occurs in A. baumannii cells cultured in dark or light conditions when grown at this temperature. The proof-of-concept experiments described here, which were conducted in triplicate and performed as described above for the light- and dark-adapted bacterial cells cultured at 24°C, showed that indeed a total of 20 proteins were differentially expressed in response to light and met the TFold test threshold established for this purpose. Nine proteins were upregulated in the light condition (Table 9) in contrast to the 11 proteins downregulated in light-adapted bacterial cells cultured at 37°C (Table 10). Although the number of differentially expressed protein in response to light is smaller than those detected in bacteria cultured at 24°C (Tables 7 and 8), they also represent different cellular functions.

Proteins with increased or decreased abundance in light compared to dark Regulatory proteins. Only one regulator-like protein was found in higher abundance in light-adapted bacteria at 37°C, the PhoB response regulator (WP_000650776.1) of the PhoBR phosphate two-component transcriptional regulatory system (168). In A. baumannii, PhoB is involved in the regulation of the pho regulon (169). Products of the pho regulon control the phosphorus assimilation response and are induced during low extracellular phosphate concentration (169), similar to the E. coli Pho system (168). Interestingly, a comparative subtractive proteomics approach based on the computational analysis of completely annotated A. baumannii genomes, which was used to identify broad-spectrum antimicrobial candidates, resulted in the identification of PhoB as a candidate target for novel antibiotics (170). This observation together with the role of this regulator in bacterial pathophysiology and its differential expression in response to light indicate that this protein could be used as an alternative target to treat A. baumannii infections, particularly those involving multi-drug isolates. Equally interesting is the CpdA phosphodiesterase (WP_000143915.1), which was found in higher abundance in bacterial cells in response to blue light (Table 9). This enzyme plays an important regulatory role in modulating the intracellular concentration of cAMP, thereby influencing cAMP-dependent processes, some of which are related to bacterial persistence and virulence (171).

135 Metabolic functions. Tables 9 and 10 show that proteins involved in fatty acid biosynthesis (WP_000815905.1, WP_000354611.1), protein production (WP_001118149.1, WP_001288187.1), polysaccharide biosynthesis (WP_000872590.1), electron transport/energy production (WP_000071881.1, WP_001274550.1) glyoxylate and dicarboxylate metabolism (WP_000532984.1), histidine degradation (WP_001016890.1) and deoxyribonucleotide biosynthesis (WP_000111540.1) were differentially expressed in response to illumination at 37°C. The identification of WP_000826956.1 as an upregulated protein that is predicted to represent a PglC ortholog is interesting. The gene coding for this protein is part of the A. baumannii K locus that is critical for the ability of the ATCC 17978 strain to produce capsule, form biofilms on abiotic surfaces and display virulence using a murine septicemia model (172). Notably, this functional group includes proteins differentially expressed by cells exposed to light at 24°C and 37°C. WP_000037737.1 represents a putative glutamine amidotransferase ortholog, which was another protein found with lower abundance in light adapted cells cultured at 24°C (Table 8). This activity is associated not only with glutamine amidotransferase, but also other biosynthetic enzymes including an anthranilate synthase, which is involved in the biosynthesis of anthranilate, pyruvate and L-glutamate, and a carbamoyl phosphate synthetase (CPSase), which is involved in arginine and pyrimidine biosynthesis. The WP_001131392.1, which represents a GTP cyclohydrolase II ortholog potentially involved in riboflavin metabolism, was identified as a protein downregulated by light-adapted cells cultured at 24°C (Table 8) but abundant at 37°C. Whether these differences in the synthesis of proteins associated with the production of flavin, particularly the precursor riboflavin, at low and high temperatures reflect different demands for this chromophore to sense and respond to light is a possibility that remains to be explored experimentally. Although a model involving a conserved RNA secondary structure named the RFN element has been implicated in the translational regulation of riboflavin biosynthetic genes in Gram-negative bacteria, including Acinetobacter calcoaceticus (173), the role of light in the expression of these genes using FAD as a chromophore is unknown. WP_000781240.1, which is related to a putative maleylacetoacetate isomerase predicted to be involved in the catabolism of phenylalanine and tyrosine, and WP_000769882.1, which represents an alanine racemase, are also proteins that are differentially expressed by light- adapted cells at both temperatures. At 24°C and 37°C, both proteins are up- and downregulated,

136 respectively, an observation that may reflect roles these enzymes play in A. baumannii’s physiology under different environmental conditions.

Adaptation-related functions. Among the 11 proteins identified as decreased in abundance in light-adapted bacteria was a TonB-dependent siderophore receptor protein (WP_005135700.1). This protein was annotated as an ortholog of FepA, the outer membrane receptor for the enterobactin siderophore (174). The genome of A. baumannii ATCC 17978 is proposed to encode 22 siderophore receptors (5), an observation that indicates that this pathogen has the capacity to use not only the siderophores it produces, but also siderophores produced by other bacteria (xenosiderophores), particularly in mixed infections involving different pathogens. Light also affected the expression of proteins that could allow bacterial cells to change cell shape by reducing the biosynthesis of a putative MreB ortholog (WP_000601379.1) (Table 10). MreB is an actin-like rod-shape determining protein in Gram-negative bacteria (175). The abundance of the replication protein DnaA (WP_000964767.1) (Table 9) was also significantly increased in the light-adapted bacteria.

137 Table 9 Proteins with increased abundance in light compared to dark at 37°C Protein ID Fold Change p value Description WP_000826956.1 2.1672537 0.00456094 Sugar transferase WP_000815905.1 1.7674221 0.00962135 3-oxoacyl-ACP reductase WP_000964767.1 1.7573768 0.03753615 DnaA WP_000071881.1 1.7299008 0.03849445 Flavodoxin family protein WP_000037737.1 1.6011717 0.0279499 Glutamine amidotransferase WP_000143915.1 1.526363 0.04032689 3',5'-cyclic-AMP phosphodiesterase CpdA WP_000532984.1 1.5230209 0.03585685 Hydroxypyruvate isomerase WP_000650776.1 1.5166688 0.02034369 PhoB WP_001131392.1 1.4696644 0.01764894 GTP cyclohydrolase II

138 Table 10 Proteins with decreased abundance in the light compared to dark at 37°C Protein ID Fold Change p value Description WP_000781240.1 1.74220486 0.01076639 Maleylacetoacetate isomerase WP_001118149.1 1.63355173 0.02336616 23S rRNA WP_000872590.1 1.62189881 0.02975696 Outer membrane protein WP_005135700.1 1.48264239 0.00997108 TonB-dependent siderophore receptor WP_001274550.1 1.46484817 0.01178564 NADH quinone oxidoreductase subunit A WP_001288187.1 1.41962171 0.008347 Ribonuclease G WP_000601379.1 1.37284044 0.01322429 MreB WP_000769882.1 1.35162418 0.00651326 Alanine racemase WP_000354611.1 1.33192966 0.00065171 Acetyl-CoA carboxylase biotin carboxyl carrier protein WP_001016890.1 1.25435024 0.00602438 Formimidoylglutamase WP_000111540.1 1.05500269 0.0004736 Ribonucleoside-diphosphate reductase subunit alpha

139 CONCLUDING REMARKS We aimed to use proteomics as a tool to understand the differential protein expression and cellular responses that occur past the transcriptional changes that have been reported previously (27). While RNA-Seq approaches are useful tools for understanding the initial global response A. baumannii has to light, they do not reveal post-transcriptional changes or protein modifications and therefore might not be an overall representation of the total cellular changes that occur after blue light illumination. Based on our preliminary analyses, we determined that light affects the expression of 124 proteins involved in global regulation, fatty acid and amino acid metabolism, DNA uptake, cell shape and virulence at 24°C. These findings are in agreement with a previous report of the light response of A. baumannii described by Muller et al. using a global transcriptional approach (27). In support of our previous observations (127), which suggested the expression of BlsA- independent light-mediated responses, we demonstrate that light regulation also occurs at 37°C, albeit to a lesser extent than at 24°C. Twenty proteins were differentially expressed in the light and were predicted to have roles in biofilm formation, capsule formation, virulence, iron acquisition and cell shape. Although the total differentially expressed proteins are fewer than those identified at 24°C, it is noteworthy that the proteins expressed at 37°C belong to different functional groups and potentially have roles specifically for A. baumannii adapted to light and dark conditions at 37°C. Further experimental validation of these observations is necessary as the proteomics experiments performed for this analysis were proof-of-concept to support our original hypothesis and report (127) that light regulation occurs at 37°C independently of BlsA. In summary, our proteomics analysis contributes to our current understanding of the ability of A. baumannii to sense and respond to blue light at 24°C, while also providing new information regarding the physiological changes that may allow this organism to persist in different environments at 37°C, such as the human host. The results presented in this report support our current understanding that blue light and BlsA act as a global regulator, but also demonstrate that A. baumannii possess light sensing and regulatory mechanisms other than the BlsA photoreceptor that are active at 37°C. Further biological assays and experiments are required to elucidate these BlsA-independent regulatory pathways, some of which could potentially be active in other bacteria including other facultative pathogens that persist outside the human host and the medical environment.

140 ACKNOWLEDGEMENTS This work was supported by National Institutes of Health Public Health grant R15GM117478-01 and Miami University research funds. We are thankful to Dr. Theresa Ramelot for her help with the liquid chromatography-tandem mass spectrometry experiments and analysis.

141 Chapter 5

Summary Acinetobacter baumannii, a causative agent of both nosocomial- and community- acquired pneumonia (13) and bacteremia due to wound infections (176) is an increasing public health concern. Due to its ability to acquire resistance to many current antibiotics, the rise of multi- and pan-drug resistant strains of A. baumannii have increased over the last decade (19, 177), demonstrating the urgent need for the development of new antimicrobial therapies. Also contributing to its success is A. baumannii’s remarkable capacity to survive on inanimate objects for long periods of time under several environmental pressures including changes in humidity, temperature, and nutrients (1, 178, 179). Its ability to sense and respond to multiple environmental cues may contribute to its success as a facultative pathogen and promote its survival outside of the human host having been isolated from soil, water samples and animals (78, 80, 81). The ability to respond to a myriad of environmental signals, such as light and temperature, may allow A. baumannii to differentially regulate biofilm and motility functions, physiological responses that are important for its interaction with its surroundings and for survival and persistence under different environmental pressures, especially outside of the human host (1, 18, 101, 180). Because the life cycle and basic ecology of A. baumannii, the most medically relevant member of the Acinetobacter genus, is poorly understood, we sought to characterize the role of light in controlling cellular responses at two different temperatures. Specifically, we hypothesized that these cellular responses do not depend solely on the activity of the BlsA photoreceptor and transcriptional regulator. Taken together, the results presented in this work provide novel insights regarding how the short photosensory protein BlsA regulates cellular responses such as motility and adherence characteristics that might contribute to virulence and persistence at two clinically relevant temperatures, 24°C and 37°C. Our findings also indicate that A. baumannii is capable of responding to blue light independent of the BlsA photoreceptor at 37°C and that these responses have a key role in the regulation of A. baumannii’s physiology.

BlsA as a photoreceptor and regulator Our work shows that BlsA, in contrast to other BLUF photoreceptors, functions as a monomer to exert its regulation of motility (Ch. 3) and biofilm (data not shown). The work

142 presented in Ch. 3 also supports our previous observations (4) that BlsA undergoes a typical red shift in its absorbance spectrum, indicative of a light-excited state (4, 109). This red shift is the result of the conserved tyrosine-glutamine pair of the BLUF domain involved in the photoactivation of BlsA. The signal is proposed to propagate through a semi-conserved tryptophan (W92) just N-terminal to the β5 strand of the protein (54). We show that without these residues, BlsA is not stable and further characterization of the protein is not possible due to solubility issues most likely due to the unique role that flavin binding plays in the overall active structure of this photoreceptor. Interestingly, other short BLUF-containing photoreceptors with mutations in analogous residues have been overexpressed and purified as soluble proteins (57, 181), whereas the AppA “long” photoreceptor depends on flavin binding for solubility (48) similar to BlsA. Whereas the function of the N-terminal BLUF domain is well conserved among the BLUF family of photoreceptors, little is known regarding the divergent C-terminal regions of these proteins. It has been proposed that each BLUF domain-containing photoreceptor has evolved protein-specific functions for protein-protein interactions such as oligomerization or DNA-binding (55, 57). For BlsA, we determined that the CT five amino acids are dispensable for its function as a regulator of motility. Deletion of 14 residues from the CT region, which includes residues K144 and K145, however, impacts the regulatory function of BlsA, indicated by a loss of light-regulated motility. It is unclear whether the K144 and K145 residues are significant for BlsA’s ability to interact with other proteins to exert its regulatory function considering that the C-terminal region of BlsA lacks recognizable DNA binding motifs. Tuttobene et al. (38) recently proposed that BlsA interacts with Fur in a light- and temperature-dependent manner to regulate genes involved in iron acquisition. More work with the K144 and K145 BlsA derivatives using the experimental conditions reported by Tuttobene et al. (38) will be necessary to determine if these BlsA residues are indeed involved in the Fur-BlsA interactions and how these interactions occur to carry out BlsA’s regulatory functions. Furthermore, point mutation of K144 or K145 had no effect on the ability of these BlsA derivatives to bind flavin or undergo a typical BLUF photocycle, comparable to the His-tagged WT BlsA derivative (Ch. 3). The K144E and K145E point-mutant derivatives could be purified to homogeneity, in contrast to the Y7A, Q51A or W92A BLUF mutant derivatives.

143 Light regulates motility, biofilm and virulence In support of our previous findings (4), analyses of the 17978 strain at 24°C demonstrated that motility and biofilm formation are enhanced in the dark compared to bacterial cells cultured under blue light. Preliminary RNA-Seq (unpublished data) and qRT-PCR analyses (Ch. 2) indicated that a type I pilus assembly system, which we named PrpABCD, is downregulated when bacterial cells are cultured under light compared to dark at 24°C. We hypothesized that this pilus system is responsible for the motility and biofilm phenotypes we observed previously (4). Mutational analysis demonstrated that a prpA mutant is nonmotile and defective for pellicle formation under blue light or dark conditions at 24°C. However, the prpA mutant strain was able to adhere to plastic, forming a biofilm that could be observed by SEM. The same biological analyses performed at 37°C revealed that the prpA mutant strain displayed a defect in motility under blue light illumination compared to the dark. These observations indicate that 17978 depends on the active expression and production of the Prp pili in response to its environment at 24°C, but uses other uncharacterized adherence and motility mechanisms that are light-regulated, independent of BlsA at higher temperatures such as 37°C (Ch. 3). The differences in biofilm and motility responses at 24°C and 37°C can potentially be explained by the observation that some pathogens use light and temperature as a signal to physiologically prepare for host invasion (24). Although the production of type I Csu pili has not been well defined in 17978, it is possible that a mechanism of pilus crosstalk results in the motility responses observed in the dark at 37°C in the prpA mutant strain through the production of an alternative unidentified pilus system (Ch. 2). The Csu type I pili are required for attachment to and biofilm formation on abiotic surfaces in the non-motile A. baumannii 19606T strain (18). However, it is not known if these pili also contribute to the biofilm phenotypes observed in the prpA mutant, or if these pili have an uncharacterized role in motility in 17978. Western blot analyses of 17978 cells cultured in blue light or darkness showed no significant difference in production of the CsuAB pilin subunit, and therefore the production of PrpABCD pili (Fig. 25, panels A and B). We also determined that the expression of csuAB is not significantly different between light and dark conditions at either 24°C or 37°C (Fig 25, panels C and D). However, further analyses of the 17978 prpA strain demonstrated that a mutation in this gene abrogates CsuAB production as indicated by western blotting (Fig. 26). Based on these observations, it is tempting to speculate that crosstalk between pili systems, as described for E. coli type I pili (182, 183) is the underlying mechanism

144 accounting for the motility phenotype in the 17978 prpA mutant at 37°C (Ch. 2). In E. coli, the global H-NS regulator along with the fimE and fimB regulatory elements, modulate fimbrial expression during temperature switches from 28°C to 37°C (184). Based on these reports and our preliminary RNA-Seq and proteomics data (unpublished) that support the presence of potential fimbrial regulators, it is possible that a similar regulatory mechanism is occurring in A. baumannii. It is also unknown if other light-dependent regulators are responsible for the light- mediated motility phenotype at temperatures higher than 30°C at which the production and photoreceptor activity of BlsA are drastically impaired (4, 39). Whether there are photoreceptor proteins functioning at these temperatures to impact the light- and temperature-dependent physiological responses of A. baumannii remain to be determined. Although our proteomic analyses of cells cultured at 37°C in darkness or illumination did not reveal the presence of traditional light sensors, the phosphodiesterase CpdA was identified as a protein that was abundant at higher levels in light-adapted bacteria compared to dark-adapted bacteria (Ch. 4). In some bacteria such as P. aeruginosa or Vibrio cholerae, CpdA has a role in the catabolism of the secondary messenger cAMP. cAMP has roles in the regulation of virulence-associated genes involved in the formation of the type III secretion system, quorum sensing and pili production (180, 185). In E. coli, cAMP is important for the production of cell surface structures for adhesion and biofilm formation (186, 187). Although little is known about cAMP and the role of the phosphodiesterase in A. baumannii, CpdA was demonstrated to regulate pellicle formation, motility, hydrophobic characteristics and the expression of the genes involved in quorum sensing (171).

145 Figure 25. Effects of light and temperature on the expression of csuAB in 17978 cells. Proteins present in equal volumes of whole lysates prepared from cells cultured in the presence

(L) or absence (D) of light at 24°C (A) or 37°C (B) in SB until they reached an OD600 of 0.8 were size fractionated by SDS-PAGE and transferred to nitrocellulose. Blotted proteins were probed with anti-CsuAB and anti-RNA polymerase (RNAP) antibodies. The image is a representative of the results collected using three different biological samples obtained and processed under the same experimental conditions. (C and D) qRT-PCR of csuAB using RNA isolated from the same 17978 samples used to test the production of CsuAB shown in panels A and B. recA was used as a constitutively expressed control gene. Error bars represent the standard error of each data set.

146

147 Figure 26. CsuAB detection in 17978 wild type and prpA mutant cells. The immunoblot is representative of three experiments from proteins present in equal volumes of whole lysates prepared from bacterial cells cultured in the absence of light at 24°C or 37°C in SB until they reached an OD600 of 0.8 in each condition. Proteins were size fractionated by SDS- PAGE and transferred to nitrocellulose. Blotted proteins were probed with anti-CsuAB antibodies. The image is a representative of the results collected using three different biological samples obtained and processed under the same experimental conditions. Lane 1, MWM; lane 2, 19606T control sample; lane 3, empty space; lane 4, 17978; lane 5, 17978 prpA mutant.

148

149 Light regulation in A. baumannii RNA-Seq data (unpublished) along with our Ch. 2 findings indicated that BlsA- independent light regulation occurs at 37°C. The discovery that the prpA mutant is motile in darkness at 37°C (Ch. 2) prompted us to investigate, using the same experimental and growth conditions as we used for the RNA isolation in Ch. 2, the proteome profile of A. baumannii after blue light exposure. Using the PatternLab program for data processing, we discovered that differential expression of proteins occurs between light and dark conditions in A. baumannii at 37°C, a result that complements our motility observations presented in Ch. 2. A total of 20 proteins were differentially expressed in response to blue light. Of particular interest were the PhoB transcriptional regulatory protein and the cAMP phosphodiesterase CpdA which were upregulated in light-adapted cells. It is also noteworthy that a LysR-type regulator was found in higher abundance in light-adapted bacteria, although the protein was filtered by its λ-stringency (L-stringency). This PatternLab parameter filters proteins due to their low abundance that satisfy both fold-change and false discovery rate cutoff values, but require more experimentation to validate their differential expression (188). The abundance of regulators in the light-adapted cells supports the observation that light regulation occurs in A. baumannii at temperatures higher than 24°C and independent of the BlsA photoreceptor. Proteins that were downregulated by cells cultured in the light included a ribonuclease, an outer membrane protein and a TonB-dependent siderophore receptor protein. Although it is possible that the presence of iron acquisition proteins, such as the TonB siderophore receptor, in the proteomic data is an artifact due to the low concentration of iron of the media (SB; 20.1± 0.6233 μg/l) (94) used in our experiments, immunoblot analyses using anti-BauA and anti-BauE antibodies do not support this idea (data not shown). Some of the proteins present in bacteria cultured either in the light or dark at 37°C were also present in the 24°C experiments, including the alanine racemase (Ch. 4). Alanine metabolism may be important for A. baumannii regardless of its lifestyle as a host-associated facultative pathogen or living outside the host in a nosocomial setting considering the role of D- alanine as a critical component of bacterial peptidoglycan (189). Alanine racemase enzymes could prove to be valuable to study in the future as drug targets as these enzymes are only found in bacteria and could be useful to target during MDR A. baumannii infections (151, 190).

150

Concluding remarks Light plays an important role on the physiology of A. baumannii. Through our work, we have shown that this ubiquitous environmental signal is involved in complex regulatory pathways that include proteins used in pilus production (Ch. 2,) and both BlsA-dependent (Ch. 3) and -independent sensory transduction mechanisms (Ch.2, Ch.4). Unfortunately, we do not have a clear understanding of the depth and complexity of the light stimulon or BlsA regulon. Further studies are needed to characterize the effects of light and temperature at both 24°C and 37° to better understand these critical pathways. Understanding the physiological changes during the life cycle of this facultative human pathogen in different environments and temperatures is important and has the potential to yield knowledge that may contribute to the development of targeted therapeutics and light-based therapies needed to combat the MDR and PDR A. baumannii strains that have been on the rise over the last decade. This knowledge should also aid the understanding of how other short BLUF photoreceptors function in response to light at different temperatures in other non-photosynthetic chemotrophic bacteria, including pathogenic organisms in which photoresponses have already been identified such as B. abortus, L. monocytogenes and P. aeruginosa (24, 34, 191).

151 References 1. Bergogne-Berezin E, Towner KJ. 1996. Acinetobacter spp. as nosocomial pathogens: microbiological, clinical, and epidemiological features. Clin Microbiol Rev 9:148-165. 2. Baumann P. 1968. Isolation of Acinetobacter from soil and water. J Bacteriol 96:39-42. 3. Peleg AY, de Breij A, Adams MD, Cerqueira GM, Mocali S, Galardini M, Nibbering PH, Earl AM, Ward DV, Paterson DL, Seifert H, Dijkshoorn L. 2012. The success of Acinetobacter species; genetic, metabolic and virulence attributes. PLoS ONE 7:e46984. 4. Mussi MA, Gaddy JA, Cabruja M, Arivett BA, Viale AM, Rasia R, Actis LA. 2010. The opportunistic human pathogen Acinetobacter baumannii senses and responds to light. J Bacteriol 192:6336-6345. 5. Eijkelkamp BA, Hassan KA, Paulsen IT, Brown MH. 2011. Investigation of the human pathogen Acinetobacter baumannii under iron limiting conditions. BMC Genomics 12:126. 6. Brisou J, Prevot AR. 1954. Etudes de systematique bacterienne. X. Revision des espéces réunies dans le genre Achromobacter. Ann Inst Pasteur (Paris) 86:722-728. 7. Juni E. 1984. Genus III. Acinetobacter. Bergey's manual of systematic bacteriology 1:303-307. 8. Harding CM, Hennon SW, Feldman MF. 2018. Uncovering the mechanisms of Acinetobacter baumannii virulence. Nat Rev Microbiol 16:91-102. 9. Berlau J, Aucken HM, Houang E, Pitt TL. 1999. Isolation of Acinetobacter spp. including A. baumannii from vegetables: implications for hospital-acquired infections. J Hosp Infect 42:201-204. 10. Fournier PE, Richet H. 2006. The epidemiology and control of Acinetobacter baumannii in health care facilities. Clin Infect Dis 42:692-699. 11. Falagas ME, Karveli EA, Kelesidis I, Kelesidis T. 2007. Community-acquired Acinetobacter infections. Eur J Clin Microbiol Infect Dis 26:857-868. 12. Anstey NM, Currie BJ, Hassell M, Palmer D, Dwyer B, Seifert H. 2002. Community- acquired bacteremic Acinetobacter pneumonia in tropical Australia is caused by diverse strains of Acinetobacter baumannii, with carriage in the throat in at-risk groups. J Clin Microbiol 40:685-686.

152 13. Rudin ML, Michael JR, Huxley EJ. 1979. Community-acquired Acinetobacter pneumonia. Am J Med 67:39-43. 14. Achar KN, Johny M, Achar MN, Menon NK. 1993. Community-acquired bacteraemic Acinetobacter pneumonia with survival. Postgrad Med J 69:934-937. 15. Chen MZ, Hsueh PR, Lee LN, Yu CJ, Yang PC, Luh KT. 2001. Severe community- acquired pneumonia due to Acinetobacter baumannii. Chest 120:1072-1077. 16. Jawad A, Seifert H, Snelling AM, Heritage J, Hawkey PM. 1998. Survival of Acinetobacter baumannii on dry surfaces: comparison of outbreak and sporadic isolates. J Clin Microbiol 36:1938-1941. 17. McConnell MJ, Actis L, Pachon J. 2013. Acinetobacter baumannii: human infections, factors contributing to pathogenesis and animal models. FEMS Microbiol Rev 37:130- 155. 18. Tomaras AP, Dorsey CW, Edelmann RE, Actis LA. 2003. Attachment to and biofilm formation on abiotic surfaces by Acinetobacter baumannii: involvement of a novel chaperone-usher pili assembly system. Microbiology 149:3473-3484. 19. Charnot-Katsikas A, Dorafshar AH, Aycock JK, David MZ, Weber SG, Frank KM. 2009. Two cases of necrotizing fasciitis due to Acinetobacter baumannii. J Clin Microbiol 47:258-263. 20. Houang ET, Sormunen RT, Lai L, Chan CY, Leong AS. 1998. Effect of desiccation on the ultrastructural appearances of Acinetobacter baumannii and Acinetobacter lwoffii. J Clin Pathol 51:786-788. 21. Nwugo CC, Arivett BA, Zimbler DL, Gaddy JA, Richards AM, Actis LA. 2012. Effect of ethanol on differential protein production and expression of potential virulence functions in the opportunistic pathogen Acinetobacter baumannii. PLoS One 7:e51936. 22. Hujer KM, Hujer AM, Hulten EA, Bajaksouzian S, Adams JM, Donskey CJ, Ecker DJ, Massire C, Eshoo MW, Sampath R, Thomson JM, Rather PN, Craft DW, Fishbain JT, Ewell AJ, Jacobs MR, Paterson DL, Bonomo RA. 2006. Analysis of antibiotic resistance genes in multidrug-resistant Acinetobacter sp. isolates from military and civilian patients treated at the Walter Reed Army Medical Center. Antimicrob Agents Chemother 50:4114-4123.

153 23. Hood MI, Skaar EP. 2012. Nutritional immunity: transition metals at the pathogen-host interface. Nat Rev Microbiol 10:525-537. 24. Swartz TE, Tseng TS, Frederickson MA, Paris G, Comerci DJ, Rajashekara G, Kim JG, Mudgett MB, Splitter GA, Ugalde RA, Goldbaum FA, Briggs WR, Bogomolni RA. 2007. Blue-light-activated histidine kinases: two-component sensors in bacteria. Science 317:1090-1093. 25. Gomelsky M, Hoff WD. 2011. Light helps bacteria make important lifestyle decisions. Trends Microbiol 19:441-448. 26. Ramirez MS, Muller GL, Perez JF, Golic AE, Mussi MA. 2015. More than just light: clinical relevance of light perception in the nosocomial pathogen Acinetobacter baumannii and other members of the genus Acinetobacter. Photochem Photobiol 91:1291-1301. 27. Müller GL, Tuttobene M, Altilio M, Martinez Amezaga M, Nguyen M, Cribb P, Cybulski LE, Ramírez MS, Altabe S, Mussi MA. 2017. Light modulates metabolic pathways and other novel physiological traits in the human pathogen Acinetobacter baumannii. J Bacteriol 199. 28. Actis LA, Tolmasky ME, Crosa LM, Crosa JH. 1993. Effect of iron-limiting conditions on growth of clinical isolates of Acinetobacter baumannii. Journal of clinical microbiology 31:2812-2815. 29. Hood MI, Mortensen BL, Moore JL, Zhang Y, Kehl-Fie TE, Sugitani N, Chazin WJ, Caprioli RM, Skaar EP. 2012. Identification of an Acinetobacter baumannii zinc acquisition system that facilitates resistance to calprotectin-mediated zinc sequestration. PLoS Pathog 8:e1003068. 30. Bagg A, Neilands JB. 1987. Ferric uptake regulation protein acts as a repressor, employing iron (II) as a cofactor to bind the operator of an iron transport operon in Escherichia coli. Biochemistry 26:5471-5477. 31. Dorsey CW, Beglin MS, Actis LA. 2003. Detection and analysis of iron uptake components expressed by Acinetobacter baumannii clinical isolates. J Clin Microbiol 41:4188-4193. 32. Penwell WF, Arivett BA, Actis LA. 2012. The Acinetobacter baumannii entA gene located outside the acinetobactin cluster is critical for siderophore production, iron acquisition and virulence. PLoS One 7:e36493.

154 33. Barends TR, Hartmann E, Griese JJ, Beitlich T, Kirienko NV, Ryjenkov DA, Reinstein J, Shoeman RL, Gomelsky M, Schlichting I. 2009. Structure and mechanism of a bacterial light-regulated cyclic nucleotide phosphodiesterase. Nature 459:1015-1018. 34. Ondrusch N, Kreft J. 2011. Blue and red light modulates SigB-dependent gene transcription, swimming motility and invasiveness in Listeria monocytogenes. PLoS One 6:e16151. 35. Eveillard M, Kempf M, Belmonte O, Pailhories H, Joly-Guillou ML. 2013. Reservoirs of Acinetobacter baumannii outside the hospital and potential involvement in emerging human community-acquired infections. Int J Infect Dis 17:e802-805. 36. McGuiness W, Vella E, Harrison D. 2004. Influence of dressing changes on wound temperature. J Wound Care 13:383-385. 37. Romanelli M, Gaggio G, Coluccia M, Rizzello F, Piaggesi A. 2002. Technological advances in wound bed measurements. Wounds 14:58-66. 38. Tuttobene MR, Cribb P, Mussi MA. 2018. BlsA integrates light and temperature signals into iron metabolism through Fur in the human pathogen Acinetobacter baumannii. Sci Rep 8:7728. 39. Abatedaga I, Valle L, Golic AE, Muller GL, Cabruja M, Moran Vieyra FE, Jaime PC, Mussi MA, Borsarelli CD. 2017. Integration of temperature and blue-light sensing in Acinetobacter baumannii through the BlsA sensor. Photochem Photobiol 93:805-814. 40. Butler WL, Norris KH, Siegelman HW, Hendricks SB. 1959. Detection, assay, and preliminary purification of the pigment controlling photoresponsive development of plants. Proc Natl Acad Sci USA 45:1703-1708. 41. Losi A, Gartner W. 2008. Bacterial bilin- and flavin-binding photoreceptors. Photochem Photobiol Sci 7:1168-1178. 42. van der Horst MA, Hellingwerf KJ. 2004. Photoreceptor proteins, "star actors of modern times": a review of the functional dynamics in the structure of representative members of six different photoreceptor families. Acc Chem Res 37:13-20. 43. Moglich A, Yang X, Ayers RA, Moffat K. 2010. Structure and function of plant photoreceptors. Annu Rev Plant Biol 61:21-47. 44. von Heijne G. 1989. Control of topology and mode of assembly of a polytopic membrane protein by positively charged residues. Nature 341:456-458.

155 45. Losi A, Gartner W. 2012. The evolution of flavin-binding photoreceptors: an ancient chromophore serving trendy blue-light sensors. Annu Rev Plant Biol 63:49-72. 46. Losi A, Mandalari C, Gartner W. 2014. From plant infectivity to growth patterns: the role of blue-light sensing in the prokaryotic world. Plants 3:70-94. 47. Walsh C. 1980. Flavin coenzymes: at the crossroads of biological redox chemistry. Acc Chem Res 13:148-155. 48. Gomelsky M, Kaplan S. 1998. AppA, a redox regulator of photosystem formation in Rhodobacter sphaeroides 2.4.1, is a flavoprotein. Identification of a novel FAD binding domain. J Biol Chem 273:35319-35325. 49. Salomon M, Eisenreich W, Durr H, Schleicher E, Knieb E, Massey V, Rudiger W, Muller F, Bacher A, Richter G. 2001. An optomechanical transducer in the blue light receptor phototropin from Avena sativa. Proc Natl Acad Sci U S A 98:12357-12361. 50. Ahmad M, Cashmore AR. 1993. HY4 gene of A. thaliana encodes a protein with characteristics of a blue-light photoreceptor. Nature 366:162-166. 51. Hitomi K, Okamoto K, Daiyasu H, Miyashita H, Iwai S, Toh H, Ishiura M, Todo T. 2000. Bacterial cryptochrome and photolyase: characterization of two photolyase-like genes of Synechocystis sp. PCC6803. Nucleic Acids Res 28:2353-2362. 52. Sancar A. 1994. Structure and function of DNA photolyase. Biochemistry 33:2-9. 53. Park HW, Kim ST, Sancar A, Deisenhofer J. 1995. Crystal structure of DNA photolyase from Escherichia coli. Science 268:1866. 54. Brust R, Haigney A, Lukacs A, Gil A, Hossain S, Addison K, Lai CT, Towrie M, Greetham GM, Clark IP, Illarionov B, Bacher A, Kim RR, Fischer M, Simmerling C, Meech SR, Tonge PJ. 2014. Ultrafast structural dynamics of BlsA, a photoreceptor from the pathogenic bacterium Acinetobacter baumannii. J Phys Chem Lett 5:220-224. 55. Gomelsky M, Klug G. 2002. BLUF: a novel FAD-binding domain involved in sensory transduction in microorganisms. Trends Biochem Sci 27:497-500. 56. Park SY, Tame JRH. 2017. Seeing the light with BLUF proteins. Biophys Rev 9:169- 176. 57. Kita A, Okajima K, Morimoto Y, Ikeuchi M, Miki K. 2005. Structure of a cyanobacterial BLUF protein, Tll0078, containing a novel FAD-binding blue light sensor domain. J Mol Biol 349:1-9.

156 58. Masuda S, Bauer CE. 2002. AppA is a blue light photoreceptor that antirepresses photosynthesis gene expression in Rhodobacter sphaeroides. Cell 110:613-623. 59. Okajima K, Yoshihara S, Fukushima Y, Geng X, Katayama M, Higashi S, Watanabe M, Sato S, Tabata S, Shibata Y, Itoh S, Ikeuchi M. 2005. Biochemical and functional characterization of BLUF-type flavin-binding proteins of two species of cyanobacteria. J Biochem 137:741-750. 60. Masuda S, Hasegawa K, Ohta H, Ono TA. 2008. Crucial role in light signal transduction for the conserved Met93 of the BLUF protein PixD/Slr1694. Plant Cell Physiol 49:1600-1606. 61. Masuda S, Hasegawa K, Ishii A, Ono TA. 2004. Light-induced structural changes in a putative blue-light receptor with a novel FAD binding fold sensor of blue-light using FAD (BLUF); Slr1694 of synechocystis sp. PCC6803. Biochemistry 43:5304-5313. 62. Fujisawa T, Masuda S. 2018. Light-induced chromophore and protein responses and mechanical signal transduction of BLUF proteins. Biophys Rev 10:327-337. 63. Iwata T, Nagai T, Ito S, Osoegawa S, Iseki M, Watanabe M, Unno M, Kitagawa S, Kandori H. 2018. Hydrogen bonding environments in the photocycle process around the flavin chromophore of the AppA-BLUF domain. J Am Chem Soc 140:11982-11991. 64. Gauden M, Yeremenko S, Laan W, van Stokkum IH, Ihalainen JA, van Grondelle R, Hellingwerf KJ, Kennis JT. 2005. Photocycle of the flavin-binding photoreceptor AppA, a bacterial transcriptional antirepressor of photosynthesis genes. Biochemistry 44:3653- 3662. 65. Jung A, Domratcheva T, Tarutina M, Wu Q, Ko WH, Shoeman RL, Gomelsky M, Gardner KH, Schlichting I. 2005. Structure of a bacterial BLUF photoreceptor: insights into blue light-mediated signal transduction. Proc Natl Acad Sci USA 102:12350-12355. 66. Lukacs A, Brust R, Haigney A, Laptenok SP, Addison K, Gil A, Towrie M, Greetham GM, Tonge PJ, Meech SR. 2014. BLUF domain function does not require a metastable radical intermediate state. J Am Chem Soc 136:4605-4615. 67. Towner KJ, Bergogne-Bérézin E, Fewson CA. 1991. Acinetobacter: Portrait of a genus, p 1-24. In Towner KJ, Bergogne-Bérézin E, Fewson CA (ed), The Biology of Acinetobacter: Taxonomy, Clinical Importance, Molecular Biology, Physiology, Industrial Relevance doi:10.1007/978-1-4899-3553-3_1. Springer US, Boston, MA.

157 68. Zemelman BV, Lee GA, Ng M, Miesenbock G. 2002. Selective photostimulation of genetically chARGed neurons. Neuron 33:15-22. 69. Boyden ES, Zhang F, Bamberg E, Nagel G, Deisseroth K. 2005. Millisecond-timescale, genetically targeted optical control of neural activity. Nat Neurosci 8:1263-1268. 70. Fenno L, Yizhar O, Deisseroth K. 2011. The development and application of optogenetics. Annu Rev Neurosci 34:389-412. 71. Levskaya A, Weiner OD, Lim WA, Voigt CA. 2009. Spatiotemporal control of cell signalling using a light-switchable protein interaction. Nature 461:997-1001. 72. Johnson HE, Toettcher JE. 2018. Illuminating developmental biology with cellular optogenetics. Curr Opin Biotechnol 52:42-48. 73. Kao YT, Saxena C, He TF, Guo L, Wang L, Sancar A, Zhong D. 2008. Ultrafast dynamics of flavins in five redox states. J Am Chem Soc 130:13132-13139. 74. Dai T, Tegos GP, Lu Z, Huang L, Zhiyentayev T, Franklin MJ, Baer DG, Hamblin MR. 2009. Photodynamic therapy for Acinetobacter baumannii burn infections in mice. Antimicrob Agents Chemother 53:3929-3934. 75. Firstenberg-Eden R, Rowley DB, Shattuck GE. 1980. Factors affecting inactivation of Moraxella-Acinetobacter cells in an irradiation process. Appl Environ Microbiol 40:480- 485. 76. Villegas MV, Hartstein AI. 2003. Acinetobacter outbreaks, 1977-2000. Infect Control Hosp Epidemiol 24:284-295. 77. Nemec A, Krizova L, Maixnerova M, Sedo O, Brisse S, Higgins PG. 2015. Acinetobacter seifertii sp. nov., a member of the Acinetobacter calcoaceticus- Acinetobacter baumannii complex isolated from human clinical specimens. Int J Syst Evol Microbiol 65:934-942. 78. Huys G, Bartie K, Cnockaert M, Hoang Oanh DT, Phuong NT, Somsiri T, Chinabut S, Yusoff FM, Shariff M, Giacomini M, Teale A, Swings J. 2007. Biodiversity of chloramphenicol-resistant mesophilic heterotrophs from Southeast Asian aquaculture environments. Res Microbiol 158:228-235. 79. Vangnai AS, Petchkroh W. 2007. Biodegradation of 4-chloroaniline by bacteria enriched from soil. FEMS Microbiol Lett 268:209-216.

158 80. Rafei R, Hamze M, Pailhories H, Eveillard M, Marsollier L, Joly-Guillou ML, Dabboussi F, Kempf M. 2015. Extrahuman epidemiology of Acinetobacter baumannii in Lebanon. Appl Environ Microbiol 81:2359-2367. 81. Endimiani A, Hujer KM, Hujer AM, Bertschy I, Rossano A, Koch C, Gerber V, Francey T, Bonomo RA, Perreten V. 2011. Acinetobacter baumannii isolates from pets and horses in Switzerland: molecular characterization and clinical data. J Antimicrob Chemother 66:2248-2254. 82. Zordan S, Prenger-Berninghoff E, Weiss R, van der Reijden T, van den Broek P, Baljer G, Dijkshoorn L. 2011. Multidrug-resistant Acinetobacter baumannii in veterinary clinics, Germany. Emerg Infect Dis 17:1751-1754. 83. Kempf M, Rolain JM, Diatta G, Azza S, Samb B, Mediannikov O, Gassama Sow A, Diene SM, Fenollar F, Raoult D. 2012. Carbapenem resistance and Acinetobacter baumannii in Senegal: the paradigm of a common phenomenon in natural reservoirs. PLoS One 7:e39495. 84. Gaddy JA, Arivett BA, McConnell MJ, Lopez-Rojas R, Pachon J, Actis LA. 2012. Role of acinetobactin-mediated iron acquisition functions in the interaction of Acinetobacter baumannii strain ATCC 19606T with human lung epithelial cells, Galleria mellonella caterpillars, and mice. Infect Immun 80:1015-1024. 85. McQueary CN, Kirkup BC, Si Y, Barlow M, Actis LA, Craft DW, Zurawski DV. 2012. Extracellular stress and lipopolysaccharide modulate Acinetobacter baumannii surface-associated motility. J Microbiol 50:434-443. 86. Bitrian M, Gonzalez RH, Paris G, Hellingwerf KJ, Nudel CB. 2013. Blue-light- dependent inhibition of twitching motility in Acinetobacter baylyi ADP1: additive involvement of three BLUF-domain-containing proteins. Microbiology 159:1828-1841. 87. Golic A, Vaneechoutte M, Nemec A, Viale AM, Actis LA, Mussi MA. 2013. Staring at the cold sun: blue light regulation is distributed within the genus Acinetobacter. PLoS One 8:e55059. 88. Sambrook JR, D.W. 2001. Molecular Cloning. A Laboratory Manual., 3rd ed. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, N.Y. 89. Barcak GJ, Chandler MS, Redfield RJ, Tomb JF. 1991. Genetic systems in Haemophilus influenzae. Methods Enzymol 204:321-342.

159 90. Dorsey CW, Tomaras AP, Actis LA. 2002. Genetic and phenotypic analysis of Acinetobacter baumannii insertion derivatives generated with a transposome system. Appl Environ Microbiol 68:6353-6360. 91. Hunger M, Schmucker R, Kishan V, Hillen W. 1990. Analysis and nucleotide sequence of an origin of DNA replication in Acinetobacter calcoaceticus and its use for Escherichia coli shuttle plasmids. Gene 87:45-51. 92. Fiester SE, Arivett BA, Schmidt RE, Beckett AC, Ticak T, Carrier MV, Ghosh R, Ohneck EJ, Metz ML, Sellin Jeffries MK, Actis LA. 2016. Iron-regulated Phospholipase C activity contributes to the cytolytic activity and virulence of Acinetobacter baumannii. PLoS One 11:e0167068. 93. Gaddy JA, Tomaras AP, Actis LA. 2009. The Acinetobacter baumannii 19606 OmpA protein plays a role in biofilm formation on abiotic surfaces and in the interaction of this pathogen with eukaryotic cells. Infect Immun 77:3150-3160. 94. Ohneck EJ, Arivett BA, Fiester SE, Wood CR, Metz ML, Simeone GM, Actis LA. 2018. Mucin acts as a nutrient source and a signal for the differential expression of genes coding for cellular processes and virulence factors in Acinetobacter baumannii. PLoS One 13:e0190599. 95. Smith MG, Gianoulis TA, Pukatzki S, Mekalanos JJ, Ornston LN, Gerstein M, Snyder M. 2007. New insights into Acinetobacter baumannii pathogenesis revealed by high-density pyrosequencing and transposon mutagenesis. Genes Dev 21:601-614. 96. Alvarez-Fraga L, Perez A, Rumbo-Feal S, Merino M, Vallejo JA, Ohneck EJ, Edelmann RE, Beceiro A, Vazquez-Ucha JC, Valle J, Actis LA, Bou G, Poza M. 2016. Analysis of the role of the LH92_11085 gene of a biofilm hyper-producing Acinetobacter baumannii strain on biofilm formation and attachment to eukaryotic cells. Virulence 7:443-455. 97. Nait Chabane Y, Marti S, Rihouey C, Alexandre S, Hardouin J, Lesouhaitier O, Vila J, Kaplan JB, Jouenne T, De E. 2014. Characterisation of pellicles formed by Acinetobacter baumannii at the air-liquid interface. PLoS One 9:e111660. 98. Caiazza NC, Merritt JH, Brothers KM, O'Toole GA. 2007. Inverse regulation of biofilm formation and swarming motility by Pseudomonas aeruginosa PA14. J Bacteriol 189:3603-3612.

160 99. Guttenplan SB, Kearns DB. 2013. Regulation of flagellar motility during biofilm formation. FEMS Microbiol Rev 37:849-871. 100. Verstraeten N, Braeken K, Debkumari B, Fauvart M, Fransaer J, Vermant J, Michiels J. 2008. Living on a surface: swarming and biofilm formation. Trends Microbiol 16:496-506. 101. Tomaras AP, Flagler MJ, Dorsey CW, Gaddy JA, Actis LA. 2008. Characterization of a two-component regulatory system from Acinetobacter baumannii that controls biofilm formation and cellular morphology. Microbiology 154:3398-3409. 102. de Breij A, Gaddy J, van der Meer J, Koning R, Koster A, van den Broek P, Actis L, Nibbering P, Dijkshoorn L. 2009. CsuA/BABCDE-dependent pili are not involved in the adherence of Acinetobacter baumannii ATCC19606T to human airway epithelial cells and their inflammatory response. Res Microbiol 160:213-218. 103. Kroger C, Kary SC, Schauer K, Cameron AD. 2016. Genetic regulation of virulence and antibiotic resistance in Acinetobacter baumannii. Genes 8. 104. Montana S, Vilacoba E, Traglia GM, Almuzara M, Pennini M, Fernandez A, Sucari A, Centron D, Ramirez MS. 2015. Genetic variability of AdeRS two-component system associated with tigecycline resistance in XDR-Acinetobacter baumannii isolates. Curr Microbiol 71:76-82. 105. Wright MS, Iovleva A, Jacobs MR, Bonomo RA, Adams MD. 2016. Genome dynamics of multidrug-resistant Acinetobacter baumannii during infection and treatment. Genome Med 8:26. 106. Peleg AY, Seifert H, Paterson DL. 2008. Acinetobacter baumannii: emergence of a successful pathogen. Clin Microbiol Rev 21:538-582. 107. Conrad KS, Manahan CC, Crane BR. 2014. Photochemistry of flavoprotein light sensors. Nat Chem Biol 10:801-809. 108. Fudim R, Mehlhorn J, Berthold T, Weber S, Schleicher E, Kennis JT, Mathes T. 2015. Photoinduced formation of flavin radicals in BLUF domains lacking the central glutamine. FEBS J 282:3161-3174. 109. Masuda S. 2013. Light detection and signal transduction in the BLUF photoreceptors. Plant Cell Physiol 54:171-179.

161 110. Dorsey CW, Tomaras AP, Actis LA. 2006. Sequence and organization of pMAC, an Acinetobacter baumannii plasmid harboring genes involved in organic peroxide resistance. Plasmid 56:112-123. 111. Bradford MM. 1976. A rapid and sensitive method for the quantitation of microgram quantities of protein utilizing the principle of protein-dye binding. Anal Biochem 72:248- 254. 112. Actis LA, Potter SA, Crosa JH. 1985. Iron-regulated outer membrane protein OM2 of Vibrio anguillarum is encoded by virulence plasmid pJM1. J Bacteriol 161:736-742. 113. Heine T, Scholtissek A, Westphal AH, van Berkel WJH, Tischler D. 2017. N-terminus determines activity and specificity of styrene monooxygenase reductases. Biochim Biophys Acta Proteins Proteom 1865:1770-1780. 114. Dorsey CW, Tolmasky ME, Crosa JH, Actis LA. 2003. Genetic organization of an Acinetobacter baumannii chromosomal region harbouring genes related to siderophore biosynthesis and transport. Microbiology 149:1227-1238. 115. Harlow E, Lane D. 1988. Antibodies: A laboratory manual. Cold Spring Harbor. New York: Cold Spring Harbor Laboratory Press. 116. Pettersen EF, Goddard TD, Huang CC, Couch GS, Greenblatt DM, Meng EC, Ferrin TE. 2004. UCSF Chimera--a visualization system for exploratory research and analysis. J Comput Chem 25:1605-1612. 117. Kim DE, Chivian D, Baker D. 2004. Protein structure prediction and analysis using the Robetta server. Nucleic Acids Res 32:W526-531. 118. Bramucci E, Paiardini A, Bossa F, Pascarella S. 2012. PyMod: sequence similarity searches, multiple sequence-structure alignments, and homology modeling within PyMOL. BMC Bioinformatics 13 Suppl 4:S2. 119. Eswar N, Webb B, Marti-Renom MA, Madhusudhan MS, Eramian D, Shen MY, Pieper U, Sali A. 2006. Comparative protein structure modeling using Modeller. Curr Protoc Bioinformatics Chapter 5:Unit-5.6. 120. Fukushima Y, Okajima K, Shibata Y, Ikeuchi M, Itoh S. 2005. Primary intermediate in the photocycle of a blue-light sensory BLUF FAD-protein, Tll0078, of Thermosynechococcus elongatus BP-1. Biochemistry 44:5149-5158.

162 121. Tanwar M, Nahar S, Gulati S, Veetil SK, Kateriya S. 2016. Molecular determinant modulates thermal recovery kinetics and structural integrity of the bacterial BLUF photoreceptor. FEBS Lett 590:2146-2157. 122. Yuan H, Anderson S, Masuda S, Dragnea V, Moffat K, Bauer C. 2006. Crystal structures of the Synechocystis photoreceptor Slr1694 reveal distinct structural states related to signaling. Biochemistry 45:12687-12694. 123. Dragnea V, Arunkumar AI, Yuan H, Giedroc DP, Bauer CE. 2009. Spectroscopic studies of the AppA BLUF domain from Rhodobacter sphaeroides: addressing movement of tryptophan 104 in the signaling state. Biochemistry 48:9969-9979. 124. Yuan H, Dragnea V, Wu Q, Gardner KH, Bauer CE. 2011. Mutational and structural studies of the PixD BLUF output signal that affects light-regulated interactions with PixE. Biochemistry 50:6365-6375. 125. Masuda S, Hasegawa K, Ono TA. 2005. Tryptophan at position 104 is involved in transforming light signal into changes of beta-sheet structure for the signaling state in the BLUF domain of AppA. Plant Cell Physiol 46:1894-1901. 126. Anderson S, Dragnea V, Masuda S, Ybe J, Moffat K, Bauer C. 2005. Structure of a novel photoreceptor, the BLUF domain of AppA from Rhodobacter sphaeroides. Biochemistry 44:7998-8005. 127. Wood CR, Ohneck EJ, Edelmann RE, Actis LA. 2018. A light-regulated type I pilus contributes to Acinetobacter baumannii biofilm, motility, and virulence functions. Infect Immun 86. 128. Masuda S, Tomida Y, Ohta H, Takamiya K. 2007. The critical role of a hydrogen bond between Gln63 and Trp104 in the blue-light sensing BLUF domain that controls AppA activity. J Mol Biol 368:1223-1230. 129. World Health Organization. 2017. WHO publishes list of bacteria for which new antibiotics are urgently needed, Geneva. 130. Aygun G, Demirkiran O, Utku T, Mete B, Urkmez S, Yilmaz M, Yasar H, Dikmen Y, Ozturk R. 2002. Environmental contamination during a carbapenem-resistant Acinetobacter baumannii outbreak in an intensive care unit. J Hosp Infect 52:259-262.

163 131. Park YK, Peck KR, Cheong HS, Chung DR, Song JH, Ko KS. 2009. Extreme drug resistance in Acinetobacter baumannii infections in intensive care units, South Korea. Emerg Infect Dis 15:1325-1327. 132. Vidal R, Dominguez M, Urrutia H, Bello H, Gonzalez G, Garcia A, Zemelman R. 1996. Biofilm formation by Acinetobacter baumannii. Microbios 86:49-58. 133. Camarena L, Bruno V, Euskirchen G, Poggio S, Snyder M. 2010. Molecular mechanisms of ethanol-induced pathogenesis revealed by RNA-sequencing. PLoS Pathog 6:e1000834. 134. Wang X, Liu W, Xin C, Zheng Y, Cheng Y, Sun S, Li R, Zhu XG, Dai SY, Rentzepis PM, Yuan JS. 2016. Enhanced limonene production in cyanobacteria reveals photosynthesis limitations. Proc Natl Acad Sci USA 113:14225-14230. 135. Liang H, Li L, Dong Z, Surette MG, Duan K. 2008. The YebC family protein PA0964 negatively regulates the Pseudomonas aeruginosa quinolone signal system and pyocyanin production. J Bacteriol 190:6217-6227. 136. Taylor RK, Hall MN, Enquist L, Silhavy TJ. 1981. Identification of OmpR: a positive regulatory protein controlling expression of the major outer membrane matrix porin proteins of Escherichia coli K-12. J Bacteriol 147:255-258. 137. Jubelin G, Vianney A, Beloin C, Ghigo JM, Lazzaroni JC, Lejeune P, Dorel C. 2005. CpxR/OmpR interplay regulates curli gene expression in response to osmolarity in Escherichia coli. J Bacteriol 187:2038-2049. 138. Tipton KA, Rather PN. 2017. An ompR-envZ two-component system ortholog regulates phase variation, osmotic tolerance, motility, and virulence in Acinetobacter baumannii strain AB5075. J Bacteriol 199. 139. Burnell JN. 2010. Cloning and characterization of Escherichia coli DUF299: a bifunctional ADP-dependent kinase--Pi-dependent pyrophosphorylase from bacteria. BMC Biochem 11:1. 140. Miles EW, Bauerle R, Ahmed SA. 1987. Tryptophan synthase from Escherichia coli and Salmonella typhimurium. Methods Enzymol 142:398-414. 141. Muramatsu MK, Brothwell JA, Stein BD, Putman TE, Rockey DD, Nelson DE. 2016. Beyond tryptophan synthase: identification of genes that contribute to Chlamydia

164 trachomatis survival during gamma interferon-induced persistence and reactivation. Infect Immun 84:2791-2801. 142. Tso WW, Adler J. 1974. Negative chemotaxis in Escherichia coli. J Bacteriol 118:560- 576. 143. Sasaki-Imamura T, Yano A, Yoshida Y. 2010. Production of indole from L-tryptophan and effects of these compounds on biofilm formation by Fusobacterium nucleatum ATCC 25586. Appl Environ Microbiol 76:4260-4268. 144. Lin HR, Shu HY, Lin GH. 2018. Biological roles of indole-3-acetic acid in Acinetobacter baumannii. Microbiol Res 216:30-39. 145. Cerqueira GM, Kostoulias X, Khoo C, Aibinu I, Qu Y, Traven A, Peleg AY. 2014. A global virulence regulator in Acinetobacter baumannii and its control of the phenylacetic acid catabolic pathway. J Infect Dis 210:46-55. 146. Bhuiyan MS, Ellett F, Murray GL, Kostoulias X, Cerqueira GM, Schulze KE, Mahamad Maifiah MH, Li J, Creek DJ, Lieschke GJ, Peleg AY. 2016. Acinetobacter baumannii phenylacetic acid metabolism influences infection outcome through a direct effect on neutrophil chemotaxis. Proc Natl Acad Sci USA 113:9599-9604. 147. Teufel R, Mascaraque V, Ismail W, Voss M, Perera J, Eisenreich W, Haehnel W, Fuchs G. 2010. Bacterial phenylalanine and phenylacetate catabolic pathway revealed. Proc Natl Acad Sci USA 107:14390-14395. 148. Ripoll F, Pasek S, Schenowitz C, Dossat C, Barbe V, Rottman M, Macheras E, Heym B, Herrmann JL, Daffe M, Brosch R, Risler JL, Gaillard JL. 2009. Non mycobacterial virulence genes in the genome of the emerging pathogen Mycobacterium abscessus. PLoS One 4:e5660. 149. Law RJ, Hamlin JN, Sivro A, McCorrister SJ, Cardama GA, Cardona ST. 2008. A functional phenylacetic acid catabolic pathway is required for full pathogenicity of Burkholderia cenocepacia in the Caenorhabditis elegans host model. J Bacteriol 190:7209-7218. 150. Walsh CT. 1989. Enzymes in the D-alanine branch of bacterial cell wall peptidoglycan assembly. J Biol Chem 264:2393-2396.

165 151. Davis E, Scaletti-Hutchinson E, Opel-Reading H, Nakatani Y, Krause KL. 2014. The structure of alanine racemase from Acinetobacter baumannii. Acta Crystallogr F Struct Biol Commun 70:1199-1205. 152. Wasserman SA, Walsh CT, Botstein D. 1983. Two alanine racemase genes in Salmonella typhimurium that differ in structure and function. J Bacteriol 153:1439-1450. 153. Clemmer KM, Bonomo RA, Rather PN. 2011. Genetic analysis of surface motility in Acinetobacter baumannii. Microbiology 157:2534-2544. 154. Gutierrez C, Devedjian JC. 1991. Osmotic induction of gene osmC expression in Escherichia coli K12. J Mol Biol 220:959-973. 155. Atichartpongkul S, Loprasert S, Vattanaviboon P, Whangsuk W, Helmann JD, Mongkolsuk S. 2001. Bacterial Ohr and OsmC paralogues define two protein families with distinct functions and patterns of expression. Microbiology 147:1775-1782. 156. Yim HH, Villarejo M. 1992. osmY, a new hyperosmotically inducible gene, encodes a periplasmic protein in Escherichia coli. J Bacteriol 174:3637-3644. 157. Eskra L, Canavessi A, Carey M, Splitter G. 2001. Brucella abortus genes identified following constitutive growth and macrophage infection. Infect Immun 69:7736-7742. 158. Ostrovsky D, Diomina G, Lysak E, Matveeva E, Ogrel O, Trutko S. 1998. Effect of oxidative stress on the biosynthesis of 2-C-methyl-D-erythritol-2,4-cyclopyrophosphate and isoprenoids by several bacterial strains. Arch Microbiol 171:69-72. 159. Heuston S, Begley M, Gahan CG, Hill C. 2012. Isoprenoid biosynthesis in bacterial pathogens. Microbiology 158:1389-1401. 160. Young DM, Parke D, Ornston LN. 2005. Opportunities for genetic investigation afforded by Acinetobacter baylyi, a nutritionally versatile bacterial species that is highly competent for natural transformation. Annu Rev Microbiol 59:519-551. 161. Traglia GM, Place K, Dotto C, Fernandez JS, Montana S, Bahiense CDS, Soler- Bistue A, Iriarte A, Perez F, Tolmasky ME, Bonomo RA, Melano RG, Ramirez MS. 2019. Interspecies DNA acquisition by a naturally competent Acinetobacter baumannii strain. Int J Antimicrob Agents doi:10.1016/j.ijantimicag.2018.12.013. 162. Romeo T, Gong M. 1993. Genetic and physical mapping of the regulatory gene csrA on the Escherichia coli K-12 chromosome. J Bacteriol 175:5740-5741.

166 163. Staerck C, Gastebois A, Vandeputte P, Calenda A, Larcher G, Gillmann L, Papon N, Bouchara JP, Fleury MJJ. 2017. Microbial antioxidant defense enzymes. Microb Pathog 110:56-65. 164. Lin JW, Chao YF, Weng SF. 2001. Riboflavin synthesis genes ribE, ribB, ribH, ribA reside in the lux operon of Photobacterium leiognathi. Biochem Biophys Res Commun 284:587-595. 165. Hecht S, Eisenreich W, Adam P, Amslinger S, Kis K, Bacher A, Arigoni D, Rohdich F. 2001. Studies on the nonmevalonate pathway to terpenes: the role of the GcpE (IspG) protein. Proc Natl Acad Sci U S A 98:14837-14842. 166. Andrews SC, Harrison PM, Guest JR. 1989. Cloning, sequencing, and mapping of the bacterioferritin gene (bfr) of Escherichia coli K-12. J Bacteriol 171:3940-3947. 167. Nwugo CC, Gaddy JA, Zimbler DL, Actis LA. 2011. Deciphering the iron response in Acinetobacter baumannii: A proteomics approach. J Proteomics 74:44-58. 168. Wanner BL. 1993. Gene regulation by phosphate in enteric bacteria. J Cell Biochem 51:47-54. 169. Gavigan JA, Marshall LM, Dobson AD. 1999. Regulation of polyphosphate kinase gene expression in Acinetobacter baumannii 252. Microbiology 145 ( Pt 10):2931-2937. 170. Ahmad S, Raza S, Uddin R, Azam SS. 2018. Comparative subtractive proteomics based ranking for antibiotic targets against the dirtiest superbug: Acinetobacter baumannii. J Mol Graph Model 82:74-92. 171. Giles SK, Stroeher UH, Eijkelkamp BA, Brown MH. 2015. Identification of genes essential for pellicle formation in Acinetobacter baumannii. BMC Microbiology 15:116. 172. Lees-Miller RG, Iwashkiw JA, Scott NE, Seper A, Vinogradov E, Schild S, Feldman MF. 2013. A common pathway for O-linked protein-glycosylation and synthesis of capsule in Acinetobacter baumannii. Mol Microbiol 89:816-830. 173. Vitreschak AG, Rodionov DA, Mironov AA, Gelfand MS. 2002. Regulation of riboflavin biosynthesis and transport genes in bacteria by transcriptional and translational attenuation. Nucleic Acids Res 30:3141-3151. 174. Earhart C. 1996. Uptake and metabolism of iron and molybdenum. Escherichia coli and Salmonella: cellular and molecular biology, 2nd ed ASM Press, Washington, DC:1075- 1090.

167 175. Wachi M, Doi M, Tamaki S, Park W, Nakajima-Iijima S, Matsuhashi M. 1987. Mutant isolation and molecular cloning of mre genes, which determine cell shape, sensitivity to mecillinam, and amount of penicillin-binding proteins in Escherichia coli. J Bacteriol 169:4935-4940. 176. Seifert H, Strate A, Pulverer G. 1995. Nosocomial bacteremia due to Acinetobacter baumannii. Clinical features, epidemiology, and predictors of mortality. Medicine 74:340- 349. 177. Zhu J, Wang C, Wu J, Jiang R, Mi Z, Huang Z. 2009. A novel aminoglycoside- modifying enzyme gene aac(6')-Ib in a pandrug-resistant Acinetobacter baumannii strain. J Hosp Infect 73:184-185. 178. Getchell-White SI, Donowitz LG, Groschel DH. 1989. The inanimate environment of an intensive care unit as a potential source of nosocomial bacteria: evidence for long survival of Acinetobacter calcoaceticus. Infect Control Hosp Epidemiol 10:402-407. 179. James GA, Korber DR, Caldwell DE, Costerton JW. 1995. Digital image analysis of growth and starvation responses of a surface-colonizing Acinetobacter sp. J Bacteriol 177:907-915. 180. McDonough KA, Rodriguez A. 2011. The myriad roles of cyclic AMP in microbial pathogens: from signal to sword. Nat Rev Microbiol 10:27-38. 181. Masuda S, Hasegawa K, Ono TA. 2005. Light-induced structural changes of apoprotein and chromophore in the sensor of blue light using FAD (BLUF) domain of AppA for a signaling state. Biochemistry 44:1215-1224. 182. Olsen PB, Klemm P. 1994. Localization of promoters in the fim gene cluster and the effect of H-NS on the transcription of fimB and fimE. FEMS Microbiol Lett 116:95-100. 183. Xia Y, Gally D, Forsman-Semb K, Uhlin BE. 2000. Regulatory cross-talk between adhesin operons in Escherichia coli: inhibition of type 1 fimbriae expression by the PapB protein. EMBO J 19:1450-1457. 184. Olsen PB, Schembri MA, Gally DL, Klemm P. 1998. Differential temperature modulation by H-NS of the fimB and fimE recombinase genes which control the orientation of the type 1 fimbrial phase switch. FEMS Microbiol Lett 162:17-23. 185. Fuchs EL, Brutinel ED, Klem ER, Fehr AR, Yahr TL, Wolfgang MC. 2010. In vitro and in vivo characterization of the Pseudomonas aeruginosa cyclic AMP (cAMP)

168 phosphodiesterase CpdA, required for cAMP homeostasis and virulence factor regulation. J Bacteriol 192:2779-2790. 186. Eisenstein BI, Beachey EH, Solomon SS. 1981. Divergent effects of cyclic adenosine 3',5'-monophosphate on formation of type 1 fimbriae in different K-12 strains of Escherichia coli. J Bacteriol 145:620-623. 187. Botsford JL, Harman JG. 1992. Cyclic AMP in prokaryotes. Microbiol Rev 56:100-122. 188. Carvalho PC, Yates JR, 3rd, Barbosa VC. 2012. Improving the TFold test for differential shotgun proteomics. Bioinformatics 28:1652-1654. 189. Hayashi H, Wada H, Yoshimura T, Esaki N, Soda K. 1990. Recent topics in pyridoxal 5'-phosphate enzyme studies. Annu Rev Biochem 59:87-110. 190. Conti P, Tamborini L, Pinto A, Blondel A, Minoprio P, Mozzarelli A, De Micheli C. 2011. Drug discovery targeting amino acid racemases. Chem Rev 111:6919-6946. 191. Davis SJ, Vener AV, Vierstra RD. 1999. Bacteriophytochromes: phytochrome-like photoreceptors from nonphotosynthetic eubacteria. Science 286:2517-2520.

169