Friends of the Ribosome: Translational Regulation in Mycobacteria
The Harvard community has made this article openly available. Please share how this access benefits you. Your story matters
Citation Fishbein, Skye. 2019. Friends of the Ribosome: Translational Regulation in Mycobacteria. Doctoral dissertation, Harvard University, Graduate School of Arts & Sciences.
Citable link http://nrs.harvard.edu/urn-3:HUL.InstRepos:42029735
Terms of Use This article was downloaded from Harvard University’s DASH repository, and is made available under the terms and conditions applicable to Other Posted Material, as set forth at http:// nrs.harvard.edu/urn-3:HUL.InstRepos:dash.current.terms-of- use#LAA Friends of the ribosome: translational regulation in mycobacteria
A dissertation presented
by
Skye Rebecca Spolski Fishbein
to
The Committee on Higher Degrees in Biological Sciences in Public Health
In partial fulfillment of the requirements
For the degree of
Doctor of Philosophy
In the subject of
Biological Sciences in Public Health
Harvard University
Cambridge, Massachusetts
April 2019
ã 2019 – Skye Rebecca Spolski Fishbein All rights reserved.
Dissertation Advisor: Dr. Eric J. Rubin, M.D., Ph.D. Skye Fishbein
Friends of the ribosome: translational regulation in mycobacteria
Abstract
Mycobacteria are responsible for a significant disease burden in the world, caused by
Mycobacterium tuberculosis (Mtb) and non-tuberculous mycobacteria (NTM). While curable with a number of well-established antibiotic regimes, the global impact of mycobacterial disease is worsening due to increasing antimicrobial resistance. Understanding how this genus regulates its growth, causes disease, and tolerates antibiotics is necessary for the development of new antibiotics. The mycobacterial ribosome is a viable drug target, not only because of its central role in protein synthesis, but also because of its unique properties as a bacterial ribosome. Work on mycobacterial translational regulation has already revealed a number of features of protein synthesis machinery unique to mycobacteria.
We were interested in understanding mycobacterial translational control through the lens of known ribosome-associated factors. In Chapter 2, we characterize the role of LepA, a mysterious, conserved translational GTPase, in mycobacterial permeability. We found that in
Mycobacterium smegmatis (Msm), LepA controls the synthesis of a family of mycobacterial porins, and this control maintains cellular permeability. Here, translational regulation during protein synthesis is used to maintain the cellular proteome, specifically the most abundant constituents of the mycobacterial membrane. In Chapter 3, we conducted a biochemical screen to find protein associations with the mycobacterial ribosome. In log-phase, we identified mycobacterial-specific secretion systems (ESX-3 and ESX-5) that have not been previously associated with the ribosome.
Our investigation of ribosome-associated factors, and their role in translational regulation, has led us to the mycobacterial membrane. We found a translational GTPase that helps to synthesize outer membrane proteins in Msm, and we found a significant proportion of ribosomes associated with a number of secretion systems. This leads us to propose a model
iii whereby mycobacteria, and likely other bacteria require a heterogeneous population of ribosomes in a single cell. For mycobacteria, who have a complex cell-wall architechture, it may be that a significant proportion of ribosomes are dedicated to membrane translation and protein export. We believe that this connection, between protein synthesis and the mycobacterial membrane, may be an Achilles’ heel for pathogenic mycobacteria, where novel drug targets could exist.
iv Table of Contents
Chapter 1: Introduction
1.1 The genus of mycobacteria: pathogens and saprophytes ...... 2
1.2 Mycobacterium tuberculosis, an ancient thread to humankind ...... 2
1.3 NTMs: the tougher cousins of Mtb ...... 3
1.4 Adaptation in mycobacteria ...... 3
1.4.1 Transcriptional adaptation to drugs ...... 4
1.4.2 Transcriptional control in the host ...... 4
1.5 Regulation at the ribosome: a renaissance of nuance ...... 5
1.5.1 The landscape of ribosome composition ...... 6
1.5.2 Decoding the message: associating RNAs to the ribosome ...... 8
1.5.3 Friends who tell the ribosome to be quiet ...... 12
1.5.4 Friends who are just helpful to the ribosome ...... 13
1.5.5 Conclusions: bacterial translation regulation ...... 15
1.6 Mycobacterial translational features: what do they mean? ...... 16
1.7 Summary of aims ...... 16
1.8 References ...... 17
Chapter 2: LepA control of mycobacterial permeability
Abstract ...... 28
Introduction ...... 28
Results ...... 30
Discussion ...... 40
v
Materials and Methods ...... 43
References ...... 53
Chapter 3: Proteomic profiling of the Mtb ribosome
Abstract ...... 59
Introduction ...... 59
Results ...... 62
Discussion ...... 75
Materials and Methods ...... 79
References ...... 83
Chapter 4: Discussion
4.1 LepA, a translational GTPase controls mycobacterial permeability
4.1.1 LepA’s role mycobacteria ...... 89
4.1.2 LepA and organismal membrane protein synthesis ...... 91
4.1.3 Why do organisms keep LepA: its essential function ...... 92
4.2 New factors and new associations in mycobacteria: a glimpse of ribosome heterogeneity ...... 93
4.2.1 Ribosome proteomics screen for translational control ...... 93
4.2.2 Phenotypic data from ribosomal associations ...... 94
4.2.3 Novel ribosome-associated proteins ...... 95
4.2.4 Optimizing the screen to uncover true ribosome associations ...... 96
4.2.5 The grass is greener on the other side of the interaction ...... 97
4.3 Conclusions: Mycobacterial translation and the maintenance of the cell wall ...... 98
vi
Appendices
Appendix 1: LepA control of mycobacterial permeability ...... 103
Appendix 2: Proteomic profiling of the Mtb ribosome ...... 139
vii List of Figures and Tables
Chapter 2
Figure 2.1: Loss of ribosome factor causes altered drug tolerance in mycobacteria ..... 31
Figure 2.2: Whole cell profiling finds mycobacterial porins altered by loss of LepA ...... 33
Figure 2.3: Loss of LepA causes membrane defects through control of MspA ...... 35
Figure 2.4: LepA affects permeability through mRNA- determinants in mspA ...... 38
Chapter 3
Figure 3.1: Screen for discovery of novel ribosome associations in Mtb ...... 62
Table 3.1: Quantitative proteomics labeling scheme ...... 64
Figure 3.2: Log-phase interactions indicate ribosome heterogeneity ...... 65
Table 3.2: ESX systems associated to ribosomes in Mtb ...... 67
Figure 3.3: Conditional associations at the ribosome ...... 69
Table 3.3: Proteins of ‘unknown function’ associated with the Mtb ribosome ...... 70
Table 3.4: Ribosome-associated factor mutants in Msm ...... 72
Table 3.5: Ribosome-associated factor mutants in Mtb ...... 72
Figure 3.4: Phenotyping of ribosome-associated protein mutants in Msm ...... 73
Figure 3.5: Phentyping of ribosome-associated protein mutants in Mtb ...... 74
Appendices
Figure A1.1: lepA defect in calcein staining correlates with altered drug tolerance in mycobacteria ...... 103
Appendix Table A1.1: Drug susceptibility of lepA mutant ...... 104
Figure A1.2: Purification of LepA ...... 105
Figure A1.3: Post-transcriptional screen to find candidates for LepA regulation ...... 106
Figure A1.4: Validation of other candidates depleted in lepA proteomics data ...... 107
viii Figure A1.5: CRISPRi sgRNA validation ...... 108
Figure A1.6: Interrogation of MspA sequence determinants ...... 109
Appendix Table A1.2: Proteomic changes in the absence of LepA in Msm ...... 111
Appendix Table A1.3: DEseq analysis of RNA sequencing in lepA strains (p<0.01) ..... 115
Appendix Table A1.4: Strain List ...... 123
Appendix Table A1.5: Plasmid List ...... 131
Appendix Table A1.6: Primer List ...... 134
Figure A2.1: Optimization of sucrose cushion analysis for ribosome associating factors in Msm ...... 139
Table A2.3: Examination of ribosome distribution of candidate mutants in Msm ...... 139
Appendix Table A2.1: Proteomic ratios for Mtb ribosome proteomics screen ...... 140
ix
Acknowledgements
I do not really have concrete words to describe how grateful I am to have had you, Dr. Eric
Rubin, as my dissertation advisor… but here goes. The things you have given me: loyalty, fearlessness, joy, frustration, disappointment, compassion, and empathy… and what a full and filling time it has been for me in the Rubin lab. You taught me how to fit in, how to think hard, how to give up, and how to push a project to its limits. I don’t know what kind of scientist I’d be without your guidance, and your unending support… in spite of my well-defined weaknesses as a scientist/person. That is leadership, thank you.
I am grateful to the members of my Dissertation committee, Dr. Sarah Fortune, Dr. Robert
Husson, and Dr. Peter Dedon for your well-structured feedback through my scientific trajectory.
It has been nothing but a pleasure to learn from each of you as I have dealt with and attempted to overcome many hurdles. I would also like to thank my Defense committee, Dr. Simon Dove,
Dr. Scarlet Shell, and Dr. Yonatan Grad for your presence, dedication, and feedback on my thesis defense. I especially would like thank Dr. Sarah Fortune, you taught me to step back, immerse myself in the broader picture of science, and to take the world head on in my career.
Oh the Rubin Lab, to the members who have come and gone, you have molded me like clay into an igloo full of memories and emotions. Dr. Hesper Rego, your guidance and acceptance of me was maybe the first step towards me growing up as a person and scientist during my Ph.D.
Dr. Cara Boutte, you taught me that biochemistry was maybe my most favorite part of science, that I never got enough of. More importantly, you taught me both in and out of the lab how to be a woman in the world of science, and how to stand up for the things that matter most. To you both, you have made me who I am today, and I am so thankful for that. I also thank Dr. Jeremy
Rock, Dr. Jackson Marakalala, Dr. Katie Wu, Dr. Kasia Baranowski, Dr. Becky Audette, and all
x those who have left the lab. I thank Jess Pinkham, Albert Wang, Chidi Akusobi, Ian Wolf,
Michelle Gardner, Shoko Wakabayashi, Emmy Dove, Francesca Tomasi, and Dr. Junhao Zhu for their continued support throughout my time in the lab.
Just next door, my thanks goes to the Fortune Lab. You guys have invigorated our science and our time at work since we moved over to HSPH. I am so very thankful to Dr. Allison Carey, for teaching me everything it means to be a scholarly woman, I will continue to admire you and your science as you move on to greater things. I am grateful to Dr. Thibault Barbier, as a fellow persister in the field of post-transcriptional regulation. I am also grateful to Nate Hicks, we became the old people who couldn’t tweet to save a life on this floor.
I am beyond thankful to my friends Jennie Moser, Becca Elwin, Eli Gerrick, Kim Yang, and
Becky Audette, who have kept me sane and laughing throughout grad school. To my friends in
Boston, you are my life blood: Jess Pinkham, Andi Sabaroff, Deb Allen, Rachel Rudlaff, Nate
Hicks, Max Schubert, Greg Babunovic, Bryan Bryson, and others who have fed/clothed me, disappeared into the wilderness with me, and generally given me comfort from lab life.
It is without a doubt that I would not be here if it wasn’t for the strength of my family. Dad, Mom,
Jackie, Sarah, Ben, Marina, and all the dog members, you guys are my heart, when there is nothing left of it from my Ph.D. Kim, Jerry, Eli, and Aaron, I am grateful to have had a home away from home with you guys. Even when I came crawling with fatigue from the city, I always felt invigorated by our adventures in nature, at the beach and in the mountains. Grandma and
Grandpa, I miss you guys every single day, and every single day, I go out into the world trying to make you proud. And of course, to Cotter and Albie, you are the future we need and need to work for. May it be yours to take and run with.
xi
Chapter 1: Introduction
1 1.1 The genus of mycobacteria: pathogens and saprophytes
Mycobacteria, a genus in the order Actinomycetales, are acid-fast bacteria named for their waxy outer layer, made up of long carbon chains of mycolic acid and arabinogalactan [1]. The genus can be divided into slow growers and fast growers. Slow growers include intracellular pathogens such as M. tuberculosis and M. leprae. Fast growers consist of opportunistic pathogens like M. abscessus, M. fortuitum and M. avium, and saprophytic bacteria like M. smegmatis and M. phlei [2]. These rod-shaped bacilli are metabolically flexible, able to growth on multiple kinds of carbon or nitrogen sources [3]. In the genus, both aerobic and anaerobic mycobacteria exist [4], though the aerobic mycobacteria pose the most serious public health threats.
1.2 Mycobacterium tuberculosis, an ancient threat to humankind
The causative agent of tuberculosis (TB), Mycobacterium tuberculosis (Mtb), is thought to have caused disease in human populations for nearly 20,000 years [5]. At present, at least nine million new cases occur every year, with a million deaths from TB ravaging populations globally [6]. TB is treatable with combination antibiotic therapy that is largely effective. However, given both the scale of the epidemic and the emergence if drug resistance, both TB and drug resistant TB pose serious threats to human populations.
While new drugs have been released for use in drug-resistant TB, the world is in desperate need of new antimicrobials for tuberculosis [7].
Easily aerosolized due to its hydrophobic wall, Mtb is inhaled into the human lung, where it infects alveolar macrophages [8]. The hallmark feature of tuberculosis is a
2 granuloma – a mixture of bacterial-infected macrophages and immune cells, walled off from the surrounding lung vasculature [9]. Within the granuloma, and specifically, in the macrophage, Mtb faces an environment with shifting nutrients, oxidative stress and other immune pressures [10].
1.3 NTMs: the tougher cousins of Mtb
While Mtb poses a larger threat world-wide, non-tuberculous mycobacterial (NTM) infection is on the rise [11, 12]. NTMs cause a range of infections from skin and soft tissue abscesses to lung infection in immunocompromised patients [13-15]. NTMs are found in the environment but spread into human populations, likely due to their ability to grow in harsh environments and resist sanitizing chemicals such as chlorine [16, 17]. In addition to their capacity to grow in rather harsh environments, NTMs are innately resistant to a number of antibiotics[18].
1.4 Adaptation in mycobacteria
Biological observations in mycobacterial systems have been largely driven by the public need to mitigate mycobacterial diseases, both tuberculosis and NTM disease. In the case of Mtb drug resistance, a number of bacterial genome-wide association studies have defined causative mutations of drug resistance. These mutations will not be discussed here, as they have been extensively characterized and reviewed elsewhere
[19, 20]. In addition, organisms like Mtb, M.smegmatis (Msm), and M. abscessus (Mab) have been studied to understand adaptive responses that mycobacteria elicit to survive in vivo conditions or tolerate drug treatment. At present, we have a strong
3 understanding of transcriptional programs that enable mycobacterial survival in a macrophage and mycobacterial drug resistance.
1.4.1 Transcriptional adaptation to drugs
Mycobacteria that cause human disease are often hard to treat with antibiotics. Both
Mtb and NTMs are intrinsically resistant to several classes of antibiotics, and treatment is further complicated by acquired drug-resistance against normally effective antimycobacterial agents. Generally, mycobacteria employ several adaptive responses through which they can resist drugs. Mycobacteria encode a number of transcriptionally- controlled efflux pumps that can clear the cell of the toxic antibiotic [21, 22].
Mycobacteria also encode a number of drug-inactivating enzymes that can detoxify the cell of everything from beta-lactams to aminoglycosides[23]. In addition to enzymes that inactivate the drugs, mycobacterial can upregulate expression of target-modifying enzymes that protect the cell from drug binding. In the case of the ribosome, these mechanisms become important for resistance to aminoglycosides and macrolides [24,
25]. For example, in fast-growing mycobacteria and slow-growing mycobacteria, WhiB7 is a transcription factor that is activated by a number of drugs including macrolides. This transcriptional program upregulates ribosome-modifying enzymes along with other cellular pumps and detoxifying enzymes that enable antibiotic resistance [26, 27].
1.4.2 Transcriptional control in the host
For Mtb infection, adaptation to the macrophage and subversion of the subsequent immune responses is imperative to successful residence. A number of studies have
4 detailed changes in the bacterial transcriptome and proteome in different infection models, from the macrophage to the guinea pig [28-30]. While each model produces a distinct bacterial expression profile [31], there are a number of transcription factors that unanimously mediate critical changes in vivo. IdeR, PhoPR, and DosRS regulate major responses of Mtb to changes in oxygen tension, metal availability, and other host pressures [32-35]. Further, the ESX secretion systems provide a major weapon inside the host cell. Mtb has 5 different ESX systems that are responsible for skewing innate immune reactions of the macrophage and acquiring resources that are limited in vivo
[36-41]. These systems are transcriptionally controlled by factors such as WhiB6, EspR and Pst-RegX, which induce responses in a macrophage environment [35, 42, 43]. Our understanding about regulatory processes that are critical to Mtb infection has defined by identification of transcriptional programs. Recent technological advances have allowed us to globally characterize RNA structure/ small RNA species, solve macromolecular structures with increasing resolution, and quantify global protein synthesis rates. Findings from these methodologies indicate that a significant amount of cellular regulation occurs after transcription, and during the process of protein synthesis
[44-48].
1.5 Regulation at the ribosome: the renaissance of nuance
When bacteriologists think “regulation,” they generally mean transcriptional regulation as a response to an external or internal stimuli. And there’s a good reason for this - there is a long history of understanding bacterial physiology through changes in the makeup of the transcriptome as a readout of adaptation. Though, the correlation
5 between transcript and protein level is imperfect and indicates another form of regulation may benefit the cell – translational regulation [49-51]. There are now an increasing number of examples where protein abundance is regulated from the ribosome, either globally or at the level of specific proteins. To understand the implications of translational regulation, we explore changes in ribosome composition, the makeup of RNAs involved in translation, and proteins that may associate to a ribosome. Throughout, this review defines examples of both homeostatic regulation that maintains the proteome, and adaptive regulation that occurs as a response to a stimulus. Bacteria use translational regulation to build, adapt and preserve cellular components during a wide range of lifestyles.
1.5.1 The landscape of ribosome composition
A bacterial ribosome is composed of approximately 55 proteins and 3 ribosomal rRNAs
[52], yet these numbers do not encapsulate the variation that occurs in the ribosomal protein composition and at the ribosomal RNA (rRNA). Changes in ribosomal protein composition and rRNA modifications are closely tied to the physiological state of the cell. [53, 54] . Variation in protein composition of the ribosome has been extensively studied in eukarya [55], but is less well understood in bacteria [56]. To date, the most apparent changes in ribosomal protein composition occur when a bacterial cell experiences changes in growth medium, either through entry into stationary phase or nutrient limitation. Organisms may encode alternative ribosomal proteins that can replace their paralogous counter parts on the ribosome. In Escherichia coli (E.coli), paralogs of large subunit proteins are found in ribosomes of cells from stationary phase
6 [57]. Additionally, in mycobacteria, zinc depletion causes the expression of ribosomal protein paralogs. The expression is a cellular response to scavenge excess zinc in the cell. These proteins, paralogs of L28, L33, S14, and S18, do not require zinc to bind the ribosome that the normal versions of the proteins do [58, 59]. Both of these alterations in ribosomal composition are more indicative of a shift in nutrient requirements than they are of any change in ribosome activity. Some data indicates that these alternative ribosomes may better bind ribosome hibernation factors [60]. By this model, a change in ribosomal protein composition would allow the cell to more rapidly preserve its protein synthesis machinery. Perhaps, under certain circumstances, the most useful activity at the ribosome is no activity.
The other key component of the ribosome is rRNA, which provides the ribosome’s enzymatic activity during peptide bond formation. A bacterial ribosome is composed of 3 rRNAs: the 23S rRNA and 5S rRNA that make up the 50S ribosomal subunit and the
16S rRNA that makes up the 30S ribosomal subunit. rRNA can be modified, mainly by methylation, both for normal ribosome function [52, 61] and as a response to a stress signal. Some modifications are critical for log-phase ribosome biogenesis, and loss of the modification decreases ribosome stability [62]. As a response to heat shock, E.coli ribosomes are methylated at a residue on the 23S rRNA. This methylation contributes to ribosome stability at high temperatures [63]. While these modifications alter how the ribosome is built, rRNA modifications are also used by bacteria as a mechanism of antibiotic resistance. High-level resistance to ribosome-targeting antibiotics due to rRNA modifications, has appeared in multiple pathogens [64-66]. Namely, the erythromycin-
7 resistance methyltransferases (Erms) are an expansive class of enzymes responsible for rRNA modification of the 23S rRNA to prevent macrolide binding [67].
Unsurprisingly, this type of modification to the rRNA does result in a fitness cost to the organism, due to altered translation rate of a number of proteins [68]. In the case of
Erms, these enzymes are often upregulated due to exposure to antibiotics [69]. Yet, there are also examples of enzymes that modify rRNA pre-emptively to protect ribosomes from antibiotics. KsgA, a 16S rRNA methyltransferase, modifies a part of the rRNA during ribosome assembly; this modification turns out to be critical to ribosome maturation, but also determines intrinsic susceptibility to a ribosome-targeting antibiotic, kasugamycin. Loss of this modification, through inactivation of KsgA, has been observed and contributes to kasugamycin and streptomycin resistance [70, 71].
Modification to the ribosome clearly plays a protective role in the cell against a number of stresses. In a single bacterial cell, what is the balance of variation in structural ribosome components that supports productive translation versus remodeling that protects, but possibly inhibits productive translation? It is likely that the more diversity in the ribosome composition, the more advantageous it is for the bacterial cell, regardless of the environment.
1.5.2 Decoding the message: associating RNAs to the ribosome
Features of the messenger RNA (mRNA) or transfer RNA (tRNA) contribute to protein expression in the cell, through changes in the translation rate. While we admit that a wealth of non-coding RNAs contribute to changes in the cellular proteome, we direct the readers to expert reviews on the matter [72, 73]. This section discusses how each
8 ribosome-associated RNA species functions normally to support production of the proteome and provide examples of how mRNA or tRNA features can be conditionally used to inform the outcome of translation at a given transcript.
A number of mRNA features exist in each coding sequence that contributes to the proper abundance and folding of the protein. For optimal initiation of any given mRNA, sequences of codons that create decreased mRNA structure at the 5’ end of a transcript contribute to increased translational efficiency [74, 75]. Signal sequences, which are specialized N-terminal peptides programming protein export, require a distinct codon usage to encourage proper export [76, 77]. It has become clear that the choice of codons in each message has been optimized to ensure proper co-translational folding of the peptide as it travels through the ribosome exit tunnel [78-80]. Further, membrane proteins require specific mRNA elements to ensure proper co-translational secretion and folding through the membrane [81]. The advancement of ribosome profiling in the field of bacteriology continues to reveal intrinsic elements of mRNA that contribute to proper synthesis of a given protein [47, 82].
Our first understanding that some level of post-transcriptional regulation existed in the mRNA came from the discovery of the leader peptide-mediated regulation of the trp operon, encoding tryptophan biosynthetic machinery. In 1981, Yanofsky showed that stalling of the ribosome on adjacent tryptophan codons induced formation of a specific mRNA secondary structure, that negatively regulated transcriptional expression of the tryptophan biosynthetic pathway [83]. Decades later, we have a broader picture of the
9 small molecules that can induce changes to the RNA structure [84] and thereby, changes in protein expression. Specifically, riboswitches are a class of RNA conformations that exist in the 5’ untranslated region (UTR) or 3’ UTR of a transcript, and can affect both transcription and translation of surrounding messages.
Riboswitches are broadly defined as mRNA structures sensitive to the presence of a molecule or metabolite. The shifting in mRNA structures, due to the metabolite, usually affects the expression of messages necessary for dealing with the physiological consequences of that molecule. In the instances of 5’ cis riboswitches, that affect translational initiation, we find that a variety of molecules such as antibiotics [85, 86], amino acids [84], and metals [87] can interact with untranslated mRNA to promote or prevent translational initiation. For example, an aminoglycoside antibiotic can facilitate a change in the RNA structure in the 5’ UTR of an aminoglycoside acetyltransferase enzyme. This structural change stimulates increased expression of the enzyme and thus, efficient detoxification of the antibiotic from the cell [85]. In addition to small molecules that can mediate mRNA structural changes, temperature can also promote alterations in protein expression through changes in mRNA conformation. Called ‘RNA thermometers’, these regions of mRNA adopt different conformations dependent on the temperature. These ‘RNA thermometers’ adapt expression during a change in temperature that may enable the synthesis of protective proteins in a changing environment or during infection [88, 89]. All of these mechanisms rely on modulation of translational initiation to help the cell to adjust its protein levels to a stimulus, in a nearly immediate manner.
10
While mRNA structure affects translation initiation rates, tRNA pools and their specific modifications change translational elongation rates. Environmental changes that the bacterial cell experiences can cause remodeling of the pool of tRNAs. These changes cause alterations in the make-up of the cellular proteome, through tRNA-related changes in translation rate. Aminoacylated tRNA pools are susceptible to change during starvation. Specifically, decreases in the availability of an amino acid may lead to differential charging across tRNAs of that amino acid. In one example from B. subtillis, rare serine codons are over-represented in the transcript of a regulator of biofilm formation, SinR. Upon serine starvation, the serine tRNA pools because unequally charged, causing slowed translation at the rare serine codons [90]. As a result, decreased levels of SinR derepresses genes involved in biofilm formation in the bacterial population. Codon-specific pauses related to changing tRNA pools are found in other organisms and likely contributes to metabolic flexibility across bacteria [91-93].
tRNA modifications are widespread and often required for growth in bacteria [94]. These
‘baseline’ tRNA modifications occur upon synthesis of the tRNA and enable the tRNA to bind all codons of their corresponding amino acid with fidelity. Additionally, some modifications promote tRNA stability, and prevent their degradation [69, 95]. More recently, it has been shown that conditions such as hypoxia and oxidative stress selectively program the modification of certain tRNAs [96]. These modifications ultimately change the rate of translation for a subset of proteins, contributing to the remodeling of the proteome as a response to stress [96, 97]. The code of codons in an
11 mRNA transcript not only helps specify the choice of amino acid, but also allows for dynamic control over protein production in multiple physiological scenarios.
1.5.3 Friends who tell the ribosome to be quiet
In times of stress, bacteria contain proteins that shutdown or destroy protein synthesis molecules in a variety of ways. One of the most conserved responses in the bacterial kingdom is the stringent response, a multi-part cascade that shuts down all synthesis and replication machinery in response to starvation/stress signals [98]. Generally, a protein called RelA structurally senses uncharged tRNAs in the ribosome, and synthesizes an alarmone, pp(p)Gpp, that depletes the cell of GTP. This alarmone binds to and inhibits cellular GTPases that are essential for protein synthesis [99, 100]. While this global response to nutrient starvation has broad consequences for energy consumption in the cell, it is also clear that specific preservation of ribosomes is necessary for bacteria to survive these environmental shifts. In fact, the stringent response induces production of hibernation factors that associate to the ribosome and inhibit translation [101, 102]. A number of ribosome-associating factors, such as hibernation promoting factor (Hpf) and ribosome modulation factor(Rmf), sequester one or two ribosomes by binding to the peptidyl-transferase-center (PTC) and preserving the nanomachine in an inactive state [44, 103]. While the mechanism of ribosome hibernation is preserved across organisms, the signals that program hibernation factor activity are a reflection of the organism’s environmental needs [104-107]. Upon the expression of ribosome hibernation factors, what percentage of ribosomes become inactive in a cell? Recently, in E.coli, it has been observed that the translation rate
12 within a cell is maintained during starvation, but accompanied by a reduction in the number of translating ribosomes [108]. Even during log-phase growth, some bacteria hibernate a smaller population of their ribosomes [109], indicating that again, ribosome heterogeneity is perhaps used to create flexibility in cells. While this class of proteins seems to preserve translational machinery, another group of proteins is responsible for both preservation and outright destruction of the protein synthesis machinery.
Toxin-antitoxin (TA) systems are pairs of a ‘toxin’ molecule, and an ‘antitoxin’ molecule, co-regulated in a manner that allows the toxin to act on its cellular target as a precise response to conditions like intracellular residence, starvation, hypoxia or antibiotics
[110-112]. A large number of the toxin-antitoxin systems target protein synthesis machinery, and these systems are hypothesized to act as brakes on bacterial growth in environments that might be deleterious [113, 114]. Like the silencing factors discussed above, some toxins bind to ribosome subunits or elongation factors to inhibit translation, but preserve the machinery [115-117]. Other toxins irreversibly damage cellular protein synthesis resources, by cleaving rRNA, tRNA pools, and either specifically or non- specifically cleaving mRNA molecules [115, 118-120]. Although it was initially shown that a toxin called MazF, selects for translation of leaderless transcripts during a stress response in E.coli [121], recent investigation has clarified that in fact, MazF cleaves ribosomal protein transcripts and rRNA [122]. This finding has reaffirmed that protein synthesis-targeting toxins inhibit and/or destroy translational machinery in the cell.
1.5.4 Friends who are just helpful to the ribosome
13 When protein synthesis is uninhibited, ribosomes require associated factors to efficiently translate messages. The essential proteins for translation, mainly initiation factors and elongation factors, are those that carry the energy to encourage structural transitions in the ribosome from subunit assembly/ disassembly to translocation [123, 124]. A number of translational GTPases are ubiquitously present and essential across bacteria, while others exhibit organism-specific requirements [125]. Here, we focus on factors that associate to the ribosome but may not be required for every translational event in a cell.
Elongation factor P (EF-P) is a ribosome-associated factor found in many organisms, but only required for translation of proteins containing strings of proline residues [126].
Unlike other elongation factors, EF-P binds a distinct location on the 70S, while still interacting with the PTC [127]. EF-P appears to contribute to efficient translation at high growth rate [128], and loss of this translation factor causes deficits related to the function of the proline-enriched protein [129]. Additionally, it may be responsible for regulating the translation of proteins that contain other motifs able to stall translation
[130]. BPI-inducible protein A (BipA) and elongation factor 4 (LepA) are also translational GTPases that have unresolved roles at the ribosome. Both conspicuously occupy the same position as EF-G on the 70S ribosome [131, 132], indicating that some translational events may require remodeling at the ribosome through alternative translation factors. That said, alternative roles for these proteins in ribosome biogenesis have been proposed, indicating that their activity at the ribosome is complex [133].
BipA’s role becomes important in adaptation to low temperature [134, 135], along with a number of other conditions unique to the organism [136, 137]. LepA, the more conserved of the two, clearly enables production of important membrane respiration
14 systems in eukaryotic mammalian and plant systems[138, 139]. In bacteria, LepA confers some benefit to the cell under conditions of low pH and altered magnesium concentrations [140, 141]. These elongation factors likely play a role in tuning expression of certain parts of organismal proteomes, such that proteins are translated to their optimal levels. In the case of EF-P, it is clear that this elongation factor is required for proteins with a certain amino acid content. For the other two elongation factors, the protein repertoire or message repertoire that they specialize in during protein synthesis remains to be seen.
1.5.5 Conclusions: bacterial translational regulation
Regulation of protein synthesis in bacteria enables control over the homeostatic level of proteins in the proteome and the conditional level of a protein that might be required as part of an adaptive response. The result of homeostatic translational control, such as innate codon bias or stabilizing modification to the ribosome, is often permanent, wired into the genome of the organism. The adaptive responses are on the other hand, more often transient changes to the cell, such as ppGpp inhibition of cellular GTPases, or changing of charged tRNA pools. Clearly, a number of stressful environments may threaten the integrity of the ribosome. Yet, the protein factors that preserve the ribosome, as discussed above, also allow for resumption of productive translation. The essence of translational regulation is the balance between dependable, efficient synthesis of the proteome and the flexibility found in the factors that alter the proteome as a timely response.
15 1.6 Mycobacterial translational features: what do they mean?
As of recent, it has become clear that mycobacteria possess a number of unique features that likely contribute to a distinct set of post-transcriptional controls on cell physiology. Mycobacteria possess an unusually large number of leaderless transcripts, mRNAs with no 5’ UTR or discernable Shine-Dalgarno sequence. Without these features, it is unclear how translation is initiated from these transcripts. In addition, some of these leaderless transcripts encode essential proteins in mycobacteria, underlining the necessity for an alternative, unknown mechanism of translational initiation [142, 143]. The mycobacterial ribosome is also unique among known bacterial ribosomes, consisting of a number of extra ribosomal proteins and rRNA extentions
[144-146]. The physiological consequences of theses structural alterations are not yet clear. In addition, as previously discussed, the mycobacterial ribosome can be structurally remodeled under zinc starvation [60, 147, 148]. Mycobacteria also encode a number of adaptive processes that are induced under hypoxia to preserve ribosomes
[149], and change the rate of translation using tRNA modifications [96]. Given the mycobacterial-specific qualities found in the mRNA and in the ribosome, we hypothesize that translational regulation in mycobacteria may reveal a number of critical processes that are imperative for growth, drug tolerance, and in vivo survival.
1.7 Summary of Aims
This dissertation aims to define mechanisms of control in mycobacterial physiology through the lens of factors associated with translation. We intend for the results of this dissertation to both provide novel avenues for theraupeutic development, and clarify the
16 critical mechanisms by which mycobacteria maintain and adapt their proteome. In
Chapter 2, we detail the pathway by which a formerly mysterious, ribosome-associated
GTPase, LepA, controls protein synthesis of a family of mycobacterial porins. We use cellular measurements of protein levels and RNA levels to define the scope of LepA control of the proteome. We find that in M. smegmatis, LepA regulates translation of specific porin mRNAs to maintain permeability through mycobacterial cell wall. We strongly believe that LepA’s function in mycobacteria is preserved across organisms. In
Chapter 3, we perform a biochemical screen for novel interacting partners of the M. tuberculosis ribosome that might be involved in ribosomal regulation. We use quantitative proteomics to assess the stoichiometry of the protein-ribosome associations in multiple Mtb physiological states. Surprisingly, we find that Mtb ribosomes are associated with a number of important membrane secretion systems. We also discover a number of previously uncharacterized ribosome-associated proteins with essential genetic requirements, perhaps due to their role at the ribosome. Integrating information from these two orthogonal approaches, we conclude that mycobacterial regulate translation to properly synthesize and build proteins into their most valuable compartment, the mycobacterial membrane.
1.8 References
1. Grange, J.M., The biology of the genus Mycobacterium. Soc Appl Bacteriol Symp Ser, 1996. 25: p. 1S-9S. 2. Fedrizzi, T., et al., Genomic characterization of Nontuberculous Mycobacteria. Sci Rep, 2017. 7: p. 45258. 3. Berney, M. and G.M. Cook, Unique flexibility in energy metabolism allows mycobacteria to combat starvation and hypoxia. PLoS One, 2010. 5(1): p. e8614. 4. Gillespie, J., L.L. Barton, and E.W. Rypka, Phenotypic changes in mycobacteria grown in oxygen-limited conditions. J Med Microbiol, 1986. 21(3): p. 251-5.
17 5. Kapur, V., T.S. Whittam, and J.M. Musser, Is Mycobacterium tuberculosis 15,000 years old? J Infect Dis, 1994. 170(5): p. 1348-9. 6. Raviglione, M. and G. Sulis, Tuberculosis 2015: Burden, Challenges and Strategy for Control and Elimination. Infect Dis Rep, 2016. 8(2): p. 6570. 7. Lohrasbi, V., et al., Trends in the discovery of new drugs for Mycobacterium tuberculosis therapy with a glance at resistance. Tuberculosis (Edinb), 2018. 109: p. 17-27. 8. Shiloh, M.U., Mechanisms of mycobacterial transmission: how does Mycobacterium tuberculosis enter and escape from the human host. Future Microbiol, 2016. 11: p. 1503-1506. 9. Ehlers, S. and U.E. Schaible, The granuloma in tuberculosis: dynamics of a host- pathogen collusion. Front Immunol, 2012. 3: p. 411. 10. Martin, C.J., A.F. Carey, and S.M. Fortune, A bug's life in the granuloma. Semin Immunopathol, 2016. 38(2): p. 213-20. 11. Lin, C., et al., Increasing Prevalence of Nontuberculous Mycobacteria in Respiratory Specimens from US-Affiliated Pacific Island Jurisdictions(1). Emerg Infect Dis, 2018. 24(3): p. 485-491. 12. Donohue, M.J. and L. Wymer, Increasing Prevalence Rate of Nontuberculous Mycobacteria Infections in Five States, 2008-2013. Ann Am Thorac Soc, 2016. 13(12): p. 2143-2150. 13. Misch, E.A., C. Saddler, and J.M. Davis, Skin and Soft Tissue Infections Due to Nontuberculous Mycobacteria. Curr Infect Dis Rep, 2018. 20(4): p. 6. 14. Gonzalez-Santiago, T.M. and L.A. Drage, Nontuberculous Mycobacteria: Skin and Soft Tissue Infections. Dermatol Clin, 2015. 33(3): p. 563-77. 15. Johnson, M.M. and J.A. Odell, Nontuberculous mycobacterial pulmonary infections. J Thorac Dis, 2014. 6(3): p. 210-20. 16. Feazel, L.M., et al., Opportunistic pathogens enriched in showerhead biofilms. Proc Natl Acad Sci U S A, 2009. 106(38): p. 16393-9. 17. Nishiuchi, Y., T. Iwamoto, and F. Maruyama, Infection Sources of a Common Non-tuberculous Mycobacterial Pathogen, Mycobacterium avium Complex. Front Med (Lausanne), 2017. 4: p. 27. 18. Falkinham, J.O., 3rd, Challenges of NTM Drug Development. Front Microbiol, 2018. 9: p. 1613. 19. Hicks, N.D., et al., Clinically prevalent mutations in Mycobacterium tuberculosis alter propionate metabolism and mediate multidrug tolerance. Nat Microbiol, 2018. 3(9): p. 1032-1042. 20. Manson, A.L., et al., Genomic analysis of globally diverse Mycobacterium tuberculosis strains provides insights into the emergence and spread of multidrug resistance. Nat Genet, 2017. 49(3): p. 395-402. 21. Szumowski, J.D., et al., Antimicrobial efflux pumps and Mycobacterium tuberculosis drug tolerance: evolutionary considerations. Curr Top Microbiol Immunol, 2013. 374: p. 81-108. 22. Black, P.A., et al., Energy metabolism and drug efflux in Mycobacterium tuberculosis. Antimicrob Agents Chemother, 2014. 58(5): p. 2491-503.
18 23. Kashyap, A., P.K. Singh, and O. Silakari, Mechanistic investigation of resistance via drug-inactivating enzymes in Mycobacterium tuberculosis. Drug Metab Rev, 2018. 50(4): p. 448-465. 24. Luthra, S., A. Rominski, and P. Sander, The Role of Antibiotic-Target-Modifying and Antibiotic-Modifying Enzymes in Mycobacterium abscessus Drug Resistance. Front Microbiol, 2018. 9: p. 2179. 25. Maurer, F.P., et al., Aminoglycoside-modifying enzymes determine the innate susceptibility to aminoglycoside antibiotics in rapidly growing mycobacteria. J Antimicrob Chemother, 2015. 70(5): p. 1412-9. 26. Morris, R.P., et al., Ancestral antibiotic resistance in Mycobacterium tuberculosis. Proc Natl Acad Sci U S A, 2005. 102(34): p. 12200-5. 27. Nash, K.A., et al., Intrinsic macrolide resistance in rapidly growing mycobacteria. Antimicrob Agents Chemother, 2006. 50(10): p. 3476-8. 28. Kruh, N.A., et al., Portrait of a pathogen: the Mycobacterium tuberculosis proteome in vivo. PLoS One, 2010. 5(11): p. e13938. 29. Cappelli, G., et al., Profiling of Mycobacterium tuberculosis gene expression during human macrophage infection: upregulation of the alternative sigma factor G, a group of transcriptional regulators, and proteins with unknown function. Res Microbiol, 2006. 157(5): p. 445-55. 30. Schnappinger, D., et al., Transcriptional Adaptation of Mycobacterium tuberculosis within Macrophages: Insights into the Phagosomal Environment. J Exp Med, 2003. 198(5): p. 693-704. 31. Ward, S.K., et al., Transcriptional profiling of mycobacterium tuberculosis during infection: lessons learned. Front Microbiol, 2010. 1: p. 121. 32. Pandey, R. and G.M. Rodriguez, IdeR is required for iron homeostasis and virulence in Mycobacterium tuberculosis. Mol Microbiol, 2014. 91(1): p. 98-109. 33. Blasco, B., et al., Virulence regulator EspR of Mycobacterium tuberculosis is a nucleoid-associated protein. PLoS Pathog, 2012. 8(3): p. e1002621. 34. Mehra, S., et al., The DosR Regulon Modulates Adaptive Immunity and Is Essential for Mycobacterium tuberculosis Persistence. Am J Respir Crit Care Med, 2015. 191(10): p. 1185-96. 35. Cao, G., et al., EspR, a regulator of the ESX-1 secretion system in Mycobacterium tuberculosis, is directly regulated by the two-component systems MprAB and PhoPR. Microbiology, 2015. 161(Pt 3): p. 477-89. 36. Wong, K.W., The Role of ESX-1 in Mycobacterium tuberculosis Pathogenesis. Microbiol Spectr, 2017. 5(3). 37. Groschel, M.I., et al., ESX secretion systems: mycobacterial evolution to counter host immunity. Nat Rev Microbiol, 2016. 14(11): p. 677-691. 38. Chen, Z., et al., Mycobacterial WhiB6 Differentially Regulates ESX-1 and the Dos Regulon to Modulate Granuloma Formation and Virulence in Zebrafish. Cell Rep, 2016. 16(9): p. 2512-24. 39. Ates, L.S., et al., The ESX-5 System of Pathogenic Mycobacteria Is Involved In Capsule Integrity and Virulence through Its Substrate PPE10. PLoS Pathog, 2016. 12(6): p. e1005696.
19 40. Wassermann, R., et al., Mycobacterium tuberculosis Differentially Activates cGAS- and Inflammasome-Dependent Intracellular Immune Responses through ESX-1. Cell Host Microbe, 2015. 17(6): p. 799-810. 41. Ates, L.S., et al., Essential Role of the ESX-5 Secretion System in Outer Membrane Permeability of Pathogenic Mycobacteria. PLoS Genet, 2015. 11(5): p. e1005190. 42. Bosserman, R.E., et al., WhiB6 regulation of ESX-1 gene expression is controlled by a negative feedback loop in Mycobacterium marinum. Proc Natl Acad Sci U S A, 2017. 114(50): p. E10772-E10781. 43. Elliott, S.R., D.W. White, and A.D. Tischler, Mycobacterium tuberculosis Requires Regulation of ESX-5 Secretion for Virulence in Irgm1-Deficient Mice. Infect Immun, 2019. 87(2). 44. Flygaard, R.K., et al., Cryo-EM structure of the hibernating Thermus thermophilus 100S ribosome reveals a protein-mediated dimerization mechanism. Nat Commun, 2018. 9(1): p. 4179. 45. Frauenfeld, J., et al., Cryo-EM structure of the ribosome-SecYE complex in the membrane environment. Nat Struct Mol Biol, 2011. 18(5): p. 614-21. 46. Estrozi, L.F., et al., Cryo-EM structure of the E. coli translating ribosome in complex with SRP and its receptor. Nat Struct Mol Biol, 2011. 18(1): p. 88-90. 47. Li, G.W., et al., Quantifying absolute protein synthesis rates reveals principles underlying allocation of cellular resources. Cell, 2014. 157(3): p. 624-35. 48. Gerrick, E.R., et al., Small RNA profiling in Mycobacterium tuberculosis identifies MrsI as necessary for an anticipatory iron sparing response. Proc Natl Acad Sci U S A, 2018. 115(25): p. 6464-6469. 49. Guimaraes, J.C., M. Rocha, and A.P. Arkin, Transcript level and sequence determinants of protein abundance and noise in Escherichia coli. Nucleic Acids Res, 2014. 42(8): p. 4791-9. 50. Maier, T., M. Guell, and L. Serrano, Correlation of mRNA and protein in complex biological samples. FEBS Lett, 2009. 583(24): p. 3966-73. 51. Taniguchi, Y., et al., Quantifying E. coli proteome and transcriptome with single- molecule sensitivity in single cells. Science, 2010. 329(5991): p. 533-8. 52. Kaczanowska, M. and M. Ryden-Aulin, Ribosome biogenesis and the translation process in Escherichia coli. Microbiol Mol Biol Rev, 2007. 71(3): p. 477-94. 53. Slavov, N., et al., Differential Stoichiometry among Core Ribosomal Proteins. Cell Rep, 2015. 13(5): p. 865-73. 54. Polikanov, Y.S., et al., Structural insights into the role of rRNA modifications in protein synthesis and ribosome assembly. Nat Struct Mol Biol, 2015. 22(4): p. 342-344. 55. Genuth, N.R. and M. Barna, Heterogeneity and specialized functions of translation machinery: from genes to organisms. Nat Rev Genet, 2018. 19(7): p. 431-452. 56. Byrgazov, K., O. Vesper, and I. Moll, Ribosome heterogeneity: another level of complexity in bacterial translation regulation. Curr Opin Microbiol, 2013. 16(2): p. 133-9.
20 57. Lilleorg, S., et al., Bacterial ribosome heterogeneity: Changes in ribosomal protein composition during transition into stationary growth phase. Biochimie, 2019. 156: p. 169-180. 58. Nanamiya, H., et al., Zinc is a key factor in controlling alternation of two types of L31 protein in the Bacillus subtilis ribosome. Mol Microbiol, 2004. 52(1): p. 273- 83. 59. Prisic, S., et al., Zinc regulates a switch between primary and alternative S18 ribosomal proteins in Mycobacterium tuberculosis. Mol Microbiol, 2015. 97(2): p. 263-80. 60. Li, Y., et al., Zinc depletion induces ribosome hibernation in mycobacteria. Proc Natl Acad Sci U S A, 2018. 115(32): p. 8191-8196. 61. Sergiev, P.V., et al., Structural and evolutionary insights into ribosomal RNA methylation. Nat Chem Biol, 2018. 14(3): p. 226-235. 62. Ishiguro, K., T. Arai, and T. Suzuki, Depletion of S-adenosylmethionine impacts on ribosome biogenesis through hypomodification of a single rRNA methylation. Nucleic Acids Res, 2019. 63. Bugl, H., et al., RNA methylation under heat shock control. Mol Cell, 2000. 6(2): p. 349-60. 64. Nash, K.A., B.A. Brown-Elliott, and R.J. Wallace, Jr., A novel gene, erm(41), confers inducible macrolide resistance to clinical isolates of Mycobacterium abscessus but is absent from Mycobacterium chelonae. Antimicrob Agents Chemother, 2009. 53(4): p. 1367-76. 65. Saribas, Z., F. Tunckanat, and A. Pinar, Prevalence of erm genes encoding macrolide-lincosamide-streptogramin (MLS) resistance among clinical isolates of Staphylococcus aureus in a Turkish university hospital. Clin Microbiol Infect, 2006. 12(8): p. 797-9. 66. Chen, J.C., et al., Report of erm(B)(+) Campylobacter jejuni in the United States. Antimicrob Agents Chemother, 2018. 62(6). 67. Weisblum, B., Erythromycin resistance by ribosome modification. Antimicrob Agents Chemother, 1995. 39(3): p. 577-85. 68. Gupta, P., et al., Deregulation of translation due to post-transcriptional modification of rRNA explains why erm genes are inducible. Nat Commun, 2013. 4: p. 1984. 69. Chancey, S.T., D. Zahner, and D.S. Stephens, Acquired inducible antimicrobial resistance in Gram-positive bacteria. Future Microbiol, 2012. 7(8): p. 959-78. 70. Phunpruch, S., et al., A role for 16S rRNA dimethyltransferase (ksgA) in intrinsic clarithromycin resistance in Mycobacterium tuberculosis. Int J Antimicrob Agents, 2013. 41(6): p. 548-51. 71. Boehringer, D., et al., Structural insights into methyltransferase KsgA function in 30S ribosomal subunit biogenesis. J Biol Chem, 2012. 287(13): p. 10453-9. 72. Hor, J., S.A. Gorski, and J. Vogel, Bacterial RNA Biology on a Genome Scale. Mol Cell, 2018. 70(5): p. 785-799. 73. Storz, G., J. Vogel, and K.M. Wassarman, Regulation by small RNAs in bacteria: expanding frontiers. Mol Cell, 2011. 43(6): p. 880-91. 74. Goodman, D.B., G.M. Church, and S. Kosuri, Causes and effects of N-terminal codon bias in bacterial genes. Science, 2013. 342(6157): p. 475-9.
21 75. Bhattacharyya, S., et al., Accessibility of the Shine-Dalgarno Sequence Dictates N-Terminal Codon Bias in E. coli. Mol Cell, 2018. 70(5): p. 894-905 e5. 76. Zalucki, Y.M., et al., Signal sequence non-optimal codons are required for the correct folding of mature maltose binding protein. Biochim Biophys Acta, 2010. 1798(6): p. 1244-9. 77. Singh, P., et al., Effect of signal peptide on stability and folding of Escherichia coli thioredoxin. PLoS One, 2013. 8(5): p. e63442. 78. Ingolia, N.T., et al., Genome-wide analysis in vivo of translation with nucleotide resolution using ribosome profiling. Science, 2009. 324(5924): p. 218-23. 79. Buhr, F., et al., Synonymous Codons Direct Cotranslational Folding toward Different Protein Conformations. Mol Cell, 2016. 61(3): p. 341-351. 80. Pechmann, S. and J. Frydman, Evolutionary conservation of codon optimality reveals hidden signatures of cotranslational folding. Nat Struct Mol Biol, 2013. 20(2): p. 237-43. 81. Fluman, N., et al., mRNA-programmed translation pauses in the targeting of E. coli membrane proteins. Elife, 2014. 3. 82. Schrader, J.M., et al., Dynamic translation regulation in Caulobacter cell cycle control. Proc Natl Acad Sci U S A, 2016. 113(44): p. E6859-E6867. 83. Yanofsky, C., Attenuation in the control of expression of bacterial operons. Nature, 1981. 289(5800): p. 751-8. 84. Caron, M.P., et al., Dual-acting riboswitch control of translation initiation and mRNA decay. Proc Natl Acad Sci U S A, 2012. 109(50): p. E3444-53. 85. Jia, X., et al., Riboswitch control of aminoglycoside antibiotic resistance. Cell, 2013. 152(1-2): p. 68-81. 86. Dar, D., et al., Term-seq reveals abundant ribo-regulation of antibiotics resistance in bacteria. Science, 2016. 352(6282): p. aad9822. 87. Dann, C.E., 3rd, et al., Structure and mechanism of a metal-sensing regulatory RNA. Cell, 2007. 130(5): p. 878-92. 88. Morita, M., et al., Heat-induced synthesis of sigma32 in Escherichia coli: structural and functional dissection of rpoH mRNA secondary structure. J Bacteriol, 1999. 181(2): p. 401-10. 89. Weber, G.G., et al., RNA thermometer controls temperature-dependent virulence factor expression in Vibrio cholerae. Proc Natl Acad Sci U S A, 2014. 111(39): p. 14241-6. 90. Subramaniam, A.R., T. Pan, and P. Cluzel, Environmental perturbations lift the degeneracy of the genetic code to regulate protein levels in bacteria. Proc Natl Acad Sci U S A, 2013. 110(6): p. 2419-24. 91. Schrader, J.M., et al., The coding and noncoding architecture of the Caulobacter crescentus genome. PLoS Genet, 2014. 10(7): p. e1004463. 92. Dittmar, K.A., et al., Selective charging of tRNA isoacceptors induced by amino- acid starvation. EMBO Rep, 2005. 6(2): p. 151-7. 93. Elf, J., et al., Selective charging of tRNA isoacceptors explains patterns of codon usage. Science, 2003. 300(5626): p. 1718-22. 94. Diwan, G.D. and D. Agashe, Wobbling Forth and Drifting Back: The Evolutionary History and Impact of Bacterial tRNA Modifications. Mol Biol Evol, 2018. 35(8): p. 2046-2059.
22 95. Kimura, S. and M.K. Waldor, The RNA degradosome promotes tRNA quality control through clearance of hypomodified tRNA. Proc Natl Acad Sci U S A, 2019. 116(4): p. 1394-1403. 96. Chionh, Y.H., et al., tRNA-mediated codon-biased translation in mycobacterial hypoxic persistence. Nat Commun, 2016. 7: p. 13302. 97. Jaroensuk, J., et al., Methylation at position 32 of tRNA catalyzed by TrmJ alters oxidative stress response in Pseudomonas aeruginosa. Nucleic Acids Res, 2016. 44(22): p. 10834-10848. 98. Boutte, C.C. and S. Crosson, Bacterial lifestyle shapes stringent response activation. Trends Microbiol, 2013. 21(4): p. 174-80. 99. Winther, K.S., M. Roghanian, and K. Gerdes, Activation of the Stringent Response by Loading of RelA-tRNA Complexes at the Ribosomal A-Site. Mol Cell, 2018. 70(1): p. 95-105 e4. 100. Wang, B., et al., Affinity-based capture and identification of protein effectors of the growth regulator ppGpp. Nat Chem Biol, 2019. 15(2): p. 141-150. 101. Tagami, K., et al., Expression of a small (p)ppGpp synthetase, YwaC, in the (p)ppGpp(0) mutant of Bacillus subtilis triggers YvyD-dependent dimerization of ribosome. Microbiologyopen, 2012. 1(2): p. 115-34. 102. Izutsu, K., A. Wada, and C. Wada, Expression of ribosome modulation factor (RMF) in Escherichia coli requires ppGpp. Genes Cells, 2001. 6(8): p. 665-76. 103. Beckert, B., et al., Structure of the Bacillus subtilis hibernating 100S ribosome reveals the basis for 70S dimerization. EMBO J, 2017. 36(14): p. 2061-2072. 104. McKay, S.L. and D.A. Portnoy, Ribosome hibernation facilitates tolerance of stationary-phase bacteria to aminoglycosides. Antimicrob Agents Chemother, 2015. 59(11): p. 6992-9. 105. Niven, G.W., Ribosome modulation factor protects Escherichia coli during heat stress, but this may not be dependent on ribosome dimerisation. Arch Microbiol, 2004. 182(1): p. 60-6. 106. Williamson, K.S., et al., Heterogeneity in Pseudomonas aeruginosa biofilms includes expression of ribosome hibernation factors in the antibiotic-tolerant subpopulation and hypoxia-induced stress response in the metabolically active population. J Bacteriol, 2012. 194(8): p. 2062-73. 107. Wada, A., et al., Structure and probable genetic location of a "ribosome modulation factor" associated with 100S ribosomes in stationary-phase Escherichia coli cells. Proc Natl Acad Sci U S A, 1990. 87(7): p. 2657-61. 108. Dai, X., et al., Reduction of translating ribosomes enables Escherichia coli to maintain elongation rates during slow growth. Nat Microbiol, 2016. 2: p. 16231. 109. Basu, A. and M.N. Yap, Ribosome hibernation factor promotes Staphylococcal survival and differentially represses translation. Nucleic Acids Res, 2016. 44(10): p. 4881-93. 110. Kedzierska, B. and F. Hayes, Emerging Roles of Toxin-Antitoxin Modules in Bacterial Pathogenesis. Molecules, 2016. 21(6). 111. Keren, I., et al., Characterization and transcriptome analysis of Mycobacterium tuberculosis persisters. MBio, 2011. 2(3): p. e00100-11.
23 112. Albrethsen, J., et al., Proteomic profiling of Mycobacterium tuberculosis identifies nutrient-starvation-responsive toxin-antitoxin systems. Mol Cell Proteomics, 2013. 12(5): p. 1180-91. 113. Yamaguchi, Y., J.H. Park, and M. Inouye, Toxin-antitoxin systems in bacteria and archaea. Annu Rev Genet, 2011. 45: p. 61-79. 114. Harms, A., et al., Toxins, Targets, and Triggers: An Overview of Toxin-Antitoxin Biology. Mol Cell, 2018. 70(5): p. 768-784. 115. Liu, M., et al., Bacterial addiction module toxin Doc inhibits translation elongation through its association with the 30S ribosomal subunit. Proc Natl Acad Sci U S A, 2008. 105(15): p. 5885-90. 116. Cruz, J.W., et al., Doc toxin is a kinase that inactivates elongation factor Tu. J Biol Chem, 2014. 289(11): p. 7788-98. 117. Arbing, M.A., et al., Crystal structures of Phd-Doc, HigA, and YeeU establish multiple evolutionary links between microbial growth-regulating toxin-antitoxin systems. Structure, 2010. 18(8): p. 996-1010. 118. Mets, T., et al., Fragmentation of Escherichia coli mRNA by MazF and MqsR. Biochimie, 2019. 156: p. 79-91. 119. Walling, L.R. and J.S. Butler, Homologous VapC Toxins Inhibit Translation and Cell Growth by Sequence-Specific Cleavage of tRNA(fMet). J Bacteriol, 2018. 200(3). 120. Schifano, J.M., et al., An RNA-seq method for defining endoribonuclease cleavage specificity identifies dual rRNA substrates for toxin MazF-mt3. Nat Commun, 2014. 5: p. 3538. 121. Vesper, O., et al., Selective translation of leaderless mRNAs by specialized ribosomes generated by MazF in Escherichia coli. Cell, 2011. 147(1): p. 147-57. 122. Culviner, P.H. and M.T. Laub, Global Analysis of the E. coli Toxin MazF Reveals Widespread Cleavage of mRNA and the Inhibition of rRNA Maturation and Ribosome Biogenesis. Mol Cell, 2018. 70(5): p. 868-880 e10. 123. Gualerzi, C.O. and C.L. Pon, Initiation of mRNA translation in bacteria: structural and dynamic aspects. Cell Mol Life Sci, 2015. 72(22): p. 4341-67. 124. Rodnina, M.V., et al., GTPases mechanisms and functions of translation factors on the ribosome. Biol Chem, 2000. 381(5-6): p. 377-87. 125. Margus, T., M. Remm, and T. Tenson, Phylogenetic distribution of translational GTPases in bacteria. BMC Genomics, 2007. 8: p. 15. 126. Doerfel, L.K., et al., EF-P is essential for rapid synthesis of proteins containing consecutive proline residues. Science, 2013. 339(6115): p. 85-8. 127. Glick, B.R., S. Chladek, and M.C. Ganoza, Peptide bond formation stimulated by protein synthesis factor EF-P depends on the aminoacyl moiety of the acceptor. Eur J Biochem, 1979. 97(1): p. 23-8. 128. Tollerson, R., 2nd, A. Witzky, and M. Ibba, Elongation factor P is required to maintain proteome homeostasis at high growth rate. Proc Natl Acad Sci U S A, 2018. 115(43): p. 11072-11077. 129. Choi, E., et al., Elongation factor P restricts Salmonella's growth by controlling translation of a Mg(2+) transporter gene during infection. Sci Rep, 2017. 7: p. 42098.
24 130. Hersch, S.J., et al., Divergent protein motifs direct elongation factor P-mediated translational regulation in Salmonella enterica and Escherichia coli. MBio, 2013. 4(2): p. e00180-13. 131. Kumar, V., et al., Structure of BipA in GTP form bound to the ratcheted ribosome. Proc Natl Acad Sci U S A, 2015. 112(35): p. 10944-9. 132. Gagnon, M.G., J. Lin, and T.A. Steitz, Elongation factor 4 remodels the A-site tRNA on the ribosome. Proc Natl Acad Sci U S A, 2016. 113(18): p. 4994-9. 133. Gibbs, M.R. and K. Fredrick, Roles of elusive translational GTPases come to light and inform on the process of ribosome biogenesis in bacteria. Mol Microbiol, 2018. 107(4): p. 445-454. 134. Choi, E. and J. Hwang, The GTPase BipA expressed at low temperature in Escherichia coli assists ribosome assembly and has chaperone-like activity. J Biol Chem, 2018. 293(47): p. 18404-18419. 135. Grant, A.J., et al., Co-ordination of pathogenicity island expression by the BipA GTPase in enteropathogenic Escherichia coli (EPEC). Mol Microbiol, 2003. 48(2): p. 507-21. 136. Farris, M., et al., BipA: a tyrosine-phosphorylated GTPase that mediates interactions between enteropathogenic Escherichia coli (EPEC) and epithelial cells. Mol Microbiol, 1998. 28(2): p. 265-79. 137. Wang, F., et al., SsTypA1, a chloroplast-specific TypA/BipA-type GTPase from the halophytic plant Suaeda salsa, plays a role in oxidative stress tolerance. Plant Cell Environ, 2008. 31(7): p. 982-94. 138. Ji, D.L., et al., CpLEPA is critical for chloroplast protein synthesis under suboptimal conditions in Arabidopsis thaliana. PLoS One, 2012. 7(11): p. e49746. 139. Zhu, P., et al., Human Elongation Factor 4 Regulates Cancer Bioenergetics by Acting as a Mitochondrial Translation Switch. Cancer Res, 2018. 78(11): p. 2813- 2824. 140. Pech, M., et al., Elongation factor 4 (EF4/LepA) accelerates protein synthesis at increased Mg2+ concentrations. Proc Natl Acad Sci U S A, 2011. 108(8): p. 3199-203. 141. Bijlsma, J.J., et al., Identification of loci essential for the growth of Helicobacter pylori under acidic conditions. J Infect Dis, 2000. 182(5): p. 1566-9. 142. Shell, S.S., et al., Leaderless Transcripts and Small Proteins Are Common Features of the Mycobacterial Translational Landscape. PLoS Genet, 2015. 11(11): p. e1005641. 143. Cortes, T., et al., Genome-wide mapping of transcriptional start sites defines an extensive leaderless transcriptome in Mycobacterium tuberculosis. Cell Rep, 2013. 5(4): p. 1121-31. 144. Yang, K., et al., Structural insights into species-specific features of the ribosome from the human pathogen Mycobacterium tuberculosis. Nucleic Acids Res, 2017. 45(18): p. 10884-10894. 145. Li, Z., et al., Cryo-EM structure of Mycobacterium smegmatis ribosome reveals two unidentified ribosomal proteins close to the functional centers. Protein Cell, 2018. 9(4): p. 384-388.
25 146. Hentschel, J., et al., The Complete Structure of the Mycobacterium smegmatis 70S Ribosome. Cell Rep, 2017. 20(1): p. 149-160. 147. Tobiasson, V., et al., Zinc depletion does not necessarily induce ribosome hibernation in mycobacteria. Proc Natl Acad Sci U S A, 2019. 116(7): p. 2395- 2397. 148. Li, Y., et al., Reply to Tobiasson et al.: Zinc depletion is a specific signal for induction of ribosome hibernation in mycobacteria. Proc Natl Acad Sci U S A, 2019. 116(7): p. 2398-2399. 149. Trauner, A., et al., The dormancy regulator DosR controls ribosome stability in hypoxic mycobacteria. J Biol Chem, 2012. 287(28): p. 24053-63
26 Chapter 2
A conserved translation factor is required for optimal synthesis of a membrane protein family in mycobacteria
Skye R.S. Fishbein1, Ian D. Wolf1, Charles L. Dulberger1, Albert Wang1, Hasmik Keshishian2,
Luke Wallace2, Steven A. Carr2, Thomas R. Ioerger3, E. Hesper Rego4, Eric J. Rubin1,
1Department of Immunology and Infectious Disease, Harvard TH Chan School of Public Health,
Boston, Massachusetts, 02115, United States
2Broad Institute of MIT and Harvard, Cambridge, 02142, United States
3Department of Computer Science and Engineering, Texas A&M University, Texas, 77843,
United States
4Department of Microbial Pathogenesis, Yale University School of Medicine, New Haven,
Connecticut, 06510, United States
Author Contributions:
Conceptualization: S.R.S.F, E.H.R, C.L.D, E.J.R; Methodology: S.R.S.F, E.H.R, C.L.D, E.J.R,
H.K.,T.R.I, S.A.C.; Investigation: S.R.S.F, E.H.R, I.D.W, A.W, L.W.,H.K.,T.R.I,; Data Curation:
S.R.S.F, H.K.,T.R.I; Writing – Original Draft: S.R.S.F, E.H.R, E.J.R.; Writing – Review & Editing:
S.R.S.F, E.H.R, E.J.R., H.K. ,T.R.I., S.A.C., C.L.D, A.C., and T.B. This work is part of a manuscript that was submitted to mBio in March 2019, at the time of writing.
27 Abstract
Ribosomes require the activity of associated GTPases to synthesize protein. Despite strong evolutionary conservation, the roles of many of these remain unknown. For example, elongation factor 4, or LepA, is a ribosome-associated GTPase found in bacteria, mitochondria, and chloroplasts, yet its physiological contribution to cell survival is not clear. Recently, we found that loss of lepA in Mycobacterium smegmatis (Msm) altered tolerance to rifampin, a drug that targets a non-ribosomal cellular process. To uncover the determinants of LepA-mediated drug tolerance, we characterized the whole-cell proteomes and transcriptomes of a lepA deletion mutant relative to a wild-type strain. We found that LepA regulates the abundance of a family of mycobacterial porins. Regulation of MspA, the most prominent porin in Msm, influences the permeability of the mycobacterial cell membrane. Disruption of this regulation results in the drug tolerance phenotype of the lepA mutant. LepA control requires a sequence motif in the 5’ region of the porin transcript. Thus, LepA controls the abundance of specific proteins, likely through regulation of translation.
Introduction
Bacterial cells rely on transcription and translation to create a homeostatic proteome that is suitable for cell growth, yet adaptable to varying environments. From our first understanding of the regulation of the trp operon [1], post-transcriptional responses were thought to provide additional input to adaptive, cellular responses. Indeed, post-transcriptional regulation can preserve the protein synthesis machinery during stress [2] or adjust the proteome to nutrient/small molecule pools [3, 4]. Recent advances in techniques such as quantitative proteomics, ribosome profiling, and cryo-electron microscopy (cryo-EM) have revealed a significant contribution of ribosomal regulation to the maintenance of the homeostatic cellular proteome [5-7]. Regardless of growth environment, bacterial cells are likely composed of a heterogeneous population of ribosomes [8, 9], including those that are dormant and those
28 dedicated to the translation of membrane porteins. [10, 11]. Associating factors and RNA species tune each ribosome, creating a cell with a spatially and temporally-regulated proteome
[8, 12].
Translational regulation is mediated by both intrinsic features in mRNA transcripts and ribosome-associated factors. Structural elements of the 5’ untranslated region (UTR) are responsive to nutrients and antibiotics, such as antibiotic-mediated riboswitches that alter translational initiation [13, 14]. Local codon bias, an indirect measure of mRNA structure and tRNA availability, along a transcript, can also control the folding and abundance of a protein [3,
15, 16]. Yet, the ribosome itself is not a static machine only sensitive to signals encoded in the mRNA. It is subject to regulation through associating factors that govern the outcome of translation. These include initiation factors [17-20] and the elongation GTPases [21-23], many of which are necessary for the initiation and progression of peptide bond formation at the ribosome during translation. A number of other factors, such as elongation factor P (EF-P), BPI-inducible protein A (BipA), and elongation factor 4 (LepA) associate with the 70S ribosome during elongation. Their role in translation appears to be message- and, in some cases, organism- specific [24-26]. These accessory factors are critical to the maintenance of physiological balance through modulation of translation and, ultimately, cell survival [27-29].
LepA is a conserved ribosome-dependent GTPase that is present in cells of almost all organisms, from bacteria to human mitochondria [26, 30]. LepA uses four classical elongation factor protein domains to contact the ribosome and hydrolyze GTP. In fact, it occupies the same position on the 70S ribosome as elongation factor G (EF-G) [31]. The C-terminal domain of
LepA makes contact with the A/P-site tRNA and, likely, alters positioning of the tRNA species on the ribosome [32-34]. While it is non-essential in most bacteria, LepA might confer some fitness benefit in certain growth conditions, such as altered cation concentrations and low pH in certain
29 bacterial species [35-40]. Despite its conservation, the physiologic role of LepA remains unclear.
Two different functions have been proposed for LepA. In E. coli, loss of LepA results in decreased polysome formation as a consequence of altered initiation rate and ribosome assembly defects [33, 40, 41]. Alternatively, structural studies of multiple bacterial LepA homologs indicate that it back-translocates the 70S ribosome during elongation by performing an action that opposes the forward translocation induced by EF-G [42-44] and perhaps participating in translational quality control.
One clue concerning LepA function comes from earlier work that defined the genetic determinants of single cell heterogeneity and drug susceptibility in mycobacteria. Mutations in several genes altered the rate of antibiotic-mediated bacterial killing, including lepA. Loss-of- function mutations in lepA resulted in increased survival in the presence of rifampin as compared to wild type Msm [45].
Here, we investigated the mechanistic basis of altered drug susceptibility in the mycobacterial lepA mutant. We find that, in mycobacteria, LepA regulates the protein levels of members of the mycobacterial porin (Msp) family during translation. LepA-dependent regulation is determined by a 5’ region in the transcript of some, but not all, Msp porins. LepA deficiency results in decreased synthesis of MspA, the major porin in Msm, and a reduction in cell permeability as measured by calcein accumulation and killing by certain antibiotics. Thus, we find that LepA acts as a transcript-specific translation factor.
Results
Loss of LepA results in pleiotropic drug tolerance in mycobacteria
In a previous screen, we found that transposon insertions in lepA were predicted to be associated with decreased accumulation of calcein AM, a dye which fluoresces when cleaved
30 intracellularly, and decreased killing by certain antibiotics [45]. To validate that loss of LepA was responsible for the observed phenotype, we constructed strains of Msm with and without lepA.
We found that strains lacking lepA exhibited an approximately 2-fold decrease in calcein signal
(Appendix Figure A1.1A). As predicted by the initial screen, loss of lepA also resulted in increased tolerance to the antibiotics rifampin and vancomycin (Figure 2.1A,B) [45] but had no effect on isoniazid or linezolid tolerance (Appendix Figure A1.1B,C) and no effect on susceptibility to a variety of translation inhibitors (Appendix Table A1.1).
a b 10 10 Rifampin Vancomycin l l *** 1 1 *** *** *** * 0.1 0.1 * 0.01 0.01 *
0.001 Proportion Surviva 0.001 Proportion Surviva
0.0001 0.0001 0 10 20 30 40 50 0 10 20 30 40 50 Time(hours) Time(hours) WT L5::empty ΔlepA L5::empty ΔlepA L5::lepA c d 1000 0.8
e 800 70S 0.6 600 polysomes
Abs. 260 0.4 50S 30S Fluorescenc 400
200 0.2
0 0.0 0 100 200 300 20 40 60 Time (minutes) Distance from top (mm) 0.03 LepA 0.15 LepA - LepA 0.09 LepA
Figure 1 - Loss of ribosome factor LepA causes altered mycobacterial drug tolerance Figure 2.1 - Loss of ribosome factor causes altered drug tolerance in mycobacteria a-b) Cells were treated with 10X MIC amounts of the corresponding drug. Proportion survival indicates colony-forming units (CFUs) normalized by corresponding CFUs at time zero. All values are mean values with error bars indicating standard error of the mean (SEM) of three biological replicates. ***P <0.001, **P<0.01, *P<0.05, calculated using a two-sided Student’s t-test. c) In vitro translation was carried out using Venus mRNA with increasing concentration of LepA. Each concentration of LepA reflects the molar
31 Figure 2.1 (continued) ratio of LepA to the ribosome. Error bars represent standard deviation across two technical replicates. d) Analysis of ribosome populations by sucrose density centrifugation and fractionation. Distance 0 corresponds to the lightest sucrose fraction. For (c-d), data is representative, and each experiment was performed at least twice.
Mycobacterial LepA aids protein synthesis in vitro and does not contribute to in vivo mycobacterial ribosome stability
To examine the link between a ribosomal factor and the phenotypes we observed, we first hypothesized that, like LepA homologs [32, 40, 46], mycobacterial LepA acts at the ribosome.
To test this, we added purified Msm LepA and purified mRNA, encoding the fluorescent protein
Venus, cell-free translation system and measured protein yield by fluorescence (Appendix
Figure A1.2). As previously observed with homologs of LepA [30, 47], addition of mycobacterial
LepA to in vitro translation reactions increased the yield of fluorescent protein (Figure 2.1C).
LepA could interact with the ribosome either during ribosome biogenesis, translation initiation, or elongation [36, 37, 41]. To determine if mycobacterial LepA plays a similar role at the ribosome, we examined ribosome populations in wild type and LepA-deficient cells. Unlike in E. coli, loss of LepA did not alter polysome formation, a marker of active translation, or distribution of ribosome subunits in Msm (Figure 2.1D). While it is possible that LepA might be involved in initiation during translation, LepA does not appear to influence global ribosome stability in Msm.
Whole cell profiling of lepA mutant reveals dysregulation of a small fraction of the mycobacterial proteome
Loss of LepA in mycobacteria could result in dysregulated translation of a number of proteins, some of which might be involved in drug tolerance. To find candidates for post-transcriptional control by LepA, we measured simultaneous steady state levels of proteins and corresponding transcripts from wild type, the lepA mutant, and complemented Msm strains. To quantify relative abundances of proteins, we used tandem-mass-tagging (TMT) labeling of peptides after tryptic
32 digestion of cell lysates coupled with LC-MS/MS. A total of 4646 proteins have been identified in the experiment with 4549 of them quantified with 2 or more peptides (Appendix Table A2.2).
a b 4 25 MSMEG_2188 porin 20 MSMEG_0546 3 MSMEG_2650 MSMEG_0546 MSMEG_5292 15 2 10 mspA MSMEG_2188 1 mspC mspB
(adjusted p-value) (adjusted p-value) 5 MSMEG_2650 MSMEG_5292 10 10 mspD g g
-lo 0 -lo 0 -1 0 1 -2 -1 0 1 2
log2(protein in ΔlepA/protein in+lepA) log2(RNA in ΔlepA/RNA in+lepA) c d * 10000 *** * *** 4 * 8000 *
ns )
6000 2
4000 (+LepA/-LepA Luminescence Ratio
Calcein Fluorescence 1 2000 *** 0 *** 0.5 WT lepA lepA Δ mspA Δ Δ MspA MspB MspC MspD mspA NanoLuc Δ Figure 2 - Whole cell profiling finds mycobacterial porins altered by loss of LepA Figure 2.2 – Whole cell profiling finds mycobacterial porins altered by loss of LepA a,b) Bolded words represent gene/protein names of the porin family. a) Ratios of protein represent fold changes in lepA knockout normalized by averaged values from the strains containing lepA. Orange dots indicate protein candidates that were significantly altered by loss of LepA. Porin’ indicates the collection of peptides that map to 4 proteins: MspA, MspB, MspC, and MspD. P-values for proteomic ratios were calculated using Student’s two-sided t-test, and adjusted for multiple testing using the Benjamini- Hochberg correction with an α of 0.05. b) Transcripts visualized correspond to protein messages significantly altered. Orange dots represent transcripts with < 1.7-fold change due to loss of LepA. Fold changes and adjusted p-values for RNA levels were generated using DEseq2.1.8. c) MFI of calcein staining across Msm strains, with error bars indicating SEM across three biological replicates. ***P<0.001, **P<0.01, *P<0.05, calculated using a two-sided Student’s t-test. d) Each point represents the average ratio of +LepA luminescence to - LepA luminescence from one experiment of three biological triplicates. ***P<0.001, *P<0.05, calculated using a one-way ANOVA and adjusted for multiple comparisons.
33 Proteins involved in membrane processes were underrepresented in the dataset (Appendix
Figure 1.3A). Of the 4549 proteins, 80 were significantly altered by the loss of LepA (Figure
2.2A). Of these proteins, multiple peptides that correspond to a highly homologous set of four porins: MspA, B, C, D, (Appendix Figure 1.3A) were significantly less abundant the absence of
LepA. As a change in protein abundance could also be caused by a change in transcript level, we used RNA-sequencing to de-couple the transcriptional contribution to the set of regulated proteins from our proteomics (Appendix Table 2.3). Considering the high rank of the ‘porin’ in our altered proteomics data (4th most decreased out of 4549 proteins), minor changes in porin transcript level (406th most decreased out of 6717 transcripts) were not enough to account for the proteomic changes (Figure 2.2B). Thus, we concluded that some of the porin genes in Msm are post-transcriptionally regulated by LepA.
LepA regulates the abundance of MspA protein
The mycobacterial porins are octameric channels built into the outer membrane and are responsible for the uptake of nutrients critical for mycobacterial growth [48]. The four porins in
Msm, encoded by mspA-mspD, are paralogs distributed across the genome. Each gene contains a classic Sec signal sequence that enables co-translational secretion of these proteins into the outer membrane [49, 50]. Strikingly, an mspA mutant, the most abundantly expressed of these porin transcripts[50], phenocopied the lepA mutant in terms of calcein staining and rifampin tolerance [45]. Thus, we hypothesized that lepA and mspA exist in the same genetic pathway. To test this, we deleted mspA in the lepA deletion background, and found that, unlike in the wild type background, deletion of mspA does not lead to a reduction in calcein accumulation (Figure 2.2C). These data suggested that LepA may be controlling the abundance of MspA and related porins. To test this, we fused a Nanoluciferase (NanoLuc) protein to each porin at its C-terminal end and integrated these reporters on the Msm chromosome. With luminescence levels as a proxy for protein abundance, we examined levels of each porin in the
34 presence (+ LepA) or absence (- LepA) of LepA. We found that LepA increases luminescence for fusions with coding sequences of MspA, MspB, and MspC, but not for MspD or NanoLuc alone (Figure 2.2D). To determine if, indeed, MspA was affected at its native membrane location, we visualized an MspA-mRFP fusion in Msm cells with or without LepA using fluorescence microscopy. We found that the fluorescent signal was localized to the membrane, and, in agreement with our luciferase data, decreased in the absence of LepA (Appendix Figure
A1.3C). Notably, loss of LepA decreases the heterogeneity found in MspA signal among cells.
These data indicate that LepA controls the abundance of a subset of outer membrane porins.
LepA appears to modulate translation of three different porins; but does LepA regulation of the abundance of a specific porin account for altered calcein staining? To answer this, we employed an inducible CRISPRi strategy to transcriptionally deplete each porin individually (Appendix
Figure A1.5) [51] and assayed calcein staining. We compared calcein staining in each porin knockdown pair (+ LepA and - LepA) relative to a ‘no sgRNA’ control set to assess LepA dependence permeability. Only depletion of mspA collapsed the LepA-dependent increase in calcein signal (Figure 2.3A).
a 4 b
mspABC lepA uninduced induced
) ** 2 + +
+ - 1 - + Calcein Ratio (+LepA/ -LepA - - 0.5
Target empty mspD mspC mspB mspA Knockdown - + - + - + - + - +
Figure 3 - Loss of LepA causes membrane defects mainly through control of MspA Figure 2.3 – Loss of LepA causes membrane defects mainly through control of MspA a) Cells were stained with calcein-AM and analyzed by flow cytometry. Calcein ratio represents the ratio of mean calcein MFI of +LepA strain relative to -LepA strain, in biological replicate. Error bars indicate
35 Figure 2.3 (continued) SEM. ‘Knockdown’ indicates overnight induction of CRISPRi system with 100 ng/μL of aTc, while ‘no knockdown’ indicates no aTc in overnight growth of the same strains. ‘Empty’ refers to strains containing the control vector with the aTc-inducible CRISPRi system and no target-specific sgRNA. **P<0.01, calculated using a two-sided Student’s t-test. b) Strains with or without lepA were depleted for mspA, mspB, and mspC using a multi-sgRNA knockdown approach. In the lepA column: ‘+’ indicates a copy of lepA expressed from Tweety site in knockout background. ‘-’ indicates an empty vector at Tweety site in knockout. In the mspABC column: ‘+’ indicates presence of control CRISPRi vector with no targeting- sgRNAs. ‘-’ indicates the presence of 3 sgRNAs targeting mspA, mspB, and mspC. ‘Uninduced’ implies growth on plates containing LB-Kan 25 ug/ml, while ‘induced’ growth on plates containing LB-Kan 25 ug/ml, 500 ng/mL aTc, and therefore, induction of CRISPRi-mediated knockdown.
As the mycobacterial porins are the major determinants of mycobacterial outer membrane homeostasis permeability [50], we hypothesized that LepA might contribute to this role in growth. We generated a CRISPRi construct that would simultaneously knockdown mspA, mspB, and mspC expression to examine the growth of Msm with and without LepA. Decreased expression of three of four mycobacterial porins resulted in a substantial growth defect, yet combinatorial loss of LepA and the three LepA-regulated porins resulted in a more pronounced defect (Figure 2.3B). The growth defect due to loss of both lepA and the three LepA-regulated
Msp proteins suggests that LepA control of these three porins is essential to membrane integrity and mycobacterial growth.
LepA regulates MspA through mRNA sequence determinants
How does LepA recognize MspA as a target? Given that Msm LepA functions during translation and may change the rate of translation [30], we hypothesized that the LepA-controlled porins have sequence features that allow for LepA control of translation rate. To define LepA-related features of the porin, either within the mRNA sequence or the protein sequence, we took advantage of the fact that LepA affects the abundance of MspA-C and not MspD. We found a 2-
3 fold increase in the use of the glycine codon ‘GGT’ in mspA-C relative to mspD (Appendix
Figure A1.6A). We initially hypothesized that this could be due to translational pausing at ‘GGT’ codons as seen in E. coli [40, 41]. However, we found that replacing these with synomous codons did not abolish the LepA-dependent increase in luminescence signal (Appendix Figure
36 1.6B,C). While the amino acid sequences of MspA and MspD have a high average pairwise- identity (82%), the nucleotide sequences are less similar (average pairwise identity of 75%).
Therefore, we hypothesized that variants in the transcript sequence may contribute to LepA- dependence. We randomly re-assigned codons in the mspA gene by sampling from two different probability distributions: the codon bias of the Msm genome (GC-rich) and the inverse codon bias relative to the Msm genome (AT-rich). Changing the transcript sequence decreased the effect of LepA on abundance (Figure 2.4A). Thus, LepA modulation of MspA abundance
37 occurs through sequence determinants in the mRNA.
a 4 b
90 135 195 258 339 mspA sequence Nanoluciferase
) mspA1-90 2 mspA1-135 mspA 1-195 mspA 1-258 1 mspA 1-339
(+LepA/-LepA DAD Luminescence Ratio 0.5 recode 1 recode 2 (GC- rich) (AT- rich) c d 2.5 4 rs=0.82 2.0 ) ) 1.5 2
1.0
Replicate 1 1 0.5 (+LepA/-LepA
(+LepA/-LepA Replicate 2 Luminescence Ratio
Luminescence Ratio 0.0 0.5
1-90 1-135 1-195 1-258 1-339 D 1-195 A A A A A 135-195 DA mspD mspA msp msp msp msp msp e f mspA mspA 1500
+ LepA MspA - LepA
) mycolic 1000 acids
AG layer PG layer PL layer Sec-translocon Calcein Ratio 500 (+LepA/ -LepA SRP
LepA
0 mRNA t n 1-195 are A P mspA mspD msp recode 2 (AT- rich)
Figure 4 - LepA affects permeability through mRNA-sequence determinants in mspA Figure 2.4 – LepA affects permeability through mRNA-sequence determinants in mspA a) mspA was recoded using a Msm codon bias (GC-rich recode), and a less optimized codon bias (AT- rich recode). Dotted lines denotes the mean ratio MspA-NanoLuc ratio (2.91), and the mean MspD- Nanoluc ratio (0.98) from Figure 2c. b) Blue parts of the gene represent regions of the protein encoded by mspA sequences, and pink parts of the gene represent regions of the protein encoded by mspD sequences. c) Luminescence was measured for fusions in (b). A correlation coefficient was determined
38 Figure 2.4 (continued) using Spearman’s rank order correlation test. d) Luminescence was measured for chimeric/fragment porin sequences. Dotted lines denote mean MspA-NanoLuc ratio (2.91), and the mean MspD-Nanoluc ratio (0.98) from Figure 3c. a,d) Replicates here refers to one experiment of reporters in biological triplicate, to determine the ratio of luminescence in +LepA reporters relative to -LepA reporters. e) Calcein staining of NanoLuc-porin reporter strains from above were used to assess porin coding sequence contribution to permeability. **P<0.01 calculated using a two-sided Student’s t-test. f) Model of LepA control. LepA-determining mRNA sequences of the porin transcript are in red, while the rest of the transcript is in green. SRP, signal recognition particle; AG, arabinogalactan; PG, peptidoglycan; PL, phospholipid
A region of the mRNA of MspA is required for LepA-mediated regulation
To determine the location of such sequences that may confer LepA control to a coding sequence, we compared MspA-D sequences in search of regions that might determine LepA control. The protein alignment of MspA-D (Appendix Figure A1.3B) suggests that the most divergent region of the protein sequence is in the signal sequence region (the first 30 amino acids). To test whether a LepA-mediated increase in porin abundance is due to the nature of the signal peptide, predicted signal peptide coding sequences were translationally fused to
NanoLuc, and expressed in the presence or absence of LepA. We found that the presence or absence of LepA did not affect production of MspA-D signal sequences through luminescence
(Appendix Figure A1.6D). To further investigate sequence determinants, we constructed chimeras of mspA and mspD to map the location in the coding region that was LepA-sensitive.
Each chimera contains mspA coding sequence at the 5’ end and mspD coding sequence at the
3’ end (Appendix Figure A1.6E, Figure 2.4B), followed by a NanoLuc fusion. The fusion of mspA1-135 or mspA1-195 (but not mspA1-90) displayed a LepA-mediated increase in luminescence
(Figure 2.4C). This indicates that a motif within the region between 90 and 195 bp was responsible for the LepA-dependent increase in expression. Additionally, we noted a reproducible drop in LepA-dependent signal in the mspA1-258 chimera indicating that regulation is complex.
39 To further define the contribution of this ~100bp region to LepA control of reporter signal, we constructed one chimera composed of mspD transcript, except for the 60bp region that was composed of mspA sequence, termed ‘DAD’. We also fused the first 195bp of the sequence to a
NanoLuc reporter. We found that 195bp of the porin transcript produce LepA dependence in our reporters (Figure 2.4D); in fact, expression of the 60bp region by itself, out of the context of the porin sequence, also produced a smaller but significant level of LepA control in the reporter.
Taken together, these data suggest that the LepA control of MspA abundance comes from an mRNA sequence near the 5’ end of the transcript, just after the end of the signal sequence. We hypothesized that the presence of this sequence-based control contributed to the synthesis of a functional porin in Msm. To test this hypothesis, we used porin reporters with a spectrum of
LepA control to examine permeability changes in cells. Modulation of functional outer membrane protein, as measured by calcein staining, correlated with the sequence elements previously described as sufficient for LepA regulation, namely MspA1-195-MspD-NanoLuc.
Unregulated porin reporters, such as the recoded MspA-NanoLuc and MspD-NanoLuc, did not increase the permeability of the mycobacterial cells (Figure 2.4E). These data suggest that, beyond simply evoking LepA control, these mRNA elements are critical for LepA-dependent modulation of permeability.
Discussion
In many organisms, LepA is dispensable. However, several studies suggest that its absence changes the abundance of membrane proteins [36, 46, 52]. We find that this is true in mycobacteria for at least one class of outer membrane proteins, the major porin family in Msm.
Mutants that lack lepA have relatively normal growth kinetics under optimal growth conditions, indicating that there is no widespread defect in membrane proteins. Further, this suggests that other proteins might have functions that overlap with that of LepA. In fact, some of the most upregulated genes in strains that lack lepA are involved in membrane-related processes.
40 Certainly, before encountering the Sec-translocon machinery that directs co-translational secretion, the membrane-bound ribosome can interact with a number of chaperone systems that are dedicated to proper membrane protein synthesis [10, 53, 54]. It is possible that many of these systems could compensate for the function of a translational GTPase such as LepA, at least under standard laboratory growth conditions. In fact, screening data suggest, that while lepA is non-essential for in vitro growth of Mycobacterium tuberculosis (Mtb), lepA- strains grow poorly during murine infection [55]. Perhaps synthesis of membrane proteins is increased or is more costly during infection, necessitating the function of LepA. Although we used large-scale methods to discover such proteins, our approach is not comprehensive. In particular, our proteomics dataset was relatively depleted of membrane proteins, likely due to issues with extraction/solubility. It is possible that additional membrane proteins might depend on LepA to achieve physiological levels in Msm. Nevertheless, LepA-mediated regulation of MspA translation does produce a distinct phenotype even in rapidly-growing Msm.
How does LepA control the abundance of MspA? We do not see the global effects on ribosome biogenesis that have been observed in E. coli [33, 41]. We found that LepA control of specific porin abundance occurs in our synthetic reporter system, where promoter and UTR elements are unchanged across reporters. Our data suggest that LepA functions during the translational elongation of mRNA messages. Specifically, we observed that approximately 100bp of the 5’ end of the open reading frame (ORF) was required for LepA-mediated modulation. Synonymous recoding of mspA across the gene abolished LepA-dependent reporter expression, suggesting the information necessary for LepA function resides in the mRNA rather than the peptide sequence. The 5’ region of a message may encode a number of signals that dictate the level of protein made from a given message, including mRNA structure, codon bias, or internal Shine
Dalgarno-like sequences [56-59]. Non-optimal codons, for example, are posited to alter
41 ribosome speed at beginning of translation to accommodate loading/assembly of multiple ribosomes onto a message [60, 61].
Changing the rate of translation at specific sites in mRNAs can have important consequences.
We hypothesize that necessity for co-translational secretion of this porin requires control over the rate of translation during this process. We found that the signal sequence alone does not confer LepA control. Rather, elements downstream of the signal sequence confer LepA control.
The signal sequence could function with additional downstream elements that control translation rate to accommodate the initiation of Sec-dependent translation as the nascent peptide emerges from the exit tunnel [62-64]. This is exemplified by LepA dependence of the mspA1-195-
NanoLuc reporter. We propose a model in which the region between 90bp and 195bp stalls translation, requiring LepA to help the ribosome back-translocate and restart, specifically at the time of signal peptide interaction with the secretion machinery (Figure 2.4F).
What, then, is the physiologic role of LepA and the functional significance of changing the rate of translation at the ribosome? It seems likely that LepA has a constitutive role, aiding the translation and folding of a set of proteins with particular characteristics. This model implies that some proteins are either difficult to make, fold or insert into their proper niche in the cell, e.g., the outer membrane. If this model is true, then LepA is a ribosomal GTPase that chaperones the synthesis of these multimeric porins into the outermost membrane of the mycobacterial cell envelope. Additionally, MspA is one of the most abundant transcripts in the cell, based on our
RNA-seq data. Improper insertion of MspA could represent a significant stress on the cell and justify a role for a more or less dedicated factor to aid in translation and folding.
Why is LepA so conserved? In other cellular systems, LepA’s function appears to be related to expression of abundant membrane proteins. In mitochondria, LepA is associated with the
42 membrane and alters the levels of inner membrane respiration complexes through its activity at the ribosome [36, 46, 52]. In chloroplasts of Arabidopsis thaliana, LepA associates to the thylakoid membrane, and its absence was associated with a reduced production of photosystems I and II components [35]. These different organisms reveal LepA’s contribution to the production of important membrane structures in a cell or organelle. The location of sequence determinants for LepA suggests that LepA may play a role in helping to coordinate translation rate with signal recognition and trafficking of the ribosome to the membrane.
Furthermore, porins, respiration complexes, and photosystems are all multi-unit complexes of proteins, which are all known to be post-transcriptionally regulated to ensure proper localization, folding, and levels of the complexes [65, 66]. Thus, LepA might be a quality-control measure that operates across phylogeny to ensure that even the most abundant and difficult proteins are produced efficiently.
Materials and Methods
Bacterial strains and growth conditions
Msm strains were inoculated from frozen stocks into Middlebrook 7H9 medium supplemented with 0.2% glycerol, 0.05% Tween-80, and ADC (5g/L bovine serum albumin, 2g/L dextrose, 3 mg/L catalase) and grown at 37°C. Appropriate antibiotics or inducing agents were used at the following concentrations in M. smegmatis: noursethricin (Nat,20 µg/ml), zeocin (Zeo, 20 µg/ml), kanamycin (Kan,25 µg/ml), hygromycin B (Hyg,50 µg/ml), anhydrous tetracycline (aTc, 100 ng/ml). Msm strains were built by transformation onto LB agar plates supplemented with the appropriate antibiotic. Unless otherwise specified for an experiment, strains were grown to log- phase (OD 0.3-0.8) without antibiotics. For cloning purposes, E. coli strains were grown in Luria
Broth (LB) or agar with antibiotics used at the following concentrations: Nat (40 µg/ml), Zeo
(50µg/ml), Kan (25 µg/ml), and Hyg (100 µg/ml).
43
Bacterial strain construction
All bacterial strains constructed in this study are found in Appendix Table A1.4. Description of the plasmids used to construct the strains can be found in Appendix Table A1.5 and primers used for construction are found in Appendix Table A1.6. Generally, all plasmids were constructed by restriction digestion of the parental vector (with the desired antibiotic resistance gene and phage integration gene for Msm propagation) and all inserts were prepared by amplifying gene fragments with 20bp-Gibson assembly overhangs. Vector and insert combinations were fused together by Gibson isothermal assembly [67].
lepA and mspA complementation vectors
The lepA mutant, ∆lepA::zeoR (HR334), and the mspA mutant, ∆mspA::zeoR (HR329) were a generous gift from E.H. Rego [45]. The complemented strain, +LepA (: SF178: ∆lepA::zeo,
L5::lepA-kan) and marked mutant, -LepA ,( SF181: ∆lepA::zeo, L5::empty-kan) were derivatives of this strain, created by transforming HR334 with two variants of a vector that integrates that the L5 phage site and is marked with a kanamycin-resistance gene. The vector contained either
300 base pairs upstream of the lepA open reading frame plus the lepA open reading frame
(+LepA, SF121) or no expressed gene (-LepA, CT242), respectively. For proteomics, a wild type strain containing the same empty L5-kanamycin vector (SF177) was used in addition to the
+LepA and -LepA strains. The ∆lepA::zeoR ∆mspA::hygR strain (SF789) was constructed using
HR334 as the parental strain, with which a deletion of mspA was created by traditional mycobacterial recombineering techniques [68, 69]. For microscopy, mspA was fused to mRFP, with no linker, constitutively expressed from PUV15-Tet, in a Tweety(Tw)-integrating vector with a noursethricin(Nat)-resistance gene. The mspA-mRFP vector was transformed into +LepA and -
LepA for localization of MspA-mRFP.
44 Candidate-luciferase fusions
Translational fusions to NanoLuc were built by fusing the reporter gene C-terminal to the candidate gene. The protein coding sequences were fused using a glycine-serine-glycine (GSG) linker. The fusions were built to express under an UV15 constitutive promoter, in a Tw- integrating vector with Nat-resistance gene. The fusion constructs were assessed for luminescence in +LepA and –LepA strain backgrounds. ‘NanoLuc’ refers to a Tw - Nat vector expressing only NanoLuc luciferase, under the previously used UV15 constitutive promoter. All fusions were created using primers with 20bp overhangs and cloned into the Tw-Nat vector
(CT250) using Gibson isothermal assembly.
Porin-knockdown constructs sgRNAs targeting each porin were built into an inducible CRISPRi vector(L5, Kan) for mycobacteria, by annealing/ligation with BsmBI-digested oligos and vector backbone as described[51]. Knockdown-vectors were co-transformed into HR334 with a Tw-Hyg vector containing lepA (SF417) expressed as above, or an empty version of this (SF418), to create
+LepA (SF432) and -LepA strains (SF433), respectively. To multiplex sgRNAs, single sgRNA expression fragments from SF419, SF424, and SF425 were amplified with SapI overhangs and cloned into CT295 using Golden Gate cloning as described[51].
Signal sequence fusions
The SignalP server [70] was used to predict the length of signal peptides in MspA-D. The first
84bp from mspA, 87bp from mspB, 96bp from mspC, and 75bp from mspD were each fused to the GSG-NanoLuc reporter as previously described, to create SF413, SF414, SF415, and
SF416, respectively.
Chimeric fusions
45 In-frame fusions of mspA and mspD were built to examine sequence contribution to LepA- dependent reporter signal. For a given fusion point, denoted by arrows in Supplementary Figure
4E, the 5’ sequence of the chimera was amplified from mspA and the 3’ sequence of the chimera was amplified from mspD. As with other reporter experiments, the chimeric sequence was fused N-terminally to GSG -NanoLuc reporter. The ‘DAD’ construct was built by amplifying
MspD1-135, MspA136-195, MspA195-624 and fused them to GSG-NanoLuc reporter as previously described.
Flow cytometry
Strains in biological triplicate were grown to log-phase and stained with 0.5 µg/ml of calcein-AM
(Invitrogen, Carlsbad,CA) for one hour. Strains were analyzed by flow cytometry on a
MACSQuant (VYB Excitation: 488nm; Emission filter: 525/50) in the same manner as previously described [45]. Median fluorescence was used from each replicate to compute an overall mean fluorescence intensity.
Kill curves
Biological replicate were grown to mid-log phase, diluted to 0.05 OD and treated with 10X MIC concentrations of the following drugs: rifampin (20 µg/ml), isoniazid (40 µg/ml), vancomycin (4
µg/ml) and linezolid (500 ng/ml). Survival was assessed over time as described previously[45].
Minimum inhibitory concentration assays
In 96-well plates, +LepA and -LepA strains were diluted to 0.005 and tested in biological triplicate in serial dilutions of tetracycline (Sigma Aldrich, St. Louis, MO), clarithromycin (Sigma
Aldrich, St. Louis, MO), chloramphenicol (Sigma Aldrich, St. Louis, MO) , amikacin (Sigma
Aldrich, St. Louis, MO), and erythromycin (Santa Cruz Biotechnology, Santa Cruz, CA). Highest
46 concentrations of each drug were: 4 µg/ml for tetracycline, 4 µg/ml for clarithromycin, 320 µg/ml for chloramphenicol, 3.2 µg/ml for amikacin, and 16 µg/ml erythromycin. Plates were agitated at
37°C for 21 hours. To determine MICs for each condition, 0.0002% resazurin was added to each well and, plates were agitated at 37°C for 3 hours. The first well with no growth (blue) in each concentration gradient was considered the MIC. A biological replicate, in this case, is considered a single preparation of drug and bacterial incubation, using bacteria from the same culture.
Purification of mycobacterial LepA
Msm LepA was fused with an N-terminal 6x His-tag and expressed from pET28a in BL21 E. coli as previously described for E. coli LepA[41]. Briefly, 1 L of log-phase culture was induced with 1 mM Isopropyl β-D-1-thiogalactopyranoside (IPTG) for 4 hours at room temperature. Cells were harvested at 5,000 g x 10 min, and pellets were frozen at -80°C overnight. The pellet was thawed with a stir bar in 30 mL of lysis buffer (50 mM Tris-HCl pH 7.6, 60 mM KCl, 5% glycerol,
6 mM β-mercaptoethanol (BME), 2 mM MgCl2, 0.2 mM phenylmethylsulfonyl fluoride (PMSF),
DNase), and lysed by French press. Cell lysates were clarified by centrifugation at 15,000 x g at
4°C. Lysate was brought up to 30 mM imidazole pH 7.6 and His-tagged LepA was extracted via batch binding with 4 mL Ni-NTA beads for 2 hours with stir bar at 4°C. Beads were collected in plastic columns with 10 mL bed volume and washed with 4x 10 mL wash buffer (50 mM Tris-
HCl pH 7.6, 300 mM KCl, 5% glycerol, 6 mM BME, 30 mM imidazole), corresponding to wash 1
– 4 in Supplementary Figure 2. 1 mL elution fractions were collected using elution buffer (50 mM
Tris-HCl pH 7.6, 40 mM KCl, 5% glycerol, 6 mM BME, 200 mM imidazole) and analyzed via
SDS-PAGE (Supplementary Figure 2). The cleanest elution fractions (2, 6, and 7) were pooled and dialyzed into 6 L (3 x 2 L) of storage buffer (50 mM Tris-HCl pH 7.6, 50 mM KCl, 5% glycerol, 6mM BME) using dialysis cassettes with 10 kDa MWCO. 10 μL aliquots were flash
47 frozen with liquid nitrogen and stored at -80°C. LepA protein concentration was calculated using the Bradford protein assay.
In vitro translation
To assess the effect of LepA on translation, in vitro translation reactions were prepared with purified mRNA. SF741 was used in a HiScribe T7 in vitro transcription kit (New England Biolabs,
Ipswich, MA) to generate Venus mRNA. A master mix of Venus mRNA and PureExpress (New
England Biolabs, Ipswich, MA) components were prepared in duplicate reactions with a range of concentrations of purified Msm LepA. When no LepA was added to the reaction, equal volume of storage buffer was added in place of protein. Reactions were carried out at 12.5 μL in a 384- well plate for 4 hours at 37°C, and fluorescence (measured at an excitation of 505 nm and an emission of 540 nm) was collected on a plate reader.
Ribosome analysis
500 ml of +LepA and –LepA strains were grown to mid-log phase, filtered over 0.22 um, 90 mm membranes(Millipore) on a fritted glass microfiltration apparatus(Kimball-Chase), and scraped into liquid nitrogen. 500 μL of lysis buffer (20mM Tris pH 8, 10 mM MgCl2, 100 mM NH4Cl, 5mM
CaCl2, 0.4% Triton-X 100, 0.1% NP-40, 34 mg/ml chloramphenicol, 100 U/ml RNase-free
DNase I) was added to cell scrapes. Frozen cells and lysis buffer were ground in a Retsch 400 mixer mill using 10ml grinding jars and 12 mm grinding balls. Cell lysates were thawed and clarified at 15,000 x g for 15 minutes at 4°C. 250 μL of lysate were layered onto a 10-40% linear sucrose gradient. The sucrose gradients were spun in a Beckman ultracentrifuge at 35,000 rpm
(150,000xg) for 2.5 hr at 4°C. The gradients were fractionated and analyzed using a gradient fractionator (BioComp Instruments, Inc., NB, Canada).
Proteomics and RNA sequencing
48 60 mls of marked-WT, +LepA, and - LepA were grown to log-phase (OD600 ~ 0.4) and the culture was split into two parts to extract protein and RNA separately. Both aliquots were spun at 4000rpm for 10 mins. For proteomics, cells were resuspended in 500 μL of urea lysis buffer
(8 M urea in 50 mM Tris pH 8.2, 75 mM NaCl, Roche Complete EDTA-free Protease Inhibitor
Cocktail tablet) and subjected to bead-beating for 4 x 45 secs on with 3 minutes on ice in between. Cell lysates were spun down at 10 min, 13,000 rpm at 4°C and the supernatant was isolated for proteomics sample preparation. For RNA sequencing, RNA was isolated as described previously[71], depleted for rRNA using RiboZero (Epicenter), and prepared for sequencing using KAPA Stranded RNA-Seq Library Preparation Kit (KAPA Biosystems).
Quantitative Proteomics
Samples for quantitative proteomics experiment were processed as described previously [72].
Briefly, three process replicates of WT, +LepA, and –LepA cell lysates were reduced with 5mM dithiothreitol (DTT), alkylated with 10mM iodoacetamide (IAA), digested with Endoproteinase
Lys-C (Wako Laboratories), for 2 hours at 1:50 enzyme to substrate ratio at 30⁰C, followed by an overnight digestion with trypsin (Promega) at 1:50 enzyme to substrate ratio at 37⁰C.
Reactions were quenched with neat formic acid (FA) for a final concentration of 1%. Digests were desalted using tC18 SepPak reversed phase cartridges (Waters, Milford, MA) following manufacturer’s protocol. Tandem mass tag (TMT) isobaric labeling strategy was used for this experiment. 50ug aliquot of each of the 9 samples were labeled by TMT10plex reagent
(ThermoFisher Scientific, Waltham, MA) following manufacturer’s protocol. Pooled reference standard was generated by mixing equal amounts of each of the nine samples and included in the tenth channel of the TMT10plex. Labeling efficiency was assessed prior to quenching the reactions. Once sufficient (>99%) labeling efficiency was achieved, reactions were quenched and samples were mixed together. Combined sample was desalted using tC18 Sep-Pac reversed phase cartridges, and the eluate was dried down completely. Sample was
49 reconstituted and fractionated on Zorbax 300 Extend-C18 4.6 x 250mm column (Agilent
Technologies, Santa Clara, CA) as described previously (Mertins, Nat Protocols, 2018).
Fractions were collected every minute during the gradient and further concatenated into a total of 24 fractions that were analyzed on Q Exactive Plus mass spectrometer (MS) coupled to
EASY-nLC 1200 ultra-high performance liquid chromatography (UHPLC) system (ThermoFisher
Scientific, Waltham, MA). One microgram of each of the fractions was injected on a 75um ID
Picofrit column (New Objective, Woburn, MA) packed with Reprosil-Pur C18-AQ 1.9um beads
(Dr. Maisch, GmbH) in-house to a length of 22cm. Sample was eluted at 200nL/min flow rate with solvent A of 0.1% FA /3% acetonitrile (ACN), solvent B of 0.1% FA / 90% ACN and a gradient of 2-6% B in 1min, 6-30% B in 84min, 30-60% B in 9min, 60-90% B in 1min, and a hold at 90% B for 5min. MS data was acquired in data-dependent mode with MS1 resolution of
70,000 and automatic gain control (AGC) of 3e6. MS/MS was performed on most intense 12 ions with a resolution of 35,000, AGC of 5e4, isolation width of 1.6amu, and normalized collision energy of 29. Data was extracted and searched against M. smegmatis database using
Spectrum Mill MS Proteomics Workbench (Agilent Technologies, Santa Clara, CA). Extracted spectra were searched using carbamidomethylation of Cysteins and TMT labeling of N-termini and lysine residues as fixed modifications and methionine oxidation, asparagine deamidation and protein N-terminal acetylation as variable modifications. Spectrum to database matching was controlled with peptide level false discovery rate (FDR) of less than 1%. Peptides were rolled into protein groups and subgroups in Spectrum Mill with protein level FDR of 0%. Protein summary export consisting of list of quantified proteins with reporter ion ratio of every TMT channel to the pooled reference channel was generated for quantitation of proteins. TMT10 reporter ion intensities were corrected for isotopic impurities in the Spectrum Mill protein/peptide summary module using the afRICA correction method which implements determinant calculations according to Cramer's Rule [Shadforth, 2005] and correction factors obtained from
50 the reagent manufacturer’s certificate of analysis
(https://www.thermofisher.com/order/catalog/product/90406) for lot number SE240163.
Proteins identified with 2 or more peptides were used for further statistical analysis. For comparison of the WT and +LepA, WT and –LepA, +LepA and -LepA samples two-sample moderated T test was performed with an adjusted p-value threshold of less than 0.05 for assessing significantly regulated proteins.
RNA Sequencing
Samples were sequenced on an Illumina HiSeq 2500 in paired-end mode with a read length of
125 bp. Approximately 4 million reads were collected for each sample. Reads were mapped to the genome sequence of M. smegmatis mc2 155 as a reference genome using BWA (Li and
Durbin, 2009). A python script was used to separate reads in .sam files that mapped to the positive strand and negative strand of the chromosome. Then reads mapping to each ORF (in a strand-specific manner) were tabulated. The raw read counts were converted to FPKMs
(fragments per kilobase per million reads) by dividing by gene length (in bp) and total reads in the sample, and scaling up by 10^9. For analyses of differential gene expression, DESeq2
[73]was used estimate log-fold-changes according to a hierarchical model based on the
Negative Binomial distribution, and p-values were calculated via a Wald test as a measure of significance. P-values were adjusted for a false-discovery rate (FDR) of 5% over all genes by the Benjamini-Hochberg procedure. The raw sequence files are deposited in SRA under accession number SRP183056, and the gene expression levels (FPKMs) are deposited in GEO under accession number GSE126130.
Luciferase assays
Strains were grown to log-phase and luciferase assays were conducted using Nanoglo
Luciferase Assay System (Promega, Fitchburg, WI). Briefly, to ascertain relative luminescence,
51 100 μL of cells were measured for OD600, and 100 μL of cells were mixed with 100 μL of
Nanoglo reagent (prepared as kit protocol described). Within 2 minutes, luminescence measurements were taken in TECAN Spark 10M plate reader with integration time of 1000ms.
These values were normalized by OD600 to obtain relative luminescence values.
Porin knockdown and contribution to LepA phenotype
Strains were grown to log-phase, diluted back into media with or without aTc, and allowed to grow for 15 hours to reach log-phase. Cells were stained with calcein and analyzed by flow cytometry in biological triplicate. Survival of knockdown strains was assessed by growing cells up to log-phase, and plating for CFU on LB containing 500 ng/μL of aTc and Kan. For mspABC knockdown, strains were grown to log-phase, diluted to OD 0.05, and serial dilutions from this resuspension on plates containing LB-Kan 25 ug/ml (uninduced) or plates containing LB-Kan 25 ug/ml, 500 ng/ml aTc (induced).
Fluorescence Microscopy and Image Analysis
+ LepA and - LepA strains containing an MspA-mRFP fusion were grown to log-phase for microscopy. Still imaging of MspA-mRFP strains was performed using a Nikon TI-E microscope at 60x magnification for image analysis and at 100x magnification to generate representative images. To quantify intensities fluorescence intensities of single mycobacterial cells, a custom semi-automated ImageJ macro was run. As mycobacteria tend to clump, a user picked single cells from phase contrast images with the ImageJ ‘point tool’. Then, in an automated way, a circular region of interest was created around each point, encompassing the bacterial cell and saved to the ROI manager. Automatic thresholding, and segmenting was performed to measure the intensity on a corresponding fluorescence image. This was repeated for each single-cell the
52 user identified and the fluorescence intensities and areas were saved to text file for further analysis.
Experimental Replicates
Unless otherwise noted all experiments were conducted at least twice, in biological triplicate.
Luciferase experiments in Fig. 2.2D, Fig. 2.4A, and D are presented as the accumulation of at least three different experiments (different days) where each experiment contained three independent biological replicates.
Data Analysis
Gene ontology (GO) term analysis was performed using custom R code that generates count data for each GO term associated with a given protein found in the proteomics dataset and found in the whole Msm proteome. Codon frequency was computed using custom R code, and specifically, the ‘seqinr’ package to analyze codon bias on a gene-by-gene and genome-wide basis. All statistical measurements and tests are specified in the figure legends.
Acknowledgements
We thank Dr. Allison Carey and Dr. Thibault Barbier for careful advice on the manuscript. We thank Dr. Sarah Fortune and the Fortune lab for insightful discussion on the story. We thank Dr.
Bill Neidermeyer, and members of the Whelan lab for allowing us to use ribosome fractionation machinery. We are grateful to the Dr. Michael Welsh and the Walker lab for allowing us to use their protein purification equipment. This research was supported by funding from the NIH (U19
AI107774 to E.J.R.).
References
53 1. Yanofsky, C., Attenuation in the control of expression of bacterial operons. Nature, 1981. 289(5800): p. 751-8. 2. Flygaard, R.K., et al., Cryo-EM structure of the hibernating Thermus thermophilus 100S ribosome reveals a protein-mediated dimerization mechanism. Nat Commun, 2018. 9(1): p. 4179. 3. Subramaniam, A.R., T. Pan, and P. Cluzel, Environmental perturbations lift the degeneracy of the genetic code to regulate protein levels in bacteria. Proc Natl Acad Sci U S A, 2013. 110(6): p. 2419-24. 4. Chionh, Y.H., et al., tRNA-mediated codon-biased translation in mycobacterial hypoxic persistence. Nat Commun, 2016. 7: p. 13302. 5. Vogel, C. and E.M. Marcotte, Insights into the regulation of protein abundance from proteomic and transcriptomic analyses. Nat Rev Genet, 2012. 13(4): p. 227- 32. 6. Brar, G.A. and J.S. Weissman, Ribosome profiling reveals the what, when, where and how of protein synthesis. Nat Rev Mol Cell Biol, 2015. 16(11): p. 651-64. 7. Razi, A., R.A. Britton, and J. Ortega, The impact of recent improvements in cryo- electron microscopy technology on the understanding of bacterial ribosome assembly. Nucleic Acids Res, 2017. 45(3): p. 1027-1040. 8. Shi, Z., et al., Heterogeneous Ribosomes Preferentially Translate Distinct Subpools of mRNAs Genome-wide. Mol Cell, 2017. 67(1): p. 71-83 e7. 9. Byrgazov, K., O. Vesper, and I. Moll, Ribosome heterogeneity: another level of complexity in bacterial translation regulation. Curr Opin Microbiol, 2013. 16(2): p. 133-9. 10. Schibich, D., et al., Global profiling of SRP interaction with nascent polypeptides. Nature, 2016. 536(7615): p. 219-23. 11. Oh, E., et al., Selective ribosome profiling reveals the cotranslational chaperone action of trigger factor in vivo. Cell, 2011. 147(6): p. 1295-308. 12. Herskovits, A.A. and E. Bibi, Association of Escherichia coli ribosomes with the inner membrane requires the signal recognition particle receptor but is independent of the signal recognition particle. Proc Natl Acad Sci U S A, 2000. 97(9): p. 4621-6. 13. Yanofsky, C., RNA-based regulation of genes of tryptophan synthesis and degradation, in bacteria. RNA, 2007. 13(8): p. 1141-54. 14. Dar, D., et al., Term-seq reveals abundant ribo-regulation of antibiotics resistance in bacteria. Science, 2016. 352(6282): p. aad9822. 15. Zhou, M., et al., Non-optimal codon usage affects expression, structure and function of clock protein FRQ. Nature, 2013. 495(7439): p. 111-5. 16. Goodman, D.B., G.M. Church, and S. Kosuri, Causes and effects of N-terminal codon bias in bacterial genes. Science, 2013. 342(6157): p. 475-9. 17. Sonenberg, N. and T.E. Dever, Eukaryotic translation initiation factors and regulators. Curr Opin Struct Biol, 2003. 13(1): p. 56-63. 18. Brandi, L., et al., Initiation of protein synthesis: a target for antimicrobials. Expert Opin Ther Targets, 2008. 12(5): p. 519-34. 19. Nakamoto, T., The initiation of eukaryotic and prokaryotic protein synthesis: a selective accessibility and multisubstrate enzyme reaction. Gene, 2007. 403(1-2): p. 1-5.
54 20. Laursen, B.S., et al., Initiation of protein synthesis in bacteria. Microbiol Mol Biol Rev, 2005. 69(1): p. 101-23. 21. Andersen, G.R., P. Nissen, and J. Nyborg, Elongation factors in protein biosynthesis. Trends Biochem Sci, 2003. 28(8): p. 434-41. 22. Yamamoto, H., et al., EF-G and EF4: translocation and back-translocation on the bacterial ribosome. Nat Rev Microbiol, 2014. 12(2): p. 89-100. 23. Maracci, C. and M.V. Rodnina, Review: Translational GTPases. Biopolymers, 2016. 105(8): p. 463-75. 24. Lassak, J., D.N. Wilson, and K. Jung, Stall no more at polyproline stretches with the translation elongation factors EF-P and IF-5A. Mol Microbiol, 2016. 99(2): p. 219-35. 25. deLivron, M.A., et al., A novel domain in translational GTPase BipA mediates interaction with the 70S ribosome and influences GTP hydrolysis. Biochemistry, 2009. 48(44): p. 10533-41. 26. Heller, J.L.E., R. Kamalampeta, and H.J. Wieden, Taking a Step Back from Back-Translocation: an Integrative View of LepA/EF4's Cellular Function. Mol Cell Biol, 2017. 37(12). 27. Yang, F., et al., Mitochondrial EF4 links respiratory dysfunction and cytoplasmic translation in Caenorhabditis elegans. Biochim Biophys Acta, 2014. 1837(10): p. 1674-83. 28. Hiramatsu, Y., et al., BipA Is Associated with Preventing Autoagglutination and Promoting Biofilm Formation in Bordetella holmesii. PLoS One, 2016. 11(7): p. e0159999. 29. Choi, E., et al., Elongation factor P restricts Salmonella's growth by controlling translation of a Mg(2+) transporter gene during infection. Sci Rep, 2017. 7: p. 42098. 30. Qin, Y., et al., The highly conserved LepA is a ribosomal elongation factor that back-translocates the ribosome. Cell, 2006. 127(4): p. 721-33. 31. Youngman, E.M. and R. Green, Ribosomal translocation: LepA does it backwards. Curr Biol, 2007. 17(4): p. R136-9. 32. Gagnon, M.G., J. Lin, and T.A. Steitz, Elongation factor 4 remodels the A-site tRNA on the ribosome. Proc Natl Acad Sci U S A, 2016. 113(18): p. 4994-9. 33. Gibbs, M.R., et al., Conserved GTPase LepA (Elongation Factor 4) functions in biogenesis of the 30S subunit of the 70S ribosome. Proc Natl Acad Sci U S A, 2017. 114(5): p. 980-985. 34. Evans, R.N., et al., The structure of LepA, the ribosomal back translocase. Proc Natl Acad Sci U S A, 2008. 105(12): p. 4673-8. 35. Ji, D.L., et al., CpLEPA is critical for chloroplast protein synthesis under suboptimal conditions in Arabidopsis thaliana. PLoS One, 2012. 7(11): p. e49746. 36. Bauerschmitt, H., S. Funes, and J.M. Herrmann, The membrane-bound GTPase Guf1 promotes mitochondrial protein synthesis under suboptimal conditions. J Biol Chem, 2008. 283(25): p. 17139-46. 37. Badu-Nkansah, A. and J.K. Sello, Deletion of the elongation factor 4 gene (lepA) in Streptomyces coelicolor enhances the production of the calcium-dependent antibiotic. FEMS Microbiol Lett, 2010. 311(2): p. 147-51.
55 38. Kida, Y., T. Shimizu, and K. Kuwano, Cooperation between LepA and PlcH contributes to the in vivo virulence and growth of Pseudomonas aeruginosa in mice. Infect Immun, 2011. 79(1): p. 211-9. 39. Shoji, S., et al., Translation factor LepA contributes to tellurite resistance in Escherichia coli but plays no apparent role in the fidelity of protein synthesis. Biochimie, 2010. 92(2): p. 157-63. 40. Liu, B. and C. Chen, Translation Elongation Factor 4 (LepA) Contributes to Tetracycline Susceptibility by Stalling Elongating Ribosomes. Antimicrob Agents Chemother, 2018. 62(8). 41. Balakrishnan, R., et al., The conserved GTPase LepA contributes mainly to translation initiation in Escherichia coli. Nucleic Acids Res, 2014. 42(21): p. 13370-83. 42. Liu, H., et al., The conserved protein EF4 (LepA) modulates the elongation cycle of protein synthesis. Proc Natl Acad Sci U S A, 2011. 108(39): p. 16223-8. 43. Liu, H., et al., Interrupted catalysis: the EF4 (LepA) effect on back-translocation. J Mol Biol, 2010. 396(4): p. 1043-52. 44. Zhang, D., et al., EF4 disengages the peptidyl-tRNA CCA end and facilitates back-translocation on the 70S ribosome. Nat Struct Mol Biol, 2016. 23(2): p. 125- 31. 45. Rego, E.H., R.E. Audette, and E.J. Rubin, Deletion of a mycobacterial divisome factor collapses single-cell phenotypic heterogeneity. Nature, 2017. 546(7656): p. 153-157. 46. Gao, Y., et al., Mammalian elongation factor 4 regulates mitochondrial translation essential for spermatogenesis. Nat Struct Mol Biol, 2016. 23(5): p. 441-9. 47. Li, J., et al., Improved cell-free RNA and protein synthesis system. PLoS One, 2014. 9(9): p. e106232. 48. Stahl, C., et al., MspA provides the main hydrophilic pathway through the cell wall of Mycobacterium smegmatis. Mol Microbiol, 2001. 40(2): p. 451-64. 49. Niederweis, M., et al., Cloning of the mspA gene encoding a porin from Mycobacterium smegmatis. Mol Microbiol, 1999. 33(5): p. 933-45. 50. Stephan, J., et al., The growth rate of Mycobacterium smegmatis depends on sufficient porin-mediated influx of nutrients. Mol Microbiol, 2005. 58(3): p. 714-30. 51. Rock, J.M., et al., Programmable transcriptional repression in mycobacteria using an orthogonal CRISPR interference platform. Nat Microbiol, 2017. 2: p. 16274. 52. Zhu, P., et al., Human Elongation Factor 4 Regulates Cancer Bioenergetics by Acting as a Mitochondrial Translation Switch. Cancer Res, 2018. 78(11): p. 2813- 2824. 53. Miller, B.K., et al., Mycobacterium tuberculosis SatS is a chaperone for the SecA2 protein export pathway. Elife, 2019. 8. 54. Kramer, G., A. Shiber, and B. Bukau, Mechanisms of Cotranslational Maturation of Newly Synthesized Proteins. Annu Rev Biochem, 2018. 55. Zhang, Y.J., et al., Tryptophan biosynthesis protects mycobacteria from CD4 T- cell-mediated killing. Cell, 2013. 155(6): p. 1296-308. 56. Liu, L., et al., Effect of codon message on xylanase thermal activity. J Biol Chem, 2012. 287(32): p. 27183-8.
56 57. Bhattacharyya, S., et al., Accessibility of the Shine-Dalgarno Sequence Dictates N-Terminal Codon Bias in E. coli. Mol Cell, 2018. 70(5): p. 894-905 e5. 58. Yu, C.H., et al., Codon Usage Influences the Local Rate of Translation Elongation to Regulate Co-translational Protein Folding. Mol Cell, 2015. 59(5): p. 744-54. 59. Li, G.W., E. Oh, and J.S. Weissman, The anti-Shine-Dalgarno sequence drives translational pausing and codon choice in bacteria. Nature, 2012. 484(7395): p. 538-41. 60. Fredrick, K. and M. Ibba, How the sequence of a gene can tune its translation. Cell, 2010. 141(2): p. 227-9. 61. Power, P.M., et al., Whole genome analysis reveals a high incidence of non- optimal codons in secretory signal sequences of Escherichia coli. Biochem Biophys Res Commun, 2004. 322(3): p. 1038-44. 62. Butkus, M.E., L.B. Prundeanu, and D.B. Oliver, Translocon "pulling" of nascent SecM controls the duration of its translational pause and secretion-responsive secA regulation. J Bacteriol, 2003. 185(22): p. 6719-22. 63. Fluman, N., et al., mRNA-programmed translation pauses in the targeting of E. coli membrane proteins. Elife, 2014. 3. 64. Mercier, E., et al., Signal recognition particle binds to translating ribosomes before emergence of a signal anchor sequence. Nucleic Acids Res, 2017. 45(20): p. 11858-11866. 65. Li, G.W., et al., Quantifying absolute protein synthesis rates reveals principles underlying allocation of cellular resources. Cell, 2014. 157(3): p. 624-35. 66. Shieh, Y.W., et al., Operon structure and cotranslational subunit association direct protein assembly in bacteria. Science, 2015. 350(6261): p. 678-80. 67. Gibson, D.G., et al., Enzymatic assembly of DNA molecules up to several hundred kilobases. Nat Methods, 2009. 6(5): p. 343-5. 68. Murphy, K.C., K. Papavinasasundaram, and C.M. Sassetti, Mycobacterial recombineering. Methods Mol Biol, 2015. 1285: p. 177-99. 69. van Kessel, J.C. and G.F. Hatfull, Mycobacterial recombineering. Methods Mol Biol, 2008. 435: p. 203-15. 70. Petersen, T.N., et al., SignalP 4.0: discriminating signal peptides from transmembrane regions. Nat Methods, 2011. 8(10): p. 785-6. 71. Gerrick, E.R., et al., Small RNA profiling in Mycobacterium tuberculosis identifies MrsI as necessary for an anticipatory iron sparing response. Proc Natl Acad Sci U S A, 2018. 115(25): p. 6464-6469. 72. Mertins, P., et al., Reproducible workflow for multiplexed deep-scale proteome and phosphoproteome analysis of tumor tissues by liquid chromatography-mass spectrometry. Nat Protoc, 2018. 13(7): p. 1632-1661. 73. Love, M.I., W. Huber, and S. Anders, Moderated estimation of fold change and dispersion for RNA-seq data with DESeq2. Genome Biol, 2014. 15(12): p. 550.
57 Chapter 3
Discovery of ribosome-associated proteins in Mtb
Skye Fishbein1, Albert Wang1, Xiaojun Li2, Qingan Sun2, Matthew J. Szucs3, Rushdy Ahmad3,
James Sacchettini2, Eric Rubin1
1Department of Immunology and Infectious Disease, Harvard TH Chan School of Public Health,
Boston, Massachusetts, 02115, United States
2Department of Biochemistry and Biophysics, Texas A&M University, Texas, 77843, United
States
3Broad Institute of MIT and Harvard, Cambridge, 02142, United States
Author Contributions:
Conceptualization: S.R.S.F., X.L.,J.S., R.A., E.J.R; Methodology: S.R.S.F., X.L., Q.S., M.S., R.A.,
E.J.R; Investigation: S.R.S.F., X.L., Q.S., M.S., R.A., A.W.; Data Curation: S.R.S.F., M.S., R.A.;
Writing: S.R.S.F.
58 Abstract
Disease and mortality caused by Mycobacterium tuberculosis (Mtb) remains a public health and socioeconomic burden, and the world is in need of new antibiotics. The mycobacterial ribosome has a distinct structure from other well-studied bacterial ribosomes, making it a viable drug target. We sought to uncover novel mechanisms of translational control through discovery of the ribosome-associated proteome. To discover these interactions, we purified the Mtb ribosome and performed quantitative proteomics to measure protein enrichment at the ribosome.
Unexpectedly, we found that known ribosomally-associated proteins (such as secretion systems and hibernation factors), were significantly associated with the log-phase 70S. Additionally, we discovered novel hypothetical proteins that are essential and associate to the ribosome under multiple condition. Our proteomics data suggests that we have found essential proteins whose association with the ribosome was unexpectedly high (ESX-3 and ESX-5 systems) and novel proteins that may alter the proteome of the cell through their interactions with the ribosome.
Future work will investigate the heterogeneity in ribosome populations based on the results of the screen.
Introduction
The scourge of tuberculosis (TB) and rise of multi-drug resistant tuberculosis remains a persistent, global pandemic [1-4]. TB disease can persist for months to years, and patients are susceptible to relapse even after the standard 6 months of drug therapy [5, 6].The causative agent of TB, Mycobacterium tuberculosis (Mtb), is a member of the mycobacterial genus, characterized by a thick, waxy mycolic acid cell wall layer [7]. Mtb grows slowly, and replicates as an intracellular pathogen inside alveolar macrophages of the human lung, causing a spectrum of disease, from asymptomatic infection to necrotic cavitary lesions [8].
59 How does Mtb adapt to the changing environment within the human lung and specifically, within the macrophage? After phagocytosis by a macrophage, Mtb produces and secretes virulence factors to skew the course of intracellular invasion [9, 10]. Mtb possesses five type VII secretion systems (ESX) and uses these systems to disrupt host membranes [11], scavenge extracellular iron[12], and skew macrophage responses [12]. Yet, the macrophage innate immune response presents a number of challenges to the bacillus that demand a bacterial response. During this host-bacterial arms race, Mtb faces a number of stresses such as reactive oxygen/nitrogen species [13, 14], changes in nutrient pools [15-17], and changes in metal availability [12, 18,
19]. Understanding the innate vulnerabilities of the pathogen M. tuberculosis, through its capacity to adapt to and survive within the macrophage, is critical to future efforts to eliminate
TB.
Determinants of Mtb infection have been thoroughly studied in vitro and in vivo through the lens of transcriptional reprogramming. In general, responses must maintain a balance between preservation of resources and active adaptation to cellular insult. Like most bacteria, Mtb employs the stringent response, a multi-part signaling cascade that globally shuts down certain cellular processes during stress [20]. A number of transcriptional regulators have been identified that reprogram bacterial physiology and likely play significant roles in Mtb survival during infection [21]. Regulators such as PhoPR [22], EspR [21] and IdeR [23] have been implicated in regulating bacterial virulence factors and iron scavenging. Additionally, DosRS is a major two- component regulator that protects the cell from damaging carbon monoxide and nitric oxide in vivo; this transcriptional program has also been implicated in surviving hypoxia [24, 25]. Mtb also uses metabolic regulators KstR and PrpR to tune its nutrient utilization [26, 27]. Yet, given the shifting environment of any bacterial lifestyle, an increase in transcript does not always result in an increase in protein [28].
60
Post-transcriptional regulation tunes the cellular proteome, protects macromolecular synthesis machinery, and produces protein at the correct location and time in the cell. The past decade of mycobacterial research has revealed that a number of post-transcriptional mechanisms may be in place to enable Mtb adaptation upon infection. Mycobacterial protein synthesis is defined by a number of unique features including leaderless transcripts and extra ribosomal proteins.
Mycobacteria have an unusually high number of leaderless transcripts, and a number of these encode essential proteins. Yet, it is not clear how these transcripts are translated; this feature indicates a gap in our understanding of how Mtb initiates translation [29]. Perhaps, a number of unkown protein factors are essential for initiation of leaderless translation in mycobacteria. The
Mtb ribosome is not only a promising drug target, but also has unique structural features not found in other well-studied bacterial ribosomes [30, 31]. The physiological implications of these extra proteins are also unresolved. Recent work on metal starvation has uncovered a number of post-transcriptional responses that may contribute to mycobacterial adaptation. Small RNAs provide the bacterium a mechanism to preserve cellular resources as a response to stressors such as metal starvation and membrane stress [32]. In addition to its unique structure, it has become clear that certain environments trigger remodeling of the ribosomal protein composition, both the structural proteins and proteins that may determine translational output [33, 34].
We hypothesize that Mtb may maintain its proteome and adapt to changing nutrients within host environments through a number of undiscovered translational mechanisms at the ribosome.
Here, we conducted a ribosome-proteomics screen, in multiple physiological states, to identify novel interacting proteins and conditional associations with the Mtb ribosome. We discovered a number of unexpected log-phase associations with the ribosome, and also found a number of putative ribosome-associating proteins that may be important for mycobacterial physiology. We characterize the fitness contributions of a set of candidate ribosome-associated proteins, and
61 found a number of instances where the associating protein may be critical for mycobacterial growth. These results indicate that ribosome heterogeneity in a cell is likely larger than previously thought, with a number of resources potentially dedicated to membrane co- translational secretion in Mtb. Future work will investigate the translational regulatory mechanisms involved in ESX-related translation in Mtb, with the aim of revealing novel targets for drug development.
Results
Proteomic screen to identify ribosome associations in Mtb a) Growth medium exposure Ribosome fractionation Quantitative proteomics
100S? 70S 50S 30S tandem-mass-tagging log-phase Mtb 260 of peptides A
1 2 3 4 5 6 7 8 stressed Mtb Bottom Top LC-MS/MS b) c) 1000 1000 log-phase stressed 800 800 0 600 0 600 70S 70S 400 400 Abs. 26 50S Abs. 26 50S 30S 30S 200 200
0 0 0 5 10 15 20 5 10 15 20 Volume from top (mL) Volume from top (mL)
Figure 3.1 – Screen for discovery of novel ribosome associations in Mtb a) Experimental schematic for discovery of ribosome-associated proteins in Mtb. Mtb cells were grown in different conditions, and quantitative proteomics was used to discover ribosome-associated proteomes. b- c) Ribosomes from both conditions were purified from cell lysates using sucrose gradients. Traces are representative of multiple gradients analyzed for each condition. b) Ribosomes purified from Mtb grown to log-phase in 7H9 media. c) Ribosomes purified from Mtb grown in HdB minimal media for two weeks after transitioning cultures from log-phase.
62
Translational regulation is often accomplished by the proteins that associate to the pool of ribosomes [35]. We hypothesized that in Mtb, we might find novel protein-ribosome interactions that contribute to mycobacterial growth and adaptation. To discover the interacting partners at the Mtb ribosome, we grew Mtb in two different mediums (rich and minimal medium) to capture the physiological state of the ribosome across environments. Mtb was grown in Hartmans-de
Bont (HdB) minimal media, lacking carbon and nitrogen, for two weeks. HdB minimal media does not have catalase, dextrose, oleic acid, or albumin, in addition to the absence of other basic nutrients. From both cell states, we harvested total cellular lysate and ribosome populations. Using tandem-mass tagging (TMT) of peptides coupled with liquid chromatography and tandem mass spectrometry (LC-MS/MS), we could capture relative protein abundances across ribosome fractions and conditions, allowing us to define biologically meaningful associations at the ribosome (Figure 3.1A).
We used a 10%-40% sucrose gradient to purify ribosomes, resulting in partial separation of each of the ribosome units. In both growth states, we identified peaks for the 70S assembled ribosome, 50S subunit and 30S subunit. Generally, we saw decreased levels of ribosomes from lysates of cells grown in minimal media (Figure 3.1B,C). The 30S subunit was the most decreased in peak height, likely due to the stressful change in growth medium. It is established that during starvation, the 16s rRNA of the 30S is degraded [36], resulting in loss of the 30S subunit. We collected and pooled fractions for each unit of the ribosome, and the additional peak that appeared to the left of the stressed 70S peak. Each ribosome unit or whole-cell lysate
63 was labeled with a different TMT reporter ion. To compare across the two TMT-10 plexes, we used log-phase 70S values as a normalizing value for each dataset (Table 3.1).
Table 3.1. Quantitative proteomics labeling scheme
TMT Label TMT 10-plex #1 TMT 10-plex #2 TMT_126 log-phase 70S log-phase 70S TMT_127N log-phase 70S log-phase 70S TMT_127C log-phase 50S log-phase 30S
TMT_128N log-phase 50S log-phase 30S TMT_128C stressed 70S stressed 30S TMT_129N stressed 70S stressed 30S TMT_129C stressed 50S log-phase whole cell lysate TMT_130N stressed 50S log-phase whole cell lysate TMT_130C stressed larger fraction stressed whole cell lysate TMT_131 stressed larger fraction stressed whole cell lysate
64
Quantitative proteomics reveals likely heterogeneous population of ribosomes
a) b)
ribosomal proteins
4 ESX-3/5 systems translation-associated
3 SecY tRNA/rRNA methytransferase RafS (p-value) 10
g 2 SRP lo - SecE1
1
-5 0 5 log2 (protein in log-phase 70S/ protein in WCL)
3
2
1
0
-1
-2
-3
log-phase log-phaselog-phase 70S (WCL) 70S (30S) 30S (WCL)
Figure 3.2 – Log-phase interactions reveal ribosome heterogeneity a) Volcano plot of proteins enriched with log-phase 70S ribosome (positive value on x-axis) relative to whole-cell lysate (negative value on x-axis). Purple points indicate structural ribosomal proteins
65 Figure 3.2 (continued) significantly enriched at 70S. Green points indicate ESX-3 and ESX-5 components significantly enriched at 70S. Blue points indicate other proteins well-known to be associated with translation (SecY, SRP, SecE1, tRNA/rRNA methyltransferase, RafS). P-value indicates results of student two-sided t-test. For total protein list, ratios and associatd p-values, see Appendix Table 2.1. b) Hierarchical clustering of enrichment ratios for translation-related proteins validates purification of 70S ribosome and subunits. Each column corresponds to a different relative protein abundance ratio, used to define association to the given ribosome unit. In the first column, ‘log-phase 70S (WCL)’ represents the ratio of protein reporter ion intensities for log-phase 70S divided by protein reporter ion intensities for log-phase whole-cell lysate (WCL). Each row corresponds to a known translation-associated protein. Red (positive ratio) indicates a strong association of the protein with the ribosome unit, where as blue (negative ratio) indicates depletion from the corresponding ribosome unit.
We hypothesized that this discovery pipeline might yield 3 kinds of protein candidates: (1) known proteins that are newly associated with the log-phase 70S, (2) new conditional associations of known proteins with the stressed ribosomes, (3) unknown proteins of interest, associating with either type of ribosome. To define enrichment at the 70S in both cell states, we used two methods to capture association: (1) whole-cell lysate normalization and (2) 30S normalization. We hypothesized that the whole-cell lysate normalization would filter out abundant proteins that may be coincidentally synthesized at the ribosome, rather than interacting with the ribosome. The 30S normalization was used to filter out any protein that non- specifically co-purified with the ribosomes in the gradient, but likely have no biological interaction with the ribosome. For a detailed formulation of each ratio, see Materials and
Methods. First, we were interested in how well we captured known associations to the ribosome. Using the log-phase 70S ribosome to whole-cell lysate ratio, we analyzed the most significantly enriched proteins at the 70S.As expected, ribosomal proteins of the 50S and 30S were highly enriched in the 70S (Figure 3.2A). Additionally, a number of secretion components and mycobacterial type VII secretion systems (ESX systems) were significantly associated with the 70S. Specifically, we found Sec-translocon components SecY, SecE1 and SRP associated with the log-phase assembled ribosome. For the mycobacterial secretion systems, we found structural components (EccB, EccC, EccD) of the ESX-3 and ESX-5 systems associated with the log-phase 70S (Table 3.2).
66 We also examined the enrichment ratios of a number of proteins associated with the translation process. Processes such as protein folding, protein secretion, and ribosome hibernation depend upon ribosome-associating proteins such as chaperones, secretion machines and hibernation factors. Although these proteins are not required for every cycle of translation, we analyzed the association of these known ribosomal-interacting proteins to define the level of sensitivity in our ribosome-proteome screen. Hierarchical clustering of known translation-associated proteins revealed that elongation factors (EF-G, EF-Tu, EF-Ts, LepA, EF-P) and termination/release factors (recycling factor, RF1, RF2) did not associate to the log-phase 70S ribosome.
Meanwhile, initiation factors (IF-1, IF-2, IF-3) were strongly associated with the 30S subunit in log-phase, as expected. Although chaperones such as GroEL did not associate with the 70S, trigger factor was associated in at least one of our enrichment ratios. RafS, a homolog of a bacterial ribosome hibernation factor was significantly associated with our log-phase 70S. Our results point to a heterogeneous population of ribosomes, consisting of active and inactive ribosomes, potentially translating both cytoplasmic proteins and proteins destined for the membrane.
Table 3.2 – ESX-systems associated to ribosomes in Mtb
Type VII Secretion Ribosome unit identified Gene Component name System with Rv0282 EccA ESX3 Stressed 70S Rv0283 EccB ESX3 Log-phase 70S Rv0284 EccC ESX3 Log-phase 70S
Rv0290 EccD ESX3 Log-phase 70S
Rv0292 EccE ESX3 Log-phase 70S
Rv1782 EccB ESX5 Log-phase 70S Rv1783 EccC ESX5 Log-phase 70S Rv1795 EccD ESX5 Log-phase 70S Rv1797 EccE ESX5 Log-phase 70S Rv3895c EccB ESX2 Log-phase 70S
67
Conditional associations at the ribosome during stress in Mtb
We were also interested in the interactions we captured at the Mtb ribosome during in vitro stress. To discern meaningful interactions at the ribosome, we incorporated our normalization methods into a ratio of ratios, comparing log-phase 70S enrichment to stressed 70S enrichment.
A number of types of processes (including translational and post-translational ones) were robustly discovered at the stressed 70S ribosome (Figure 3.3A). We found a number of chaperones and peptidases that were conditionally associated with the 70S during stress
(PepA, PepN, PepQ, HtpG) in addition to both Mtb methionine aminopeptidases (MapA and
MapB). All of these proteins likely deal with nascent peptides coming out of the ribosome in a manner which protects the cell from proteotoxic stress [37, 38]. We also examined how these enrichments of certain factors compared to log-phase 70S. We compared ratios of enriched translation factors in log-phase and during stress (Figure 3.3B) We noted an increased enrichment of elongation factors at the ribosome, indicating that we were able to purify these kind of associating factors. We also found a number of members of cold-shock proteins (CspA,
CspB), which act as RNA-binding proteins, or RNA chaperones. Recently, these proteins have been linked to enabling translation to continue under non-optimal conditions [39]. Surprisingly, we found a number of amino acid tRNA synthetases (glutamyl tRS, histidyl tRS, tryptophanyl tRS) enriched at the 70S during stress. In general, the associations discussed here indicate a general need to prevent proteotoxicity in the cell (peptidases, HtpG), while also potentially preserving translation (elongation factors, SecE2).
68 translation-associated a) 4 membrane-associated proteases chaperones
3 (p-value) 10
g 2 lo -
1
-4 -2 0 2 4
log2 (protein in log-phase 70S/ protein in WCL) b) 3 stressed 70S (30S normalization) 2 starved 70S (log-phase 70S) 1 starved 70S (WCL normalization) 0
log-phase 30 (WCL) -1 log-phase 70S (WCL) -2 log-phase 70S (30S) -3 EF-P HtpG PepA PepN CspA CspB PepQ Zmp1 MapA EF-Ts MapB EF-Tu EF-G2 EF-G1 SecE2 EccA3 histidyl-tRS glutamyl-tRS
tryptophanyl-tRS
Figure 3.3 – Conditional associations at the ribosome a) Histogram of protein enrichment ratios at the stressed 70S ribosome, normalized by whole-cell lysate. Purple dots indicate represent MapB and EF-Ts. Green dots indicate PepA, PepQ, a Clp subunit, and Zmp1, a metalloprotease. Blue dots indicate SecE2 and EccA3 (ESX-3 system). Red dots indicate GroEL2, HtpG, and DnaK. b) Heatmap of enrichment ratios for selected proteins, with values as described in Figure 3. Four different enrichment ratios were generated to discover conditional association
69
Figure 3.3 (continued) at the 70S, and are represented in the top four rows. Each column represents a protein, or a representative candidate of processes found with stressed 70S ribosome. tRS: tRNA synthetase.
Discovery of previously uncharacterized ribosomal factors
Table 3.3 - Proteins of 'unknown function' associated with Mtb ribosome
Ribosome unit Gene identified with Possible function Essentiality Regulator of protease In Rv1488 Stressed 70S NE activity HflC P-loop NTPase/DEAD-box Rv2024c Stressed 70S NE helicase Log-phase 70S/ Rv2629 peptide-release factor NE stressed 70S Rv2689c Stressed 70S TrmA NE nucleotide binding/ NE, conditional Rv2837c Stressed 70S ribosomal protein L9 essentially Rv2974c Stressed 70S Obg-like protein NE Rv3030 Stressed 70S RNA methytransferase essential Log-phase 70S/ log- formyl methionine Rv3404c growth advantage phase 50S transferase growth disadvantage, Rv3422c Stressed 70S ATP-binding protein/TsaE conditionally essential Sassetti et al 2003 addition to known proteins that interact with the ribosomes, we were interested in expanding our understanding of factors that associate with the ribosome in Mtb. We used a bioinformatics approach to assign putative function to unannotated proteins that we found associated to the mycobacterial ribosome. For each distribution of enrichment ratios, we investigated unknown proteins, classified as ‘conserved’ or ‘hypothetical’ with enrichment ratios at least one standard deviations above the mean ratio. Using the conserved domain database (CDD) of NCBI, we searched for functional domains in the list of these unknown proteins associated with various ribosome units. This search yielded a number of putative ribosome-interacting partners (Table
3.3). Among these candidates was a putative ribosomal protein, Rv2837c, that had an L9-like ribosomal protein domain. This protein was associated to the ribosome during stress. The gene
70 exists in an operon with infB and rpfA (two known ribosomal genes), also indicating it may function at the ribosome. Genetic screening data reveals that this gene is conditionally required for growth on cholesterol [40], during murine infection [41], and for drug susceptibility [42]. We also found a number of predicted copies of known translation factors (by homology modeling):
Rv2629, Rv3404, and Rv2974, indicating that there are additional versions of these proteins in the Mtb proteome. Our association of these proteins to ribosome units in Mtb, coupled with relevant genetic essentiality data, may provide new avenues for understanding the tenets of translational regulation in mycobacteria.
Defining the fitness impact of ribosome-interacting partners in mycobacteria
Factors that affect Mtb physiology through their association to the ribosome could prove critical to cell survival in vivo. From our screen, we identified a number of factors that may contribute to ribosome remodeling. Yet, ribosome-interacting partner stoichiometry (with the ribosome) exists on a spectrum in the cell. Essential processes, mediated by factors such as chaperones or alternative elongation factors, exist at both ends of the spectrum [43]. We hypothesized that both types of interactions could be important to the cell, and therefore used a wide-range of enrichment values as motivation to pursue candidate ribosome-association factors. Because we did not first validate localization to the ribosome, we selected factors that were likely to associate to the ribosome based on protein domain homology. We chose proteins to further characterize based on any combination of the following criteria: proteins with a putative function linked to translation, proteins that had conditional genetic requirements in Mtb, or proteins with transcriptional signatures of conditional requirements. Although we associated the proteins to the Mtb ribosome, we made mutants in both Msm (Table 3.4) and Mtb (Table 3.5) to perform initial phenotyping experiments.
71 Table 3.4 – Ribosome-associated factor mutants in Msm
Ribosome unit identified Gene with Possible function Msm number heat shock-like family Rv0215c Stressed 70S protein MSMEG_0424 Rv0734/MapA Stressed 70S nascent peptide processing MSMEG_1485 Ribosome-associated Rv0120c/EF-G2 Stressed 70S GTPase MSMEG_6535 Ribosome-associated Rv1112/EngD Stressed 70S GTPase MSMEG_5222 Rv0208 Log-phase 70S tRNA methyltransferase MSMEG_0252 Rv2725/HflX Log-phase 50S GTP-binding protein MSMEG_2736 Stressed 70S/ Log-phase Rv2861c/MapB 70S nascent peptide processing MSMEG_2587 Putative formyl-methionine Rv3404c Log-phase 50S transferase MSMEG_0014
Table 3.5 – Ribosome-associated factor candidates in Mtb
Ribosome unit identified Gene Possible function Essentiality with conditionally essential in Rv0120c Stressed 70S Ribosome-associated GTPase mouse
conditionally essential in Rv1112/EngD Stressed 70S Ribosome-associated mouse GTPase
contains a peptide release Rv2629 Stressed 70S/log 70S DosR regulon member factor domain
conditionally essential Log-phase 30S/stressed Rv3228/RsgA P-loop GTPase during growth in 30S cholesterol
In Msm, we examined the ribosome-associated mutants for general growth defects and condition-specific growth defects. Examination of log-phase growth rates indicated that a mutant
72 in a methionine aminopeptidase (MSMEG_2587/Rv2861c) was sick (Figure 3.4a). a) b) 0.4 1
0.3 0 0 0.2 OD60 OD60 0.1 WT 0.1 MSMEG_0424 MSMEG_1485 0.0 0 10 20 30 40 50 0 10 20 30 40 50 MSMEG_2587 Time (hours) Time (hours) MSMEG_6535 c) d) 10 10 MSMEG_5222 MSMEG_0252 l l MSMEG_2736 1 MSMEG_0014 1
0.1 Proportion Suriva Proportion Suriva
0.1 0.01 0 5 10 15 20 0 10 20 30 Time (days) Time (hours)
Figure 3.4 – Phenotyping of ribosome-associated protein mutants in Msm
Growth curve of mutants for ribosome-associated candidates in Msm in a) rich 7H9 media and b) HdB minimal media with carbon and nitrogen. c) Colony forming units (CFU) were assessed for cell survival in HdB minimal media with no carbon or nitrogen sources over time. Proportion survival indicates CFU normalized by CFU at t=0. d) Proportion survival was assessed for growth at 50 degrees °C.
There are two methionine aminopeptidases in Msm, and we suspect that the MSMEG_2587 may be the major methionine aminopeptidase, responsible for cleaving initiator methionine residues off of nascent polypeptide chains [44]. We wanted to test the hypothesis that cellular fitness during growth media transition might be tied to remodeling of the ribosome. Using a number of different assessments of cell survival (OD600 and CFU), and a number of different media (rich 7H9, HdB minimal media with carbon and nitrogen, and HdB minimal media without carbon and nitrogen), we could not find any conditional phenotypes in any of the gene candidates we selected (Figure 3.4B-D). For one ribosome-associated factor, Rv2629, there was no Msm homolog. To test if this associated factor had any cost to Msm growth, we expressed a copy of Rv2629 in Msm, to observe any toxic effects of the protein. We found that
73 Msm cells with Rv2629 had a significantly delayed transition into exponential growth (Figure
3.5A). a) b)
0.8 10
0.6 0 0 0.4 1
WT OD60 OD60 0.2 WT L5::Rv2629
0.0 0.1 0 50 100 150 200 250 0 2 4 6 8 Time (minutes) Time (days) c) d) 10 PBS-T Growth l ph 4.5 Killing l 1 1 0.1
0.01 0.1 0.001 Proportion Surviva
Proportion Surviva 0.0001 0.01 0 10 20 30 40 0.00001 0 2 4 6 8 Time (days) Time (days) WT Rv0120c Rv1112 Rv2629 Rv3228
Figure 3.5 - Phenotyping of ribosome-associated protein mutants in Mtb a) Growth curve of strains with and without Rv2629 in Msm. b) Growth curve of Mtb mutants for ribosome-associated candidates in 7H9. c) Survival analysis of Mtb mutants in PBS-tyloxapol over time. Proportion survival indicates CFU normalized by CFU at t=0. d) Survival analysis of Mtb mutants in PBS- tyloxapol over time. Proportion survival indicates CFU normalized by CFU at t=0.
In Mtb, we made mutants in ribosome-associated proteins that we deemed high-priority due to genetic essentiality or novel associations to the ribosome. Rv0120c, and Rv1112c are both conditionally essential in the mouse, while Rv3228c and Rv1112c are both conditionally required in vitro in acidified media [45]. Rv0120c, Rv1112c, and Rv3228c were also all
GTPases likely to function during translation. The fourth target, Rv2629, was a hypothetical
74 peptide release factor, that is transcriptional controlled by DosR. We assessed growth rates of mutants in rich 7H9 media by growth curve and found no significant differences between any of the mutants (Figure 3.5B). To assess survival of the mutants in a minimal media, we chose to grow mutants in PBS-T rather than HdB minimal media, given Mtb’s propensity to clump in HdB media (personal communication with Jessica Pinkham). We measured survival in the candidate mutants, grown in PBS-Tyloxapol, over the course of 6 weeks. We found little variation between mutant survival and WT survival in this starvation media (Figure 3.5C). From the genetic screening data, we had evidence that during acid stress, Mtb may require some of these candidate factors for survival. We profiled mutant survival under acid-killing (7H9 pH 4.5), and found that again, mutants had no obvious fitness defect relative to WT. Although we associated these proteins to the Mtb ribosomes by proteomics, we could not find physiological conditions where these ribosome-associated proteins may benefit the cell.
Discussion
To discover mechanisms of proteome control at the ribosome, we used quantitative proteomics to define new interacting partners with the ribosome in multiple cellular states. We found that purification of the ribosome, through sucrose density centrifugation, yielded both biologically meaningful associations and artifacts of our purification method. The three major limitations of this screen are: (1) an inability to detect some known ribosome-associated factors, (2) artefactual ribosome interactions, and (3) an inadequate separation of each of the ribosome units.
Even among the known ribosome-associated proteins, we could not enrich for the major elongation factors during log-phase protein synthesis (EF-G, EF-Tu). Likely, the sucrose gradient buffer conditions dissociated any elongation factor-ribosome interactions. Enrichment
75 of EF-Tu with E.coli ribosomes has been observed by a similar sucrose gradient/proteomics method using a different salt conditions than used in our study[46]. Additionally, a Caulobacter crescentus Obg protein (a GTPase) was co-purified with the ribosome using sucrose gradients with GTP/GDP in the buffer [47]. Separate from the initial screen, we were able to detect epitope-tagged elongation factors in a sucrose cushion purified ribosome (Appendix Figure 2.1).
A worthwhile validation would be to track the co-purification of a number of known ribosome- interacting partners and known non-interacting partners, while varying ammonium and magnesium salt concentrations in the gradient buffers. Due to the effect of salt concentration on protein-protein interactions, there is likely an empirical balance between enriching for interacting factors and biochemically depleting the purification of non-specific proteins[48].
We also found a number of proteins whose putative function suggests that they should not be associated with the ribosome. In example, Rv0432 is predicted to be a periplasmic superoxide dismutase and has an enrichment ratio equivalent to those of ribosomal protein ratios. In a living cell, the ribosome does not make contact with the periplasm. We found other examples of these types of enrichments: cytochrome oxidase C components, NADH dehydrogenase components.
Other ribosome proteomics attempts have not documented the number of proteins that non- specifically co-purify with the ribosome using sucrose density centrifugation [46, 49]; we suspect this percentage of the proteome is large. We hypothesize that performing proteomics on a cytoplasmic fraction of the sucrose gradient (the top most layer) might have served as a reasonable control for proteins that migrate close to the ribosome. Future work will incorporate modifications to the gradient conditions and employ controls like the cytoplasmic fraction.
Increasing our sensitivity and specificity in detecting true ribosome-association will be necessary for future discovery efforts.
76 Finally, we note that the separation of 50S-associated proteins and 70S- associated proteins was not as clean as the separation of 30S- and 70S-associated proteins (Figure 3.1). For some of our candidates, they are associated with both the 70S and 50S. While we admit this association could have biological meaning, we also admit that 70S particles could have contaminated 50S fraction. Certainly, it would be worth redoing the screen with a longer sucrose gradient, to better separate ribosome units (10-30% sucrose vs. 10-40% sucrose).
Despite these limitations, our proteomics dataset tells us a number of interesting physiological observations about mycobacterial ribosomes. In log-phase, ribosomes were populated with a number of expected and unexpected proteins. Two proteins, RafS, and RsfS, considered hibernation/silencing factors on the ribosomes, were found heavily associated with the log- phase 70S and log-phase 50S [50, 51]. We find that the RafS association increases at the stressed 70S ribosome, as expected from a homolog of a hibernation promoting factor [52]. We speculate that during log-phase, Mtb may use a number of silencing factors to maintain a population of inactive ribosomes. In other organisms, there is a population of ribosomes associated with hibernation factors during log-phase growth [53]. Our data suggests that a larger than expected population of ribosomes are associated with these hibernation factors, or at least all the hibernation factors in the cell are associated with ribosomes. A measurable population of 70S particles co-purify with RafS, indicating that a significant number of ribosomes are not translating. The role of these ‘silent’ ribosomes in log-phase is unclear.
Additionally, we found that a significant proportion of ribosomes are associated with Sec- translocon machinery (either SecY or Ffh) and most ESX- 3 and ESX-5 (type VII secretion systems) components. Again, admitting the caveats of the screen, our data suggests that at any given time, there are a significant proportion of ribosome associated to the type VII secretion
77 systems. What is the proportion of cytoplasmic ribosomes vs. membrane-bound ribosomes in
Mtb at any given time? In mycobacteria, ESX-associated proteins do not contain Sec or Tat signals in their amino acid sequence, but do possess C-terminal secretion signals[54, 55]. The mechanism of how these ‘virulence’ factors are produced, trafficked and secreted outside the cell from the ribosome is an important gap in our understanding of mycobacterial secretion, and is likely critical to pathogenesis.
Future endeavors will validate these two associations to the log-phase ribosome, RafS and Sec translocon/type VII secretion components. We suggest using an antibody co- immunoprecipitation to determine if any ribosomal proteins are precipitated with the candidate proteins. Alternatively, bacterial ribosomes tagged with fluorescent proteins [56] could be co- localized with our candidate proteins using fluorescent microscopy. We hypothesize that these associations will also change based on condition. SecE2 was increased in enrichment at the stressed 70S, and we hypothesize that SecE1 and SecE2 may be required for Sec-translocon activity at different times. Understanding both log-phase ribosome heterogeneity and condition- dependent changes in the heterogeneity, using the protein markers discussed above, will be important experiments in the discovery of ribosome factors that are critical for mycobacteria.
We chose to study mostly non-essential candidates that associate to the ribosome. At the time of candidate selection, we did not incorporate the use of p-values, because we only had biological duplicates. Specifically for the candidates we pursued in Mtb, we chose genes with conditional essentiality during murine infection or during in vitro alternative media conditions.
We hypothesized that these ribosome-associated factors would be conditionally required at the ribosome, and therefore contribute to cell growth and survival in such conditions. Although we associated these factors to the ribosome units, we did not discover the condition in which the
78 gene product is most essential to the cell. In the case of some candidates (Rv0120c, Rv1112), an Mtb infection condition might be the place to find a phenotype for cells lacking these genes.
To understand processes critical to the mycobacterial ribosome, we suggest the pursuit of only essential candidates or conditionally essential candidates. In addition, we suggest that recently published genetic screening data be used to triage candidates with meaningful conditional phenotypes. Rv2837, for example, is a strong candidate to pursue, as it becomes essential for survival in most in vitro drug treatments assayed in the chemical genetic screen [42].
This work is the first biochemical screen of ribosome interactions in Mtb in multiple cell states.
While we have found many meaningful associations at the log-phase ribosome and the stressed ribosome, we also admit the challenges that remain to detect true ribosome associations in Mtb.
Future efforts will engage both genetic and biochemical experiments in Mtb to understand targetable vulnerabilities in the regulation of protein synthesis machinery.
Materials and Methods
Bacterial growth conditions
For M. tuberculosis mc27000 (a biosafety level 2 + auxotrophic TB strain), cultures were grown in 7H9 containing 10% oleic albumin dextrose catalase, 0.5% glycerol and 0.05% Tween-80 with 50 ug/ml of pantothenic acid. One culture was washed and resuspended in Hartman’s de
Bont (HdB) minimal media [57] with 0.05% tyloxapol and no carbon or nitrogen source. Cells were allowed to grow in the media for two weeks before cell lysis. For M. smegmatis experiments, cells were grown in 7H9 liquid media containing ADC and 0.05% Tween-80 and strains were built by plating on LB-agar. For M. tuberculosis H37Rv experiments, cells were grown in 7H9 liquid media, containing OADC and 0.05% Tween-80. Knockout strains were constructed by plating cells on 7H10 solid media with OADC and 0.05% Tween-80. Appropriate
79 antibiotics or inducing agents were used at the following concentrations in M. smegmatis and
M.tuberculosis: zeocin (Zeo,20 µg/ml), kanamycin (Kan,25 µg/ml), hygromycin B (Hyg,50
µg/ml), anhydrous tetracycline (aTc, 500 ng/ml), isovaleronitrile (IVN, 0.01 nM)
Isolation of ribosomes
Ribosomes were purified as described previously [51]. Briefly, Mtb mc27000 auxotroph was grown to late-log phase (OD600 0.8-1.0). To generate stressed ribosome, late-log phase cells were washed and transitioned to HdB minimal media. Cells were collected by centrifugation at
4°C, resuspended in lysis buffer (20 mM Tris-HCl pH 7.5, 100 mM NH4Cl, 10 mM MgCl2, 0.5 mM EDTA, 6 mM b-mercaptoethanol) in a bead beater (BioSpec). Cell lysate was clarified by centrifugation at 30,000 × g for 1 hr. Aliquots of this whole-cell lysate were saved for proteomics. Sucrose gradients were made by manually layering fractions of 10%, 20%, 30% and 40% sucrose in gradient buffer (5mM HEPES, 10mM NH4Cl, 50mM KCl, 10mM MgCl2, and
6mM b-mercaptoethanol); gradients were left to equilibrate at 4°C. 80 ODs of cell lysate
(~300ul) were loaded onto 10%–40% linear sucrose gradients, centrifuged in a Beckman SW28 rotor at 19,000 rpm for 19 hr. Gradients were analyzed by FPLC, and 70S, 50S, 30S and other ribosome species were collected. For each condition and replicate, 6 gradients were fractionated and ribosome units were pooled. Ribosome units were concentrated and sucrose gradient buffer was exchanged with 100 mM ammonium bicarbonate.
Proteomics
Samples were prepared for quantitative proteomics as previously described [32]. Briefly, samples were Lys-C and trypsin digested overnight and labeled with TMT-10 reagents (Thermo
Fisher Scientific). Label incorporation was confirmed by LC-MS/MS. To decrease sample complexity, labeled and digested samples were combined and basic reverse phase (bRP)
80 fractionated as described [58] The sample proteome was acquired using a Q-Exactive+ mass spectrometer (Thermo Fisher Scientific), with peptide/protein identification performed in
Spectrum Mill (Agilent), and peptide FDRs calculated as previously described [59]. Only proteins with > 2 unique peptides, positive TMT quantification, and mapping to the
Mycobacterium tuberculosis H37Rv proteome were used for analysis.
Data analysis
The two TMT-10plex proteins reports were analyzed as follows. 86% of proteins discovered in the two TMT-10plexs were found in both datasets. Only those proteins were further analyzed.
Reporter ion intensities for duplicates were averaged, and standard deviations used for later error propagation. TMT-10plexes were combined by dividing each channel reporter ion intensity by that of the log 70S channel reporter ion intensity (averaged TMT-126 and TMT-127N). Using propagation of error, standard deviation for these linker-normalized ratios (! ) was equaled to ! ∙
# & # & % $ + ( , where ! = $ . We define each protein p as a vector of reporter ion %$ %( %( intensity ratios (vL70S, vL50S, vL30S, vS70S, vS50S, vS30S, vLWCL, vSWCL), where vL70S=1. To quantify enrichment (ratio of r) of protein p with a given ribosome unit (70S, 50S, or 30S), r=vunit,p/vWCL,p, where the corresponding whole cell lysate proteome is used as the normalizing factor for the given unit . To discover conditionally-associated ribosomal proteins, three enrichment ratios were used to find candidates: a whole-cell lysate normalization strategy, rw, where rw=(vS70S/ vSWCL)/(vL70S/ vLWCL), a subunit normalization strategy, rsub, where rsub=(vS70S/( vS50S+ vS30S))/(vL70S/( vL50S+ vL30S)), and strategy with no normalization, rs, where rs=(vS70S/vL70S). Proteins that were at least one standard deviation above the mean enrichment ratio for each strategy were considered as final candidates. P-values were calculated using Welsh’s t-test with the means of each ratio and the derived standard deviation.
81 Candidate selection
Proteins were selected as candidates through mining the literature to understand: 1) if the protein had a structure that was likely to be found associated with the ribosome 2) if the gene had any genetic screen data indicative of conditional requirements in mycobacteria 3) if the protein/gene had novel features in structure or putative function.
Construction of Msm mutants
To delete genes encoding proteins that were associated to the Mtb ribosome, we used a novel genetic recombineering method called oligonucleotide-mediated recombineering followed by
Bxb1 integrase targeting (ORBIT). An Msm strain carrying a RecT,Bxb1 plasmid was grown to log-phase and recombination machinery was induced with 500 ng/mL a for three hours at 37 degrees C. Cells were made competent and co-transformed with oligonucleotides targeting each gene candidate and a vector containing a HygR marker (pKM446). Colonies were screened as previously described [60].
Expression of Rv2629 in Msm
Rv2629 was cloned into an L5-integrating vector (using Gibson Assembly) under a strong,constitutive promoter, and transformed in mc2155[61].
Phenotyping of Msm mutants
To determine if any of the candidate ribosome factors contributed to fitness in mycobacteria, we examined deletion mutants for growth defects in a number of growth scenarios. For starvation, cells were grown to log-phase and wash 3x in the corresponding media. For growth curves, strains were diluted to OD 0.05 and grown in 96-well plates using the TECAN Spark 10M plate reader, in 7H9 liquid culture (rich media) or HdB media with 0.005% glycerol and 1.5 mM
82 (NH4)2SO4 (HdB reduced media). To measure death related to loss of any putative ribosome- associated factors, strains were grown to log-phase, washed 3x with HdB minimal media containing no carbon or nitrogen, or starvation media, and resuspended at OD 0.05. Cells were grown in starvation media and plated for CFU approximately once a day for 15 days.
Construction of Mtb mutants
Gene candidates encoding proteins associated to the ribosome were deleted in Mtb H37Rv using traditional mycobacterial recombineering methods [62, 63]. All deletion mutants were
HygR-marked after recombineering.
Phenotyping of Mtb mutants
Mtb mutants in genes with conditionally essential phenotypes were examine for growth defects during starvation. Cells were washed 3x in PBS-0.05% Tyloxapol (PBS-T) and resuspended at
OD 0.05. Cell were grown in PBS-T for 6 weeks and plated for CFU analysis once a week. For survival at low pH, cells were grown to log-phase, diluted to OD 0.05 washed 3x in 7H9-0.05%
Tyloxapol ph 4.5 and resuspended in 7H9-0.05% Tyloxapol ph 4.5. Cultures were plated for
CFU analysis over the course of a week.
References
1. Silva, D.R., et al., New and repurposed drugs to treat multidrug- and extensively drug-resistant tuberculosis. J Bras Pneumol, 2018. 44(2): p. 153-160. 2. Makhado, N.A., et al., Outbreak of multidrug-resistant tuberculosis in South Africa undetected by WHO-endorsed commercial tests: an observational study. Lancet Infect Dis, 2018. 18(12): p. 1350-1359. 3. Berry, C., J. Achar, and P. du Cros, WHO recommendations for multidrug- resistant tuberculosis. Lancet, 2016. 388(10057): p. 2234. 4. Gulland, A., More cases of tuberculosis than previously thought, WHO reports. BMJ, 2016. 355: p. i5562. 5. Chen, M.Y., et al., Recurrence after Successful Treatment of Multidrug-Resistant Tuberculosis in Taiwan. PLoS One, 2017. 12(1): p. e0170980.
83 6. Verver, S., et al., Rate of reinfection tuberculosis after successful treatment is higher than rate of new tuberculosis. Am J Respir Crit Care Med, 2005. 171(12): p. 1430-5. 7. Nataraj, V., et al., Mycolic acids: deciphering and targeting the Achilles' heel of the tubercle bacillus. Mol Microbiol, 2015. 98(1): p. 7-16. 8. Lin, P.L. and J.L. Flynn, The End of the Binary Era: Revisiting the Spectrum of Tuberculosis. J Immunol, 2018. 201(9): p. 2541-2548. 9. Huang, L., E.V. Nazarova, and D.G. Russell, Mycobacterium tuberculosis: Bacterial Fitness within the Host Macrophage. Microbiol Spectr, 2019. 7(2). 10. VanderVen, B.C., et al., The Minimal Unit of Infection: Mycobacterium tuberculosis in the Macrophage. Microbiol Spectr, 2016. 4(6). 11. Wong, K.W., The Role of ESX-1 in Mycobacterium tuberculosis Pathogenesis. Microbiol Spectr, 2017. 5(3). 12. Tinaztepe, E., et al., Role of Metal-Dependent Regulation of ESX-3 Secretion in Intracellular Survival of Mycobacterium tuberculosis. Infect Immun, 2016. 84(8): p. 2255-2263. 13. Libardo, M.D.J., et al., Phagosomal Copper-Promoted Oxidative Attack on Intracellular Mycobacterium tuberculosis. ACS Infect Dis, 2018. 4(11): p. 1623- 1634. 14. Corleis, B., et al., Escape of Mycobacterium tuberculosis from oxidative killing by neutrophils. Cell Microbiol, 2012. 14(7): p. 1109-21. 15. Lamprecht, D.A., et al., Turning the respiratory flexibility of Mycobacterium tuberculosis against itself. Nat Commun, 2016. 7: p. 12393. 16. Lee, W., et al., Intracellular Mycobacterium tuberculosis exploits host-derived fatty acids to limit metabolic stress. J Biol Chem, 2013. 288(10): p. 6788-800. 17. Martinot, A.J., et al., Mycobacterial Metabolic Syndrome: LprG and Rv1410 Regulate Triacylglyceride Levels, Growth Rate and Virulence in Mycobacterium tuberculosis. PLoS Pathog, 2016. 12(1): p. e1005351. 18. Pandey, M., et al., Iron homeostasis in Mycobacterium tuberculosis is essential for persistence. Sci Rep, 2018. 8(1): p. 17359. 19. Chao, A., et al., Iron Acquisition in Mycobacterium tuberculosis. Chem Rev, 2018. 20. Prusa, J., D.X. Zhu, and C.L. Stallings, The stringent response and Mycobacterium tuberculosis pathogenesis. Pathog Dis, 2018. 76(5). 21. Schnappinger, D., et al., Transcriptional Adaptation of Mycobacterium tuberculosis within Macrophages: Insights into the Phagosomal Environment. J Exp Med, 2003. 198(5): p. 693-704. 22. Walters, S.B., et al., The Mycobacterium tuberculosis PhoPR two-component system regulates genes essential for virulence and complex lipid biosynthesis. Mol Microbiol, 2006. 60(2): p. 312-30. 23. Pandey, R. and G.M. Rodriguez, IdeR is required for iron homeostasis and virulence in Mycobacterium tuberculosis. Mol Microbiol, 2014. 91(1): p. 98-109. 24. Voskuil, M.I., et al., Inhibition of respiration by nitric oxide induces a Mycobacterium tuberculosis dormancy program. J Exp Med, 2003. 198(5): p. 705-13.
84 25. Shiloh, M.U., P. Manzanillo, and J.S. Cox, Mycobacterium tuberculosis senses host-derived carbon monoxide during macrophage infection. Cell Host Microbe, 2008. 3(5): p. 323-30. 26. Hicks, N.D., et al., Clinically prevalent mutations in Mycobacterium tuberculosis alter propionate metabolism and mediate multidrug tolerance. Nat Microbiol, 2018. 3(9): p. 1032-1042. 27. Kendall, S.L., et al., Cholesterol utilization in mycobacteria is controlled by two TetR-type transcriptional regulators: kstR and kstR2. Microbiology, 2010. 156(Pt 5): p. 1362-71. 28. Liu, Y., A. Beyer, and R. Aebersold, On the Dependency of Cellular Protein Levels on mRNA Abundance. Cell, 2016. 165(3): p. 535-50. 29. Shell, S.S., et al., Leaderless Transcripts and Small Proteins Are Common Features of the Mycobacterial Translational Landscape. PLoS Genet, 2015. 11(11): p. e1005641. 30. Li, Z., et al., Cryo-EM structure of Mycobacterium smegmatis ribosome reveals two unidentified ribosomal proteins close to the functional centers. Protein Cell, 2018. 9(4): p. 384-388. 31. Shasmal, M. and J. Sengupta, Structural diversity in bacterial ribosomes: mycobacterial 70S ribosome structure reveals novel features. PLoS One, 2012. 7(2): p. e31742. 32. Gerrick, E.R., et al., Small RNA profiling in Mycobacterium tuberculosis identifies MrsI as necessary for an anticipatory iron sparing response. Proc Natl Acad Sci U S A, 2018. 115(25): p. 6464-6469. 33. Li, Y., et al., Zinc depletion induces ribosome hibernation in mycobacteria. Proc Natl Acad Sci U S A, 2018. 115(32): p. 8191-8196. 34. Trauner, A., et al., The dormancy regulator DosR controls ribosome stability in hypoxic mycobacteria. J Biol Chem, 2012. 287(28): p. 24053-63. 35. Genuth, N.R. and M. Barna, The Discovery of Ribosome Heterogeneity and Its Implications for Gene Regulation and Organismal Life. Mol Cell, 2018. 71(3): p. 364-374. 36. Zundel, M.A., G.N. Basturea, and M.P. Deutscher, Initiation of ribosome degradation during starvation in Escherichia coli. RNA, 2009. 15(5): p. 977-83. 37. Jankiewicz, U. and W. Bielawski, The properties and functions of bacterial aminopeptidases. Acta Microbiol Pol, 2003. 52(3): p. 217-31. 38. Garcie, C., et al., The Bacterial Stress-Responsive Hsp90 Chaperone (HtpG) Is Required for the Production of the Genotoxin Colibactin and the Siderophore Yersiniabactin in Escherichia coli. J Infect Dis, 2016. 214(6): p. 916-24. 39. Zhang, Y., et al., A Stress Response that Monitors and Regulates mRNA Structure Is Central to Cold Shock Adaptation. Mol Cell, 2018. 70(2): p. 274-286 e7. 40. Griffin, J.E., et al., High-resolution phenotypic profiling defines genes essential for mycobacterial growth and cholesterol catabolism. PLoS Pathog, 2011. 7(9): p. e1002251. 41. Sassetti, C.M. and E.J. Rubin, Genetic requirements for mycobacterial survival during infection. Proc Natl Acad Sci U S A, 2003. 100(22): p. 12989-94.
85 42. Xu, W., et al., Chemical Genetic Interaction Profiling Reveals Determinants of Intrinsic Antibiotic Resistance in Mycobacterium tuberculosis. Antimicrob Agents Chemother, 2017. 61(12). 43. Mohapatra, S., et al., Spatial Distribution and Ribosome-Binding Dynamics of EF- P in Live Escherichia coli. MBio, 2017. 8(3). 44. Bradshaw, R.A., W.W. Brickey, and K.W. Walker, N-terminal processing: the methionine aminopeptidase and N alpha-acetyl transferase families. Trends Biochem Sci, 1998. 23(7): p. 263-7. 45. Zhang, Y.J., et al., Tryptophan biosynthesis protects mycobacteria from CD4 T- cell-mediated killing. Cell, 2013. 155(6): p. 1296-308. 46. Jiang, M., et al., Identification of novel Escherichia coli ribosome-associated proteins using isobaric tags and multidimensional protein identification techniques. J Bacteriol, 2007. 189(9): p. 3434-44. 47. Lin, B., D.A. Thayer, and J.R. Maddock, The Caulobacter crescentus CgtAC protein cosediments with the free 50S ribosomal subunit. J Bacteriol, 2004. 186(2): p. 481-9. 48. Mehta, P., et al., Ribosome purification approaches for studying interactions of regulatory proteins and RNAs with the ribosome. Methods Mol Biol, 2012. 905: p. 273-89. 49. Butland, G., et al., Interaction network containing conserved and essential protein complexes in Escherichia coli. Nature, 2005. 433(7025): p. 531-7. 50. Mishra, S., et al., Structures of Mycobacterium smegmatis 70S ribosomes in complex with HPF, tmRNA, and P-tRNA. Sci Rep, 2018. 8(1): p. 13587. 51. Li, X., et al., Structure of Ribosomal Silencing Factor Bound to Mycobacterium tuberculosis Ribosome. Structure, 2015. 23(12): p. 2387. 52. Polikanov, Y.S., G.M. Blaha, and T.A. Steitz, How hibernation factors RMF, HPF, and YfiA turn off protein synthesis. Science, 2012. 336(6083): p. 915-8. 53. Ueta, M., C. Wada, and A. Wada, Formation of 100S ribosomes in Staphylococcus aureus by the hibernation promoting factor homolog SaHPF. Genes Cells, 2010. 15(1): p. 43-58. 54. Feltcher, M.E., J.T. Sullivan, and M. Braunstein, Protein export systems of Mycobacterium tuberculosis: novel targets for drug development? Future Microbiol, 2010. 5(10): p. 1581-97. 55. Champion, P.A., et al., C-terminal signal sequence promotes virulence factor secretion in Mycobacterium tuberculosis. Science, 2006. 313(5793): p. 1632-6. 56. Nikolay, R., et al., Fluorescence-based monitoring of ribosome assembly landscapes. BMC Mol Biol, 2015. 16: p. 3. 57. Smeulders, M.J., et al., Adaptation of Mycobacterium smegmatis to stationary phase. J Bacteriol, 1999. 181(1): p. 270-83. 58. Mertins, P., et al., Integrated proteomic analysis of post-translational modifications by serial enrichment. Nat Methods, 2013. 10(7): p. 634-7. 59. Elias, J.E. and S.P. Gygi, Target-decoy search strategy for mass spectrometry- based proteomics. Methods Mol Biol, 2010. 604: p. 55-71. 60. Kenan C. Murphy, S.J.N., Subhalaxmi Nambi, Kadamba Papavinasasundaram, Christina E. Baer, Christopher M. Sassetti, ORBIT: a new paradigm for genetic engineering of mycobacterial chromosomes. BioRxiv, 2018.
86 61. Gibson, D.G., Synthesis of DNA fragments in yeast by one-step assembly of overlapping oligonucleotides. Nucleic Acids Res, 2009. 37(20): p. 6984-90. 62. Murphy, K.C., K. Papavinasasundaram, and C.M. Sassetti, Mycobacterial recombineering. Methods Mol Biol, 2015. 1285: p. 177-99. 63. van Kessel, J.C. and G.F. Hatfull, Mycobacterial recombineering. Methods Mol Biol, 2008. 435: p. 203-15.
87
Chapter 4
Discussion
88 4.1 LepA, a translational GTPase, controls mycobacterial permeability
4.1.1 LepA’s role in mycobacteria
To understand the role of accessory translational GTPases at the mycobacterial ribosome, we characterized LepA-mediated control of the synthesis of a family of abundant outer membrane porins in Msm. We conducted a proteomics-transcriptomics screen to discover what proteins
LepA might be controlling post-transcriptionally. To validate candidates from the screen, candidates were tagged with a C-terminal luciferase reporter and expressed from a strong constitutive promoter (pUV15) in cells with and without LepA. Only a subset of the porins showed a robust 2-3 fold increase in reporter signal due to LepA. Although the screen identified other candidates that might be regulated by LepA, we did not find any as robustly altered as the porins. Correspondingly, the DlepA phenotype was largely due to LepA regulation of MspA. We tested a number of other candidates uncovered from the screen, using our luciferase assay.
While we may have detected small LepA-dependent changes in at least one of the candidates, these changes were not as robust as those found in the porins (Appendix Figure 1.4). Why was the LepA influence so specific to the porins?
We suspect that LepA must act on other ribosome-bound transcripts in the cell. Our inability to find other candidates could be due to a number of factors involving the magnitude of the LepA effect, our detection method, or the condition assayed. From absolute proteomic estimates in
Mtb, LepA is not highly abundant like other elongation factors[1]. We hypothesize that its activity at the ribosome is likely determined by transcript level. Therefore, a highly abundant transcript like mspA, may be the major target of LepA in log-phase growing Msm, simply because of the ratio of LepA to translating ribosomes. Is it possible that in a DmspA background, we might find increased LepA effect on other candidates because of an increased availability of LepA?
Alternatively, perhaps the structure of the porin mRNA at the ribosome is the major determinant
89 of LepA recruitment, and we did not assess the conditions in which other mRNA structures might require LepA activity during translation. This question merits further investigation, as more validated candidates for LepA regulation would allow us to define motifs that enable LepA activity at a given mRNA transcript. Future efforts concerning the discovery of LepA effects in mycobacteria should focus on the use of ribosome profiling, to directly identify transcript-specific differences in ribosome occupancy due to LepA.
Loss of LepA causes upregulation of a number of members of the WhiB7 regulon. The most upregulated transcript in cells without LepA is an integral membrane protein, regulated by
WhiB7. WhiB7 is a transcriptional regulator in actinomycetes that is responsive to redox- changes in the cell and upregulates genes involved in antibiotic resistance [2]. In particular,
WhiB7 programs intrinsic resistance to ribosome-targeting antibiotics. Why is WhiB7 upregulated in the absence of LepA? The simplest explanation in this case, is likely that transcriptional changes in DlepA are a result of porin loss at the outermembrane. We hypothesize that without LepA, Msm cells experience defects in nutrient, metal ion, and phosphate uptake due to loss of MspA [3-5]. Hypothetically, metal or phosphate loss might alter redox homeostasis in the cell, perhaps causing activation of WhiB7-relate genes. To test this, one could measure WhiB7-regulon activation in DmspA, DlepA, or DlepADmspA backgrounds to ask if activation of WhiB7 also occurs due to the loss of mycobacterial porin. If this altered
WhiB7-activation is specific to the DlepA background, it suggests activation due to effects of other LepA-regulated proteins or perhaps other downstream effects of ribosome stalling in the absence of LepA.
Finally, DlepA strains are conditionally sick in Mtb during mouse infection [6] (personal communication with Allison Carey), and at a growth advantage in sub-MIC levels of isoniazid[7].
90 Is this growth advantage synonymous to the growth advantage, or drug tolerance, that Msm
DlepA displays during rifampin killing. If so, it highlights a synonymous effect from loss of LepA.
We hypothesize that LepA may control protein synthesis of TB-specific cell membrane processes that manifest in pathogenesis defects and altered drug tolerance. Future work will address the regulon of LepA in TB and its importance during pathogenesis/drug treatment.
4.1.2 LepA and organismal membrane protein synthesis
In other organisms, the biochemistry of LepA at the ribosome has been extensively characterized [8-13]. Yet, the physiological function of LepA in a cell or organelle remains debated. The earliest literature identified lepA as a membrane-associated protein, in an operon with lepB, an E.coli gene encoding a signal peptidase protein, responsible for maturing membrane/secreted proteins [14]. Data from both chloroplasts and mitochondria have shown that loss of LepA alters the levels of membrane proteins/complexes [15, 16]. Here, our data adds to the growing body of evidence that LepA controls the quality of membrane-protein synthesis. Retrospectively, our data is in line with data from E. coli, indicating that a DlepA strain had altered antibiotic tolerance. In E.coli, DlepA cells are less susceptible to kanamycin, oxolinic acid, and ampicillin [17]. Again, Msm DlepA cells are less susceptible to rifampin and vancomycin. Taken together, with our understanding of LepA control of MspA synthesis, we hypothesize that drug tolerance changes in DlepA cells are the result of less efficient translation of membrane proteins (through LepA) and results in different cellular permeability to drugs. In addition, LepA regulation of MspA in mycobacteria is analogous to the alterations in membrane- localized respiration complexes that appear in lepA- mitochondria and chloroplasts. We were unable to localize LepA to the mycomembrane (data not shown), but if our model is correct, a
LepA-Sec translocon interaction through the ribosome is likely to exist. This question remains
91 an important one in the pursuit of understanding the interactions of LepA with ribosomes in both bacteria and eukarya.
4.1.3 Why do organisms keep LepA: its essential function
One of the most compelling questions regarding LepA is why is the protein conserved across phyla? Cells survive without LepA, which could indicate functional redundancy in the cell. And yet, LepA has not been lost in many organisms [18]. In our model, we propose that LepA back- translocates the ribosome at a time during MspA translation when the signal peptide emerges from the exit tunnel, and requires recognition proteins (SRP) to enable membrane co- translational secretion. The purported back-translocation activity by LepA could change the rate of translation around a translational arrest, protecting the nascent peptide from being prematurely produced and aggregated. We suspect if this model is preserved across organisms, cells possess multiple kinds of membrane translocon-associating proteins that enable translational pausing around signal peptide recognition; these kinds of factors would provide redundancy over LepA function[19]. In eukaryotes, the signal recognition particle contacts the elongation factor binding region of the ribosome, arresting translation during signal peptide recognition[20]. While this mechanism has not been observed in bacteria, it is clear that recognition of a peptide destined for the membrane requires precise changes in translation rate.
Alternatively, LepA iteratively associates to transcripts with a particular mRNA structure and enables the coordinated co-translational folding of the membrane protein. Human mitochondrial respiration complexes made in organelles without LepA are not just less abundant, they are also less active [21]. If LepA’s role is involved in membrane translation of MspA, perhaps the synthesis of this porin octamer requires iterative LepA back-translocation events, enabling folding of the stable octamer into the mycobacterial outer-membrane. If this model is true, cells could employ a host of chaperones to compensate for loss of LepA activity [22]. We suspect
92 that in either case, LepA’s conservation hints at an important cellular process. To discover such redundant systems in LepA, we propose transposon-sequencing (TnSeq), in a DlepA mutant, in
Msm and Mtb, to identify conserved processes, synthetically lethal with lepA that might overlap with LepA function at the mycobacterial ribosome. Membrane protein synthesis is a complex task that requires a number of proteins at the exit tunnel, and perhaps, at the peptidyl- transferase-center, to properly build proteins into membranes.
4.2 New factors and new associations in mycobacteria: a glimpse of ribosome heterogeneity
4.2.1 Ribosome proteomics screen for translational control
We were interested in finding new protein associations with the ribosome to help uncover Mtb ribosomal regulation. Using an auxotrophic strain of Mtb (mc27000), we examined ribosome- associated proteomes in cells grown in rich (log-phase ribosomes) and minimal media (stressed ribosomes). We then purified the ribosomes and performed quantitative proteomics to detect interacting proteins. At the log-phase ribosome, we found unexpected enrichment of secretion systems and ribosome hibernation factors. To our knowledge, this is the first evidence of a ribosome-type VII secretion system association. Specifically, the mechanism of how ESX effector proteins are produced and secretion is not well understood. The ribosomes were robustly associated with EccB, EccC, and EccD of both ESX-3 and ESX-5 system. While EccB and EccD are likely integral membrane components of the ESX system, EccC has ATPase domains that make contact with the cytoplasm[23, 24]. If there is direct binding of the ribosome to an ESX-system, we hypothesize it would be between the 70S ribosome and EccC. The only known signal required for secretion through an ESX system is a C-terminal motif [25]. Future experiments delving into the nature of the ribosome-type VII secretion system relationship will hinge upon the orthogonal validation of this interaction.
93
At the stressed ribosome, we found a number of proteins involved in post-translational activity.
Of note, we found both chaperones such as HtpG, DnaK, and GroEL2. Additionally, we found a number of peptidases and proteases. The presence of both chaperones and protein degradation machinery indicates that stressed Mtb cells may be dealing with proteotoxicity or stalled ribosomes. The association of ‘heat-shock’ chaperones to the ribosome has been observed as a response to acute stress in some bacteria[22, 26]. In addition, a stalled ribosome, or a tmRNA-tagged ribosome may require protease activity to destroy the incompletely synthesized nascent peptide[27]. In either case, it is tempting to speculate that upon acute stress, the Mtb ribosome may be surrounded by upregulated chaperones and protein degradation systems, protecting the cells from proteotoxicity.
4.2.2 Phenotypic data from ribosomal associations
We were interested in proteins that might play a critical role in remodeling the Mtb ribosome either during log-phase or during stress. To determine which proteins might be critical to Mtb physiology, we explored the fitness cost of a number of ribosome-associated candidates in Msm and Mtb. We made gene knockouts for protein candidates associated to the ribosome, specifically choosing proteins with previous association to some translational process in bacteria. We also chose candidates with an emphasis on proteins of biological interest i.e. those with essentiality requirements from TnSeq data, transcriptional regulons of interest, hypothetical proteins.
We found one gene whose loss contributed to fitness defects in mycobacteria. A deletion mutant in mapB (MSMEG_2587/Rv2861c), one of the methionine aminopeptidases in Msm, had a slower doubling time in log-phase than wildtype Msm. Two methionine aminopeptidases,
94 MapA and MapB, were associated with the stressed ribosome in Mtb; we also found MapB associated with the log-phase 70S ribosome. Methionine aminopeptidases are universally essential enzymes that cleave off the formyl-methionine from the nascent peptide [28]. Rv2861c is non-essential in Mtb growth in vitro in rich 7H9 [29]. Our growth curve data indicates that
MapB may be the major methionine aminopeptidase in Msm. Yet, the unanswered question remains, why do these enzymes become associated with the stressed 70S ribosome? A number of chaperones and proteases also associate to the stressed ribosome in the proteomics enrichment data, indicating that protein degradation is potentially occurring in the vicinity of the
70S ribosome in stressed Mtb. We hypothesize that perhaps increased proteolysis occurring at the ribosome may require increased cleavage of the formyl-methionine. The association of
Maps to the ribosome indicates the importance of peptide processing at the ribosome, as the two Map enzymes in Mtb have previously been explored as potential drug targets [30].
4.2.3 Novel ribosome-associated proteins
We associated a hypothetical protein, Rv2629, to both the log-phase 70S ribosome and the stressed 70S ribosome. Rv2629 only exists in Mtb, and the knockout strain had no observable defects in the conditions we assayed. Yet, when we expressed a copy of this protein in Msm, the strain had a lag-phase growth defect relative to wildtype. This gene has been defined as a member of the DosR regulon, a group of genes upregulated during hypoxia [1, 31, 32]. Also under the control of DosR, RafH and RafS are ribosome-associated proteins that help preserve the ribosome during hypoxia [33]. We hypothesize that Rv2629, which resembles a peptidyl release factor, may be responsible for inhibiting translation as a response to hypoxia. This activity would explain the Msm data, where cells expressing extra of the protein had a delayed transition into exponential phase. Future efforts would first comparatively assess survival of WT
Mtb and DRv2629. Given that it is a member of the DosR regulon, we hypothesize that the
95 mutant will suffer a growth defect either transitioning into or out of hypoxia. If this phenotype is validated, one could assess the effect of this protein on ribosome populations, using sucrose gradient analysis, Presumably, a release factor would shift the balance of assembled ribosome
(70S) relative to ribosomal subunits (50S and 30S). This novel ribosome-associated factor, regulated by DosR in Mtb, is worth further characterization to understand its role in translation and its role in ribosomal regulation during hypoxia.
In log-phase, we found a number of novel proteins associated to the ribosome that we did not follow up with the production of gene knockouts and phenotypic characterization. Each of these protein’s putative function was determined from in silico domain homology [34, 35]. Specifically, we discovered a putative ribosomal protein (Rv2837c), that is conditionally essential in a number of Mtb drug conditions [7]. We also discovered a putative formyl methionine transferase
(Rv3404c), that is non-essential and may provide a growth advantage to the cell [29]. Finally, we found a putative ATP-binding protein/ribosome biogenesis factor (Rv3422c); transposon mutants in this gene have a growth disadvantage in Mtb. To test the hypothesis that these novel factors directly interact with the translation process, we first suggest biochemical purification and characterization of each of these proteins using a series of in vitro assays to define participation in translation. Each candidate could be tested in an in vitro translation system to gather readouts of protein synthesis and ribosome population distribution. These readouts could reveal the protein’s capacity to inhibit/help translation, and the protein’s capacity to stabilize/destabilize particular ribosome conformations [36]. These biochemical experiments should be performed in coordination with genetic experiments in Mtb that will assess and prioritize fitness defects in candidate gene mutants. A combination of biochemical and genetic experiments for candidates like these could reveal a novel protein or protein function at the Mtb ribosome.
4.2.4 Optimizing the screen to uncover true ribosome associations
96 Although we associated many proteins to the Mtb ribosome, we believe that optimizing a number of variables would increase the sensitivity and specificity of the screen. Our hypothesis remains: novel drug targets exist in protein factors that associate to the ribosome in Mtb. While we used sucrose gradients to purify ribosome-associated factors, we admit the limitations of this technique. We have previously mentioned these limitations: (1) inadequate separation of each of the ribosome units, (2) non-specific co-sedimentation of proteins with ribosomes, (3) decreased association of translational elongation factors. We predict that a number of other methods may be useful in identifying critical proteins that association to the ribosome in mycobacteria. Translating ribosome affinity purification (TRAP) is a method that relies on immunoprecipitation to enrich for associating factors to the protein of interest, or machine of interest in the case of the ribosome [37]. In mycobacteria, we propose the purification of ribosome-associated proteins using a set of aptamer-tagged essential mRNAs, and use mass spectrometry to identify associated proteins with the ribosomes. Given the large proportion of the transcriptome that is leaderless, affinity purification of leaderless transcript-bound ribosomes might provide us a set of interacting proteins that are required for leaderless translation.
4.2.5 The grass is greener on the other side of the interaction
While we pulled down the ribosome to identify ribosome-associated proteins, would it be more specific to pull down an associating protein to characterize the activity of the interacting- subpopulation of ribosomes? We use the ESX association from our dataset as an example. It is tempting to speculate that a subpopulation of ribosomes are constantly producing ESX substrates to be secreted out of the Mtb cell. To define the translational output of ESX- associated ribosomes, we propose experiments around ESX-3 substrate synthesis, as this system has a tractable in vitro phenotype in terms of iron homeostasis. EccC3 is a membrane associated ATPase that is hypothesized to be involved in protein secretion[38]. Given that
97 secretion systems often require ATPases to enable susbtrate translocation across membranes
[19], we hypothesize that this ATPase or any of its homologs in the other ESX systems, may be most likely to directly interact with the ribosome. First, co-immunoprecipitation of ESX-3 components or ESX-5 components should be used to pull down ribosomal proteins, validating the proteomics data. Then, to identify the translational output of ESX-3 associated ribosomes,
EccC3 could be used to co-immunoprecipitate ribosomes and perform ribosome profiling on the subpopulation of ribosomes[39]. Ribosome profiling could provide details of the types of transcripts that are translated by the ribosome, and the important features of these transcripts that are necessary for translation. It would be informative to perform such profiling in both iron- rich and iron-limiting conditions, as this is relevant to Mtb pathogenesis and ESX-3 activity. We hypothesize that this approach could identify novel translation control around type VII secretion systems [40].
4.3 Conclusion: Mycobacterial translation and the mycomembrane
This work has examined a myriad of ribosome-associated factors through their conditional association to the ribosome, impact on the proteome, and influence on mycobacterial fitness.
While we have characterized the costly implications of losing certain ribosome associations
(such as LepA and MapB), we have also revealed other associations whose physiological weight is yet to be understood (ESX systems, RafS, Rv2837c).
We found that a protein of low abundance in mycobacteria, LepA, causes a drastic change in the mycomembrane permeabilty through its control over MspA translation. Dedication to the proper synthesis of these outer membrane porins in mycobacteria requires a factor such as
LepA to act in coordination with the mycobacterial ribosome. While it was surprising to see LepA have such a specific effect on mycobacterial porin abundance, it indicated to us that
98 translational control through LepA was tied to the energetic demands of the mycobacterial cell i.e. synthesis of abundant membrane proteins that determine nutrient uptake in mycobacteria.
Additionally, we conducted a screen for ribosome-associated proteins in Mtb. We found a number of processes associated to the Mtb 70S ribosome, ranging from ribosomal maturation factors to peptide processing factors. These data reveal a potential for ribosome heterogeneity, where ribosome specialization comes from both external cues (DosR-regulated Rv2629) and internal cues (SsrA-binding proteins). Strikingly, we found ESX-3 and ESX-5 associated to log- phase 70S ribosomes in addition to multiple Sec-translocon related factors. It is possible that a significant amount of translational control in mycobacteria is dedicated to membrane/secretion processes. For these proteins and other essential/conditionally-essential proteins found at the ribosome, we hypothesize that these factors may underscore mechanisms that Mtb uses to maintain its proteome and ribosome populations. Future work will focus on elucidating the heterogeneity of ribosome function within a mycobacterial cell and its implications for bacteria fitness.
References
1. Schubert, O.T., et al., Absolute Proteome Composition and Dynamics during Dormancy and Resuscitation of Mycobacterium tuberculosis. Cell Host Microbe, 2015. 18(1): p. 96- 108. 2. Burian, J., et al., The mycobacterial transcriptional regulator whiB7 gene links redox homeostasis and intrinsic antibiotic resistance. J Biol Chem, 2012. 287(1): p. 299-310. 3. Jones, C.M. and M. Niederweis, Role of porins in iron uptake by Mycobacterium smegmatis. J Bacteriol, 2010. 192(24): p. 6411-7. 4. Wolschendorf, F., M. Mahfoud, and M. Niederweis, Porins are required for uptake of phosphates by Mycobacterium smegmatis. J Bacteriol, 2007. 189(6): p. 2435-42. 5. Stephan, J., et al., The growth rate of Mycobacterium smegmatis depends on sufficient porin-mediated influx of nutrients. Mol Microbiol, 2005. 58(3): p. 714-30. 6. Zhang, Y.J., et al., Tryptophan biosynthesis protects mycobacteria from CD4 T-cell- mediated killing. Cell, 2013. 155(6): p. 1296-308. 7. Xu, W., et al., Chemical Genetic Interaction Profiling Reveals Determinants of Intrinsic Antibiotic Resistance in Mycobacterium tuberculosis. Antimicrob Agents Chemother, 2017. 61(12).
99 8. Balakrishnan, R., et al., The conserved GTPase LepA contributes mainly to translation initiation in Escherichia coli. Nucleic Acids Res, 2014. 42(21): p. 13370-83. 9. Liu, H., et al., The conserved protein EF4 (LepA) modulates the elongation cycle of protein synthesis. Proc Natl Acad Sci U S A, 2011. 108(39): p. 16223-8. 10. Pech, M., et al., Elongation factor 4 (EF4/LepA) accelerates protein synthesis at increased Mg2+ concentrations. Proc Natl Acad Sci U S A, 2011. 108(8): p. 3199-203. 11. Liu, H., et al., Interrupted catalysis: the EF4 (LepA) effect on back-translocation. J Mol Biol, 2010. 396(4): p. 1043-52. 12. Evans, R.N., et al., The structure of LepA, the ribosomal back translocase. Proc Natl Acad Sci U S A, 2008. 105(12): p. 4673-8. 13. Youngman, E.M. and R. Green, Ribosomal translocation: LepA does it backwards. Curr Biol, 2007. 17(4): p. R136-9. 14. March, P.E. and M. Inouye, Characterization of the lep operon of Escherichia coli. Identification of the promoter and the gene upstream of the signal peptidase I gene. J Biol Chem, 1985. 260(12): p. 7206-13. 15. Ji, D.L., et al., CpLEPA is critical for chloroplast protein synthesis under suboptimal conditions in Arabidopsis thaliana. PLoS One, 2012. 7(11): p. e49746. 16. Gao, Y., et al., Mammalian elongation factor 4 regulates mitochondrial translation essential for spermatogenesis. Nat Struct Mol Biol, 2016. 23(5): p. 441-9. 17. Li, L., et al., Ribosomal elongation factor 4 promotes cell death associated with lethal stress. MBio, 2014. 5(6): p. e01708. 18. Qin, Y., et al., The highly conserved LepA is a ribosomal elongation factor that back- translocates the ribosome. Cell, 2006. 127(4): p. 721-33. 19. Green, E.R. and J. Mecsas, Bacterial Secretion Systems: An Overview. Microbiol Spectr, 2016. 4(1). 20. Halic, M., et al., Structure of the signal recognition particle interacting with the elongation-arrested ribosome. Nature, 2004. 427(6977): p. 808-14. 21. Zhu, P., et al., Human Elongation Factor 4 Regulates Cancer Bioenergetics by Acting as a Mitochondrial Translation Switch. Cancer Res, 2018. 78(11): p. 2813-2824. 22. Preissler, S. and E. Deuerling, Ribosome-associated chaperones as key players in proteostasis. Trends Biochem Sci, 2012. 37(7): p. 274-83. 23. Rosenberg, O.S., et al., Substrates Control Multimerization and Activation of the Multi- Domain ATPase Motor of Type VII Secretion. Cell, 2015. 161(3): p. 501-512. 24. Phan, T.H. and E.N.G. Houben, Bacterial secretion chaperones: the mycobacterial type VII case. FEMS Microbiol Lett, 2018. 365(18). 25. Champion, P.A., et al., C-terminal signal sequence promotes virulence factor secretion in Mycobacterium tuberculosis. Science, 2006. 313(5793): p. 1632-6. 26. Kirstein-Miles, J. and R.I. Morimoto, Ribosome-associated chaperones act as proteostasis sentinels. Cell Cycle, 2013. 12(15): p. 2335-6. 27. Choy, J.S., L.L. Aung, and A.W. Karzai, Lon protease degrades transfer-messenger RNA-tagged proteins. J Bacteriol, 2007. 189(18): p. 6564-71. 28. Jankiewicz, U. and W. Bielawski, The properties and functions of bacterial aminopeptidases. Acta Microbiol Pol, 2003. 52(3): p. 217-31. 29. DeJesus, M.A., et al., Comprehensive Essentiality Analysis of the Mycobacterium tuberculosis Genome via Saturating Transposon Mutagenesis. MBio, 2017. 8(1). 30. Zhang, X., et al., Expression and characterization of two functional methionine aminopeptidases from Mycobacterium tuberculosis H37Rv. Curr Microbiol, 2009. 59(5): p. 520-5. 31. Minch, K.J., et al., The DNA-binding network of Mycobacterium tuberculosis. Nat Commun, 2015. 6: p. 5829.
100 32. Rustad, T.R., et al., Mapping and manipulating the Mycobacterium tuberculosis transcriptome using a transcription factor overexpression-derived regulatory network. Genome Biol, 2014. 15(11): p. 502. 33. Trauner, A., et al., The dormancy regulator DosR controls ribosome stability in hypoxic mycobacteria. J Biol Chem, 2012. 287(28): p. 24053-63. 34. Marchler-Bauer, A., et al., CDD: NCBI's conserved domain database. Nucleic Acids Res, 2015. 43(Database issue): p. D222-6. 35. Bennett-Lovsey, R.M., et al., Exploring the extremes of sequence/structure space with ensemble fold recognition in the program Phyre. Proteins, 2008. 70(3): p. 611-25. 36. Li, X., et al., Structure of Ribosomal Silencing Factor Bound to Mycobacterium tuberculosis Ribosome. Structure, 2015. 23(12): p. 2387. 37. Dougherty, J.D., The Expanding Toolkit of Translating Ribosome Affinity Purification. J Neurosci, 2017. 37(50): p. 12079-12087. 38. Serafini, A., et al., The ESX-3 secretion system is necessary for iron and zinc homeostasis in Mycobacterium tuberculosis. PLoS One, 2013. 8(10): p. e78351. 39. Oh, E., et al., Selective ribosome profiling reveals the cotranslational chaperone action of trigger factor in vivo. Cell, 2011. 147(6): p. 1295-308. 40. Tinaztepe, E., et al., Role of Metal-Dependent Regulation of ESX-3 Secretion in Intracellular Survival of Mycobacterium tuberculosis. Infect Immun, 2016. 84(8): p. 2255- 2263.
101
Appendices
102
Appendix 1: Supplementary Material from Chapter 2 a b ** 4000 *** 10
l Linezolid 3000 1
2000 0.1 1000 Proportion Surviva
Calcein Fluorescence 0.01 0 0 10 20 30 40 50 Time(hours) WT L5::empty ΔlepA L5::empty ΔlepA L5::lepA
c 10 Isoniazid
al 1 v i rv 0.1 ion Su rt 0.01 opo r P 0.001 0 10 20 30 40 50 Time(hours)
AppendixSupplementary Figure FigureA1.1- lepA 1- lepA defect defect in calcein in calcein staining staining correlates correlates with with altered altered drug drug tolerance tolerance in in mycobacteriamycobacteria a) Calcein staining across M. smegmatis strains with and without lepA. b-c) Antibiotic killing was measured as in Figure 1 with 10x MIC concentrations of vancomycin and isoniazid. All values in (a-c) are mean values with error bars indicating standard error of the mean (SEM) of three biological replicates. ***P <0.001, **P<0.01, calculated using a two-sided Student’s t-test.
103 Appendix Table A1.1 – Drug susceptibility of lepA mutant Drug + LepA - LepA Tetracycline 0.250 0.250 Clarithromycin 0.125 0.125 Chloramphenicol 10 10 Amikacin 0.200 0.200 Erythromycin 0.500 0.500
Susceptibility of Msm strains was determined using an Alamar blue MIC assay. Values are averages of three replicates, in µg/ml concentrations of drug.
104
Appendix Figure A1.2- Purification of LepA
SDS-PAGE gel displaying Ni-NTA column purification of Msm LepA. Lane 1: cell lysate; Lane 2: cell pellet; Lane 3: wash 1 (30 mM imidazole); Lane 4: wash 4; Lane 5: elution fraction 2; Lane 6: elution fraction 3; Lane 7: elution fraction 4; Lane 8: elution fraction 5; Lane 9: elution fraction 6; Lane 10: elution fraction 7; Lane 11: BenchMark Pre-stained ladder; Lane 12: elution fraction 3 + DTT
105
Appendix Figure A1.3 – Post-transcriptional screen to find candidates for LepA regulation a) Enrichment of the most abundant gene ontology (GO) term processes in the proteomics dataset relative to the whole Msm proteome. b) Protein alignment of mycobacterial porin family in Msm. Black boxes indicate unique peptides that mapped to a mycobacterial porin. Colored amino acids indicate differences between porins. c) On the left, representative images of MspA-mRFP localization in cells with or without LepA. On the right, quantification of average fluorescence across a single cell (N=100) from each strain. Scale bar represents 5 microns. ***P<0.001 calculated using a Mann-Whitney test.
106 ** ns ns
108
107
6
Luminescence (RLU/OD600) 10
MSMEG_0114MSMEG_1679 MSMEG_2650
Appendix Figure A1.4 – Validation of other candidates depleted in lepA proteomics data
Y-axis indicates relative luminescence normalized by OD600 of cells. Bars indicate mean with error bars indicating SEM of three biological replicates. Luminescence reporters in +LepA (filled bars) and –LepA (open bars). Data are representative of multiple experiments. **P<0.01, calculated using Student’s t-test.
107
Appendix Figure A1.5 – CRISPRi sgRNA validation
Three sgRNAs with different PAM strengths [1] were chosen to achieve maximal knockdown of each porin. Black lines around a bar indicate selection of that sgRNA for future experiments. Fold repression, quantified using RT-qPCR, indicates the difference in knockdown between induced and uninduced strains.
108
Appendix Figure 1.6 – Interrogation of MspA sequence determinants a) Codon frequencies were calculated for all four porin genes, compared to genomic codon frequencies. b) The mspA gene was recoded at 8 glycine codons, and LepA control was tested in the luciferase reporter system. c) The glycine-recoded mspA was fused to a NanoLuc reporter, and constructs were analyzed for LepA-dependent luminescence in biological triplicate. Each glycine codon was selected
109 Appendix Figure 1.6 (continued) based on the presence of a ‘GGT’ in mspA-C and a corresponding ‘GGC’ in mspD. Each number denotes the location of the T to C mutation made at the end of the glycine codon. d) Data is representative of multiple experiments. Significance was calculated using a two-sided Student’s t-test; error bars represent SEM. e) Coding sequences for mspA and mspD were aligned to define points of fusion for chimeras in Figure 5D. Each fusion point was selected, such that the codon on either side only contained synonymous differences between either porin sequence.
110 Appendix Table 1.2 – Proteomic changes in the absence of LepA in Msm
log2 (Protein in +LepA/ No. of Adjusted p- protein Unique Accession value in -LepA) Gene Peptide number entry_name 1.622333 5.92936E-08 333 lepA 39 A0R0Y9 Elongation factor 4 Integral membrane 0.000138695 -0.635 MSMEG_2188 2 A0QUF5 protein A0QPU4|A0 0.710833 R3I3|A0QR2 0.000925143 333 mspB 3 9 Porin MspB Uncharacterized 0.00106377 -0.7425 MSMEG_0546 5 A0QPW9 protein 0.453833 Acyl-CoA 0.001929096 333 MSMEG_2650 13 A0QVP9 dehydrogenase Uncharacterized 0.002715154 1.0585 MSMEG_5292 2 A0R300 protein ABC transporter 0.422166 oligopeptide binding 0.004018933 667 MSMEG_4385 18 A0R0H2 protein Acyl-CoA 0.432166 dehydrogenase-family 0.005213354 667 MSMEG_4391 20 A0R0H7 protein Extracellular solute- 0.558333 binding protein, family 0.005419054 333 MSMEG_0114 16 A0QNP2 protein 3 0.326333 Enoyl-CoA 0.005942338 333 MSMEG_4299 19 A0R088 hydratase/isomerase - 0.383166 N-acetyltransferase 0.006079036 667 eis 21 A0QY29 Eis Uncharacterized 0.006079036 0.431 MSMEG_2446 2 A0QV48 protein - 0.608333 Uncharacterized 0.006440274 333 MSMEG_1419 7 A0QSB4 protein Uncharacterized 0.006751491 0.2445 MSMEG_4729 13 A0R1F0 protein Monooxygenase, 0.433666 NtaA/SnaA/SoxA 0.007028886 667 MSMEG_4389 14 A0R0H6 family protein - 0.253666 0.007028886 667 MSMEG_3914 11 A0QZ65 Beta-lactamase TROVE domain 0.007259192 -0.6025 MSMEG_1193 6 A0QRP7 protein 0.318666 0.00826972 667 MSMEG_1679 18 A0QT17 AmiB 0.452333 Uncharacterized 0.011345341 333 MSMEG_1042 6 A0QRA1 protein
111 Appendix Table A1.2 (Continued) Phosphate-binding 0.012163454 -0.399 pstS 17 A0R4C3 protein PstS Alkanesulfonate monooxygenase 0.012163454 0.3725 MSMEG_3850 4 A0QZ04 family protein Alkanesulfonate monooxygenase 0.013056015 0.308 MSMEG_2651 22 A0QVQ0 family protein Putative sugar ABC 0.306833 transporter ATP- 0.013056015 333 MSMEG_3269 13 A0QXE2 binding protein 0.013056015 0.295 MSMEG_5519 16 A0R3L8 Monooxygenase 0.591333 Uncharacterized 0.013056015 333 MSMEG_2754 12 A0QVZ7 protein Metallo-beta- 0.322833 lactamase family 0.013056015 333 MSMEG_4342 9 A0R0C8 protein - 0.315666 ABC transporter, ATP- 0.013056015 667 MSMEG_5659 13 A0R404 binding protein 0.013056015 0.332 MSMEG_4985 7 A0R248 Carbonic anhydrase - 0.326166 0.013056015 667 MSMEG_4972 6 A0R235 Acetyltransferase 0.588833 Uncharacterized 0.013056015 333 MSMEG_5919 2 A0R4Q7 protein ABC transporter, ATP- 0.013279501 0.2495 MSMEG_6309 14 A0R5T8 binding protein 0.358833 0.013279501 333 MSMEG_4714 6 A0R1D6 MoaC domain protein 0.014306976 0.2485 derK 26 A0QXE4 D-erythrulose kinase Sensor histidine 0.014306976 -0.2055 mtrB 18 A0QTK3 kinase MtrB - 0.334666 Immunogenic protein 0.014306976 667 MSMEG_5412 11 A0R3B5 MPT63 - ESX-1 secretion- 0.231666 associated protein 0.014306976 667 espE 6 A0QNI5 EspE Acyl-CoA dehydrogenase, short- 0.014752742 0.4265 MSMEG_0323 4 A0QP98 chain specific, putative 0.251166 0.01771159 667 MSMEG_5192 5 A0R2P9 Aldo/keto reductase Aldehyde 0.189333 dehydrogenase, 0.023211622 333 MSMEG_6687 14 A0R6V7 thermostable 0.210333 Alcohol 0.024336178 333 MSMEG_2079 18 A0QU52 dehydrogenase 0.025433671 -0.189 MSMEG_1516 30 A0QSL0 Thioredoxin reductase 0.182666 Ferredoxin sulfite 0.026342306 667 MSMEG_4527 54 A0R0W1 reductase
112
Appendix Table A1.2 (Continued)
0.222666 N-acetylmuramoyl-L- 0.026385818 667 MSMEG_6281 6 A0R5R2 alanine amidase Methylmalonate- 0.220333 semialdehyde 0.026521336 333 mmsA 33 A0QSJ2 dehydrogenase 0.187333 Aldehyde 0.026521336 333 MSMEG_0327 10 A0QPA2 dehydrogenase - 0.226333 ABC transporter ATP- 0.026521336 333 MSMEG_5660 20 A0R405 binding protein 0.206666 Choloylglycine 0.029068545 667 MSMEG_5454 15 A0R3F5 hydrolase, putative Taurine import ATP- 0.029533452 0.4105 MSMEG_0116 10 A0QNP3 binding protein TauB Histidine N-alpha- 0.030217284 -0.2435 egtD 16 A0R5M8 methyltransferase - 0.196333 Uncharacterized 0.03058158 333 MSMEG_0067 30 A0QNJ7 protein - 0.212166 Putative transcriptional 0.0316866 667 MSMEG_0091 14 A0QNL9 regulator SN-glycerol-3- phosphate ABC transporter, ATP- 0.033287594 0.2905 MSMEG_3270 10 A0QXE3 binding protein 0.176333 Isochorismatase 0.033486139 333 MSMEG_4396 13 A0R0I3 hydrolase 0.225333 Aminopeptidase, 0.034375232 333 MSMEG_2167 11 A0QUD3 putative Enoyl-CoA hydratase, 0.035032625 0.2365 MSMEG_4852 3 A0R1S0 putative 0.563333 Uncharacterized 0.036375308 333 MSMEG_1887 6 A0QTL5 protein Phosphoglycerate 0.036933061 0.183 pgk 30 A0QWW3 kinase 0.226166 0.036933061 667 MSMEG_6907 13 A0R7G8 MmcI protein Nitrilotriacetate monooxygenase 0.039380599 0.1885 MSMEG_4085 15 A0QZM9 component A Nitrilotriacetate 0.212666 monooxygenase 0.039557974 667 ssuD 12 A0QZN0 component A Uncharacterized 0.039557974 -0.3125 MSMEG_5708 7 A0R453 protein Transcriptional 0.039557974 -0.353 whiB7 5 A0QTT1 regulator WhiB7 0.545166 0.039557974 667 MSMEG_1840 3 A0QTH2 Rubredoxin Immunogenic protein 0.039666082 -0.3125 MSMEG_0828 8 A0QQP4 MPT63
113
Appendix Table A1.2 (Continued)
ABC transporter ATP- 0.039966771 0.1925 MSMEG_5571 21 A0R3R8 binding protein - 0.263666 Aspartate 0.039966771 667 MSMEG_0688 22 A0QQA8 aminotransferase Probable monoacyl - phosphatidylinositol 0.219666 tetramannoside- 0.039966771 667 MSMEG_5130 19 A0R2I8 binding protein LpqW - 0.331833 ABC transporter ATP- 0.039966771 333 MSMEG_5102 11 A0R2G3 binding protein Dihydrofolate 0.040512742 -0.213 MSMEG_0308 4 A0QP85 reductase - Hypoxanthine 0.206166 phosphoribosyltransfer 0.041570634 667 hpt 11 A0R593 ase Acyl-CoA 0.257666 dehydrogenase family 0.041614248 667 MSMEG_1497 27 A0QSJ1 protein member 8 - 0.283666 0.041614248 667 hflX 22 A0QVY1 GTPase HflX 0.149833 Uncharacterized 0.043763222 333 MSMEG_2381 15 A0QUY5 protein Acetyltransferase, 0.043763222 -0.2495 MSMEG_2691 9 A0QVT8 gnat family protein 0.226666 Subtilase family 0.043763222 667 MSMEG_0624 8 A0QQ47 protein L-carnitine 0.185666 dehydratase/bile acid- 0.044076655 667 MSMEG_1992 9 A0QTW8 inducible protein F 0.167166 Linear gramicidin 0.045258089 667 MSMEG_4511 37 A0R0U5 synthetase subunit B 0.258333 Lactoylglutathione 0.047535368 333 MSMEG_6743 4 A0R713 lyase - 0.304666 0.048416822 667 MSMEG_6583 7 A0R6K6 Antigen 85-C
114 Appendix Table A1.3 - DEseq Analysis of RNA sequencing in lepA strains (p<0.01)
log2(RNA in Gene Number Base Mean +LepA/-LepA) Adjusted p-value MSMEG_5187 555.9234919 -1.077395912 4.08E-31 MSMEG_4556 1663.18363 13.68168962 1.58E-27 MSMEG_1419 249.6840839 -1.314410069 4.91E-23 MSMEG_0546 251.361143 -1.462779327 8.03E-23 MSMEG_1193 396.1786513 -0.899171669 3.77E-16 MSMEG_4961 149.3585384 -1.283684416 2.99E-14 MSMEG_3137 231.7225157 -0.923895856 3.27E-13 MSMEG_0422 585.9941489 -0.677306473 6.74E-13 MSMEG_0406 7133.844033 0.963793059 1.22E-12 MSMEG_5119 1291.577886 1.28976664 8.72E-12 MSMEG_5612 740.861454 -0.724963063 1.40E-11 MSMEG_6249 596.6736 -0.6637911 1.40E-11 MSMEG_6511 3506.521 0.9225746 2.48E-11 MSMEG_6232 391.2278 -1.022042 2.48E-11 MSMEG_5649 192.1557612 1.473423034 9.68E-11 MSMEG_6250 635.3328 -0.6233834 1.22E-10 MSMEG_1026 217.5367706 -0.926121521 1.39E-10 MSMEG_2736 1773.21895 -0.55532801 1.76E-10 MSMEG_2188 455.0195556 -0.741531455 2.04E-10 MSMEG_4351 285.7428843 -1.167963445 3.10E-10 MSMEG_0076 985.1950771 -0.517330837 3.96E-10 MSMEG_2306 208.9374095 -0.902984917 7.18E-10 MSMEG_6881 392.6356 1.387739 7.93E-10 MSMEG_3276 224.7687476 -0.814807991 9.45E-10 MSMEG_0965 20978.38266 0.511325753 1.55E-09 MSMEG_6728 54.60162 -1.382211 1.76E-09 MSMEG_1951 299.0952499 -0.803275548 6.04E-09 MSMEG_6213 496.7844 -0.7092595 7.35E-09 MSMEG_4558 438.5617754 0.705640484 1.19E-08 MSMEG_5782 359.444889 -0.60394365 1.28E-08 MSMEG_1114 226.4410756 -1.041995515 1.28E-08 MSMEG_6177 101.7577812 -1.392945955 1.33E-08 MSMEG_5788 808.0946504 0.837128975 1.62E-08 MSMEG_3735 361.6887477 0.862518964 1.65E-08 MSMEG_3680 894.5568991 -0.876980829 2.40E-08 MSMEG_2754 1209.298335 0.442399252 3.29E-08
115
Appendix Table A1.3 (Continued)
MSMEG_3582 232.7882226 -0.980403725 4.15E-08 MSMEG_0120 695.6580479 0.760205481 4.51E-08 MSMEG_5650 192.4039827 1.140016293 5.08E-08 MSMEG_2989 208.1542614 1.097389432 6.25E-08 MSMEG_6209 185.0561 1.256628 8.15E-08 MSMEG_6212 209.1845 -0.9536396 9.40E-08 MSMEG_1821 3081.380405 1.005863446 9.48E-08 MSMEG_5280 317.3122406 1.206390939 1.05E-07 MSMEG_0688 831.4512672 -0.612475487 2.73E-07 MSMEG_5866 763.7785493 0.823309581 4.51E-07 MSMEG_0449 286.6473249 -0.886947139 4.51E-07 MSMEG_2303 417.3836797 0.59044678 4.53E-07 MSMEG_0064 931.7321414 -0.489355612 7.02E-07 MSMEG_1790 299.6192467 -0.710404899 7.91E-07 MSMEG_5660 695.5971406 -0.499809001 8.86E-07 MSMEG_6467 1366.585 -0.746644 1.08E-06 MSMEG_6365 1586.324 -0.3587234 1.25E-06 MSMEG_6163 354.1235632 0.77719741 1.56E-06 MSMEG_3541 50.0232044 -1.149452307 1.57E-06 MSMEG_1566 480.2029685 0.70485306 1.73E-06 MSMEG_1017 1285.291021 0.52104239 2.16E-06 MSMEG_0684 166.1876031 -1.104112697 2.29E-06 MSMEG_1646 25.99559579 -1.553655443 2.29E-06 MSMEG_0671 342.4183886 -0.635402647 2.54E-06 MSMEG_2925 258.5959049 -0.950306419 2.62E-06 MSMEG_0669 182.8348215 -1.257783458 3.60E-06 MSMEG_6454 751.4091 -0.525617 3.87E-06 MSMEG_5086 1628.703579 -0.311470043 4.83E-06 MSMEG_5363 772.2811449 0.515515921 5.03E-06 MSMEG_5739 207.3428677 -0.632840573 5.25E-06 MSMEG_4536 712.1141361 1.792567598 5.48E-06 MSMEG_3198 103.6459068 1.377530378 6.94E-06 MSMEG_5576 1690.33941 0.547950096 7.75E-06 MSMEG_3305 83.17954514 -0.882583103 8.33E-06 MSMEG_5483 907.038959 -0.613365549 1.05E-05 MSMEG_5816 450.1062682 -0.692856538 1.60E-05 MSMEG_2791 192.736839 -0.898308529 1.60E-05
116
Appendix Table A1.3 (Continued)
MSMEG_5100 544.1244444 -0.41264357 1.64E-05 MSMEG_0403 2426.680086 0.455980035 2.31E-05 MSMEG_1169 201.4851542 -0.670170237 2.41E-05 MSMEG_1513 1338.622017 -0.497747068 2.54E-05 MSMEG_1303 86.66656619 -0.798126535 2.55E-05 MSMEG_0305 165.9632001 0.784030461 2.57E-05 MSMEG_1597 298.9667466 -1.008106236 2.69E-05 MSMEG_0402 31265.06856 0.282842515 3.05E-05 MSMEG_6366 1119.138 -0.3672279 3.43E-05 MSMEG_1762 175.0016482 -0.698646717 3.55E-05 MSMEG_1037 1696.397456 -0.348901214 3.90E-05 MSMEG_1789 174.6239209 -0.678458919 4.15E-05 MSMEG_2121 375.4336531 0.679224136 5.20E-05 MSMEG_5887 753.6320575 0.641598881 5.32E-05 MSMEG_0216 7092.085478 0.456476128 5.94E-05 MSMEG_0986 742.3532511 -0.355004209 5.94E-05 MSMEG_5270 3513.295035 0.365176969 6.09E-05 MSMEG_6447 92.15906 -0.776895 6.47E-05 MSMEG_1768 56.06776863 -0.978210624 6.47E-05 MSMEG_3619 2146.320211 0.517378962 7.38E-05 MSMEG_1822 1029.869934 0.501442339 7.53E-05 MSMEG_4688 1130.480168 -0.729596287 8.25E-05 MSMEG_6245 24.90217 -1.338081 8.40E-05 MSMEG_4344 70.92825174 1.084734029 8.52E-05 MSMEG_4388 51.16871756 1.280631773 9.76E-05 MSMEG_6730 158.6589 -0.8424713 9.98E-05 MSMEG_6291 183.3377 0.7683858 0.000103264 MSMEG_6329 2191.514 0.3581882 0.000110493 MSMEG_3811 1465.972479 -0.575246946 0.000110627 MSMEG_3513 1214.737584 -0.319144618 0.000116057 MSMEG_2598 285.6163086 -0.640516833 0.000130431 MSMEG_5400 123.1662263 -0.767305409 0.000135771 MSMEG_2177 61.94984254 -0.87488817 0.000135771 MSMEG_4728 362.293293 0.553129536 0.000140667 MSMEG_1755 95.77806147 -0.909885149 0.000149619 MSMEG_5549 176.0051948 0.929417034 0.000151206 MSMEG_3583 76.44601762 -0.892554959 0.000151206
117
Appendix Table A1.3 (Continued)
MSMEG_5362 1770.864892 0.394290174 0.000151433 MSMEG_5659 321.3814509 -0.480675512 0.000151433 MSMEG_0866 1998.027779 0.277878919 0.000155469 MSMEG_5729 37.05949014 -1.201407372 0.000175176 MSMEG_4329 9687.910116 -0.480906418 0.000193305 MSMEG_5421 38.50358816 -1.012942162 0.000195755 MSMEG_2009 262.106795 0.540185022 0.000196922 MSMEG_2454 163.1609345 -0.582102671 0.000196922 MSMEG_2092 769.7285831 -0.333776632 0.00019695 MSMEG_0493 42.90430266 -1.09867519 0.000208704 MSMEG_2651 466.5163895 0.72004809 0.00021557 MSMEG_0078 482.2817635 -0.515403697 0.000226985 MSMEG_2906 55.49293277 -1.058127263 0.00024787 MSMEG_6753 202.188 -0.6745614 0.000254606 MSMEG_0121 495.3757068 0.5042155 0.00028239 MSMEG_1980 944.2511053 -0.344640469 0.000324331 MSMEG_5558 141.2650026 -1.107991745 0.000324331 MSMEG_0059 1121.341602 -0.36805474 0.000334346 MSMEG_1802 257.6407606 -0.497098707 0.000346504 MSMEG_6230 132.437 -0.6196334 0.0003817 MSMEG_1769 112.4585336 -1.160090053 0.000390157 MSMEG_0904 435.1159929 0.416762318 0.000393099 MSMEG_1741 509.8748374 0.409143862 0.000402564 MSMEG_1749 166.7303629 -0.971373062 0.000402564 MSMEG_1666 395.3385775 0.438716897 0.000403642 MSMEG_0575 100.7322701 -0.96886685 0.000403642 MSMEG_1782 125.3714206 -0.671471486 0.000435046 MSMEG_3410 46.06618608 -1.056433184 0.000436841 MSMEG_2597 365.5778229 -0.573647609 0.000447143 MSMEG_4499 789.2260682 -0.621351926 0.000448138 MSMEG_1498 621.9421982 0.510950991 0.000544965 MSMEG_3203 95.3515168 -0.678250378 0.000579507 MSMEG_0303 77.30617155 -0.730939223 0.000670922 MSMEG_6286 2416.361 0.3009888 0.000673667 MSMEG_2078 5155.274086 -0.3365247 0.000678953 MSMEG_2115 292.1183098 -0.652627646 0.000678953 MSMEG_0024 6597.648792 0.220677139 0.000685782
118
Appendix Table A1.3 (Continued)
MSMEG_5154 761.2666041 0.444918966 0.000768267 MSMEG_1677 1914.638059 0.390263307 0.000775071 MSMEG_2655 265.7837384 0.564151877 0.000833187 MSMEG_3231 791.2222104 -0.496079098 0.000876208 MSMEG_4531 772.8023111 0.645969874 0.000938661 MSMEG_1809 4018.84468 0.273481214 0.000938661 MSMEG_5167 39.82975215 -0.93777637 0.000960099 MSMEG_5279 214.0121567 0.752590142 0.001018124 MSMEG_0779 226.2475581 -0.797183381 0.001024894 MSMEG_6248 197.0415 -0.4867931 0.001066434 MSMEG_1947 768.7756248 -0.443149916 0.001114998 MSMEG_2671 821.9400465 0.353306708 0.001115653 MSMEG_3897 3094.613009 0.265623197 0.001121755 MSMEG_5022 152.5904758 0.647116244 0.001122776 MSMEG_0728 280.8201975 -0.607855654 0.001175458 MSMEG_6615 47.1535 -0.9657178 0.001185462 MSMEG_0114 234.1978962 0.786987571 0.001242318 MSMEG_1544 373.5096128 0.596480104 0.001258259 MSMEG_6756 1543.908 0.2698918 0.001267787 MSMEG_0139 1528.20328 0.338352932 0.001273313 MSMEG_0060 874.2962186 -0.341156593 0.001282171 MSMEG_2347 176.9772618 -0.592543822 0.00130422 MSMEG_3636 303.7351914 -0.459315046 0.001321287 MSMEG_0711 2944.167854 -0.599514552 0.001321287 MSMEG_0055 220.1266479 -0.573751618 0.001326058 MSMEG_2648 1290.06551 -0.268603619 0.001338136 MSMEG_6159 22834.17342 0.342255178 0.00133834 MSMEG_0572 22.19420403 -1.619321845 0.00133834 MSMEG_1548 915.5025453 0.598207119 0.001341535 MSMEG_6141 629.976742 0.353686865 0.001342818 MSMEG_6747 127.2943 -0.5725166 0.001348206 MSMEG_2086 1294.286874 0.279124538 0.001358604 MSMEG_5273 6456.340309 -0.272202304 0.001366829 MSMEG_2297 1282.129882 0.424751414 0.001374286 MSMEG_1772 148.3987198 -0.733464767 0.001389403 MSMEG_6933 3250.9 -0.2836119 0.001424778 MSMEG_4121 289.2537454 0.507894792 0.001558288
119
Appendix Table A1.3 (Continued)
MSMEG_6878 346.5984 -0.698435 0.001558288 MSMEG_1286 213.7352013 -0.470078654 0.001562877 MSMEG_1112 291.8414091 -0.671170628 0.001562877 MSMEG_2410 2568.997516 0.27121177 0.001613991 MSMEG_5231 101.9538007 -0.650072565 0.001695792 MSMEG_2650 197.0953054 0.789264339 0.001738616 MSMEG_3906 1471.851032 0.342140042 0.001755034 MSMEG_2958 105.9080117 -0.711910857 0.001755034 MSMEG_2593 2156.113624 -0.273120677 0.001878599 MSMEG_6246 323.8206 -0.4056636 0.001977853 MSMEG_4387 27.93107587 1.2589955 0.001999306 MSMEG_0290 167.8913246 -0.5216634 0.002030026 MSMEG_5613 209.9276339 -0.456751558 0.002141666 MSMEG_6197 156.0313826 0.515511462 0.002212004 MSMEG_5969 1937.277644 0.342599557 0.002310184 MSMEG_3277 159.0231108 0.583814477 0.002328864 MSMEG_6877 356.9947 -0.5643473 0.002336537 MSMEG_6643 165.1639 0.5831196 0.002357697 MSMEG_0559 24673.38308 0.232281245 0.002406997 MSMEG_0255 4495.604103 -0.269692329 0.0024397 MSMEG_4732 84.24975736 0.696429611 0.002457643 MSMEG_3409 90.75641839 -0.816622039 0.002457643 MSMEG_5664 2704.652933 0.231805698 0.002508326 MSMEG_3022 678.1735853 -0.609920124 0.002613641 MSMEG_2175 52.5520361 -0.948950402 0.002618889 MSMEG_1023 429.4236462 0.586014603 0.002707919 MSMEG_3914 131.9523795 -0.788370051 0.002707919 MSMEG_2654 3036.80611 0.42806434 0.002735988 MSMEG_5995 84.3467774 0.800610102 0.002829306 MSMEG_4465 60.501362 -0.900433582 0.002894295 MSMEG_3417 77.95690109 -0.830016269 0.003052271 MSMEG_0710 1725.914267 -0.614917828 0.003098144 MSMEG_2647 2219.846682 -0.248237666 0.003209866 MSMEG_5047 339.0725387 -0.492893136 0.00321709 MSMEG_2456 327.9544223 -0.561525675 0.00321709 MSMEG_0132 1150.908401 0.378064885 0.003242338 MSMEG_0431 87.92519 -0.806845938 0.003243468
120
Appendix Table A1.3 (Continued)
MSMEG_6638 2265.659 0.3056965 0.003284117 MSMEG_0372 3518.432059 0.503068237 0.003284117 MSMEG_3297 536.2800886 0.393157255 0.003318867 MSMEG_6764 93.63236 0.7784197 0.003355196 MSMEG_1743 956.7111728 -0.505511458 0.003550818 MSMEG_6307 1374.173 0.2896654 0.003621892 MSMEG_5575 522.2438958 0.411171129 0.003652878 MSMEG_0709 7384.3366 -0.588175519 0.003776107 MSMEG_5051 400.3731163 0.493321039 0.003808579 MSMEG_0920 430.7501171 0.342933921 0.003821292 MSMEG_6247 262.7591 -0.3953051 0.003835665 MSMEG_5248 6696.067351 0.225932678 0.003962448 MSMEG_5896 1003.851469 0.272824101 0.003988485 MSMEG_6130 885.0629633 0.33577961 0.004133594 MSMEG_2694 805.9039083 0.428918829 0.0041433 MSMEG_2015 26.74990883 -1.25930165 0.004165285 MSMEG_1604 1310.794758 0.327811639 0.004173809 MSMEG_3468 292.7897257 -0.372494474 0.004354723 MSMEG_0903 7882.480984 0.241232678 0.004468389 MSMEG_3057 276.701513 0.419597505 0.004621183 MSMEG_1448 185.5542501 -0.578436547 0.004684588 MSMEG_0953 3308.773794 0.2163013 0.00474696 MSMEG_0134 1196.627616 0.386682643 0.004912682 MSMEG_2122 140.2856884 0.656589968 0.004933284 MSMEG_3605 191.0292024 -0.577479626 0.004933284 MSMEG_4920 1705.043382 -0.23453588 0.004952256 MSMEG_0061 1484.775042 -0.246629582 0.004956696 MSMEG_3673 238.3724928 -0.412133783 0.004956696 MSMEG_6472 18.00746 -1.310017 0.004990463 MSMEG_6858 68.24188 -1.032022 0.005020045 MSMEG_6879 387.8499 -0.6528202 0.005028493 MSMEG_4533 4760.747394 0.558916004 0.005214403 MSMEG_5178 193.8973635 0.517680462 0.005214403 MSMEG_0091 674.4961671 -0.332642143 0.005214403 MSMEG_0113 57.09891741 0.842489656 0.005383007 MSMEG_0713 566.1562349 -0.699892229 0.005737429 MSMEG_1075 767.7477567 -0.35286434 0.005799101
121
Appendix Table A1.3 (Continued)
MSMEG_1764 150.6906838 -0.495859932 0.005799101 MSMEG_2014 29.95363445 -1.097359566 0.005799101 MSMEG_4970 100.4240172 -0.52751597 0.005941033 MSMEG_0756 291.6098304 0.369887694 0.00596239 MSMEG_5185 159.3039751 -0.435011047 0.005968789 MSMEG_3479 6696.99105 0.289434911 0.006231671 MSMEG_5406 968.1202865 0.341148738 0.006326989 MSMEG_4715 187.6049703 -0.755692354 0.00662419 MSMEG_3689 8072.461667 -0.230962162 0.006695546 MSMEG_4532 788.6412975 0.659402792 0.006732795 MSMEG_2607 385.2856269 0.482052215 0.006799703 MSMEG_6742 86.46427 0.7729921 0.006858598 MSMEG_6210 127.8584 0.7195624 0.00707514 MSMEG_2542 64.98921241 -1.03321546 0.007302997 MSMEG_1130 253.1856347 0.405607321 0.007324814 MSMEG_2317 1637.411692 -0.299740111 0.007329298 MSMEG_5966 126.7995474 0.54637823 0.007394314 MSMEG_0404 557.5481905 0.301726447 0.007394314 MSMEG_6362 261.5581 -0.4548715 0.007451214 MSMEG_0637 185.2434515 -0.513534635 0.007451214 MSMEG_5401 85.81318421 -0.593834486 0.007451214 MSMEG_6608 32.22466 -0.9687755 0.007512003 MSMEG_1808 1139.623694 0.257340784 0.007512003 MSMEG_0595 396.1960634 0.500138388 0.007662444 MSMEG_2659 240.3265731 0.557679444 0.00864938 MSMEG_0083 725.5695453 -0.32692769 0.008805553 MSMEG_6178 70.09383915 -0.807925198 0.008865298 MSMEG_1543 3436.502163 0.463641664 0.009028789 MSMEG_5556 319.2758622 -0.483526754 0.009256082 MSMEG_3581 33.72385203 -1.03496973 0.009483393 MSMEG_2670 2988.308234 0.313267513 0.009494119 MSMEG_2453 257.7177486 -0.466419093 0.009647091 MSMEG_3384 166.5343197 -0.791440164 0.009811692 MSMEG_1670 4013.173067 -0.204857284 0.009842149 MSMEG_4918 1321.638796 -0.314131452 0.009901325 MSMEG_6208 646.1012 0.4629201 0.009912802 MSMEG_2087 808.5292617 0.337521244 0.009980636
122
Appendix Table A1.4 – Strain List
Strain Parental Plasmid at Plasmid at Antibi Experiment Number Background L5 Tweety otic (Strain Integration Integration Mark number) Site (Plasmid Site (Plasmid ers number) Number)
HR334 mc2155 None None ZeoR Characterizati on of LepA phenotype
SF177 mc2155 p-empty- None KanR Characterizati KanR(CT242) on of LepA phenotype
SF178 ∆lepA::ZeoR pPnatp-lepA- None ZeoR, Characterizati (HR334) KanR(SF121) KanR on of LepA phenotype
SF181 ∆lepA::ZeoR p-empty- None ZeoR, Characterizati (HR334) KanR(CT242) KanR on of LepA phenotype
SF290 ∆lepA::ZeoR pPnatp-lepA- pPUV15- ZeoR, Valdiation of (HR334) KanR(SF121) mspA-Nluc- KanR, proteomics NatR(283) NatR SF291 ∆lepA::ZeoR p-empty- pPUV15- ZeoR, Valdiation of (HR334) KanR(CT242) mspA-Nluc- KanR, proteomics NatR(283) NatR SF302 ∆lepA::ZeoR pPnatp-lepA- pPUV15- ZeoR, Valdiation of (HR334) KanR(SF121) Nluc- KanR, proteomics NatR(287) NatR SF303 ∆lepA::ZeoR p-empty- pPUV15- ZeoR, Valdiation of (HR334) KanR(CT242) Nluc- KanR, proteomics NatR(287) NatR SF461 ∆lepA::ZeoR pPnatp-lepA- pPUV15- ZeoR, Sequence (HR334) KanR(SF121) mspAss-Nluc- KanR, determinants NatR(413) NatR of LepA regulation
SF462 ∆lepA::ZeoR pPnatp-lepA- pPUV15- ZeoR, Sequence (HR334) KanR(SF121) mspBss-Nluc- KanR, determinants NatR(414) NatR of LepA regulation
SF463 ∆lepA::ZeoR pPnatp-lepA- pPUV15- ZeoR, Sequence (HR334) KanR(SF121) mspCss-Nluc- KanR, determinants NatR(414) NatR of LepA regulation
123
Appendix Table A1.4 (Continued)
SF464 ∆lepA::ZeoR pPnatp-lepA- pPUV15- ZeoR, Sequence (HR334) KanR(SF121) mspDss-Nluc- KanR, determinants NatR(416) NatR of LepA regulation
SF465 ∆lepA::ZeoR p-empty- pPUV15- ZeoR, Sequence (HR334) KanR(CT242) mspAss-Nluc- KanR, determinants NatR(413) NatR of LepA regulation
SF466 ∆lepA::ZeoR p-empty- pPUV15- ZeoR, Sequence (HR334) KanR(CT242) mspBss-Nluc- KanR, determinants NatR(414) NatR of LepA regulation
SF467 ∆lepA::ZeoR p-empty- pPUV15- ZeoR, Sequence (HR334) KanR(CT242) mspCss-Nluc- KanR, determinants NatR(415) NatR of LepA regulation
SF468 ∆lepA::ZeoR p-empty- pPUV15- ZeoR, Sequence (HR334) KanR(CT242) mspDss-Nluc- KanR, determinants NatR(417) NatR of LepA regulation
SF432 ∆lepA::ZeoR None pPnatp-lepA- ZeoR, Porin (HR334) HygR(417) HygR Knockdown, LepA alleles
SF433 ∆lepA::ZeoR None p-emtpy- ZeoR, Porin (HR334) HygR(418) HygR Knockdown, LepA alleles
SF434 ∆lepA::ZeoR pPUV15-Tet- pPnatp-lepA- ZeoR, Porin (HR334) dCas9- HygR(417) HygR, Knockdown sgRNA1 KanR (mspA1)- KanR(419) SF435 ∆lepA::ZeoR pPUV15-Tet- pPnatp-lepA- ZeoR, Porin (HR334) dCas9- HygR(417) HygR, Knockdown sgRNA2 KanR (mspA2)- KanR(420) SF436 ∆lepA::ZeoR pPUV15-Tet- pPnatp-lepA- ZeoR, Porin (HR334) dCas9- HygR(417) HygR, Knockdown sgRNA3 KanR (mspA3)- KanR(421)
124 Appendix Table A1.4 (Continued)
SF437 ∆lepA::ZeoR pPUV15-Tet- pPnatp-lepA- ZeoR, Porin (HR334) dCas9- HygR(417) HygR, Knockdown sgRNA4 KanR (mspB1)- KanR(422) SF438 ∆lepA::ZeoR pPUV15-Tet- pPnatp-lepA- ZeoR, Porin (HR334) dCas9- HygR(417) HygR, Knockdown sgRNA5(msp KanR B2)- KanR(423) SF439 ∆lepA::ZeoR pPUV15-Tet- pPnatp-lepA- ZeoR, Porin (HR334) dCas9- HygR(417) HygR, Knockdown sgRNA6 KanR (mspB3))- KanR(424) SF440 ∆lepA::ZeoR pPUV15-Tet- pPnatp-lepA- ZeoR, Porin (HR334) dCas9- HygR(417) HygR, Knockdown sgRNA7 KanR (mspC1)- KanR(425) SF441 ∆lepA::ZeoR pPUV15-Tet- pPnatp-lepA- ZeoR, Porin (HR334) dCas9- HygR(417) HygR, Knockdown sgRNA8 KanR (mspC2)- KanR(426) SF442 ∆lepA::ZeoR pPUV15-Tet- pPnatp-lepA- ZeoR, Porin (HR334) dCas9- HygR(417) HygR, Knockdown sgRNA9 KanR (mspC3)- KanR(427) SF443 ∆lepA::ZeoR pPUV15-Tet- pPnatp-lepA- ZeoR, Porin (HR334) dCas9- HygR(417) HygR, Knockdown sgRNA10 KanR (mspD1)- KanR(428) SF444 ∆lepA::ZeoR pPUV15-Tet- pPnatp-lepA- ZeoR, Porin (HR334) dCas9- HygR(417) HygR, Knockdown sgRNA11 KanR (mspD2)- KanR(429) SF445 ∆lepA::ZeoR pPUV15-Tet- pPnatp-lepA- ZeoR, Porin (HR334) dCas9- HygR(417) HygR, Knockdown sgRNA12 KanR (mspD3)- KanR(430) SF568 ∆lepA::ZeoR pPnatp-lepA- pPUV15- ZeoR, Sequence (HR334) KanR(SF121) mspA1-90- KanR, determinants mspD- NatR of LepA NatR(544) regulation
125 Appendix Table A1.4 (Continued)
SF569 ∆lepA::ZeoR pPnatp-lepA- pPUV15- ZeoR, Sequence (HR334) KanR(SF121) mspA1-135- KanR, determinants mspD- NatR of LepA NatR(545) regulation
SF570 ∆lepA::ZeoR pPnatp-lepA- pPUV15- ZeoR, Sequence (HR334) KanR(SF121) mspA1-195- KanR, determinants mspD- NatR of LepA NatR(546) regulation
SF571 ∆lepA::ZeoR pPnatp-lepA- pPUV15- ZeoR, Sequence (HR334) KanR(SF121) mspA1-258- KanR, determinants mspD- NatR of LepA NatR(547) regulation
SF572 ∆lepA::ZeoR pPnatp-lepA- pPUV15- ZeoR, Sequence (HR334) KanR(SF121) mspA1-339- KanR, determinants mspD- NatR of LepA NatR(548) regulation
SF576 ∆lepA::ZeoR p-empty- pPUV15- ZeoR, Sequence (HR334) KanR(CT242) mspA1-90- KanR, determinants mspD- NatR of LepA NatR(544) regulation
SF577 ∆lepA::ZeoR p-empty- pPUV15- ZeoR, Sequence (HR334) KanR(CT242) mspA1-135- KanR, determinants mspD- NatR of LepA NatR(545) regulation
SF578 ∆lepA::ZeoR p-empty- pPUV15- ZeoR, Sequence (HR334) KanR(CT242) mspA1-195- KanR, determinants mspD- NatR of LepA NatR(546) regulation
SF579 ∆lepA::ZeoR p-empty- pPUV15- ZeoR, Sequence (HR334) KanR(CT242) mspA1-258- KanR, determinants mspD- NatR of LepA NatR(547) regulation
SF580 ∆lepA::ZeoR p-empty- pPUV15- ZeoR, Sequence (HR334) KanR(CT242) mspA1-339- KanR, determinants mspD- NatR of LepA NatR(548) regulation
SF388 ∆lepA::ZeoR pPnatp-lepA- pPUV15- NatR, Valdiation of (HR334) KanR(SF121) mspB-Nluc- ZeoR, proteomics NatR(384) KanR
126 Appendix Table A1.4 (Continued)
SF389 ∆lepA::ZeoR p-empty- pPUV15- NatR, Valdiation of (HR334) KanR(CT242) mspB(MSME ZeoR, proteomics G_0520)- KanR Nluc- NatR(384) SF390 ∆lepA::ZeoR pPnatp-lepA- pPUV15- NatR, Valdiation of (HR334) KanR(SF121) mspC(MSME ZeoR, proteomics G_5483)- KanR Nluc- NatR(383) SF391 ∆lepA::ZeoR p-empty- pPUV15- NatR, Valdiation of (HR334) KanR(CT242) mspC(MSME ZeoR, proteomics G_5483)- KanR Nluc- NatR(383) SF392 ∆lepA::ZeoR pPnatp-lepA- pPUV15- NatR, Valdiation of (HR334) KanR(SF121) mspD(MSME ZeoR, proteomics G_6057)- KanR Nluc- NatR(385) SF393 ∆lepA::ZeoR p-empty- pPUV15- NatR, Valdiation of (HR334) KanR(CT242) mspD(MSME ZeoR, proteomics G_6057)- KanR Nluc- NatR(385) SF902 ∆lepA::ZeoR pPnatp-lepA- pPUV15- ZeoR, Localization of (HR334) KanR(SF121) mspA-mRFP- HygR, mspA NatR(896) KanR, NatR SF903 ∆lepA::ZeoR p-empty- pPUV15- ZeoR, Localization of (HR334) KanR(CT242) mspA-mRFP- HygR, mspA NatR(896) KanR, NatR SF626 ∆lepA::ZeoR pPnatp-lepA- pPUV15- ZeoR, Sequence (HR334) KanR(SF121) mspD1-135- KanR, determinants mspA136- NatR of LepA 195- regulation mspD196- 624-Nluc- NatR(596) SF627 ∆lepA::ZeoR p-empty- pPUV15- ZeoR, Sequence (HR334) KanR(CT242) mspD1-135- KanR, determinants mspA136- NatR of LepA 195- regulation mspD196- 624-Nluc- NatR(596) SF630 ∆lepA::ZeoR pPnatp-lepA- pPUV15- ZeoR, Sequence (HR334) KanR(SF121) mspA135- KanR, determinants 195-Nluc- NatR of LepA NatR(599) regulation
127 Appendix Table A1.4 (Continued) SF631 ∆lepA::ZeoR p-empty- pPUV15- ZeoR, Sequence (HR334) KanR(CT242) mspA135- KanR, determinants 195-Nluc- NatR of LepA NatR(599) regulation
SF924 ∆lepA::ZeoR pPnatp-lepA- pPUV15- ZeoR, Sequence (HR334) KanR(SF121) mspA1-195- HygR, determinants Nluc- KanR, of LepA NatR(908) NatR regulation
SF925 ∆lepA::ZeoR p-empty- pPUV15- ZeoR, Sequence (HR334) KanR(CT242) mspA1-195- HygR, determinants Nluc- KanR, of LepA NatR(908) NatR regulation
SF789 ∆lepA::ZeoR/ None None ZeoR, Calcein ∆mspA::Hyg HygR contribution R HR329 ∆mspA::Zeo None None ZeoR Calcein R contribution
SF447 ∆lepA::ZeoR pPUV15-Tet- p-emtpy- ZeoR, Porin (HR334) dCas9- HygR(418) HygR, Knockdown sgRNA1 KanR (mspA1)- KanR(419) SF448 ∆lepA::ZeoR pPUV15-Tet- p-emtpy- ZeoR, Porin (HR334) dCas9- HygR(418) HygR, Knockdown sgRNA2 KanR (mspA2)- KanR(420) SF449 ∆lepA::ZeoR pPUV15-Tet- p-emtpy- ZeoR, Porin (HR334) dCas9- HygR(418) HygR, Knockdown sgRNA3 KanR (mspA3)- KanR(421) SF450 ∆lepA::ZeoR pPUV15-Tet- p-emtpy- ZeoR, Porin (HR334) dCas9- HygR(418) HygR, Knockdown sgRNA4 KanR (mspB1)- KanR(422) SF451 ∆lepA::ZeoR pPUV15-Tet- p-emtpy- ZeoR, Porin (HR334) dCas9- HygR(418) HygR, Knockdown sgRNA5(msp KanR B2)- KanR(423) SF452 ∆lepA::ZeoR pPUV15-Tet- p-emtpy- ZeoR, Porin (HR334) dCas9- HygR(418) HygR, Knockdown sgRNA6 KanR (mspB3))- KanR(424)
128 Appendix Table A1.4 (Continued)
SF453 ∆lepA::ZeoR pPUV15-Tet- p-emtpy- ZeoR, Porin (HR334) dCas9- HygR(418) HygR, Knockdown sgRNA7 KanR (mspC1)- KanR(425) SF454 ∆lepA::ZeoR pPUV15-Tet- p-emtpy- ZeoR, Porin (HR334) dCas9- HygR(418) HygR, Knockdown sgRNA8 KanR (mspC2)- KanR(426) SF455 ∆lepA::ZeoR pPUV15-Tet- p-emtpy- ZeoR, Porin (HR334) dCas9- HygR(418) HygR, Knockdown sgRNA9 KanR (mspC3)- KanR(427) SF456 ∆lepA::ZeoR pPUV15-Tet- p-emtpy- ZeoR, Porin (HR334) dCas9- HygR(418) HygR, Knockdown sgRNA10 KanR (mspD1)- KanR(428) SF457 ∆lepA::ZeoR pPUV15-Tet- p-emtpy- ZeoR, Porin (HR334) dCas9- HygR(418) HygR, Knockdown sgRNA11 KanR (mspD2)- KanR(429) SF458 ∆lepA::ZeoR pPUV15-Tet- p-emtpy- ZeoR, Porin (HR334) dCas9- HygR(418) HygR, Knockdown sgRNA12 KanR (mspD3)- KanR(430) SF502 ∆lepA::ZeoR pPUV15-Tet- p-natp-lepA- ZeoR, Porin (HR334) dCas9- HygR(417) HygR, Knockdown sgRNA null- KanR KanR(CT295) SF503 ∆lepA::ZeoR pPUV15-Tet- p-emtpy- ZeoR, Porin (HR334) dCas9- HygR(418) HygR, Knockdown sgRNA null- KanR KanR(CT295) SF504 ∆lepA::ZeoR pPUV15-Tet- p-natp-lepA- ZeoR, Porin (HR334) dCas9- HygR(417) HygR, Knockdown sgRNA1- KanR sgRNA6- sgRNA7- KanR(484) SF505 ∆lepA::ZeoR pPUV15-Tet- p-emtpy- ZeoR, Porin (HR334) dCas9- HygR(418) HygR, Knockdown sgRNA1- KanR sgRNA6- sgRNA7- KanR(484) SF740 pET26-PT7- None None KanR Purification of His-lepA- LepA KanR
129 Appendix Table A1.4 (Continued)
SF969 ∆lepA::ZeoR pPnatp-lepA- pPUV15- ZeoR, Sequence (HR334) KanR(SF121) mspA(T162C, KanR, determinants T213C,T306C NatR of LepA ,T357C,T417, regulation T540C,T552C ,T594C)-Nluc- NatR(944) SF970 ∆lepA::ZeoR p-empty- pPUV15- ZeoR, Sequence (HR334) KanR(CT242) mspA(T162C, KanR, determinants T213C,T306C NatR of LepA ,T357C,T417, regulation T540C,T552C ,T594C)-Nluc- NatR(944) SF574 ∆lepA::ZeoR pPnatp-lepA- pPUV15-Tet- ZeoR, Sequence (HR334) KanR(SF121) mspA recode KanR, determinants (GC-rich)- NatR of LepA Nluc- regulation NatR(554) SF582 ∆lepA::ZeoR p-empty- pPUV15-Tet- ZeoR, Sequence (HR334) KanR(CT242) mspA recode KanR, determinants (GC-rich)- NatR of LepA Nluc- regulation NatR(554) SF575 ∆lepA::ZeoR pPnatp-lepA- pPUV15-Tet- ZeoR, Sequence (HR334) KanR(SF121) mspA recode KanR, determinants (AT-rich)- NatR of LepA Nluc- regulation NatR(555) SF583 ∆lepA::ZeoR p-empty- pPUV15-Tet- ZeoR, Sequence (HR334) KanR(CT242) mspA recode KanR, determinants (AT-rich)- NatR of LepA Nluc- regulation NatR(555)
130 Appendix Table A1.5 – Plasmid List
Plasmid Contents Antibiotic Parental Notes Number Selection Vector/ Source
121 pPnatp-lepA-KanR Kan CT94
283 pPUV15-mspA- Nat CT250 Nanoluciferase(Nluc)- NatR 287 pPUV15-Nluc-NatR Nat CT250
383 pPUV15- Nat CT250 mspC(MSMEG_5483)- Nluc-NatR 384 pPUV15- Nat CT250 mspB(MSMEG_0520)- Nluc-NatR 385 pPUV15- Nat CT250 mspD(MSMEG_6057)- Nluc-NatR 413 pPUV15-mspAss-Nluc- Nat CT250 NatR 414 pPUV15-mspBss-Nluc- Nat CT250 NatR 415 pPUV15-mspCss-Nluc- Nat CT250 NatR 416 pPUV15-mspDss-Nluc- Nat CT250 NatR 417 pPnatp-lepA-HygR Hyg CT204 418 p-emtpy-HygR Hyg CT204
419 pPUV15-Tet-dCas9- Kan CT295 sgRNA1 (mspA1)- KanR 420 pPUV15-Tet-dCas9- Kan CT295 sgRNA2 (mspA2)- KanR 421 pPUV15-Tet-dCas9- Kan CT295 sgRNA3 (mspA3)- KanR 422 pPUV15-Tet-dCas9- Kan CT295 sgRNA4 (mspB1)- KanR 423 pPUV15-Tet-dCas9- Kan CT295 sgRNA5(mspB2)- KanR 424 pPUV15-Tet-dCas9- Kan CT295 sgRNA6 (mspB3)- KanR 425 pPUV15-Tet-dCas9- Kan CT295 sgRNA7 (mspC1)- KanR 426 pPUV15-Tet-dCas9- Kan CT295 sgRNA8 (mspC2)- KanR
131 Appendix Table A1.5 (Continued)
427 pPUV15-Tet-dCas9- Kan CT295 sgRNA9 (mspC3)- KanR 428 pPUV15-Tet-dCas9- Kan CT295 sgRNA10 (mspD1)- KanR 429 pPUV15-Tet-dCas9- Kan CT295 sgRNA11 (mspD2)- KanR 430 pPUV15-Tet-dCas9- Kan CT295 sgRNA12 (mspD3)- KanR 484 pPUV15-Tet-dCas9- Kan CT295 Used sgRNA1-sgRNA6- sgRNAs sgRNA7-KanR from p419, p424, 425 to create multi- sgRNA knockdown of mspABC 1-90 544 pPUV15-mspA -mspD- Nat CT250 NatR 1-135 545 pPUV15-mspA - Nat CT250 mspD-NatR 1-195 546 pPUV15-mspA - Nat CT250 mspD-NatR 1-258 547 pPUV15-mspA - Nat CT250 mspD-NatR 1-339 548 pPUV15-mspA - Nat CT250 mspD-NatR 554 pPUV15-Tet-mspA recode Nat CT250 (GC-rich)-Nluc-NatR
555 pPUV15-Tet-mspA recode Nat CT250 (AT-rich)-Nluc-NatR 1-135 596 pPUV15-mspD - Nat CT250 mspA136-195-mspD196- 624-Nluc-NatR 135-195 599 pPUV15-mspA - Nat CT250 Nluc-NatR 711 pET26-PT7-6xHisTag- Kan pET26 lepA-KanR 741 pET21-PT7-Venus- Amp pET21 AmpR 896 pPUV15-mspA-mRFP- Nat CT250 NatR 1-195 908 pPUV15-mspA -Nluc- Nat CT250 NatR 944 pPUV15- Nat CT250 mspA(T162C,T213C,T 306C,T357C,T417,T54 0C,T552C,T594C)- Nluc-NatR
132 Appendix Table A1.5 (Continued)
CT204 pPTB21-eGFP-wag31- Hyg Lab Backbone HygR vector of cloning vector for Tweety- integrating lepA alleles
CT242 p-empty-KanR Kan CT94
CT250 pPUV15-eGFP-NatR Nat Lab Backbone vector of cloning vector for Tweety- integrating porin alleles
CT295 pPUV15-Tet-dCas9- Kan Courteous gift of sgRNA null-KanR Jeremy Rock
CT94 pPUV15-Tet-eGFP-KanR Kan Lab Backbone vector of cloning vector for L5- integrating lepA alleles
133 Appendix Table A1.6: Primer List
Primer Name Sequence Purpose (vector Number number or description) P315 CT133_nat_prom_4556_F TTTGCGTTTAATACTGCATGCAC 121 Tgaatcgagcatatgccagcgg P422 4556_R_redo CTAGGGTCCCCAATTAATTAGCT 121 AAtcacttcttgggcttgtcgccc P626 UV15-mspA_F cttaattaagaaggagatatacatatgaaggca 283;413;544;545;54 atcagtcgggtgctg 6;547;599;908;944 P627 mspA-Nluc_R ggctcacgtctgtTCCAGAACCgttcatgtt 283;944 ccagggttcgccg P628 mspA-Nluc_F aaccctggaacatgaacGGTTCTGGAac 283 agacgtgagccgaaagattcg P629 Nluc-Ct250-R GGGTCCCCAATTAATTAGCTAAT 283;287;383;384;38 CACAGGTCTTCCTCGCTGATCAG 5;413;414;415;416;5 CTTCTGCTCggcgaggatgcgctcgc 44;545;546;547;596; 599;908;944 P632 UV15-Nluc_F gcttaattaagaaggagatatacatatgacaga 287 cgtgagccgaaagattcg P844 Nluc-F gtcttcaccctggaggacttcgtcg 383;384;385;413;41 4;415;416;544;545;5 46;547;596;599;908 P772 UV15_MspC_F cttaattaagaaggagatatacatatgaaggca 384;415 atcagtcgggtgctg P773 MspC_Nluc_R cctccagggtgaagacTCCAGAACCgttc 384 atgttccagggttcgccg P774 UV15_MspB_F ttaattaagaaggagatatacatatgacggcatt 383;414 caagcgggtgct P775 MspB_Nluc_R cctccagggtgaagacTCCAGAACCgttc 383 atattccaaggctcgccgta P776 UV15_MspD_F ttaattaagaaggagatatacatgtgcgctacct 385;417 cgtcatgatg P777 MspD_Nluc_R tccagggtgaagacTCCAGAACCgttcat 385;544;545;546;54 gttccagggctcgcc 7;548;596 P817 mspAss_R acgaagtcctccagggtgaagacgcctgcgtg 413 agaggtgcctgtgc P818 mpsBss_R cgaagtcctccagggtgaagaccgccgcgcc 414 cgcactcac P819 mspCss_R acgaagtcctccagggtgaagacacccgcgt 415 gagaggtgcccg P820 mspDss_R acgaagtcctccagggtgaagaccaccgcgtt 416 ggcagggcgg P842 CT204_lepA_F actgtttaaactctagaaatattattgcgggtgcg 417 ttgttcgc P843 CT204_lepA_R tcgtcgccaccaatccccatatgTCGTACG 417 CTAGTTAACTACGTCGACA P807 mspA_R AAACAATGGTGTGTTCCCCCTGG 419 AC
134 Appendix Table A1.6 (Continued)
P808 mspA_F GGGAGTCCAGGGGGAACACACC 419 ATT P826 MspA_R2 AAACGTCGACGTCTCCGGCGCC 420 GAGGGT P827 MspA_F2 GGGAACCCTCGGCGCCGGAGAC 420 GTCGAC P828 MspA_R3 AAACCTGGACCGCAACCGTCTTA 421 CCCGT P829 MspA_F3 GGGAACGGGTAAGACGGTTGCG 421 GTCCAG P811 mspB_R AAACTGGTCGCTGGGTGTGGGC 422 ATC P812 mpsB_F GGGAGATGCCCACACCCAGCGA 422 CCA P830 MspB_R2 AAACGTCGACGTCTCGGGTCCT 423 GC P831 MspB_F2 GGGAGCAGGACCCGAGACGTCG 423 AC P832 MspB_R3 AAACGTTGCTCGCAGGCACCAC 424 GGGAAT P833 MspB_F3 GGGAATTCCCGTGGTGCCTGCG 424 AGCAAC P813 mspC_R AAACGTTGGCTGCGGCCGTCGC 425 GGGGT P814 mspC_F GGGAACCCCGCGACGGCCGCA 425 GCCAAC P834 MspC_R2 AAACGTCGACGTCTCGGGTCCC 426 GCAGGC P835 MspC_F2 GGGAGCCTGCGGGACCCGAGAC 426 GTCGAC P836 MspC_R3 AAACTCGACGACGGTGACATCAC 427 CGGT P837 MspC_F3 GGGAACCGGTGATGTCACCGTC 427 GTCGA P809 mspD_R AAACTGGTCATTGGGCGTCGGC 428 ATC P810 mspD_F GGGAGATGCCGACGCCCAATGA 428 CCA P838 MspD_R2 AAACGTGGACGTGAAGGGCGCG 429 AAAGGAGC P839 MspD_F2 GGGAGCTCCTTTCGCGCCCTTC 429 ACGTCCAC P840 MspD_R3 AAACTCGACGGAGGCGACATCA 430 CC P841 MspD_F3 GGGAGGTGATGTCGCCTCCGTC 430 GA
135 Appendix Table A1.6 (Continued)
P905 MspA(1-90)- MspD_F GCACCTCTCACGCAGGCCTGGA 544 CAATCAGCTCAGCGTGGTCGAC GGCC P906 MspA(1-90)- MspD_R GGCCGTCGACCACGCTGAGCTG 544 ATTGTCCAGGCCTGCGTGAGAG GTGC P907 MspA(1-135)- MspD_F CCAGGACCGCACCCTCACCGTG 545 CAGCAAGCCGAGACATTCCTCAA CGGC P908 MspA(1-135)- MspD_R GCCGTTGAGGAATGTCTCGGCTT 545 GCTGCACGGTGAGGGTGCGGTC CTGG P909 MspA(1-195)-MspD-F CCTGGACCGCAACCGTCTTACC 546 CGTGAGTGGTTTCACTCCGGCC GCG P910 MspA(1-195)-MspD-R CGCGGCCGGAGTGAAACCACTC 546 ACGGGTAAGACGGTTGCGGTCC AGG P911 MspA(1-258)-MspD-F CCCCGGTGCCGACGAGTTCGAG 547 GGCACGCTCGAACTCGGGTATC AGG P912 MspA(1-258)-MspD-R CCTGATACCCGAGTTCGAGCGT 547 GCCCTCGAACTCGTCGGCACCG GGG P913 MspA(1-339)-MspD-F CATCAACTTCAGCTACACCACCC 548 CGAACATCCTCATCGACGGAGG CGACATCACC P914 MspA(1-339)-MspD-R GGTGATGTCGCCTCCGTCGATG 548 AGGATGTTCGGGGTGGTGTAGC TGAAGTTGATG P933 MspD135_MspA_R TTGAGGAAGGTGTCCCACTGCT 596 GCACGGTCAGCGTGCGACCTTG GC P934 MspD135_MspA_F GCCAAGGTCGCACGCTGACCGT 596 GCAGCAGTGGGACACCTTCCTC AA P935 MspD195_MspA_R GCGCGACCGGAGTGGAACCACT 596 CACGGGTCAGTCGGTTCCGGTC GA P936 MspD195_mspA_F TCGACCGGAACCGACTGACCCG 596 TGAGTGGTTCCACTCCGGTCGC GC P950 UV15_MspA135-195_F tgcttaattaagaaggagatatacatATGCA 599 GCAGTGGGACACCTTCCTCAATG P951 MspA135-195_Nluc_R agggtgaagacTCCAGAACCACGGG 599 TAAGACGGTTGCGGTCCA P917 UV15_MspA- tgcttaattaagaaggagatatacatatgaagg 554 recode_genome_F cgatttcgcgggtcctg
136 Appendix Table A1.6 (Continued)
P918 MspA- agggtgaagacTCCAGAACCgttcatgttc 554 recode_genome_Nluc_R cacggctcgcc P919 UV15_MspA- tgcttaattaagaaggagatatacatatgaaag 555 recode_inverse_F ccatttcaagggttttgatagct P920 MspA- agggtgaagacTCCAGAACCattcatattc 555 recode_inverse_Nluc_R catggttcaccatacgtcgt P1188 MspA1-195_Nluc-R CCTCCAGGGTGAAGACTCCAGA 908 ACCACGGGTAAGACGGTTGCGG TCC P1139 MspA/C_mRFP_R CTTGATGACGTCCTCGGAGGAG 896 GCGTTCATGTTCCAGGGTTCGCC GT P1144 mRFP_CT250_R GGTCCCCAATTAATTAGCTAAAG 896 CTTCAGCGGCCCTCGGCGCGCT CGTACT P1143 mRFP_F GCCTCCTCCGAGGACGTCATCA 896 A P1042 pET26_venus_F actttaagaaggagatatacatATGGTCTC 741 GAAGGGCGAGGAGCTG P1043 pET26_venus_R gtggtggtggtggtgctcgatcaGTGATGG 741 TGATGGTGATGACACTTGTACAG C P845 mspA_F1_RT CAGATGTGATCCTCTTAGATCTC RT-qPCR of porins C P846 mspA_R1_RT CTCAGGCCCACCTGTTT RT-qPCR of porins P847 mspA_F2_RT TAGATCTCCGAAGTCTCTGAA RT-qPCR of porins P848 mspA_R2_RT GTTCTCCCTAACTGTATCGC RT-qPCR of porins P849 mspB_F1_RT ATTCAAGCGGGTGCTGAT RT-qPCR of porins P850 mspB_R1_RT ACGAGGCTCAATTCGTTGT RT-qPCR of porins P851 mspB_F2_RT ACGGGAATGTTCGTGAGTG RT-qPCR of porins P852 mspB_R2_RT GCTGCACGGTCAGAGTT RT-qPCR of porins P853 mspC_F1_RT GAACGTAAGAACAGAGTGTGGA RT-qPCR of porins G P854 mspC_R1_RT CACCGATGTTGTGGTGTAGG RT-qPCR of porins P855 mspC_F2_RT TTTAATTCGCGCCACATGAC RT-qPCR of porins P856 mspC_R2_RT GGGTCCACCGATGTTGT RT-qPCR of porins P857 mspD_F1_RT CGTCATGATGTTCGCTCTACTC RT-qPCR of porins P858 mspD_R1_RT TTGAGGAATGTCTCGGCTTG RT-qPCR of porins P859 mspD_F2_RT CGCTACCTCGTCATGATGTT RT-qPCR of porins P860 mspD_R2_RT AATCAGCTCAGCGTGGTC RT-qPCR of porins SF37 RTPCR_sigA_F gactacaccaagggctacaag RT-qPCR of porins SF38 RTPCR_sigA_R ttgatcacctcgaccatgtg RT-qPCR of porins
137 Appendix Table A1.6 (Continued)
P1017 pET28_LepA_myco_R agtggtggtggtggtggtgctcgactacttcttgg 711 gcttgtcgcccgc P1018 pET28_LepA_myco_F ttaactttaagaaggagatataccatgggcagc 711 agccatcatcatcatcatcaccgcttgacccac gctcaccag
138 Appendix 2: Supplementary Material from Chapter 3
sucrose cushion S9 Ami3 supernatant LepA ribosome pellet EF-G2
Figure A2.1- Optimization of sucrose cushion analysis for ribosome-associating factors in Msm Ribosomes were purified from strains carrying epitope-tagged versions of Ami3, LepA, and EF-G2. Each lane contains signal from corresponding antibody. For S9, this was an antibody to Mtb ribosomal S9 protein. For Ami3, LepA, and EF-G2, this was an antibody to the Myc tag.
2.0
1.5 WT ΔMSMEG_6535 (EF-G like) ΔMSMEG_5222( EngD/GTPase)
Abs. 260 1.0 ΔMSMEG_1889 (RsgA, 30S-associated)
0.5
20 40 60 Distance (mm from top of gradient)
Figure A2.2 – Examination of ribosome distribution of candidate mutants in Msm
139 Figure A2.2 (continued) Ribosomes from Msm strains lacking various ribosome-associated candidate genes were analyzed using sucrose gradient density centrifugation.
Appendix Table 2.1 – Candidate factor ratio information
Ribosome unit ratio Protein Possible Function found with L70S/LWCL, L50S/LWCL, L30S/LWCL, S70S/L70S, Rv1488 Regulator of protease activity HflC S70S_WCL Rv2024c P-loop NTPase/DEAD-box helicase S70S_30S Rv2629 peptide-release factor L70S/LWCL, S70S/L70S S70S_30S, S70S/L70S, Rv2689c TrmA S70_WCL Rv2837c nucleotide binding/ ribosomal protein L9 S70S_WCL Rv2974c Obg-like protein S70S_30S Rv3030 RNA methytransferase S70S_WCL Rv3422c ATP-binding protein/TsaE S70S_S30S
Rv0282 EccA S70_WCL
Rv0283 EccB L70S_L30,L50S/L30S L70/L30S, L70S/LWCL, Rv0284 EccC L50S/LWCL,L50S/L30S L70/L30S, L70S/LWCL, L50S/LWCL,L50S/L30S, Rv0290 EccD S70S/SWCL L70/L30S, L70S/LWCL, L50S/LWCL, L30S/LWCL,L50S/L30S, Rv0292 EccE S70S/SWCL
L70/L30S, L70S/LWCL, Rv1782 EccB L50S/LWCL, L50S/L30S
L70/L30S, L70S/LWCL, Rv1783 EccC L50S/LWCL, L50S/L30S
L70/L30S, L70S/LWCL, Rv1795 EccD L50S/LWCL, L50S/L30S
L70/L30S, L70S/LWCL, Rv1797 EccE L50S/LWCL, L50S/L30S
L70/L30S, L70S/LWCL, Rv3895c EccB L50S/LWCL, L50S/L30S
Rv0251c heat shock-like family protein L30S/LWCL, S70S/L70S
140
Appendix Table A2.1 (Continued) S70S_30S, S70S/L70S, Rv0734 MapA, nascent peptide processing S70_WCL Rv0120c EF-G2, ribosome-associated GTPase S70S_30S,S70S/L70S Rv1112 EngD, ribosome-associated GTPase S70S_WCL
Rv0208 tRNA methyltransferase L70S/L30S
Rv2725 HflX,GTP-binding protein L50S/LWCL, L50S/L30S Rv2861c MapB, nascent peptide processing S70S_WCL, S70S_30S Rv3404c putative formyl-methionine transferase L50S/L30S RsgA,small subunit ribosome maturation Rv3228 factor L30S/LWCL, S30S/SWCL L=log-phase; S= stressed. ‘S70S_S30S’ indicates ratio of ratios using 30S as normalizing factor. ‘S70S/S30S’ indicates a simple ratio where protein abundance in S70S is divided by protein abundance in S30S.
Materials and Methods
Sucrose cushion-western blotting Cells were grown to log-phase and lysed by bead beating in lysis buffer (20 mM Tris-HCl, pH 7.5, 100 mM NH4Cl, 10.5 mM MgCl2, 0.5 mM EDTA, 6mM β-mercaptoethanol, 2 mM DTT). 1- 5ml of cell lysate was layered onto 13 mL of a sucrose cushion (37.7% sucrose, 20 mM Tris- HCl, ph 7.5, 0.5M NH4Cl, 10.5 mM MgCl2, 0.5 mM EDTA, 2 mM DTT, 6mM β-mercaptoethanol) in ultracentrifuge tubes. Sucrose cushions were spun in an ultracentrifuge using 70.1 Ti Rotor at 40K, 4⁰C for 22 hours. Use lysis buffer to wash ribosome pellet at bottom of cushion. 1ml of storage buffer (25 mM HEPES-KOH pH 7.5, 100 mM NH4OAc, 10 mM Mg(OAc), 2 mM DTT, 6mM β-mercaptoethanol) was used to resuspend the pellet overnight at 4⁰C. The remaining supernatant from the pellet was concentrated to the same volume as the pellet. Equal sample was loaded from supernatants and pellets into an SDS PAGE gel. Western blotting was performed using a primary a-Myc antibody, to detect tagged proteins, and a a-Mtb ribosomal protein S9 antibody.
Ribosome analysis Ribosomes were purified from Msm strains as described previously [2].
References
1. Rock, J.M., et al., Programmable transcriptional repression in mycobacteria using an orthogonal CRISPR interference platform. Nat Microbiol, 2017. 2: p. 16274. 2. Gomez, J.E., et al., Ribosomal mutations promote the evolution of antibiotic resistance in a multidrug environment. Elife, 2017. 6.
141