New Tools for Studying Mal Gene Regulation in Yersinia Pestis Carrie Jo Oster Iowa State University

Iowa State University Capstones, Theses and Graduate Theses and Dissertations Dissertations

2012 New tools for studying mal gene regulation in Yersinia pestis Carrie Jo Oster Iowa State University

Follow this and additional works at: https://lib.dr.iastate.edu/etd Part of the Genetics Commons, Microbiology Commons, and the Molecular Biology Commons

Recommended Citation Oster, Carrie Jo, "New tools for studying mal gene regulation in Yersinia pestis" (2012). Graduate Theses and Dissertations. 12426. https://lib.dr.iastate.edu/etd/12426

This Thesis is brought to you for free and open access by the Iowa State University Capstones, Theses and Dissertations at Iowa State University Digital Repository. It has been accepted for inclusion in Graduate Theses and Dissertations by an authorized administrator of Iowa State University Digital Repository. For more information, please contact [email protected]. New tools for studying mal gene regulation in Yersinia pestis

Carrie Jo Oster

A thesis submitted to the graduate faculty

in partial fulﬁllment of the requirements for the degree of

MASTER OF SCIENCE

Major: Molecular, Cellular, and Developmental Biology

Program of Study Committee:

Gregory J. Phillips, Major Professor

F. Chris Minion

Adam Bogdanove

Iowa State University

Ames, Iowa

2012

TABLE OF CONTENTS

LIST OF TABLES ...... iv

LIST OF FIGURES ...... v

ACKNOWLEDGEMENTS ...... vi

ABSTRACT ...... vii

CHAPTER 1. GENERAL INTRODUCTION ...... 1

1.1 Introduction and Literature Review ...... 1

1.2 Thesis Organization ...... 7

CHAPTER 2. VECTORS FOR LIGATION-INDEPENDENT CONSTRUC-

TION OF lacZ GENE FUSIONS AND CLONING OF PCR PRODUCTS

USING A NICKING ENDONUCLEASE ...... 10

Abstract ...... 10

2.1 Introduction ...... 11

2.2 Materials and Methods ...... 13

2.2.1 Bacterial strains, growth media, and enzymes ...... 13

2.2.2 Construction of pLacCOs1 and pKanCOs1 ...... 14

2.2.3 PCR ampliﬁcation of promoters ...... 15

2.2.4 Promoter cloning and characterization ...... 15

2.3 Results and Discussion ...... 17

2.3.1 Development of a nicking endonuclease system for lacZ gene fusions . . 17

2.3.2 Construction of lacZ gene fusions ...... 18

2.3.3 Reporter gene expression ...... 19

2.3.4 PCR amplicon cloning ...... 20 iii

2.3.5 Conclusions ...... 20

CHAPTER 3. TEMPERATURE-REGULATED EXPRESSION OF mal GENES

IN Yersinia pestis IS MEDIATED BY THE TRANSCRIPTIONAL AC-

TIVATOR MalT, AS REVEALED BY NEW METHODS FOR CON-

STRUCTION OF SINGLE-COPY lacZ GENE FUSIONS ...... 22

Abstract ...... 22

3.1 Introduction ...... 23

3.2 Materials and Methods ...... 25

3.2.1 Bacterial strains, growth media, and enzymes ...... 25

3.2.2 Construction of malT deletion mutants in E. coli and Y. pestis ..... 26

3.2.3 Cloning of malT ...... 28

3.2.4 Construction of lacZ gene fusions expressed in single copy ...... 28

3.2.5 Transfer of lacZ gene fusions by conjugation ...... 31

3.2.6 Compatibility of F0 lac pro with Y. pestis virulence plasmids ...... 33

3.2.7 Construction and use of the helper plasmid pCLIP-ﬂp ...... 33

3.2.8 β-galactosidase Assays ...... 34

3.3 Results ...... 34

3.3.1 Streamlined method for construction of new F0s carrying lacZ gene fusions 34

3.3.2 Characterization of PmalK -lacZ gene fusions ...... 35 3.3.3 Eﬀect of quorum sensing on mal gene expression ...... 37

3.3.4 Role of MalT in gene expression ...... 37

3.3.5 Expression of malT is independent of temperature ...... 40

3.4 Discussion ...... 40

CHAPTER 4. GENERAL CONCLUSIONS ...... 45

BIBLIOGRAPHY ...... 50 iv

LIST OF TABLES

Table 2.1 PCR and DNA sequencing primers used in this study...... 16

Table 3.1 Bacterial strains and plasmids used in this study ...... 27

Table 3.2 PCR primers used in this study ...... 29

Table 4.1 Plasmids built using no-ligase cloning system ...... 46

Table 4.2 Gene deletions in Y. pestis constructed by recombineering ...... 48 v

LIST OF FIGURES

Figure 2.1 Strategy for construction of lacZ gene fusions using the nicking endonu-

clease Nt.BspQI...... 12

Figure 2.2 pLacCOs1 and pKanCOs1 plasmids ...... 14

Figure 2.3 Temperature regulation of mal genes as measured by lacZ gene fusions

expressed from pLacCOs1 ...... 19

Figure 3.1 Plasmids constructed for genetic engineering of Y. pestis ...... 30

Figure 3.2 Scheme used to construct lacZ gene fusions expressed from F0 factors . 32

Figure 3.3 β-galactosidase activity of malK -lacZ fusions ...... 36

Figure 3.4 Incompatibility of F0 lac with Y. pestis plasmids ...... 37

0 Figure 3.5 Expression of Φ(malK Yp-lacZ )Hyb fusions in the absence of quorum sensing ...... 38

Figure 3.6 Comparison of amino acid sequences of MalT from Y. pestis and E. coli

K-12 ...... 39

Figure 3.7 β-galactosidase activity in malT deletion backgrounds ...... 41

Figure 3.8 β-galactosidase activity of Φ(malT 0-lacZ )Hyb ...... 42 vi

ACKNOWLEDGEMENTS

I want to express my gratitude to everyone who has helped me throughout the process of researching and writing this thesis. In particular, I would like to thank my advisor, Dr. Gregory

Phillips, for all of his encouragement and assistance during my time in his lab. I appreciate all the time he has spent helping me complete this project. I would also like to thank the other members of my thesis committee, Dr. F. Chris Minion and Dr. Adam Bogdanove, for their input and advice. Thank you to Dr. Jenni Boonjakuakul and Dr. Joanna Kuehn for listening to my questions and providing me with a sounding board for my ideas. I’ve enjoyed working with all of you.

Thank you to the Oﬃce of Naval Research (grant ONR-MURI N00014-06-1-1176) and NIH

(grant NIAID R21 AI074591) for funding this research.

Most of all, thank you to my family — especially my best friend, my husband Zach. I could never have done this without your love and support. vii

ABSTRACT

Y. pestis, the etiologic agent of plague, is a signiﬁcant human pathogen because of its historic role as the cause of three major pandemics and its current role as a potential bioterrorism agent.

Gene regulation in this organism is complex and multiple microarray studies have shown that the expression of several genes is aﬀected by quorum sensing as well as temperature. Maltose utilization genes are of particular interest because they are regulated by both of these systems.

In order to perform genetic studies on mal genes, and to better understand gene expression in general in Y. pestis, we developed a number of new tools that comprise the majority of the work in this thesis.

To take advantage of recombineering as a useful system for constructing new mutants, we improved the protocol to promote a higher efficiency of survival and recombination in electroporated Y. pestis cells. We also designed a ligase-independent cloning system to build lacZ gene fusions more quickly and efficiently. We then took advantage of this system to design a method for building single-copy lacZ fusions that are conveniently transferred to Y. pestis by conjugation. These methods were used in combination to examine temperature regulation and quorum sensing effects on malK expression. We determined that a malK -lacZ translational gene fusion is strongly downregulated at 37◦C in Y. pestis and slightly downregulated in similar conditions in E. coli. Fusions to a constitutively expressed promoter, however, did not exhibit this kind of expression. In contrast, the gene fusion did not show significant differences in a quorum sensing negative strain, suggesting that AHL quorum sensing does not contribute to regulation of the expression of mal genes under the conditions of the experiment. When these fusions were assayed in strains deleted for malT but complemented with malT expressed on low-copy plasmids, we observed that Y. pestis MalT produced a more significant change in response to temperature than did E. coli MalT. Because the amino acid sequences are mostly similar between both species, future work will focus on what differences in MalT produce such viii dissimilar temperature regulation effects in these bacteria. The systems we have developed will improve future work with this organism to determine the molecular mechanisms of gene regulation in this species. 1

CHAPTER 1. GENERAL INTRODUCTION

1.1 Introduction and Literature Review

Historically, Yersinia pestis is an important bacterial pathogen, most notably as the cause of three pandemics including the Black Death (Parkhill et al., 2001), which was responsible for the death of up to one-third of the European population in the 14th century. In 1894 Alexandre

Yersin identiﬁed Y. pestis as the causative agent of bubonic plague; however, a link between this pathogen and the Black Death was largely based on historical accounts from the time that described bubo-like swellings in victims’ armpits and groins (Callaway, 2011). Although modern plague has remained in existence for the past seven centuries, no outbreaks have been as deadly as the Black Death. This led some skeptics to question the etiology of the 14th century plague, rather attributing the disease outbreak to other bacterial or viral agents (Duncan and Scott,

2005). The use of “suicide PCR” to conﬁrm the identity of Black Death as Y. pestis (Raoult et al., 2000) has also been criticized because of its lack of positive controls and the possibility of contamination in ancient DNA samples (Wood and DeWitte-Avi˜na, 2003; Prentice et al., 2004;

Gilbert et al., 2004). Most recently, these alternative explanations have been largely silenced by the use of next-generation DNA sequencing technology that has enabled sequencing of

Y. pestis chromosomal and plasmid DNA isolated directly from corpses that apparently died of plague (Bos et al., 2011; Schuenemann et al., 2011; Haensch et al., 2010). Further studies using the latest methods in DNA sequencing and molecular biology should allow us to identify the genetic diﬀerences between extant strains of Y. pestis and strains that may help explain details of the Black Death.

While in modern times the plague has not posed a severe threat to large populations, it is a concern for public health because of its potential as a bioterrorism agent (Kortepeter and 2

Parker, 1999). The microorganism is relatively easy to grow and can be genetically manipulated, leading to its inclusion on the Center for Disease Control’s Select Agent list of biological agents determined to have the potential to pose a severe threat to human and animal health.

Genome sequence analysis has also led to a greater understanding of how Y. pestis evolved into a highly pathogenic microorganism. Comparative genomics have been used to suggest that

Y. pestis evolved from its closest relative Y. pseudotuberculosis around 20,000 years ago (Acht- man et al., 1999). While Y. pseudotuberculosis generally results in self-limiting gastrointestinal disorders, infection with Y. pestis is usually fatal if left untreated, resulting from systemic infection throughout the body. A number of genetic differences were identified that likely con- tributed to these differences in pathobiology. Specifically, Y. pestis has horizontally acquired two plasmids (pPCP1 and pMT1) that encode factors known to contribute to virulence (Zhou and Yang, 2009; Brubaker, 2007). These factors include the murine toxin and F1 capsular antigen, encoded on the Y. pestis pMT (pFra) plasmid (Lindler et al., 1998), and the plasminogen activator and immunity proteins found on the pPCP1 (pPla) plasmid (Lindler, 2004). The murine toxin is a phospholipase D that causes vascular collapse in mice during the septicemic stage of plague and also possesses β-andrenergic blocking activity (Hinnebusch et al., 2002).

Since this gene is more highly expressed at lower temperatures, it may be required primarily for bacterial survival in the flea host (Hinnebusch et al., 2002). pPla encodes the plasminogen activator that allows bacterial adhesion and damage to mammalian host cells (Lähteenmäki et al., 1998).

In addition to acquisition of new genes, gene inactivation has also led to altered virulence.

For example, the main adhesin and invasin in Y. pseudotuberculosis, yadA and inv, are not functional in Y. pestis (Zhou and Yang, 2009). While these gene products assist in Y. pseudotuberculosis invasion of epithelial cells, their loss may contribute to the dissemination of

Y. pestis from primary infection sites. Other Y. pseudotuberculosis genes are inactivated in

Y. pestis, including lpxL, which helps the bacteria avoid the inﬂammatory response by alter- ing lipopolysaccharide (LPS), allowing the bacteria to inhibit TLR4 (Zhou and Yang, 2009).

Y. pestis is also missing genes from the O-antigen cluster, resulting in production of rough LPS deﬁcient in the O antigen, an adaptation that might contribute to systemic infection (Zhou 3 and Yang, 2009). The glyoxylate bypass pathway is also constitutive in Y. pestis, which may contribute to the ability of the pathogen to grow on acetate and fatty acids provided by the host (Chain et al., 2004).

Despite the numerous genetic diﬀerences between Y. pestis and Y. pseudotuberculosis, both species, along with the more distantly related enteric pathogen Y. enterocolitica (Perry et al.,

1998) also share virulence factors. For example, all three species possess a large virulence plasmid, known as pCD1 in Y. pestis. This plasmid encodes the type III secretion system

(T3SS), which is one of the best characterized virulence factors in Gram negative pathogens.

The importance of this virulence factor delivery system is underscored by the observation that

Y. pestis mutants cured of pCD1 are highly attenuated in animal models (Perry et al., 1998).

Upon contact with a target cell, Y. pestis bacteria use a needle apparatus, characteristic of

T3SS, to inject Yop (“Yersinia outer proteins”) virulence proteins across the bacterial inner and outer membranes and across the cell membrane of the eukaryotic cells they infect (Marenne et al., 2004; Torruellas et al., 2005). The Yop eﬀector proteins assist in the progression of

Yersinia infection by disrupting fundamental signaling pathways leading to interruption of phagocytosis, inhibition of inflammation, and interference with the cytoskeleton by binding and depolymerizing actin (Cornelis, 2002; Marenne et al., 2004). Studies using confocal microscopy with FRET technology revealed that Yop effector proteins are preferentially targeted to host immune cells, including dendritic cells, macrophages and neutrophils, which helps to explain the inability of the host to mount an effective immune response to Y. pestis (Marketon et al.,

2005).

The T3SS was ﬁrst discovered by the observation that growth of Y. pestis ceased upon culturing at 37◦C in a low-calcium environment (Yother et al., 1986; Skrzypek and Straley,

1995). This so-called “low calcium response” (LCR) was later determined to be the result of induction of expression of yop genes and other genes whose products are important for the

T3SS (Marenne et al., 2004; Cornelis, 2002; Torruellas et al., 2005). Apparently, a shift to 37◦C in the absence of calcium is a key trigger of virulence in Y. pestis. Expression of the T3SS is highly regulated at the transcriptional and translational levels, by gene products that include

YopN, TyeA, and LcrG (Marenne et al., 2004). Several of the Yop proteins are also part of 4 the needle apparatus, including LcrV. LcrV represents the yersinial V antigen that appears to possess anti-inﬂammatory activity and is required for the low-calcium response because of its role in the secretion of YopB and YopD (Brubaker, 2003; Sarker et al., 1998; Skrzypek and Straley, 1995). The LcrV protein is signiﬁcant in that antibodies against this protein are protective in animal models (Wang et al., 2004; DeBord et al., 2006; Chichester et al., 2009).

Other Yops play roles other than as eﬀectors of signal transduction pathways. For example,

YopE and YopH, along with the F1 antigen encoded on the pFra plasmid (Lindler et al., 1998), help Y. pestis avoid phagocytosis by macrophages by inhibiting receptor interactions with the phagocytes (Du et al., 2002).

In addition to its virulence properties, Y. pestis possesses a natural ecology that includes an insect vector, which further distinguishes it from its closest relatives. In nature, Y. pestis is picked up by fleas that feed on an infected animal. During flea infection, Y. pestis forms a biofilm that blocks the proventriculus of the flea, impeding passage of the flea’s blood meal into the midgut for digestion (Jarrett et al., 2004; Hinnebusch, 2004). The starving fleas attempt to feed with greater and greater frequency and, upon feeding, regurgitate their previous blood meal

ﬁlled with Y. pestis bacilli into their mammalian hosts, hence spreading the infection (Jarrett et al., 2004; Hinnebusch, 2004). When new ﬂeas feed on the infected mammal, they also acquire the bacteria, which subsequently block their digestive tract, and the cycle continues.

As might be expected for a bacterial pathogen with a complex transmission cycle, expression of virulence factors is highly regulated. As just described, regulation of the T3SS by the

LCR mimics environmental cues that signal the transition from the ﬂea vector to mammalian host, where growth temperature appears to play an important role. This observation has been exploited as an alternative strategy to identify new virulence factors and other genes important for ﬂea-borne transmission of Y. pestis. For example, microarray studies have been conducted to identify genes induced by changes to the growth environment of Y. pestis.A change in growth temperature was shown to induce expression of T3SS genes, in addition to other virulence factors, including the F1 capsule, pH 6 antigen genes, hemin storage system, and catalase-peroxidase genes (Han et al., 2004; Motin et al., 2004). Several other genes whose function in virulence is not understood were also found to be temperature regulated in Y. pestis, 5 including a proposed type VI secretion system and the hypothetical protein YPO1996 (Han et al., 2004; Motin et al., 2004; Parkhill et al., 2001).

In addition to environmental temperature, quorum sensing has also been investigated for its potential significance in Y. pestis gene expression. Quorum sensing is the process by which bacteria sense and respond to the production of autoinducing chemical signals in a population- dependent manner, often leading to the production of virulence factors, biofilm formation, and other cooperative cell processes (Davies et al., 1998; Waters and Bassler, 2005). In Gram negative bacteria, the AI-1 systems originally identified in Vibrio fischeri commonly respond to the production of acyl-homoserine lactones that have species-specific varying side chain struc- tures (Waters and Bassler, 2005). Y. pestis possesses two of these systems, Ysp and Yps, that respond to N -(3-oxooctanoyl)-L-homoserine lactone, N -(3-oxohexanoyl)-L-homoserine lactone, N -hexanoyl homoserine lactone, and N -octanoyl homoserine lactone (Kirwan et al., 2006; Yu et al., manuscript submitted). Many bacteria, including Y. pestis, also have an AI-2 LuxS system that may act as a universal bacterial communication system (Schauder et al., 2001;

Xavier and Bassler, 2003).

Analysis of the data from several microarray studies (Han et al., 2004; Motin et al., 2004) led us to identify the maltose (mal) regulon of Y. pestis to be diﬀerentially regulated by multiple signaling pathways, including both growth temperature and quorum sensing. mal gene expression is under control of the AHL quorum sensing system, as evidenced by studies with Y. pestis mutants deleted for yspIR and ypsIR. In contrast to most virulence factors, however, transcription of the mal genes increases at lower temperatures (26–30◦C) as compared to higher growth temperatures (37◦C) (Han et al., 2004; Motin et al., 2004). It was further observed that Y. pestis mutants defective in quorum sensing were unable to grow on maltose as the sole carbon source, leading to the proposal that quorum sensing in Y. pestis may be used to regulate genes for the metabolism of suboptimal carbon sources (LaRock et al., manuscript submitted). These authors proposed that quorum sensing may be used in the ﬂea portion of the

Y. pestis life cycle where bacteria grow to a quorum in a nutrient-rich environment and then increase gene expression for the metabolism of other sugars as glucose becomes depleted. The interplay between quorum sensing and maltose metabolism includes additional features that 6 are not currently understood. For example, Yu et al. observed that while mal gene expression increased upon addition of AI-2 and AHLs together, expression decreased when AHLs alone were added, suggesting that the distinct QS pathways work in concert in undeﬁned ways.

Furthermore, addition of AHLs elevated mal gene expression at 30◦C, but not at 37◦C, revealing a role for temperature in the function of QS pathways (Yu et al., manuscript submitted).

The mal regulon has been most well characterized in E. coli, and includes malQPT, malEFG, malK lamB malM, and malS (Boos and Shuman, 1998; Boos and B¨ohm, 2000). These genes encode proteins important for the uptake and metabolism of maltose (Boos and Shuman, 1998), as well as enzymes that play a role in the conversion of maltose into glucose and glucose α-1- phosphate (Boos and B¨ohm, 2000).

The mal genes are regulated by MalT, a transcriptional activator that responds to the presence of maltose and maltodextrins to regulate the expression of these genes (Richet and

Raibaud, 1987; Boos and Böhm, 2000). ATP and maltotriose are necessary for this activation (Richet and Raibaud, 1989), and some of the mal operons are also controlled by catabolite repression (Chapon, 1982; Chapon and Kolb, 1983; Boos and Böhm, 2000). MalT is a member of the “STAND” (Signal Transduction ATPases with Numerous Domains) family of transcription factors, which combine input from multiple regulatory systems and transmit a new signal of their own (Marquenet and Richet, 2007; Danot, 2010). STAND proteins are monomeric at rest and self-associate with the addition of (d)ATP (Marquenet and Richet, 2007). They contain a nucleotide organization domain (NOD) that acts as an on-off switch and sensor and effector domains (Danot, 2010). For MalT, the binding of maltotriose in the sensor domain in the presence of ATP promotes the shift to active form and allows MalT to bind cooperatively to its targets (Marquenet and Richet, 2007; Danot, 2010) to promote transcription of mal genes.

Along with the structural genes for maltose metabolism, malF and malG encode components of an ATP-dependent transport complex for the uptake of maltose from the periplasmic space. In this complex, MalK hydrolyzes ATP to facilitate maltose transport (Boos and B¨ohm,

2000). MalK can also serve a regulatory role as elevated levels of the protein inhibit expression of the mal regulon by aﬀecting the binding of MalT (Reyes and Shuman, 1988; Panagiotidis et al., 1998; Joly et al., 2004; Richet et al., 2005). Another well-characterized component of 7 maltose utilization is the product of the lamB gene, which encodes a maltoporin protein localized to the outer membrane (Bavoil et al., 1980). In E. coli, LamB also serves as the receptor for bacteriophage λ (Gehring et al., 1987).

Interestingly, while mal genes play a universal role in metabolism of maltose, these genes may also be related to virulence in other bacteria related to Y. pestis, including Vibrio cholerae and Yersinia enterocolitica (L˚anget al., 1994; Brzostek and Raczkowska, 2001). In these studies, the addition of maltose to V. cholerae cultures inhibited the extracellular secretion of cholera toxin and soluble hemagglutinin-protease and increased the production of toxin-coregulated pilus (L˚anget al., 1994). Mutants of malF in V. cholerae were unable to synthesize or secrete cholera toxin (L˚anget al., 1994). Also, exposing Y. enterocolitica to maltose decreased the production of Yop proteins, leading to the proposal that this eﬀect was related to the interaction of the maltose transport system and the Yop secretion system (Brzostek and Raczkowska, 2001).

Because of the ill-defined role that maltose metabolism has on virulence in other bacterial pathogens, and as a result of the multiple levels of regulation that include temperature and quorum sensing, we have taken a genetic approach to better understand the nature of mal gene regulation in Y. pestis. In performing these studies, we identified the need for improved genetic tools to genetically manipulate Y. pestis. Consequently, we have improved the methods to construct Y. pestis mutants that do not require permanent addition of antibiotic resistance markers, and developed an efficient pipeline to clone and express genes in single copy. While these methods have been applied for the study of quorum sensing and temperature regulation, they should prove useful for several other applications.

1.2 Thesis Organization

Following this introduction are three additional chapters describing my studies. Chapter2 discusses a new ligase-independent technique for the construction of lacZ gene fusions and recombinant plasmids generated from PCR products. For this, we took advantage of the activity of the nicking endonuclease Nt.BspQI to generate relatively long single-strand tails of

DNA that eﬃciently anneal with complementary sequences representing PCR products. Using a combination of homologous recombination and recombinant DNA, we constructed vectors 8 with appropriately positioned Nt.BspQI recognition sites that facilitate the quick and eﬃcient generation of lacZ fusions, as well as direct cloning of PCR products. Using these plasmids we constructed several lacZ gene fusions using promoters from both Y. pestis and E. coli. In addition, we cloned malT from both Y. pestis and E. coli directly from PCR products.

Chapter3 describes the development of improved methods for constructing new Y. pestis strains expressing lacZ gene fusions and their use to study mal gene regulation. For this, we relied on recombineering technology that takes advantage of the highly eﬃcient Red homologous recombination system of bacteriophage λ. The use of bacteriophage λ Red genes have revolutionized genetic engineering in many species of bacteria (Court et al., 2002; Datsenko and Wanner, 2000; Datta et al., 2006; Derbise et al., 2003; Gerlach et al., 2007; Murphy and

Campellone, 2003; Poteete, 2001; Yu et al., 2000), and here I show its value in genome engineering of Y. pestis. Recombineering allows for the production of targeted gene knockouts using a Red helper plasmid that expresses the λ genes exo, beta, and gam under arabinose

(araBAD) control (Datsenko and Wanner, 2000). These genes allow linear PCR products to recombine with regions that possess high homology to the PCR product, which usually includes a selectable antibiotic marker with long ends matching the gene to be deleted (Copeland et al.,

2001). To eliminate the antibiotic markers from recombineered strains, we constructed plasmids expressing the site-specific recombinase Flp that are based on the origin of replication of bacteriophage λ (Boyd and Sherratt, 1995). Flp functions to eliminate the antibiotic resistance cassette used for selection of recombinants by allowing the deletion of the cassette through recombination at specific sites, termed FRT sites (Cherepanov and Wackernagel, 1995). This plasmid serves the dual function of forcing out pSC101-derivative plasmids by an unknown mechanism of incompatibility (Ekaterinaki and Prentki, 1991). We effectively used this system while constructing multiple gene deletions in Y. pestis, including malT mutants that were in- strumental in understanding the role of this activator protein in temperature-dependent gene regulation.

We also used recombineering in combination with our ligation-independent cloning method to construct lacZ fusions under control of selected bacterial promoters and express them in single copy from F0 lac-pro elements. The new F0 factors are easily introduced to Y. pestis by 9 conjugation for genetic analysis of the promoters expressed at physiologically relevant levels.

This system also takes advantage of E. coli auxotrophic for diaminopimelic acid (DAP) (De- marre et al., 2005) as a conjugation donor that can be used with no requirement for antibiotic selection, minimizing the number of steps of manipulating and culturing Y. pestis.

Using these new tools, we studied the eﬀects of quorum sensing and temperature on the expression of mal genes in Y. pestis. To characterize gene expression, we performed β- galactosidase assays on bacteria expressing lacZ from the malK-lamB promoter (PmalK) to conﬁrm this operon is temperature regulated in Y. pestis (Han et al., 2004; Motin et al., 2004), but not in E. coli K-12. In a Y. pestis background, expression of PmalK-lacZ fusions was temperature regulated, while the same fusion was expressed independently of temperature in an

E. coli background. Because MalT functions as a transcription factor of mal genes (Richet and Raibaud, 1987; Boos and B¨ohm, 2000), we tested the hypothesis that temperature regulation was mediated through this activator. While a malT-lacZ fusion was not temperature regulated, we did observe a key role for functional MalT in thermoregulation of the mal genes in Y. pestis. In combination with malT knockout mutants, we observed that expression of

MalT from Y. pestis conferred temperature-dependent expression of the PmalK-lacZ fusions in Y. pestis, but not E. coli. In contrast, expression of the same fusion was altered at a much lower level by growth temperature in strains when MalT from E. coli was expressed. However, we observed only minimal effects of quorum sensing on mal gene expression indicating temperature is the most significant environmental condition that affects expression of the mal genes.

Chapter4 is a brief summary of my accomplishments and suggestions for future studies.

In conclusion, the tools that we have developed have streamlined several genetic and molecular manipulations used to characterize gene expression in Y. pestis. We have applied these tools to examine temperature and QS regulation in Y. pestis leading to the determination that the decreased expression of genes in the mal regulon at 37◦C is mediated in part by MalT.

While this effect is specific to Y. pestis, future studies could determine the significance of MalT- mediated gene regulation in other bacterial pathogens. The genetic tools developed here will also be useful for these studies. 10

CHAPTER 2. VECTORS FOR LIGATION-INDEPENDENT CONSTRUCTION OF lacZ GENE FUSIONS AND CLONING OF PCR PRODUCTS USING A NICKING ENDONUCLEASE

A paper published in Plasmid

Carrie J. Oster1 and Gregory J. Phillips2

Abstract

Several ligation-independent cloning methods have been developed that oﬀer advantages for construction of recombinant plasmids at high eﬃciency while minimizing cloning artifacts. Here we report new plasmid vectors that use the nicking endonuclease Nt.BspQI to generate extended single stranded tails for direct cloning of PCR products. The vectors include pLacCOs1, a

ColE1-derivative plasmid imparting resistance to ampicillin, which allows facile construction of lacZ translational fusions, and pKanCOs1, a pSC101-derivative cloning vector that imparts resistance to kanamycin, for cloning of PCR amplicons from genomic DNA as well as from ampicillin-based plasmids. We have successfully used these plasmids to directionally clone and characterize bacterial promoters that exhibit temperature regulated expression, as well as for cloning a variety of PCR products. In all cases, constructs with the correct configurations were generated at high efficiency and with a minimal number of manipulations. The cloning vectors can also be easily modified to incorporate additional reporter genes or to express epitope-tagged gene products.

1Graduate student, Molecular, Cellular, and Developmental Biology Program, Iowa State University; primary researcher and author. 2Professor, Department of Veterinary Microbiology and Preventive Medicine, Iowa State University; corre- sponding author. 11

2.1 Introduction

Gene fusions between promoters and reporter genes are a fundamental tool for the genetic analysis of bacterial gene expression and protein localization (Hand and Silhavy, 2000; Hughes and Maloy, 2007; Silhavy, 2000; Silhavy and Beckwith, 1985; Slauch and Silhavy, 1991). Multi- ple reporter genes have been described with lacZ (encoding β-galactosidase) (Hand and Silhavy,

2000; Silhavy and Beckwith, 1985), gfp (green ﬂuorescent protein) (Phillips, 2001; Southward and Surette, 2002), lux (luciferase) (Engelbrecht et al., 1985; Stewart and Williams, 1992), and phoA (alkaline phosphatase) (Manoil et al., 1990) being among the most widely used.

Promoter fusions are frequently constructed using standard techniques of recombinant DNA, which includes the use of restriction endonucleases to generate DNA fragments that are joined by T4 DNA ligase. In constructing multiple gene fusions, however, this approach can be time- consuming and subject to artifacts inherent to recombinant DNA methodology. In response, a variety of ligation-independent cloning methods have been reported to improve the efficiency of constructing recombinant molecules (Aslanidis and de Jong, 1990; Huan et al., 1992; Li and Evans, 1997; Quan and Tian, 2011; Sharan et al., 2009; Yang et al., 2010), including specialized plasmid vectors for purifying epitope tagged proteins (Bardóczyet al., 2008; Berrow et al., 2007; Chanda et al., 2006; Curiel et al., 2011; Du et al., 2011). Multiple strategies have been developed, such as use of homologous or site-specific recombination in vivo (Sharan et al.,

2009; Zhang et al., 2002) or enzymatic manipulation (PCR, T4 DNA polymerase, or uracil-DNA glycoslase for example) to join DNA molecules in vitro (Quan and Tian, 2011; Tachibana et al.,

2009). More recently, nicking endonucleases that cut only a single strand of a double-stranded

DNA recognition site (Chan et al., 2011; Jeltsch et al., 1996; Zhang et al., 2010; Zheleznaya et al., 2009) have been used to generate relatively long complementary single-stranded tails that can be eﬃciently linked in the absence of T4 DNA ligase (Vroom and Wang, 2008; Yang et al., 2010; Du et al., 2011), as well as to study a model DNA replication fork (Ishikawa et al.,

2009, 2011).

We reasoned that this approach could also be used to eﬃciently construct lacZ reporter gene fusions. To demonstrate this, we modiﬁed a plasmid carrying a promoterless lacZY 12

Figure 2.1 Strategy for construction of lacZ gene fusions using the nicking endonuclease Nt.BspQI. Top. The site of insertion into pLacCOs1 is shown, including the StuI restriction enzyme recognition site (shown in bold) for linearizing the vector and the two Nt.BspQI recognition sites (shown in bold italics). The arrows indicate the positions where the DNA strands are cut by each endonuclease. Bottom. The compatible single-stranded overhangs, generated as described in Section 2.2, of both vector and PCR product are shown. The location of a promoter (P) is shown within the PCR product.

operon (Jain, 1993) to enable direct fusion of PCR products in-frame with lacZ. For this, we positioned a recognition site for the nicking endonuclease Nt.BspQI immediately upstream of lacZ. Nt.BspQI recognition sites were also incorporated into PCR primers used to amplify speciﬁc promoter regions from bacterial genomic DNA. Treatment of linearized vector DNA and PCR products with Nt.BspQI prior to heating the molecules resulted in single-stranded regions on both plasmid and insert with suﬃcient homology to generate recombinant plasmids without the need for T4 DNA ligase (Fig. 2.1).

Here we report the use of this strategy to construct a series of lacZ gene fusions to characterize and compare expression of genes from Yersinia pestis and Escherichia coli. Historically,

Yersinia pestis has been a significant human pathogen, and it remains a concern as a potential bioterrorism agent. Expression of several chromosomally-encoded Y. pestis genes is known to vary with temperature, which reflects its unique natural reservoirs that include both fleas and 13 mammalian hosts (Han et al., 2004; Motin et al., 2004). To better understand the genetic basis for temperature-regulated gene expression in Y. pestis, we sought to construct several lacZ gene fusions. The ligation-independent cloning system described here was found to be an efficient method to generate lacZ reporter gene fusions to multiple bacterial promoters. The cloning system is relatively simple in that PCR products are cloned directly into the vectors without the need for expensive enzymes, recombinases, or additional oligonucleotide linkers.

The success of this method further prompted us to construct a separate plasmid for direct cloning of PCR products. Further modiﬁcations to these plasmids can be made for additional applications such as creating fusions to alternative reporter genes and epitope tagging of PCR products.

2.2 Materials and Methods

2.2.1 Bacterial strains, growth media, and enzymes

Promoters were cloned from E. coli K-12 strains, including LMG194 (F− mcrA, ∆[mrr- hsdRMS-mcrBC ] Φ80lacZ ∆M15, ∆lacX74, recA1, araD139, ∆(ara-leu)7697, galU, galK, rpsL, endA1, nupG) (obtained from Invitrogen, Inc., Carlsbad, CA), or derivatives of Y. pestis strain

CO92 (Parkhill et al., 2001) deleted for either the pgm locus (∆pgm)(Fetherston et al., 1992) or pCD1, the large virulence plasmid required for pathogenesis (Lcr−)(Perry et al., 1998), enabling manipulation under BSL-2 laboratory conditions. E. coli NEB 5-alpha (fhuA2 ∆(argF- lacZ )U169 phoA glnV44 Φ80 ∆(lacZ )M15 gyrA96 recA1 relA1 endA1 thi-1 hsdR17 ) was obtained from New England Biolabs (NEB) (Ipswich, MA) and used as a host for all cloning experiments. Y. pestis mutants were grown on Tryptic Soy Blood Agar Base #2 (TSB) from

Difco (obtained through Thermo Scientiﬁc, Waltham, MA), while E. coli was cultured using

LB medium (Miller, 1992b). When appropriate, media were supplemented with antibiotics including ampicillin (Amp, 100 µg/ml), chloramphenicol (Cam, 20 µg/ml), or kanamycin (Kan,

30 µg/ml) (Sigma-Aldrich, St. Louis, MO).

All enzymes required for recombinant DNA, including the nicking endonuclease Nt.BspQI, restriction endonucleases, and T4 DNA ligase, were obtained from NEB. PCR was performed 14

Figure 2.2 pLacCOs1 and pKanCOs1 plasmids. Shown are sizes and relevant features of each plasmid. pLacCOs1: ColE1ori, ColE1 origin of replication; bla, β-lactamase imparting ampicillin resistance gene; ‘lacZ, β-galactosidase, which requires a promoter, ribosome binding site, and initiation codon for expression; lacY, lactose permease, which is expressed coordinately with lacZ. pKanCOs1; repA, replication initiation protein for the medium copy number, ColE1-like compatible, pSC101-derivative plasmid (Peterson and Phillips, 2008); kan, kanamycin resistance. Also shown for each vector are the locations of the Nt.BspQI and StuI recognition sites.

using polymerase master mixes that do not add additional bases on the 30 end of the amplification products, including Thermo Scientific Extensor Master Mix (Thermo Scientific, Waltham,

MA).

2.2.2 Construction of pLacCOs1 and pKanCOs1

To construct pLacCOs1 (Fig. 2.2), pLacZY2 (Jain, 1993) was ﬁrst modiﬁed by eliminating a unique SapI recognition site. The plasmid was digested with SapI (an isoschizomer of BspQI) and then treated with the Quick Blunting Kit (NEB) before recircularizing the vector with T4

DNA ligase.

To introduce recognition sites for Nt.BspQI, PCR was used to amplify a CamR cassette from pKD3 (Datsenko and Wanner, 2000) using primers BspQI-KD3.S and BspQI-KD3.AS 15

(Table 2.1), obtained from Integrated DNA Technologies (IDT, Coralville, IA). PCR was performed by an initial step of 2 min at 94◦C, followed by 30 cycles of: 94◦C for 30 s., 55◦C for 30 s, 72◦C for 1 min 30 s, and completed with a ﬁnal incubation at 72◦C for 10 min. The PCR product was cloned into pCR2.1 (Invitrogen, Carlsbad, CA), and colonies were selected for

Cam and Kan resistance. The resulting plasmid DNA was digested with HindIII and a 1.3-kbp fragment was purified using a QIAQuick Gel Extraction Kit (Qiagen, Valencia, CA) and cloned into pLacZY2(SapI◦) digested with the same enzyme. The resulting plasmid was then digested with StuI, which deleted the CamR cassette while maintaining the flanking Nt.BspQI sites, and religated yielding pLacCOs1 (Fig. 2.2). pKanCOs1 was constructed by modifying pJPK12 (Pe- terson and Phillips, 2008) in a similar manner. Plasmid DNA was first digested with EagI and religated to eliminate a unique BspQI site. The resulting vector was then digested with EcoRI and ligated to the 1.3-kbp CamR cassette flanked by BspQI recognition sites originating from the pCR2.1-derivative plasmid, as just described. Finally, the resulting plasmid was digested with StuI and religated to delete the CamR cassette, resulting in pKanCOs1. pKanCOs1 is compatible with other plasmids with ColE1-like replication origins, including pMB1 (pBR322 and pUC-derivative vectors) and p15A (pACYC177 and pACYC184 vectors).

2.2.3 PCR ampliﬁcation of promoters

PCR primers designed to amplify Y. pestis and E. coli promoters are shown in Table 2.1.

Each primer was designed to amplify a region of the bacterial chromosome containing a promoter of interest, as well as include a region of homology to pLacCOs1 and Nt.BspQI recognition sites, as summarized in Fig. 2.1. Genomic DNA was isolated from either Y. pestis CO92

(∆pgm or Lcr−) or E. coli LMG194 using the Epicentre MasterPure Complete DNA and RNA

Puriﬁcation Kit (Epicentre, Madison, WI) and used as a template for PCR. PCR products were puriﬁed using the QIAQuick Gel Extraction Kit.

2.2.4 Promoter cloning and characterization

To prepare pLacCOs1 for cloning, plasmid DNA was digested with StuI and nicked in a subsequent reaction with Nt.BspQI for 3 h at 50◦C. Puriﬁed PCR products were likewise 16 are shown by triplet lacZ TGTGTAGGCTGGAGCTGCTTCG CATATGAATATCCTCCTTAG CTC TAG CAG CTT ATT GGC TTG C GGC GAG AGT TAT TAG TAC GGG TAA G TGA CGG ACT TCT ACC AAT AAG CC AGG CAC AAT AAA CGC CTG CGA TG CGT TGG GCC ATT CAT CTT GGT TGT ACT CAC CGG ATG AAT GCG GTA CAA GGG TTG CAG AGT AAA GTC GAT AC CAA ATG CGC CAG TTG TAG CAC TTC CAT TCAAATCCTCTCGATATCAACGAGGTTATCATTCACTG CCCACGCGAAGTCATTATATAGC GACAGGATTTCAGTAGGGT GTTGAAATCGCCGATCAGGTATC AATGGCAGCTATGTTCCGCCATTC TTCTGTTTATGCACCGATGGCAGC CGTGGTTAACGCCGAAAGAGCC TCCGGATGCTCAACGGTGACTTTA GGGAACGAGATAAATACCCAGC CATGGCAGTCTTTCGACTGAGCCTTTC GAAACCCAACACCGGTAAATCC GGTTGTGACATAGGAGTTCCAC TTCCAGGTCGAATATCACAGC CATGGAGCTTAGTACGTTAGCCATGAGG GGATAATTTAACCAGCAGGCGGTC GAAAGAGGTGAAACGCTACCTG TGCCGCTTTCTTGCCGGTTGATCTTTC CTTATACTCTTTCCCTG ACTCTCGCAGGCCT TCGACATTCAGGCCT I sites are underlined. The regions of selected genes joined in-frame to Stu TG AAG AGC TG AAG AGC TG AAG AGC TG AAG AGC TG AAG AGC TG AAG AGC TG AAG AGC TG AAG AGC TG AAG AGC GGAAGAGC GGAAGAGC GGAAGAGC GGAAGAGC GGAAGAGC GGAAGAGC GGAAGAG GGAAGAGC GGAAGAG GGAAGAGC GGAAGAGC GGAAGAGC GGAAGAGC a TGAAGAGC TGAAGAGC TGAAGAGC TGAAGAGC ) 0 GCTCTTCA GCTCTTCC to 3 0 Table 2.1 PCR and DNA sequencing primers used in this study. QI recognition sites are shown in bold and the Bsp YpP0498-BspQI.SYpP0498-BspQI.ASYpPkatY-BspQ.S TCGACATTCAGG YpPkatY-BspQ.AS A CTC TCG CAGYPO1996P-BspQI.S G TCGACATTCAGG YPO1996P-BspQI.AS A CTC TCG TCGACATTCAGG YpPpsaC.BspQI.S CAG A G CTC TCGYpPpsaC.BspQI.AS CAG G YpPpsaE.BspQI.S TCGACATTCAGG AYpPpsaE.BspQI.AS CTC TCG CAGYpPmalK-BspQI.S G TCGACATTCAGG AYpPmalK-BspQI.AS CTC TCG CAGYpPlsrA-BspQI.S G TCGACATTCAGG A CTCYpPlsrA-BspQI.AS TCG CAG G EcPmalK-BspQI.S TCGACATTCAGG EcPmalK-BspQI.AS A CTC TCGEcPrpoHP3.-BspQI.S CAG G TCGACATTCAGG AEcPrpoHP3-BspQI.AS CTC TCG TCGACATTCAGG CAG G YplsrR-BspQI.S A CTC TCG CAGYplsrR-BspQI.AS G TCGACATTCAGG ACTCTCGCAGG PrimerBspQI-KD3.SBspQI-KD3.AS AAGCTTGCA TAGATCTTC Sequence (5 bla-BspQI.S TCGACATTCAGG bla-BspQI.AS ACTCTCGCAGG YpmalT.BspQI.S TCGACATTCAGG YpmalT.BspQI.AS ACTCTCGCAGG malT-Ec-BspQI.S TCGACATTCAGG malT-Ec-BspQI.AS ACTCTCGCAGG LacZRev2 ACAAACGGCGGATTGACCGTAATG Nt. a spacing within the appropriate antisense (AS) primer sequences. 17 treated with Nt.BspQI and both vector and insert were then incubated at 65◦C for 5 min.

DNA was immediately treated using the PCR puriﬁcation/DNA concentration protocol from the Qiagen QIAExII Gel Extraction Kit. The DNA was quantiﬁed using a Nanodrop ND-

1000 Spectrophotometer. The vector and insert were mixed at a 3:1 ratio in a 20 µl reaction with NEB Buffer 2 (50 mM NaCl, 10 mM TrisHCl, 10 mM MgCl2, 1 mM dithiothreitol), containing 100× bovine serum albumin (NEB) at a final concentration of 1×. The reaction was incubated at 65◦C for 10 min and then allowed to cool slowly to room temperature. DNA was then transformed directly into NEB 5-alpha Competent E. coli (NEB) and plated on LB agar containing Amp and spread with ∼100 µl of 10 mg/ml 5-bromo-4-chloro-3-indolyl-β-D- galactopyranoside (X-gal) (Gold Biotechnology, Inc., St. Louis, MO). Blue colonies were chosen for plasmid extraction and sequencing. All lacZ fusions were confirmed by DNA sequencing using LacZRev2 primer (Table 2.1) at Iowa State University’s DNA Facility.

To study gene expression, plasmid DNA was introduced to E. coli K-12 strains or Y. pestis

(∆pgm or Lcr−) by electroporation (Murphy and Campellone, 2003). β-galactosidase assays were performed as described (Miller, 1992b). DNA sequences of pLacCOs1 and pKanCOs1 have been deposited in GenBank with accession numbers JF837312 and JF837313, respectively.

2.3 Results and Discussion

2.3.1 Development of a nicking endonuclease system for lacZ gene fusions

Development of a more efficient cloning system was prompted by our efforts to clone and characterize several promoters of Y. pestis that are differentially regulated by growth temperature. A survey of published microarray experiments revealed multiple chromosomally encoded genes whose transcription was elevated upon a shift from 30◦C to 37◦C, including katY, YPO1996 and psaAC, as well as genes whose expression was reduced at 37◦C, including

YPO0498 and the mal operon (Han et al., 2004; Motin et al., 2004). Our approach to this end was to construct plasmid-borne lacZ translational gene fusions representing selected genes from both Y. pestis and E. coli and measure β-galactosidase activity in both bacterial species. 18

To facilitate construction of multiple lacZ gene fusions, we developed a system that allowed direct cloning of PCR products, representing promoters and their ﬂanking regions, directly in-frame with lacZ without the use of ligase and conventional restriction endonucleases. The strategy, as summarized in Fig. 2.1, uses pLacCOs1 (Fig. 2.2) to generate a cloning vector with extended single stranded DNA overhangs that easily assemble into desired recombinant molecules in vitro. Since the sequence of each overhang is distinct, inserts are cloned in the proper orientation for direct in-frame fusion with the lacZ reporter gene. Lactose permease

(lacY ) is also expressed from this plasmid, enabling genetic selections for altered promoter activity (Shuman and Silhavy, 2004).

2.3.2 Construction of lacZ gene fusions

To construct lacZ gene fusions using pLacCOs1, selected promoters were PCR ampliﬁed using primer pairs with homology to the bacterial chromosome in addition to Nt.BspQI recognition sites (Table 2.1 and Fig. 2.1). Antisense primers were designed so that when inserted into pLacCOs1 a translational fusion was made between the ﬁrst few amino acids of the target gene and lacZ (Table 2.1). pLacCOs1 was prepared for cloning by linearization with StuI before digestion with Nt.BspQI to generate DNA molecules with two nicks on opposite DNA strands. Both plasmid DNA and PCR products were then heated to 65◦C to yield molecules with relatively long single-stranded overhangs.

To generate recombinant plasmids, both linearized vector and PCR products were puriﬁed and mixed, as described in Section 2.2. The DNA mixtures were then used to directly transform chemically competent E. coli with selection on LB Amp plates containing X-gal. Several hundred transformants were typically obtained from each cloning reaction and 75–95% of the colonies were blue on X-gal. In each case, all of the blue colonies tested (typically 4–6) were shown by DNA sequencing to encode the predicted construct with the target promoter driving expression of a lacZ translational fusion.

In constructing gene fusions with this method, the PCR amplicons were typically 0.5–1.5 kbp, and all eﬃciently yielded gene fusions. In constructing a fusion between the E. coli heat shock promoter rpoHP3 and lacZ, however, we designed primers that yielded only a very short 19

Figure 2.3 Temperature regulation of mal genes as measured by lacZ gene fusions expressed from pLacCOs1. Translational fusions between the malEFG promoter, from both Y. pestis and E. coli K-12, and lacZ were constructed in pLacCOs1 as described in Section 2.2. Each gene fusion was expressed in both Y. pestis (∆pgm or Lcr−) and E. coli K-12 as shown. β-galactosidase activity (Miller units) was measured in bacterial cultures grown at 30◦C and 37◦C, revealing that mal gene expression was temperature regulated only in Y. pestis.

(∼200 bp) PCR product. In this case, cloning was successful with only a PCR clean up reaction without gel elution.

2.3.3 Reporter gene expression

To test the utility of the lacZ gene fusions, we tested each for temperature-dependent gene regulation in both Y. pestis and E. coli. While a more complete analysis of these studies will be reported elsewhere, results for mal operon expression are shown in Fig. 2.3. We observed that while expression of the mal genes from Y. pestis or E. coli showed no change in regulation at 37◦C when expressed in E. coli, expression of the loci was signiﬁcantly reduced at 37◦C when expressed in Y. pestis. While the basis for temperature-dependent gene expression of the mal genes in Y. pestis is not yet clear, the plasmids constructed in this study are currently being used to identify the genetic basis for this regulation. We are especially interested in the potential role of the transcriptional activator MalT in regulating gene expression in response to growth temperature. 20

2.3.4 PCR amplicon cloning

Because of the eﬃciency at which lacZ gene fusions were made using pLacCOs1, we also constructed a vector to allow direct cloning of any PCR product. pKanCOs1 (Fig. 2.2) was made by modifying pJPK12, a pSC101-derivative cloning vector that replicates with elevated copy number and imparts KanR (Peterson and Phillips, 2008). This vector oﬀers advantages for cloning PCR products in that the replicon is compatible with ColE1 plasmids, allowing recombinants to be immediately transformed into strains with compatible plasmids. Also, since pKanCOs1 imparts KanR, use of this plasmid eliminates background colonies that typically arise when cloning PCR products generated from AmpR-based plasmid templates.

For an initial test, we PCR amplified bla from pUC18 and cloned the resulting AmpR gene cassette into pKanCOs1 using the Nt.BspQI nicking endonuclease cloning system. Transfor- mants were first selected for KanR and then tested for AmpR. The cloning efficiency was similar to that of pLacCOs1 with 70–90% of the KanR transformants also conferring AmpR.

These eﬃciencies were consistently obtained with several other genes from E. coli and Y. pestis, including malT, with a length of over 3-kbp.

In selecting genes for cloning, including malT, we observed that BspQI sites were occasionally found within the desired regions. Nevertheless, the presence of these sites did not interfere with cloning eﬃciency using the methods described here. In all of these cases, however, the

BspQI site was greater than a few hundred base pairs from the end of the PCR product. In the case where it is not possible to avoid a BspQI site near the end of the PCR amplicon, we predict that the denaturing step to generate single stranded overlapping regions could be lowered to minimize denaturation of the PCR products. It should also be possible to build similar cloning vectors that contain recognition sites for other commercially available nicking enzymes.

2.3.5 Conclusions

We have further developed a ligation independent cloning method that uses nicking endonucleases for construction of recombinant plasmids, including lacZ gene fusions. We found 21 this method to be a reliable approach that requires relatively few steps to progress from PCR amplification to confirmation of recombinants by DNA sequencing. Although the method was specifically developed for construction of lacZ translational gene fusions, pLacCOs1 can easily be modified to also construct lacZ transcriptional fusions. Since the cloning system also allows for directional insertion of PCR products, it is ideal for generating additional reporter gene fusions, or adding epitope tags to cloned genes. It should be also be well suited for applications that require construction of a large number of recombinant plasmids, such as promoter libraries (van Dyk et al., 2001; Zaslaver et al., 2008), or cloning and expressing gene products for structural analysis (Betton, 2004). As additional nicking enzymes become commercially available, new vectors can be constructed to expand the choices of enzymes used constructing recombinant plasmids.

Acknowledgements

This work was funded by grants from the Oﬃce of Naval Research (ONR-MURI N00014-

06-1-1176) and NIH (NIAID R21 AI074591). 22

CHAPTER 3. TEMPERATURE-REGULATED EXPRESSION OF mal GENES IN Yersinia pestis IS MEDIATED BY THE TRANSCRIPTIONAL ACTIVATOR MalT, AS REVEALED BY NEW METHODS FOR CONSTRUCTION OF SINGLE-COPY lacZ GENE FUSIONS

A paper to be submitted to Journal of Bacteriology

Carrie J. Oster1 and Gregory J. Phillips2

Abstract

As Yersinia pestis progresses through its unique infection cycle, it routinely experiences a change in temperature from the ambient temperature (26–30◦C) of a ﬂea vector to the temperature of a mammalian host (37◦C). In contrast to several known virulence genes, expression of the mal regulon is signiﬁcantly reduced at 37◦C when compared to 30◦C. To better understand the mechanism of this regulation we have developed a system to use ligation-independent cloning and recombineering to construct lacZ fusions expressed in low copy number from new

F0 lac elements. The F0 lac factors are easily introduced to Y. pestis by conjugation to express the gene fusions at physiologically relevant levels. This system has been used to compare the regulation of mal genes in Y. pestis with that of E. coli. We have identiﬁed MalT, the positive regulator of mal transcription, as a key component of temperature-dependent regulation in

Y. pestis.

3.1 Introduction

Y. pestis, the causative agent of bubonic plague, is well recognized as a highly virulent bacterial pathogen. Y. pestis is also distinct from other species of Yersinia in that infection of a mammalian host is accomplished through the bite of a ﬂea vector (Hinnebusch, 2004, 2005;

Perry and Fetherston, 1997). Studies reveal that Y. pestis can facilitate this transmission step by blocking the ﬂea’s proventriculus through formation of bioﬁlms (Darby et al., 2002;

Hinnebusch, 2004; Jarrett et al., 2004). Fleas “blocked” in this manner are unable to digest their blood meals, as the blood cannot reach the midgut. As a result, the insects begin to starve and attempt to feed on mammals more frequently until the blockage is regurgitated and the infection is spread (Hinnebusch, 2004; Jarrett et al., 2004).

Bacterial passage from ﬂea to mammal triggers the expression of several genes, including those for the virulence factors responsible for its extreme pathogenesis in mammalian hosts that is associated with Y. pestis. For example, genes encoding components of the type III secretion system (T3SS) have been shown to be induced in the mammalian host (Marenne et al., 2004; Skrzypek and Straley, 1995; Torruellas et al., 2005; Yother et al., 1986). In addition to formation of the needle complex that delivers eﬀector proteins, the products of these genes disrupt the normal functioning of the host immune cells in part by interfering with signal transduction pathways (Finlay and McFadden, 2006; Marenne et al., 2004). The result of this coordinated gene expression includes evasion of the host immune system by the neutralization of phagocytes (Finlay and McFadden, 2006).

A key environmental signal that triggers changes in gene expression in Y. pestis is the temperature shift associated with the transition from the ambient temperature of the ﬂea

(≈26–30◦C) to ≈37◦C in the mammal. The effect of temperature on gene expression has been investigated by multiple microarray studies, which consistently reveal that several virulence factors, including plasmid-encoded T3SS genes as well as chromosomally encoded genes, are differentially regulated by growth temperature (Han et al., 2004; Motin et al., 2004). In addition to known virulence factors, microarray studies showed that the genes comprising the mal (maltose utilization) operon are among the most highly regulated genes responding to 24 temperature (Han et al., 2004; Motin et al., 2004). In contrast to the T3SS genes, however, expression of the mal genes is highest at 30◦C and significantly decreased at 37◦C(Han et al.,

2004; Motin et al., 2004).

The genes for maltose utilization in Y. pestis closely resemble those of E. coli, and include malK lamB malM and malEFG, which are divergently transcribed from a pair of promoters lo- cated between base pairs 4,147,316 and 4,147,896 on the Y. pestis CO92 chromosome (Parkhill et al., 2001). Y. pestis can utilize maltose as a carbon source, indicating that the genes are fully functional (Zhou et al., 2004). In E. coli, mal genes are regulated by MalT, which acts as a transcriptional activator in the presence of maltose and maltodextrins plus ATP (Boos and B¨ohm, 2000; Boos and Shuman, 1998; Richet et al., 2005). MalT is a member of the

STAND (Signal Transduction ATPases with Numerous Domains) family of transcriptional activators that compile multiple signals from diﬀerent regulatory systems throughout the bacterial cell and in turn transmit new signals to alter gene expression (Danot, 2010; Marquenet and

Richet, 2007). STAND activators generally exist as monomers at rest, but with the addition of (d)ATP they self-associate to bind speciﬁc targets, with Nucleotide Organization Domains

(NODs) acting as on-oﬀ switches for this function (Danot, 2010). In the case of MalT, the binding of maltotriose encourages conversion to a multimeric active form that can promote the transcription of the mal genes (Danot, 2010; Marquenet and Richet, 2007).

Interestingly, in addition to temperature regulation, in Y. pestis the mal genes also appear to be regulated by quorum sensing (QS) (Yu et al., manuscript submitted). QS is characterized by cell-to-cell communication that can lead to population-dependent regulation of genes in multiple bacteria. When bacterial cells reach a critical mass, sufficient levels of QS molecules, which include acyl-homoserine lactones and furanone derivatives, are produced. In response, bacteria differentially regulate genes important for a variety of cellular processes, including virulence, biofilm production, motility, and carbon source utilization (Davies et al., 1998; Waters and Bassler, 2005). In Y. pestis, three quorum sensing systems have been identified, including a

LuxS (AI-2)-dependent system as well as two distinct AHL systems, Ysp and Yps. Recent studies indicate that Y. pestis possesses QS systems responsive to the AHLs N -(3-oxooctanoyl)-L- homoserine lactone, N -(3-oxohexanoyl)-L-homoserine lactone, N -hexanoyl homoserine lactone, 25 and N -octanoyl homoserine lactone (Kirwan et al., 2006). Expression of the Y. pestis mal genes appears to be under control of the AHL QS system (LaRock et al., manuscript submitted; Yu et al., manuscript submitted).

The induction of the Y. pestis T3SS genes by temperature and the mechanism of yop gene expression can be explained biologically. Translation of the transcription factor VirF, encoded by the large virulence plasmid of Y. pestis, appears to be elevated by increased temperature as a result of localized unfolding of the 50 untranslated region of the mRNA (Waldminghaus et al., 2007). In turn, VirF induces expression of many of the T3SS genes, including the yop genes (Marenne et al., 2004). The product of the chromosomal gene ymoA also participates in upregulating yop genes at 37◦C by functioning as a modulator of supercoiling that may work to make promoters more accessible to VirF (Cornelis, 2002; Marenne et al., 2004; Torruellas et al.,

2005). In contrast, how mal gene expression, as well as the expression of other chromosomal genes, responds to temperature is unknown.

In order to approach these fundamental questions about how and why mal genes are regulated by temperature and quorum sensing in Y. pestis, we have taken a genetic approach to help identify genes whose products are necessary for mal regulation in Y. pestis. In performing these studies we noted a dearth of genetic tools for gene expression studies in Y. pestis. To overcome this limitation, we developed a method to construct lacZ gene fusions for expression at physiologically relevant levels from a very low copy number plasmid. We have also developed improved methods for recombineering (Court et al., 2002) to assess the role of MalT in temperature and QS regulation of the mal genes. Here we report signiﬁcant diﬀerences between the role of MalT in gene regulation in Y. pestis in comparison to E. coli.

3.2 Materials and Methods

3.2.1 Bacterial strains, growth media, and enzymes

Bacterial strains used in this study are shown in Table 3.1. The E. coli K-12 strain LMG194 was used for comparison with the attenuated Y. pestis CO92 derivatives R88 and R114, which were cultured under BSL-2 laboratory conditions. E. coli NEB 5-alpha was used as a host for 26 recombinant DNA and propagation of plasmids, and CSH100 was used for single copy gene expression.

E. coli was grown on LB medium (Miller, 1992a), while Y. pestis was grown on Tryptic

Soy Blood Agar Base No. 2 (TSB) (Thermo Scientiﬁc, Waltham, MA) or in Heart Infusion broth (HI) (Remel, Thermo Fisher Scientiﬁc, Waltham, MA). MacConkey agar base (Difco,

Thermo Fisher Scientific, Waltham, MA) supplemented with maltose at a final concentration of 30 mM (Fisher Scientific, Waltham, MA) was used to test strains for their ability to use maltose. Antibiotics used included ampicillin (Amp, 100 µg/ml), kanamycin (Kan, 30 µg/ml), or chloramphenicol (Cam, 20 µg/ml), (Sigma-Aldrich, St. Louis, MO). All enzymes for recombinant DNA were obtained from New England Biolabs (NEB, Ipswich, MA). 5-bromo-4-chloro-3- indolyl-β-D-galactopyranoside (X-gal) was obtained from Gold Biotechnologies, St. Louis, MO and L-arabinose from Sigma Aldrich, St. Louis, MO.

3.2.2 Construction of malT deletion mutants in E. coli and Y. pestis

E. coli and Y. pestis ∆malT mutants were constructed by recombineering (Datsenko and

Wanner, 2000). For E. coli, primers Ec-malT-KD3.S and Ec-malT-KD3.AS (Table 3.1) were used to amplify a CamR cassette from pKD3 (Datsenko and Wanner, 2000), and introduced to the chromosome of LMG194 using pRedET (Gene Bridges, Heidelberg) to express the Red recombinases of bacteriophage λ. For Y. pestis, primers Yp-malT-KD.S and Yp-malT-KD.AS

(Table 3.1) were used to generate a PCR product using pKD4 (Datsenko and Wanner, 2000) as a template. The recombineering protocol (Datsenko and Wanner, 2000) was modiﬁed for

Y. pestis to double the length of the initial growth phase and the cells were allowed to recover for at least 6 hours after electroporation before plating. After purifying antibiotic resistant recombinants from both E. coli and Y. pestis, the drug markers were removed by site-specific recombination using pCLIP-flp, as described below. The malT deletion mutants were confirmed by streaking recombinants on maltose-MacConkey agar; Mal− colonies appeared yellow on this agar while Mal+ colonies were red. A red colony color was restored after introducing plasmids expressing malT + as described below. 27 ( Miller , 1992a ) This study R. Perry, University of Kentucky R. Perry, University of Kentucky Invitrogen, Inc. ( Datsenko and, Wanner 2000 ) ( Datsenko and, Wanner 2000 ) This study ( Oster and, Phillips 2011 ) This study Pharmacia, Inc. ( Wang and Kushner , 1991 ) This study This study 10 ::Tn leu 714 ara Φ80∆ New England Biolabs, Ipswich, MA II) ∆ Pvu Plasmids ( R Bacterial strains phoA relA1 spoT1 thi-1 ∆ − )Hyb This study λ Y. pestis E. coli )Hyb)Hyb This study This study R 5 )Hyb This study ) R lacZ R )Hyb This study + U169 phoA glnV44 - ) lacZ lacZ from from - - flp 0 Yp lacZ plasmid, Amp ) This study , - cat lacZ + + ypsIR 0 Yp 0 Yp - , Cam , Kan :: R gpt-lac R gyrA96 recA1 relA1 endA1 thi-1 hsdR17 0 Yp and Amp ∆ 0 Yp galE thi rpsL ori ori ∆( R ) This study malK R malT R malT malT dapA malK malK argF-lacZ )M15 malK pgm Table 3.1 Bacterial strains and plasmids used in this study X74 malK yspIR ∆( malT lac lacZ ara-600 ( ∆ − − fhuA2 pCLIP18, Amp pWSK30, pWSK30, Yp Ec malT malT flp CSH100 F R88 CO92 ∆ R114 R88, ∆ Strain ID Relevant Genotype or DescriptionCOs15COs13COs14 LMG194 Φ( R88 Φ( Source or Reference R114 Φ( COs12 LMG194, COs10COs10.1pKD3 (R88∆ COs10 Φ( R6Kgamma NEB5-alpha LMG194COs11.1 F COs11 Φ( COs11 (LMG194∆ pKD4pCLIP- pLacCOs1pKanLacCOs1pNeopWSK30 ColE1, R6Kgamma Kan ColE1, Amp Low-copy Amp ColE1, Neo pWSK30- pWSK30- 28

3.2.3 Cloning of malT

The malT gene was cloned from both Y. pestis and E. coli into the pSC101-derivative plasmid pKanCOs1 using the ligation-independent method described previously (Oster and

Phillips, 2011). For this, the malT coding region was PCR ampliﬁed from Y. pestis using primers YpmalT.BspQI.S and YpmalT.BspQI.AS and from E. coli using primers malT -Ec-

BspQI.S and malT -Ec-BspQI.AS (Table 3.2). The malT genes from both bacterial species were then transferred into the lower copy number AmpR plasmid pWSK30 (Wang and Kushner,

1991). For this, the pKanCOs1-derivative plasmids carrying malT were digested with PstI and

DrdI and DNA fragments of ≈ 4-kbp (E. coli) and 4.2-kbp (Y. pestis) were gel puriﬁed and inserted into pWSK30 digested with the same enzymes. Both recombinant plasmids were tested for their ability to restore a Mal+ phenotype to the E. coli and Y. pestis malT mutants

(Table 3.1) by plating on maltose MacConkey agar plates.

3.2.4 Construction of lacZ gene fusions expressed in single copy

To introduce the lacZ gene fusions in single copy, we used the ligation-independent cloning system described previously (Oster and Phillips, 2011) to construct gene fusions. To facilitate expression of the gene fusions in single copy, we ﬁrst modiﬁed pLacCos1 (Oster and Phillips,

2011) to generate pKanLacCOs1 (Fig. 3.1). For this, a gene cassette was constructed where the KanR cassette from pKD4 was ﬂanked by transcriptional terminators from pQE30 (Qia- gen Inc., Valencia, CA). The gene cassette was cloned into pLacCos1 at ApaI-SalI, creating pKanLacCOs1 (Fig. 3.1).

Translational gene fusions were generated by cloning PCR products in pKanLacCOs1 as described (Oster and Phillips, 2011). All primers for PCR are summarized in Table 3.1 and were obtained from IDT (Coralville, IA). Genomic DNA was prepared using the Epicentre Master-

Pure Complete DNA & RNA Purification Kit (Epicentre, Madison, WI). PCR was performed using Thermo Scientific Extensor Master Mix (Thermo Scientific, Waltham, MA). The malK- lacZ gene fusions were constructed as follows: a 560-bp region (Y. pestis) and a 940-bp region

(E. coli) containing the divergent promoters of the malK-lamB operon and the malEFG operon 29 a Table 3.2 PCR primers used in this study ) 0 to 3 0 -BspQI.S-BspQI.AS TCGACATTCAGGGGAAGAGCCGTGGTTAACGCCGAAAGAGCC ACTCTCGCAGGTGAAGAGCACTCACCGGATGAATGCGGTACAA -BspQI.S-BspQI.AS TCGACATTCAGGGGAAGAGCTCCGGATGCTCAACGGTGACTTTA ACTCTCGCAGGTGAAGAGCCAAATGCGCCAGTTGTAGCACTTC .BspQI.S.BspQI.AS ACTCTCGCAGGTGAAGAGCGGATAATTTAACCAGCAGGCGGTC TCGACATTCAGGGGAAGAGCGAAACCCAACACCGGTAAATCC -Ec-BspQI.S-Ec-BspQI.AS TCGACATTCAGGGGAAGAGCGGTTGTGACATAGGAGTTCCAC ACTCTCGCAGGTGAAGAGCGAAAGAGGTGAAACGCTACCTG malK malK malK malK malT malT Primers with names containing “BspQI” were designed for use in no-ligase cloning reactions. a PrimerEc-malT-KD3.SEc-malT-KD3.ASYp-malT-KD.S ATGCTGATTCCGTCAAAACTAAGTCGTCCGGTTCGACTCGACCATACCGTGTGTAGGCTGGAGCTGCTTCG Yp-malT-KD.AS ATCATCTTCAGCAATTGCTGGGCGTGTTGTACCGCATCCTGGCGATGGGCCATATGAATATCCTCCTTAG YpP YpP ATGCTGATTCCATCCAAACTGAGCCGCCCGGTACGGCTGCAAAATACCGTGTGTAGGCTGGAGCTGCTTC Sequence (5 GACATATCCCATCATCTGCAATAAACGTTGGGCTTGTTGTACGGCTTCATATGAATATCCTCCTTAG EcP EcP Yp-malT-BspQI.SYp-malT-BspQI.ASpneo-BspQI.S TCGACATTCAGGGGAAGAGCGAGCAGATAACTGTTCGGCTAACG ACTCTCGCAGGTGAAGAGCCCTCCGCGACGAATAAGAGATTAC pneo-BspQI.ASlacI.SlacI.AS ACTCTCGCAGGTGAAGAGCGTCATAGCCGAATAGCCTCTCC dapA:KD3.S TCGACATTCAGGGGAAGAGCCATAAACTGCCAGGAATTGGG dapA:KD3.ASYp Yp malT ATGTTCACGGGAAGTATTGTCGCGATTGTTACTCCGATGGATGAAAAAGGTGTGTAGGCTGGAGCTGCTTC TTACAGCAAACCGGCATGCTTAAGCGCCGCTCTGACCGTCTCACGACCATATGAATATCCTCCTTAG GGCTGATCATTAACTATCCGCTGG malT CCGTAATGGGATAGGTTACGTTGG 30

Figure 3.1 Plasmids constructed for genetic engineering of Y. pestis. A. pLacCOS1-kan vector for building lacZ fusions. T1, transcriptional terminator from E. coli rrnB operon; t0, transcriptional terminator from bacteriophage λ; frt, recognition sites for Flp site-specific recombinase; bla, AmpR; kan, KanR; lacZ and lacY, β-galactosidase and permease; BspQI and StuI sites indicate the cloning site. B. pCLIP-flp helper plasmid — araC, activator for arabinose induction; flp, flp recombinase; bla, AmpR; O and P, bacteriophage λ replication genes. 31 was generated by PCR from both Y. pestis and E. coli using primers YpPmalK -BspQI.S and

YpPmalK -BspQI.AS or EcPmalK -BspQI.S and EcPmalK -BspQI.AS, respectively. A malT- lacZ fusion was constructed using Yp-malT-BspQI.S and Yp-malT-BspQI.AS primers with

Y. pestis genomic DNA to amplify a ≈1000-bp region between malT and malP containing the malT promoter.

A control gene fusion where expression of lacZ was from the constitutive promoter of the neo

(neomycin resistance) gene was also made in a similar manner by amplifying the promoter for neo from pNeo (originally obtained from Pharmacia, Inc.) using the primer set pneo-BspQI.S and pneo-BspQI.AS (Table 3.1) and cloning it into the pKanLacCOs1 as described (Oster and

Phillips, 2011).

After sequencing to confirm the correct DNA sequences, the resulting plasmids were used as a template for PCR with the primer pair lacI.S and lacZ.AS (Table 3.1). The amplification product was agarose gel purified and treated with DpnI to reduce the amount of intact template DNA. The purified product was used for recombineering into F0 lac-pro in CSH100, as summarized in (Fig. 3.2). Recombinants were selected on LB+Kan plates containing X-gal to identify functional lacZ fusions.

3.2.5 Transfer of lacZ gene fusions by conjugation

To introduce the lacZ gene fusions to Y. pestis, the F0 factors were ﬁrst transferred to the

E. coli dapA mutant COs12 (Table 3.1). COs12 was constructed by using recombineering to replace dapA with the CamR cassette from pKD3 (Datsenko and Wanner, 2000). Using pKD3 as a template, the primer pair dapA:KD3.S and dapA:KD3.AS (Table 3.2) was used to generate a PCR product that was recombined into LMG194. Recombinants were selected on LB plates supplemented with Kan and 0.3 mM diaminopimelic acid (DAP), as described (Demarre et al.,

2005; Ferri`ereset al., 2010). After conﬁrming the DAP requirement for growth, COs12 was used as a recipient in a conjugation with each of the CSH100 F0 derivatives carrying lacZ gene fusions.

Conjugations were performed by patch matings where donor and recipient were mixed on antibiotic-free plates, grown overnight at 30◦C, and then exconjugates were selected on 32

Figure 3.2 Scheme used to construct lacZ gene fusions expressed from F0 factors. A. Pro- moter cloning by ligation-independent recombinant DNA as described in (Oster and Phillips, 2011). B. Introduction of lacZ fusions to F0 lac-pro in E. coli by recombineering. C. Transfer of F0 lac to E. coli K-12 and Y. pestis strains by conjugation with a dapA donor. See text for additional details. 33

LB plates with Kan for selection and Cam for counterselection in the presence of DAP. The resulting COs12-derivatives served as donors to subsequently transfer the F0 lac elements to

Y. pestis. When Y. pestis was the recipient, TSB plates were used rather than the LB plates used for E. coli. The resulting Y. pestis strains are listed in Table 3.1.

3.2.6 Compatibility of F0 lac pro with Y. pestis virulence plasmids

The Epicentre colony Fast-Screen PCR kit (Epicentre, Madison, WI) was used to prepare

DNA from R88 Y. pestis and from Y. pestis containing F0 lac. PCR was performed on these colonies using primers speciﬁc to each of the three Y. pestis plasmids identiﬁed in (Ni et al.,

2008) to test for loss of virulence plasmids.

3.2.7 Construction and use of the helper plasmid pCLIP-ﬂp

pCLIP-ﬂp (Fig. 3.1) was constructed by ﬁrst inserting a ≈2.4 kbp fragment encoding araC- bla from pKD46 (Court et al., 2002) into the unique EcoRI site of pCLIP19 (Boyd and Sherratt,

1995). The plasmid pBAD-flp, encoding flp under control of araC, was constructed by PCR amplification of EL250 (Lee et al., 2001), a strain where araC -PBADflpe had been integrated into the E. coli chromosome. To complete construction of the helper plasmid, pCLIP-araC-bla was digested with BssHII and BsaAI (NEB, Ipswich, MA), and pBAD-flp was digested with

BssHII and PmeI (NEB, Ipswich, MA). The appropriate DNA fragments (≈4 kb and 2.5 kb, respectively) were eluted from an agarose gel using the QIAQuick Gel Extraction Kit (Qiagen,

Valencia, CA) and then ligated using T4 DNA ligase. The ligation reactions were transformed into DH5α with selection for AmpR to yield pCLIP-ﬂp (Fig. 3.1).

To use the pCLIP-ﬂp, the plasmid was transformed by electroporation into E. coli or

Y. pestis that had been previously recombineered using pKD3 or pKD4 as PCR templates and cured of the pRed helper plasmid. Overnight cultures were then grown in appropriate medium supplemented with Amp and 25 mM L-arabinose at 30◦C for Y. pestis or 37◦C for

E. coli. Cultures were diluted and plated on LB or TSB plates with no antibiotics to obtain single colonies. The colonies obtained were then patched on plates supplemented with Amp or

Kan to test for the loss of the antibiotic region and also the loss of pCLIP-ﬂp. 34

3.2.8 β-galactosidase Assays

All β-galactosidase assays were performed using the method described by Miller(1992a).

Assays were performed in triplicate and the values averaged.

3.3 Results

3.3.1 Streamlined method for construction of new F0s carrying lacZ gene fusions

To better understand regulation of the mal genes in Y. pestis, we sought to construct translational gene fusions to monitor and compare gene expression in both E. coli and Y. pestis.

We have previously shown that construction of new F0 factors, which replicate at very low copy numbers, by recombineering was a convenient method to express genes at physiologically relevant levels (Peterson and Phillips, 2008), and we reasoned that this approach would also be useful to construct and express lacZ gene fusions.

The strategy for the construction of lacZ gene fusions is summarized in Fig. 3.2, which includes cloning a promoter along with the 50 end of the adjacent gene in frame with lacZ.

The gene fusion is ﬁrst constructed on a ColE1-derivative plasmid using a ligation-independent method that takes advantage of relatively long single-stranded DNA tails generated by the nicking endonuclease BspQI (Oster and Phillips, 2011). To adapt the ligation-independent cloning method to better accommodate subsequent recombineering steps, we constructed pLacCOs- kan (Fig. 3.1). After cloning PCR products representing a selected promoter, pLacCOs-kan- derivative plasmids were used as PCR templates to generate amplicons representing the lacZ fusion with a portion of lacI immediately upstream as well as a KanR marker. This PCR product was then introduced to F0 lac pro in E. coli CSH100 by recombineering (Fig. 3.2).

To complete the construction of new strains expressing the gene fusions, the new F0 lac elements were transferred to either diﬀerent E. coli K-12 strains or to Y. pestis by conjugation.

For convenience, and to avoid introduction of unnecessary antibiotic markers to Y. pestis, we constructed an E. coli donor strain that does not require a counterselection marker. Strain

COs12 (Table 3.1) was constructed by disrupting dapA by recombineering (Court et al., 2002), 35 generating a strain that required DAP for growth (Demarre et al., 2005). Following conjugation,

KanR exconjugants were selected in the absence of DAP to prevent growth of the donor cells.

Using the cloning and recombination system shown in Fig. 3.2, we constructed lacZ gene fusions by joining the promoter regions and the ﬁrst few amino acids of malK from Y. pestis to lacZ (PmalK -lacZ ). As a constitutively expressed control, we also used the system to construct a lacZ fusion driven by the neo promoter (Pneo-lacZ ). Although not a part of this study, we have further demonstrated the utility of the cloning/ recombineering system by constructing lacZ gene fusions to other Y. pestis genes whose expression is also temperature regulated (data not shown). In all cases we were able to construct strains of Y. pestis expressing lacZ gene fusions at a very low copy number.

3.3.2 Characterization of PmalK -lacZ gene fusions

β-galactosidase assays were performed with the Y. pestis exconjugants to test the response of the PmalK -lacZ fusions under conditions known to alter mal gene expression. Consistent with prior microarray studies (Han et al., 2004; Motin et al., 2004), temperature regulation was observed, i.e., expression was strongly reduced at 37◦C (Fig. 3.3a). In addition, the gene fusion was induced by maltose to a concentration of 50 µM, although the addition of maltose did not overcome the strong downregulation at 37◦C.

To compare the response of the malK promoter from Y. pestis with that of the well- characterized mal system in E. coli, we also monitored expression in an E. coli K-12 background.

As shown in Fig. 3.3b, in E. coli, while malK responds to maltose, it is not temperature regulated at the level observed in Y. pestis. As a control, we measured expression of a Pneo- lacZ fusion, however in contrast to PmalK -lacZ expression, it displayed an increase in expression at 37◦C (Fig. 3.3c).

One caveat with the introduction of F factors into Y. pestis is that the large virulence plasmid pCD1 is known to be incompatible with F factor plasmids apparently due to a shared partitioning system (Bakour et al., 1983; Perry et al., 1998). Consequently, we tested the

Y. pestis F0 lac strains to determine if pCD1 had been lost from the exconjugants. Using a

PCR assay for pCD1, we conﬁrmed that the virulence plasmid had indeed been cured from 36

Figure 3.3 β-galactosidase activity of: a. Y. pestis R88 Φ(malK 0-lacZ )Hyb (COs13) b. E. coli Φ(malK 0-lacZ )Hyb (COs15) c. neo-lacZ Φ(neo0-lacZ )Hyb. 37

Figure 3.4 Incompatibility of F0 lac with Y. pestis plasmids. PCR was performed on Y. pestis R88 (lanes 1–3) and Y. pestis R88∆malT containing F0 lac (lanes 5–7), as described in Materials and Methods, to detect each of the three virulence plasmids. Plasmids detected include: pPCP1 (lanes 1 and 5); pCD1 (lanes 2 and 6); and pMT (lanes 3 and 7).

strains expressing the lacZ gene fusions (Fig. 3.4). No loss of the other virulence plasmids, pMT1 and pPCP1 in Y. pestis R88, was observed.

3.3.3 Eﬀect of quorum sensing on mal gene expression

Previous microarray studies have suggested that mal genes in Y. pestis are also under quorum sensing control. To test this, we introduced the PmalK -lacZ gene fusion to the Y. pestis strains R88 and R114 (deleted for both the Ysp and Yps QS systems, Table 3.1), and performed assays on the overnight cultures. Under these conditions, we observed no change in

β-galactosidase activity in the absence of the AHL quorum sensing systems (Fig. 3.5). In contrast, temperature-dependent expression of malK was not aﬀected.

3.3.4 Role of MalT in gene expression

In E. coli, MalT is responsible for mal gene regulation in response to maltose (Boos and

B¨ohm, 2000; Boos and Shuman, 1998; Panagiotidis et al., 1998; Richet et al., 2005). We proposed that this positive regulator could also be responsible for temperature regulation in 38

0 Figure 3.5 Expression of Φ(malK Yp-lacZ )Hyb fusions in the absence of quorum sensing. Tem- perature regulation of mal occurs in Y. pestis strain R88 (wild type for quorum sensing) as well as in R114 (mutant defective in Ysp and Yps quorum sensing systems) (Table 3.1).

Y. pestis. Inspection of the MalT amino acid sequences of Y. pestis and E. coli showed them to be similar throughout the linear amino acid sequence with a few exceptions (Fig. 3.6), most notably a region near the center of the protein where three adjacent amino acids have been introduced to the Y. pestis protein. Other changes include multiple positions where substitutions of charged amino acids for uncharged occur.

To determine if MalT from Y. pestis is functionally distinct from that of E. coli, and is responsible for temperature-dependent mal expression, we constructed recombinant plasmids expressing malT from both E. coli and Y. pestis and compared their eﬀects on PmalK -lacZ expression in both hosts. We also constructed malT deletion mutants in both species to express each malT clone as homozygotes.

As anticipated, expression of PmalK -lacZ was abolished in the malT deletion mutants in both species, and expression was restored when malT from either Y. pestis (MalTYp) and

E. coli (MalTEc) was expressed from recombinant plasmids (Fig. 3.7a). A functional diﬀerence between MalTYp and MalTEc was noted, however, when promoter activity was measured at 30◦C and 37◦C. The decrease in expression was more pronounced at 37◦C in Y. pestis when

MalTYp was expressed in comparison to when MalTEc was present (Fig. 3.7a). 39 K-12. The E. coli and Y. pestis boxed region is markedly diﬀerent between these species. Figure 3.6 Comparison of amino acid sequences of MalT from 40

To determine if MalTYp was the sole determinant of temperature-dependent regulation, we also performed the converse set of experiments. In E. coli, no signiﬁcant diﬀerence in expression

◦ ◦ of PmalK -lacZ was observed at 30 C and 37 C when MalTYp was expressed, while an increase

◦ in β-galactosidase activity was observed at 37 C in the strain expressing MalTEc (Fig. 3.7b).

The PmalK -lacZ fusion maintained its response to maltose in both backgrounds (Fig. 3.7c).

3.3.5 Expression of malT is independent of temperature

Since MalT is an important determinant of temperature-dependent gene expression in

Y. pestis, we also tested if its expression was influenced by growth temperature. We constructed a PmalT -lacZ fusion and determined that expression in Y. pestis was not significantly affected by temperature beyond what would be expected from the increased rate of metabolism

(Fig. 3.8).

3.4 Discussion

In our initial eﬀorts to study gene expression in Y. pestis, we attempted to use relatively high copy number ColE1 plasmids to express lacZ gene fusions. However, expression of the fusions consistently resulted in growth defects in Y. pestis, making accurate measurements of

β-galactosidase activity impossible. As a result, we developed a system that combines ligation- independent cloning (Oster and Phillips, 2011) and recombineering to construct new F0 lac elements to express lacZ fusions at a very low copy number.

We found the lacZ fusions made with this system to be a great improvement over the use of high copy number plasmids. The lower expression of the β-galactosidase fusions did not significantly affect growth of Y. pestis and yielded reliable assay data. We have constructed a plasmid (Fig. 3.1) that facilitates construction of any lacZ gene fusion using a highly efficient ligation-independent cloning system (Oster and Phillips, 2011) and then seamlessly introduces the construct into F0 lac in E. coli.

To transfer the gene fusions to Y. pestis, we took advantage of the self-transmission of F0 lac from E. coli. To facilitate this we constructed a dapA mutant of E. coli to serve as a donor 41

Figure 3.7 β-galactosidase activity in malT deletion backgrounds. These strains are ∆malT complemented with malT on low-copy pWSK30 plasmids. a. Y. pestis strain COs10.1 b. E. coli strain COs11.1 c. COs11.1 fusion with and without maltose. 42

Figure 3.8 β-galactosidase activity of Φ(malT 0-lacZ )Hyb

that obviated the need for a counterselection marker since the donor is incapable of growth in the absence of DAP (Demarre et al., 2005; Ferri`eres et al., 2010).

A caveat with the introduction of F0 factors to Y. pestis is the incompatibility with the large virulence plasmid, pCD1. This plasmid uses an IncFIIA replication origin whose partitioning system is apparently incompatible with the F factor (Bakour et al., 1983; Perry et al., 1998).

Indeed, by using a PCR assay to monitor presence of the three virulence plasmids (Ni et al.,

2008) in Y. pestis R88, we conﬁrmed that strains that harbored F0 lac had lost pCD1 (Fig. 3.4).

However, the other two virulence plasmids, pMT and pPCP1, were still present in R88. While this incompatibility would present a problem in virulence studies, it poses no complication for our studies of expression of genes that map to the Y. pestis chromosome.

In performing recombineering in Y. pestis, we routinely remove the drug resistant maker

(KanR) used for selection by site-specific recombination. We consistently experienced difficulty in curing the Y. pestis recombinants of the helper plasmid (pCP1) used to express the Flp recombinase (Datsenko and Wanner, 2000). Even though the vector is based on a pSC101 temperature sensitive replicon and is poorly maintained in E. coli, for reasons not understood, it is very stable in Y. pestis. To improve the efficiency of removing the helper plasmid, we constructed derivatives of pCLIP19, whose replicon is based on the origin of replication of bacteriophage λ (Boyd and Sherratt, 1995). This vector lacks dedicated partitioning machinery and is readily lost in the absence of selection. While versions of this plasmid that expressed the 43

Red recombinase proteins dramatically reduced survival of Y. pestis following electroporation

(data not shown), a pCLIP19-derivative plasmid expressing ﬂp (Fig. 3.1) was successfully used in Y. pestis. Following removal of the selectable marker, pCLIP-ﬂp is readily lost in the population by overnight growth in the absence of ampicillin.

Consistent with previous microarray studies (Han et al., 2004; Motin et al., 2004), we

◦ observed a significant decrease in expression of the PmalK -lacZ fusion at 37 C (Fig. 3.3). This effect was much more pronounced in Y. pestis than in E. coli K-12, consistent with the difference in mal gene regulation of expression between these species. Expression of the gene fusion was induced by maltose in both species. In contrast, a Pneo-lacZ control fusion showed higher expression at 37◦C in both Y. pestis and E. coli, indicating that the effect of growth temperature was specific for the mal genes.

We also examined the role of quorum sensing on mal regulation, as previous studies have indicated that the mal genes are also under control of the AHL-dependent quorum sensing system. No significant change in expression of PmalK -lacZ was observed in a mutant deleted for the two quorum sensing systems in Y. pestis. The reason for this discrepancy is not clear, but could be related to the growth conditions used in the microarray experiments versus the conditions used in this study. Microarrays measure transcriptional activity only, so the use of a translational gene fusion could also account for the different results. Prior microarray studies indicated that the different quorum sensing systems of Y. pestis appear to interact in complex ways, so it is possible that there may be other factors that have yet to be identified that influence mal gene expression in vivo.

Because MalT is a positive activator of the mal regulon, we tested if it is responsible for the temperature regulation effect in Y. pestis. For this, we constructed malT deletion mutants in both Y. pestis and E. coli and complemented the mutations with cloned copies of malT from each species. While expression of the PmalK -lacZ fusion was eliminated in the malT mutants, expression was elevated and responsive to maltose when malT was expressed from low-copy number plasmids (Fig. 3.7a and 3.7b). Consistent with the previous results, we found that in a Y. pestis ∆malT background, the presence of Y. pestis MalT produced a strong decrease in expression at 37◦C, while E. coli MalT resulted in only a modest change. A comparison of the 44 linear amino acid sequences of MalT from Y. pestis and E. coli revealed that while the proteins share significant homology, some regions of differences exist (Fig. 3.6). These changes could account for functional differences we observe between the two species. Differential regulation of malT cannot account for these differences, however, as a PmalT -lacZ gene fusion did not shown temperature-dependent expression in Y. pestis (Fig. 3.8). The MalT positive activator is not the sole determinant of temperature regulation of the mal genes, however: E. coli did not gain the ability to regulate mal gene expression in response to temperature when malT from Y. pestis was expressed.

While we have demonstrated the importance of MalT in the downregulation of mal gene expression in Y. pestis, it remains to be determined what other cellular factors are working in combination with MalT to bring about this regulation. Future studies will seek to understand how the differences in protein sequence and structure of the protein in both species lead to the functional differences. Experiments to “swap” or mutate regions of the genes that correspond to the differences can be performed to determine if this changes the function of MalT. Epitope tagging and proteomics can also be used to identify potential proteins that interact with MalT in Y. pestis. Understanding the mechanism of mal gene regulation in Y. pestis will also likely inform us of the significance of temperature regulation in the biology of this pathogen. Perhaps, for example, downregulating the mal genes helps Y. pestis evade host immunity as bacteria enter the mammalian host. Alternatively, increased expression may contribute to survival in the flea. The genetic tools developed for this initial study will be useful to better understand gene regulation in Y. pestis.

Acknowledgements

This work was funded by grants from the Oﬃce of Naval Research (ONR-MURI N00014-

06-1-1176) and NIH (NIAID R21 AI074591). 45

CHAPTER 4. GENERAL CONCLUSIONS

From the onset of this project to genetically characterize gene expression and function in

Y. pestis, I identiﬁed the lack of tools for working with this microorganism as a considerable limitation. The need to generate Y. pestis mutants defective at several loci, and to construct multiple lacZ fusions, encouraged the development of a pipeline using recombineering and recombinant DNA. The description and use of these techniques comprise the majority of the work in this thesis.

To develop a streamlined method to construct lacZ gene fusions for expression analysis, we ﬁrst developed the ligase-independent cloning system discussed in Chapter2. This system represented a distinct advantage over building lacZ fusions using traditional cloning in that the resulting plasmids were assembled with a minimal number of steps and the ﬁnal constructs were consistently correct. We have taken advantage of this system to clone several Y. pestis and E. coli promoters that have been described in this thesis, as well as others that will be used in future studies. The promoters used for construction of lacZ gene fusions are listed in

Table 4.1. In developing this cloning method, we also recognized its utility beyond constructing gene fusions and also used it as a general method to clone PCR products, as demonstrated by the cloning of the malT gene from Y. pestis and E. coli.

Initially, the lacZ fusions made by the ligase-independent system were expressed from

ColE1-derivative plasmids that replicate at a relatively high copy number in E. coli and

Y. pestis. Early characterization of these plasmids suggested that MalT controlled temperature regulation in Y. pestis and that MalT from Y. pestis could confer temperature regulation onto E. coli. However, we were concerned that this striking observation could also be due to artifacts related to the expression of lacZ gene fusions as well as MalT on multiple copy number plasmids. Indeed, growth defects were occasionally observed with these strains. Consequently, 46 ) lacZ ) lacZ gene fusion to ) lacZ Y. pestis ) ) ) lacZ ) ) lacZ lacZ ) ) lacZ lacZ lacZ lacZ Type VI secretion system component fused to fusion to fusion to clone) fusion to clone) Y. pestis pH6 antigen fusion to catalase gene fusion to pH6 antigen fusion to quorum sensing ATPase fusion to Y. pestis malK Y. pestis malT E. coli malK Y. pestis malT E. coli malT ( ( Y. pestis Y. pestis Y. pestis Y. pestis + Yp + Ec )Hyb (Hypothetical temperature regulated )Hyb (Putative )Hyb ( )Hyb ( )Hyb ( )Hyb (Heat shock gene fusion to )Hyb ( )Hyb ( )Hyb ( )Hyb ( malT malT lacZ lacZ lacZ - - lacZ lacZ 0 0 - - - lacZ lacZ lacZ lacZ - - 0 Yp 0 Ec - 0 Yp - lacZ 0 0 0 0 - 0 Table 4.1 Plasmids built using no-ligase cloning system malT katY lsrA rpoH psaA psaC ypo1996 malK malK ypo0498 Φ( Φ( Φ( Φ( Φ( Φ( (Yp) pKanCOs1, (Eco) pKanCOs1, (YP)(Eco) Φ( Φ( malT malT malT katY lsrA rpoH psaA psaC malK malK pLacCOs1- pKanCOs1- pLacCOs1- pLacCOs1- pLacCOs1- pKanCOs1- pLacCOs1- pLacCOs1- pLacCOs1-1996 Φ( PlasmidpLacCOs1- pLacCOs1- Description pLacCOs1-0498 Φ( 47 to address this issue we constructed a Y. pestis pcnB mutant to lower the plasmid copy number (Table 4.2). While plasmid copy numbers were lowered, we still observed problems with growth of Y. pestis cultures. To lower expression of the gene fusions even further, we combined the ligase-independent cloning system with recombineering to generate F0 factors carrying the individual lacZ constructs, allowing us to express gene fusions at a very low copy number. We accomplished this by modifying our pLacCOs1 plasmid for use as a template for recombineering into F0 lac-pro in CSH100. This strain was conjugated with an E. coli dapA auxotroph that we constructed. The resulting exconjugant served, in turn, as a donor to transfer the modiﬁed F0 into Y. pestis by selecting for kanamycin and without the need for introducing another antibiotic resistance marker for counterselection. To further avoid growth defects, we cloned malT from both E. coli and Y. pestis onto a low-copy number vector. Collectively, these strategies allowed us to generate E. coli and Y. pestis strains without noticeable growth defects and where expression of the gene fusions was at or near physiological levels.

Chapter3 describes how I used the combination of the ligation-independent cloning with a modified recombineering protocol to study regulation of the mal genes in Y. pestis. While recombineering has been used in Y. pestis, I found that optimization was necessary to take full advantage of this powerful recombination system for genetic engineering. Since Y. pestis cells grow more slowly than E. coli, and fewer cells tend to survive the electroporation, I modified the standard protocol by increasing initial growth phases and post-electroporation recovery times, as well as using richer growth media. Use of increased amounts of DNA compared to the amounts used for E. coli was also necessary for efficient recovery of recombinants. A list of Y. pestis deletion mutants I have built over the course of this study is shown in Table 4.2.

An additional problem in recombineering Y. pestis is that the wild-type bacteria are highly virulent and the Select Agent program allows only a limited number of antibiotic resistance markers to be introduced to this microorganism. As in other systems, we took advantage of the ﬂp recombinase to remove the antibiotic markers used for selecting recombineered mutants.

However, I (and others) have observed that the helper plasmid used to express both Red recombinases as well as the Flp recombinase are diﬃcult to cure from Y. pestis. While the reason for this is unknown, we sought to express the recombinases from a plasmid that was inherently 48

Table 4.2 Gene deletions in Y. pestis constructed by recombineering (frt represents the scar sequence generated by site-speciﬁc recombination of antibiotic resistance)

Mutant Description malT::frt mal transcriptional activator malQ::kan amylomaltase lsrR::kan Quorum sensing gene — lsr operon repressor lsrK::kan Quorum sensing gene — phosphorylase of AI-2 ytbI::frt AHL synthase gene in Y. pestis pcnB::frt Plasmid copy number regulator lacZ::kan β-galactosidase unstable in Y. pestis to permit easier recovery of recombinants cured of the helper plasmids.

For this we chose pCLIP19, a cloning vector using the origin of replication of bacteriophage

λ. This plasmid lacks determinants of plasmid stability and is easily lost from its host in the absence of selection. We constructed pCLIP-derivative plasmids that expressed both the

Red and Flp recombinases under control of the araBAD promoter and operator. While the pCLIP-Red plasmid worked well for recombineering in E. coli, it was not suitable for Y. pestis.

However, the plasmid pCLIP-ﬂp was eﬀective for generating antibiotic-sensitive recombinants and was readily lost from Y. pestis. This plasmid should be of use to Y. pestis researchers for construction of new mutants void of antibiotic resistance.

The initial focus of our studies focused on the role of quorum sensing in Y. pestis biology.

For this, I used recombineering to construct several mutants defective at different levels of quorum sensing. These studies led to confirm the observation previously made by microarray studies that the mal operon was regulated by growth temperature. Since this latter observation is unique and is not shared with other well-characterized bacteria such as E. coli, we focused our efforts on better understanding this phenomenon. Using the new tools we developed, we constructed lacZ gene fusions to examine regulation of mal genes in Y. pestis. Consistent with microarray studies, expression of the gene fusions was significantly reduced at higher growth temperatures of 37◦C, suggesting reduced expression of this regulon is important when Y. pestis enters one of its mammalian hosts. In our hands, however, the effect of quorum sensing on expression of the malK-lacZ fusion was very modest, and we focused our continued efforts on characterizing the effect of temperature on mal gene expression. 49

Characterization of the lacZ fusions in both Y. pestis and E. coli led to the conclusion that MalT is responsible for the temperature regulation we observe in Y. pestis. Inspection of the MalT primary amino acid sequence reveals overall similarity with E. coli, with some significant differences that may account for the temperature regulation of mal gene expression in Y. pestis. While understanding how MalT mediates temperature-dependent gene expression will require additional studies, these studies suggest testable hypotheses that can be applied to this question. For example, future work will include construction of chimeric malT clones to determine if specific domains within MalT of Y. pestis are important for its unique properties.

In addition, biochemical and genetic tests can be performed to identify proteins in Y. pestis that may interact with MalT. The genetic tools described in this thesis will continue to be an asset for future studies. 50

BIBLIOGRAPHY

Achtman, M., Zurth, K., Morelli, G., Torrea, G., Guiyoule, A., Carniel, E., 1999. Yersinia

pestis, the cause of plague, is a recently emerged clone of Yersinia pseudotuberculosis. PNAS

USA 96, 14043–14048.

Aslanidis, C., de Jong, P., 1990. Ligation-independent cloning of PCR products (LIC-PCR).

Nucleic Acids Res. 18, 6069–6074.

Bakour, R., Laroche, Y., Cornelis, G., 1983. Study of the incompatibility and replication of the

70-kb virulence plasmid of Yersinia. Plasmid 10, 279–289.

Bardóczy, V., Géczi, V., Sawasaki, T., Endo, Y., Mészáros, T., 2008. A set of ligation-

independent in vitro translation vectors for eukaryotic protein production. BMC Biotechnol.

8, 32–38.

Bavoil, P., Hofnung, M., Nikaido, H., 1980. Identiﬁcation of a cytoplasmic membrane-associated

component of the maltose transport system of Escherichia coli. J. Biol. Chem. 255, 8366–

8369.

Berrow, N., Alderton, D., Sainsbury, S., Nettleship, J., Assenberg, R., Rahman, N., Stuart,

D., Owens, R., 2007. A versatile ligation-independent cloning method suitable for high-

throughput expression screening applications. Nucleic Acids Res. 35, e45.

Betton, J. M., 2004. High throughput cloning and expression strategies for protein production.

Biochemie 86, 601–605.

Boos, W., B¨ohm,A., 2000. Learning new tricks from an old dog: MalT of the Escherichia coli

maltose system is part of a complex regulatory network. TIG 16, 404–409.

Boos, W., Shuman, H., 1998. Maltose/maltodextrin system of Escherichia coli: Transport,

metabolism, and regulation. Microbiol. Mol. Biol. Rev. 62, 204–229.

Bos, K., Schuenemann, V., Golding, G., Burbano, H., Waglechner, N., Coombes, B., McPhee,

J., DeWitte, S., Meyer, M., Schmedes, S., Wood, J., Earn, D., Herringa, D., Bauer, P., 51

Poinar, H., Krause, J., 2011. A draft genome of Yersinia pestis from victims of the Black

Death. Nature 478, 506–510.

Boyd, A., Sherratt, D., 1995. The pCLIP plasmids: versatile cloning vectors based on the

bacteriophage lambda origin of replication. Gene 153, 57–62.

Brubaker, R., 2003. Interleukin-10 and inhibition of innate immunity to Yersiniae: Roles of

Yops and LcrV (V Antigen). Infect. Immun. 71, 3673–3681.

Brubaker, R., 2007. How the structural gene products of Yersinia pestis relate to virulence.

Future Microbiol. 2, 377.

Brzostek, K., Raczkowska, A., 2001. The level of Yop proteins secreted by Yersinia enterocolitica

is changed in maltose mutants. FEMS Lett. 204, 95–100.

Callaway, E., 2011. The Black Death decoded. Nature 478, 444–446.

Chain, P., Carniel, E., Larimer, F., Lamerdin, J., Stoutland, P., Regala, W., Georgescu, A.,

Vergez, L., Land, M., Motin, V., Brubaker, R., Fowler, J., Hinnebusch, J., Marceau, M.,

Medigue, C., Simonet, M., Chenal-Francisque, V., Souza, B., Dacheux, D., Elliott, J., Der-

bise, A., Hauser, L., Garcia, E., 2004. Insights into the evolution of Yersinia pestis through

whole-genome comparison with Yersinia pseudotuberculosis. PNAS USA 101, 13826–13831.

Chan, S.-H., Stoddard, B., Xu, S.-Y., 2011. Natural and engineered nicking endonucleases —

from cleavage mechanism to engineering of strand-speciﬁcity. Nucleic Acids Res. 39, 1–18.

Chanda, P., Edris, W., Kennedy, J., 2006. A set of ligation-independent expression vectors for

co-expression of proteins in Escherichia coli. Protein Exp. Purif. 47, 217–224.

Chapon, C., 1982. Expression of malT, the regulator gene of the maltose regulon in E. coli, is

limited at both transcription and translation. EMBO J. 1, 369–374.

Chapon, C., Kolb, A., 1983. Action of CAP on the malT promoter in vitro. J. Bacteriol. 156,

1135–1143. 52

Cherepanov, P., Wackernagel, W., 1995. Gene disruption in Escherichia coli: TcR and KmR

cassettes with the option of Flp-catalyzed excision of the antibiotic-resistance determinant.

Gene 158, 9–14.

Chichester, J., Musiychuk, K., Farrance, C., Mett, V., Lyons, J., Mett, V., Yusibov, V., 2009.

A single component two-valent LcrV-F1 vaccine protects non-human primates against pneu-

monic plague. Vaccine 27, 3471–3474.

Copeland, N., Jenkins, N., Court, D., 2001. Recombineering: A powerful new tool for mouse

functional genomics. Nat. Rev. Genet. 2, 769–779.

Cornelis, G., 2002. Yersinia Type III secretion: send in the eﬀectors. J. Cell Biol. 158, 401–408.

Court, D., Sawitzke, J., Thomason, L., 2002. Genetic engineering using homologous recombi-

nation. Annu. Rev. Genet. 36, 361–388.

Curiel, J., de las Rivas, B., Manche˜no,J., Mu˜noz,R., 2011. The pURI family of expression

vectors: A versatile set of ligation independent cloning plasmids for producing recombinant

His-fusion proteins. Protein Exp. Purif. 76, 44–53.

Danot, O., 2010. The inducer maltotriose binds in the central cavity of the tetratricopeptide-

like sensor domain of MalT, a bacterial STAND transcription factor. Mol. Microbiol. 77,

628–641.

Darby, C., Hsu, J., Ghori, N., , Falkow, S., 2002. Caenorhabditis elegans: plague bacteria

bioﬁlm blocks food intake. Nature 417, 243–244.

Datsenko, K., Wanner, B., 2000. One-step inactivation of chromosomal genes in Escherichia

coli K-12 using PCR products. Proc. Natl. Acad. Sci. USA 97, 6640–6645.

Datta, S., Costantino, N., Court, D., 2006. A set of recombineering plasmids for Gram-negative

bacteria. Gene 379, 109–115.

Davies, D., Parsek, M., Pearson, J., Iglewski, B., Costerton, J., Greenberg, E., 1998. The

involvement of cell-to-cell signals in the development of a bacterial bioﬁlm. Science 280,

295–298. 53

DeBord, K., Anderson, D., Marketon, M., Overheim, K., DePaolo, R., Ciletti, N., Jabri, B.,

Schneewind, O., 2006. Immunogenicity and protective immunity against bubonic plague and

pneumonic plague by immunization of mice with the recombinant V10 antigen, a variant of

LcrV. Infect. Immun. 74, 4910–4914.

Demarre, G., Gu´erout,A.-M., Matsumoto-Mashimo, C., Rowe-Magnus, D., Marli`ere,P., Mazel,

D., 2005. A new family of mobilizable suicide plasmids based on broad host range R388 plas-

mid (IncW) and RP4 plasmid (IncPα) conjugative machineries and their cognate Escherichia

coli host strains. Res. Microbiol. 156, 245–255.

Derbise, A., Lesic, B., Dacheux, D., Ghigo, J., Carniel, E., 2003. A rapid and simple method for

inactivating chromosomal genes in Yersinia. FEMS Immunol. Med. Microbiol. 38, 113–116.

Du, R., Li, S., Zhang, X., 2011. A modiﬁed plasmid vector pCMV-3Tag-LIC for rapid, reliable,

ligation-independent cloning of polymerase chain reaction products. Anal. Biochem. 408,

357–359.

Du, Y., Rosqvist, R., Forsberg, A.,˚ 2002. Role of fraction 1 antigen of Yersinia pestis in

inhibition of phagocytosis. Infect. Immun. 70, 1453–1460.

Duncan, C., Scott, S., 2005. What caused the black death? Postgrad. Med. J. 81, 315–320.

Ekaterinaki, N., Prentki, P., 1991. Incompatibility between pSC101 and λdv replicons. Res.

Microbiol. 142, 381–387.

Engelbrecht, J., Simon, M., Silverman, M., 1985. Measuring gene expression with light. Science

227, 1345–1347.

Ferri`eres,L., Hmery, G., Nham, T., Gurout, A., Mazel, D., Beloin, C., Ghigo, J., 2010. Silent

mischief: bacteriophage Mu insertions contaminate products of Escherichia coli random

mutagenesis performed using suicidal transposon delivery plasmids mobilized by broad-host-

range RP4 conjugative machinery. J. Bacteriol. 192, 6418–6427. 54

Fetherston, J., Schuetze, P., Perry, R., 1992. Loss of the pigmentation phenotype in Yersinia

pestis is due to the spontaneous deletion of 102 kb of chromosomal DNA which is ﬂanked by

a repetitive element. Mol. Microbiol. 6, 2693–2704.

Finlay, B., McFadden, G., 2006. Anti-immunology: evasion of the host immune system by

bacterial and viral pathogens. Cell 124, 767–782.

Gehring, K., Charbit, A., Brissaud, E., Hofnung, M., 1987. Bacteriophage λ receptor site on

the Escherichia coli K-12 LamB protein. J. Bacteriol. 169, 2103–2106.

Gerlach, R., H¨olzer,S., J¨ackel, D., Hensel, M., 2007. Rapid engineering of bacterial reporter

gene fusions by using Red recombination. Appl. Environ. Microbiol. 73, 4234–4242.

Gilbert, M., Cuccui, J., White, W., Lynnerup, N., Titball, R., Cooper, A., Prentice, M., 2004.

Absence of Yersinia pestis-speciﬁc DNA in human teeth from ﬁve European excavations of

putative plague victims. Microbiol. 150, 341–354.

Haensch, S., Bianucci, R., Signoli, M., Rajerison, M., Schultz, M., Kacki, S., Vermunt, M.,

Weston, D., Hurst, D., Achtman, M., Carniel, E., Bramanti, B., 2010. Distinct clones of

Yersinia pestis caused the Black Death. PLoS Pathog. 6, e1001134.

Han, Y., Zhou, D., Pang, X., Song, Y., Zhang, L., Bao, J., Tong, Z., Wang, J., Guo, Z., Zhai,

J., Du, Z., Wang, X., Zhang, X., Wang, J., Huang, P., Yang, R., 2004. Microarray analysis

of temperature-induced transcriptome of Yersinia pestis. Microbiol. Immunol. 48, 791–805.

Hand, N., Silhavy, T., 2000. A practical guide to the construction and use of lac fusions in

Escherichia coli. Methods Enzymol. 326, 11–35.

Hinnebusch, B., 2004. Yersinia: Molecular and Cellular Biology. Horizon Bioscience, Norfolk,

UK, Ch. The evolution of ﬂea-borne transmission in Yersinia pestis, pp. 49–73.

Hinnebusch, B., 2005. The evolution of ﬂea-borne transmission in Yersinia pestis. Curr. Issues

Mol. Biol. 7, 197–212. 55

Hinnebusch, B., Rudolph, A., Cherepanov, P., Dixon, J., Schwan, T., Forsberg, A.,˚ 2002. Role

of Yersinia murine toxin in survival of Yersinia pestis in the midgut of the ﬂea vector. Science

296, 733–735.

Huan, R., Serventi, I., Moss, J., 1992. Rapid, reliable ligation-independent cloning of PCR

products using modiﬁed plasmid vectors. Biotechniques 13, 515–518.

Hughes, K., Maloy, S., 2007. Use of operon and gene fusions to study gene regulation in

Salmonella. Methods Enzymol. 421, 140–158.

Ishikawa, K., Handa, N., Kobayashi, I., 2009. Cleavage of a model DNA replication fork by a

Type I restriction endonuclease. Nucleic Acids Res. 37, 3531–3544.

Ishikawa, K., Handa, N., Sears, L., Raleigh, E., Kobayashi, I., 2011. Cleavage of a model DNA

replication fork by a methyl-speciﬁc endonuclease. Nucleic Acids Res. 39, 5489–5498.

Jain, C., 1993. New improved lacZ gene fusion vectors. Gene 133, 99–102.

Jarrett, C., Deak, E., Isherwood, K., Oyston, P., Fischer, E., Whitney, A., Kobayashi, S.,

DeLeo, F., Hinnebusch, B., 2004. Transmission of Yersinia pestis from an infectious bioﬁlm

in the ﬂea vector. J. Infect. Dis. 190, 783–792.

Jeltsch, A., Wenz, C., Wende, W., Selent, U., Pingoud, A., 1996. Engineering novel restriction

endonucleases: principles and applications. Trends Biotechnol. 14, 235–238.

Joly, N., B¨ohm,A., Boos, W., Richet, E., 2004. MalK, the ATP-binding cassette component

of the Escherichia coli maltodextrin transporter, inhibits the transcriptional activator MalT

by antagonizing inducer binding. J. Biol. Chem. 279, 33123–33130.

Kirwan, J., Gould, T., Schweizer, H., Bearden, S., Murphy, R., Churchill, M., 2006. Quorum-

sensing signal synthesis by the Yersinia pestis acyl-homoserine lactone synthase YspI. J.

Bacteriol. 188, 784–788.

Kortepeter, M., Parker, G., 1999. Potential biological weapons threats. Emerg. Infect. Dis. 5,

523–527. 56

Lähteenmäki,K., Virkola, R., Sarén,A., Emödy, L., Korhonen, T., 1998. Expression of plas-

minogen activator Pla of Yersinia pestis enhances bacterial attachment to the mammalian

extracellular matrix. Infect. Immun. 66, 5755–5762.

L˚ang,H., Jonson, G., Holmgren, J., Palva, E., 1994. The maltose regulon of Vibrio cholerae

aﬀects production and secretion of virulence factors. Infect. Immun. 62, 4781–4788.

Lee, E., Yu, D., de Velasco, J. M., Tessarollo, L., Swing, D., Court, D., Jenkins, N., Copeland,

N., 2001. A highly eﬃcient Escherichia coli-based chromosome engineering system adapted

for recombinogenic targeting and subcloning of BAC DNA. Genomics 73, 56–65.

Li, C., Evans, R., 1997. Ligation independent cloning irrespective of restriction site compati-

bility. Nucleic Acids Res. 25, 4165–4166.

Lindler, L., 2004. Yersinia, Molecular and Cellular Biology. Horizon Bioscience, Norfolk, UK,

Ch. The Yersinia pestis-speciﬁc plasmids pFra and pPla.

Lindler, L., Plano, G., Burland, V., Mayhew, G., Blattner, F., 1998. Complete DNA sequence

and detailed analysis of the Yersinia pestis KIM5 plasmid encoding murine toxin and capsular

antigen. Infect. Immun. 66, 5731–5742.

Manoil, C., Mekalanos, J., Beckwith, J., 1990. Alkaline phosphatase fusions: sensors of subcel-

lular location. J. Bacteriol. 172, 515–518.

Marenne, M.-N., Mota, L., Cornelis, G., 2004. Yersinia, Molecular and Cellular Biology. Hori-

zon Bioscience, Norfolk, UK, Ch. The pYV plasmid and the Ysc-Yop Type III secretion

system, pp. 319–348.

Marketon, M., DePaolo, R., DeBord, K., Jabri, B., Schneewind, O., 2005. Plague bacteria

target immune cells during infection. Science 309, 1739–1741.

Marquenet, E., Richet, E., 2007. How integration of positive and negative regulatory signals

by a STAND signaling protein depends on ATP hydrolysis. Mol. Cell 28, 187–199.

Miller, J., 1992a. Experiments in Molecular Genetics. Cold Spring Harbor Laboratory, Cold

Spring Harbor, NY, Ch. Procedures for working with lac, pp. 71–80. 57

Miller, J., 1992b. A short course in bacterial genetics. Cold Spring Harbor Laboratory, Cold

Spring Harbor, NY.

Motin, V., Georgescu, A., Fitch, J., Gu, P., Nelson, D., Mabery, S., Garnham, J., Sokhansanj,

B., Ott, L., Coleman, M., Elliott, J., Kegelmeyer, L., Wyrobek, A., Slezak, T., Brubaker, R.,

Garcia, E., 2004. Temporal global changes in gene expression during temperature transition

in Yersinia pestis. J. Bacteriol. 186, 6298–6305.

Murphy, K., Campellone, K., 2003. Lambda Red-mediated recombinogenic engineering of en-

terohemorrhagic and enteropathogenic E. coli. BMC Mol. Biol. 4, 11.

Ni, B., Du, Z., Guo, Z., Zhang, Y., Yang, R., 2008. Curing of four diﬀerent plasmids in Yersinia

pestis using plasmid incompatibility. Lett. Appl. Microbiol. 47, 235–240.

Oster, C., Phillips, G., 2011. Vectors for ligation-independent construction of lacZ gene fusions

and cloning of PCR products using a nicking endonuclease. Plasmid 66, 180–185.

Panagiotidis, C., Boos, W., Shuman, H., 1998. The ATP-binding cassette subunit of the maltose

transporter MalK antagonizes MalT, the activator of the Escherichia coli mal regulon. Mol.

Microbiol. 30, 535–546.

Parkhill, J., Wren, B., Thomson, N., Titball, R., Holden, M., Prentice, M., Sebaihia, M., James,

K., Churcher, C., Mungall, K., Baker, S., Basham, D., Bentley, S., Brooks, K., Cerd˜eo-

T´arraga,A., Chillingworth, T., Cronin, A., Davies, R., Davis, P., Dougan, G., Feltwell, T.,

Hamlin, N., Holroyd, S., Jagels, K., Karlyshev, A., Leather, S., Moule, S., Oyston, P., Quail,

M., Rutherford, K., Simmonds, M., Skelton, J., Stevens, K., Whitehead, S., Barrell, B., 2001.

Genome sequence of Yersinia pestis, the causative agent of plague. Nature 413, 523–527.

Perry, R., Fetherston, J., 1997. Yersinia pestis — etiologic agent of plague. Clin. Microbiol.

Rev. 10, 35–66.

Perry, R., Straley, S., Fetherston, J., Rose, D., Gregor, J., Blattner, F., 1998. DNA sequencing

and analysis of the low-Ca2+-response plasmid pCD1 of Yersinia pestis KIM5. Infect. Immun.

66, 4611–4623. 58

Peterson, J., Phillips, G., 2008. New pSC101-derivative cloning vectors with elevated copy

numbers. Plasmid 59, 193–201.

Phillips, G., 2001. Green ﬂuorescent protein — a bright idea for the study of bacterial protein

localization. FEMS Microbiol. Lett. 204, 9–18.

Poteete, A., 2001. What makes the bacteriophage lambda Red system useful for genetic engi-

neering: Molecular mechanism and biological function. FEMS Microbiol. Lett. 201, 9–14.

Prentice, M., Gilbert, T., Cooper, A., 2004. Was the black death caused by Yersinia pestis?

Lancet Infect. Dis. 4, 72.

Quan, J., Tian, J., 2011. Circular polymerase extension cloning for high-throughput cloning of

complex and combinatorial DNA libraries. Nat. Protoc. 6, 242–251.

Raoult, D., Aboudharam, G., Crub´ezy, E., Larrouy, G., Ludes, B., Drancourt, M., 2000.

Molecular identiﬁcation by “suicide PCR” of Yersinia pestis as the agent of Medieval Black

Death. PNAS USA 97, 12800–12803.

Reyes, M., Shuman, H., 1988. Overproduction of MalK protein prevents expression of the

Escherichia coli mal regulon. J. Bacteriol. 170, 4598–4602.

Richet, E., Joly, N., Danot, O., 2005. Two domains of MalT, the activator of the Escherichia

coli maltose regulon, bear determinants essential for anti-activation by MalK. J. Mol. Biol.

347, 1–10.

Richet, E., Raibaud, O., 1987. Puriﬁcation and properties of the MalT protein, the transcription

activator of the Escherichia coli maltose regulon. J. Biol. Chem. 262, 12647–12653.

Richet, E., Raibaud, O., 1989. MalT, the regulatory protein of the Escherichia coli maltose

system, is an ATP-dependent transcriptional activator. EMBO J. 8, 981–987.

Sarker, M., Neyt, C., Stainier, I., Cornelis, G., 1998. The Yersinia Yop virulon: LcrV is required

for extrusion of the translocators YopB and YopD. J. Bacteriol. 180, 1207–1214. 59

Schauder, S., Shokat, K., Surette, M., Bassler, B., 2001. The LuxS family of bacterial autoin-

ducers: biosynthesis of a novel quorum-sensing signal molecule. Mol. Microbiol. 41, 463–476.

Schuenemann, V., Bos, K., DeWitte, S., Schmedes, S., Jamieson, J., Mittnik, A., Forrest,

S., Coombes, B., Wood, J., Earn, D., White, W., Krause, J., Poinar, H., 2011. Targeted

enrichment of ancient pathogens yielding the pPCP1 plasmid of Yersinia pestis from victims

of the Black Death. PNAS USA 108, E746–E752.

Sharan, S., Thomason, L., Kuznetsov, S., Court, D., 2009. Recombineering: a homologous

recombination-based method of genetic engineering. Nat. Protoc. 4, 206–223.

Shuman, H., Silhavy, T., 2004. The art and design of genetic screens: Escherichia coli. Nat.

Rev. Genet. 4, 419–431.

Silhavy, T., 2000. Gene fusions. J. Bacteriol. 182, 5935–5938.

Silhavy, T., Beckwith, J., 1985. Uses of lac fusions for the study of biological problems. Micro-

biol. Rev. 49, 398–418.

Skrzypek, E., Straley, S., 1995. Diﬀerential eﬀects of deletions in lcrV on secretion of V antigen,

regulation of the low-Ca2+ response, and virulence of Yersinia pestis. J. Bacteriol. 177, 2530–

2542.

Slauch, J., Silhavy, T., 1991. Genetic fusions as experimental tools. Methods Enzymol. 204,

213–248.

Southward, C., Surette, M., 2002. The dynamic microbe: green ﬂuorescent protein brings

bacteria to light. Mol. Microbiol. 45, 1191–1196.

Stewart, G., Williams, P., 1992. Lux genes and the applications of bacterial bioluminescence.

J. Gen. Microbiol. 138, 1289–1300.

Tachibana, A., Tohiguchi, K., Ueno, T., Setogawa, Y., Harada, A., Tanabe, T., 2009. Prepa-

ration of long sticky ends for universal ligation-independent cloning: Sequential T4 DNA

polymerase treatments. J. Biosci. Bioeng. 107, 668–669. 60

Torruellas, J., Jackson, M., Pennock, J., Plano, G., 2005. The Yersinia pestis type III secretion

needle plays a role in the regulation of Yop secretion. Mol. Microbiol. 57, 1719–1733. van Dyk, T., Wei, Y., Hanafey, M., Dolan, M., Reeve, M., Rafalski, J., Rothman-Denes, L.,

LaRossa, R., 2001. A genomic approach to gene fusion technology. Proc. Natl. Acad. Sci.

USA 98, 2555–2560.

Vroom, J., Wang, C., 2008. Modular construction of plasmids through ligation-free assembly

of vector components with oligonucleotide linkers. Biotechniques 44, 924–926.

Waldminghaus, T., Heidrich, N., Brantl, S., Narberhaus, F., 2007. FourU: a novel type of RNA

thermometer in Salmonella. Mol. Microbiol. 65, 413–424.

Wang, R., Kushner, S., 1991. Construction of versatile low-copy-number vectors for cloning,

sequencing and gene expression in Escherichia coli. Gene 100, 195–199.

Wang, S., Heilman, D., Liu, F., Giehl, T., Joshi, S., Huang, X., Chou, T., Goguen, J., Lu,

S., 2004. A DNA vaccine producing LcrV antigen in oligomers is eﬀective in protecting mice

from lethal mucosal challenge of plague. Vaccine 22, 3348–3357.

Waters, C., Bassler, B., 2005. Quorum sensing: cell-to-cell communication in bacteria. Annu.

Rev. Cell Dev. Biol. 21, 319–346.

Wood, J., DeWitte-Avi˜na,S., 2003. Was the Black Death yersinial plague? Lancet Infect. Dis.

3, 327–328.

Xavier, K., Bassler, B., 2003. LuxS quorum sensing: more than just a numbers game. Curr.

Opin. Microbiol. 6, 191–197.

Yang, J., Zhang, Z., Zhang, X., Luo, Q., 2010. A ligation-independent cloning method using

nicking DNA endonuclease. Biotechniques 49, 817–821.

Yother, J., Chamness, T., Goguen, J., 1986. Temperature-controlled plasmid regulon associated

with low calcium response in Yersinia pestis. J. Bacteriol. 165, 443–447. 61

Yu, D., Ellis, H., Lee, E., Jenkins, N., Copeland, N., Court, D., 2000. An eﬃcient recombination

system for chromosome engineering in Escherichia coli. PNAS USA 97, 5978–5983.

Zaslaver, A., Bren, A., Ronen, M., Itzkovitz, S., Kikoin, I., Shavit, S., Liebermeister, W.,

Surette, M., Alon, U., 2008. A comprehensive library of ﬂuorescent transcriptional reporters

for Escherichia coli. Nat. Met. 3, 623–628.

Zhang, P., Li, M., Elledge, S., 2002. Towards genetic genome projects: genomic library screening

and gene-targeting vector construction in a single step. Nat. Genet. 30, 31–39.

Zhang, P., Too, P., Samuelson, J., Chan, S., Vincze, T., Doucette, S., B¨ackstr¨om,S., Pota-

mousis, K., Schramm, T., Forrest, D., Schwartz, D., Xu, S., 2010. Engineering BspQI nicking

enzymes and application of N.BspQI in DNA labeling and production of single-strand DNA.

Protein Exp. Purif. 69, 226–234.

Zheleznaya, L., Kachalova, G., Artyukh, R., Yunusova, A., Perevyazova, T., Matvienko, N.,

2009. Nicking endonucleases. Biochemistry (Mosc) 74, 1457–1466.

Zhou, D., Han, Y., Song, Y., Huang, P., Yang, R., 2004. Comparative and evolutionary ge-

nomics of Yersinia pestis. Microbes Infect. 6, 1226–1234.

Zhou, D., Yang, R., 2009. Molecular Darwinian evolution of virulence in Yersinia pestis. Infect.

Immun. 77, 2242–2250.