Role of the Pdul Enzyme in the Function of Primitive Bacterial Organelles Found in Salmonella Yu Liu Iowa State University

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

2014 Role of the PduL enzyme in the function of primitive bacterial organelles found in Salmonella Yu Liu Iowa State University

Follow this and additional works at: https://lib.dr.iastate.edu/etd Part of the Allergy and Immunology Commons, Biochemistry Commons, Genetics Commons, Immunology and Infectious Disease Commons, and the Medical Immunology Commons

Recommended Citation Liu, Yu, "Role of the PduL enzyme in the function of primitive bacterial organelles found in Salmonella" (2014). Graduate Theses and Dissertations. 14214. https://lib.dr.iastate.edu/etd/14214

This Dissertation 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].

Role of the PduL enzyme in the function of primitive bacterial organelles found in Salmonella

Yu Liu

A dissertation submitted to the graduate faculty

in partial fulfillment of the requirements for the degree of

DOCTOR OF PHILOSOPHY

Major: Immunobiology

Program of Study Committee: Thomas A. Bobik, Co-major Professor Qijing Zhang, Co-major Professor Marian Kohut Alan Dispirito Michael Cho

Iowa State University

Ames, Iowa

2014

TABLE OF CONTENTS

LIST OF FIGURES……………………………………………………………………………….v

LIST OF TABLES………………………………………………………………………………vii

ABSTRACT…………………………………………………………………………...... viii

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

1.1.Salmonella and Salmonella Infection...... 1

1.2.Bacterial Microcompartment (MCP) ………………………………………………………5

1.2.1. Discovery…………………………………………………………………………….5

1.2.2. Distribution and Diversity……………………………………………………………6

1.2.3. Genetics……………………………………………………………………………..11

1.2.4. Functional Models…………………………………………………………………..16

1.2.5. BMC Domain and MCP Shell………………………………...... 22

1.2.6. Stability……………………………………………………………………...... 27

1.2.7. Association of MCPs with Salmonella Pathogenesis……………………………….28

1.2.8. Applications…………………………………………………………...... 30

1.2.9. MCP Shell vs. Viral Capsid……………………………………………...... 32

1.3.Vitamin B12…………………………………………………………………...... 34

1.3.1. Control of the cob/pdu Regulon…………………………………………………….34

1.3.2. Structure…………………………………………………………………………….36

1.3.3. Biosynthesis…………………………...... 37

1.3.4. Cobalamin-Dependent Reactions………………………………………...... 39

1.3.5. A Cobalamin Recycling System in the Pdu MCP………………………………….42

1.3.6. B12-Related Diseases…………………………………………………...... 43

iii

1.4.Coenzyme A (HS-CoA)… ………………………………………………...... 44

1.4.1. Discovery and Structure……………………………………………………………44

1.4.2. Biosynthesis…………………………………………………………………...... 45

1.4.3. CoA and its Thioester Derivatives…………………………………………...... 47

1.4.4. Regulation of CoA Biosynthesis…………………………………………………...47

1.4.5. Functions of CoA and CoA Derivatives…………………………………...... 52

1.4.6. HS-CoA-Related Diseases…………………………………………………………55

1.5.Protein Targeting Mechanisms………………………………………………..…...... 56

1.5.1. Target Recognition of the PDZ Domains…………………………………………..56

1.5.2. Protein Targeting to the Sec Complex………………………………………...... 58

1.5.3. Protein Encapsulation into Encapsulin-a Bacterial Nanocompartment…...... 61

1.5.4. Vault Ribonucleoprotein Particles……………………………………………...... 63

1.6.Research Overview…………………………………………………………...... 64

1.7.Thesis Organization……………………………………………………………………....65

CHAPTER 2. PDUL IS AN EVOLUTIONARY DISTINCT PHOSPHOTRANSACYLASE INVOLVED IN B12-DEPENDENT 1,2-PROPANEDIOL DEGRADATION BY SALMONELLA ENTERICA SEROVAR TYPHIMURIUM LT2...... 67

2.1. Abstract………………………………………………………………………………….68

2.2. Introduction……………………………………………………………………………...68

2.3. Materials and Methods………………………………………………………...... 71

2.4. Results…………………………………………………………………………………...77

2.5. Discussion…………………………………………………………………………...... 88

2.6. Acknowledgements……………………………………………………………………...90

CHAPTER 3. PDUL RECYLES COENZYME A IN PROPANEDIOL UTILIZATION MICROCOMPARTMENT IN SALMONELLA ENTERICA …………………………………..91

3.1. Abstract………………………………………………………………………………….92

3.2. Introduction…………………………………………………………………...... 92

3.3. Materials and Methods………………………………………………………………….96

3.4. Results………………………………………………………………...... 98

3.5. Discussion…………………………………………………………...... 107

3.6. Acknowledgements……………………………………………………...... 111

CHAPTER 4. PROTEIN-PROTEIN INTERACTIONS IN THE PDU MCP…...... 112

4.1. Abstract……………………………………………………………………...... 113

4.2. Introduction…………………………………………………………………………....113

4.3. Materials and Methods………………………………………………………………...115

4.4. Results………………………………………………………………………...... 117

4.5. Discussion……………………………………………………………...... 121

4.6. Acknowledgements……………………………………………………...... 122

CHAPTER 5. SUMMARY AND CONCLUSIONS………………………………………...123

REFERENCES…………………………………………………………...... 125

ACKNOWLEDGEMENTS…………………………………...... 141

NOMENCLATURE……………………………………………………………...... 143

LIST OF FIGURES

Figure 1.1 Electron micrograph of MCPs ...... 6

Figure 1.2 Electron micrograph of dividing Cyanobacterium and carboxysome ...... 8

Figure 1.3 Electron micrograph of the Pdu MCP of S. enterica ...... 9

Figure 1.4 Genomic organization of carboxysomes ...... 12

Figure 1.5 The pdu operon of S. enterica ...... 14

Figure 1.6 The eut operon of S. enterica ...... 15

Figure 1.7 Carbon concentration and carboxysome ...... 17

Figure 1.8 Model for 1,2-PD degradation by S. enterica ...... 20

Figure 1.9 Model for ethanolamine catabolism and MCP function ...... 21

Figure 1.10 A model for assembly of bacterial MCPs ...... 25

Figure 1.11 Control of the cob/pdu regulon in S. enterica ...... 35

Figure 1.12 Structure of cobalamin ...... 37

Figure 1.13 Biosynthesis of precorrin 6 ...... 38

Figure 1.14 Proposed pathway for the conversion of CNCbl into AdoCbl ...... 39

Figure 1.15 The proposed model for B12 recycling in Pdu MCP in S. enterica ...... 42

Figure 1.16 The structure of coenzyme A ...... 44

Figure 1.17 Schematic diagram of a universal pathway for CoA biosynthesis ...... 46

Figure 1.18 Chemical structures of CoA thioester derivatives ...... 48

Figure 1.19 Possible PDZ interaction modes ...... 59

Figure 2.1 Model for 1,2-PD degradation by S. enterica ...... 70

Figure 2.2 Growth of a pduL null mutant on 1,2-PD ...... 82

Figure 2.3 Ectopic expression of pduL corrects the growth defect of a pta mutant.... 83

Figure 2.4 Propionyl-CoA is produced by the PduL reaction ...... 85

Figure 2.5 SDS-PAGE analysis of PduL-His8 purification ...... 87

Figure 3.1 A role of PduL associated with Pdu MCP during 1,2-PD degradation ...... 94

Figure 3.2 Sequence analysis and structure prediction of PduL N-terminus ...... 99

Figure 3.3 PduL is a component of the Pdu MCP ...... 100

Figure 3.4 PduL-eGFP fusion protein was detected in Pdu MCPs ...... 101

Figure 3.5 N-terminal deletion affects encapsulation of PduL-Western blots ...... 102

Figure 3.6 PduL N-terminal 20 amino acids showed targeting effect ...... 103

Figure 3.7 Genetic effects of preventing MCP formation on pduL mutants ...... 104

Figure 3.8 Slow growth of a pta mutant was partially corrected by MCP- mutant ... 106

Figure 3.9 N-terminal deletion of PduL showed targeting defect-growth curve ...... 107

Figure 4.1 Pull-down indicated interaction of PduL and PduBB’ ...... 119

Figure 4.2 Western blots supported that PduL coeluted with PduBB’ ...... 120

vii

LIST OF TABLES

Table 2.1 Bacterial strains used in PduL characterization study ...... 72

Table 2.2 Phosphotransacylase activity in cell extracts from S. enterica mutants .... 80

Table 2.3 Kinetic constants for purified recombinant PduL-His8...... 86

Table 3.1 Bacterial strains used in cofactor recycling and targeting study ...... 97

Table 3.2 N-terminal deletion of pduL showed targeting defect on acetate/PD ...... 108

Table 4.1 Bacterial strains used in protein-protein interaction study ...... 116

Table 4.2 Two-hybrid assay to test the interaction with PduL ...... 118

Table 4.3 Two-hybrid assay to test the interaction between PduU and PduV ...... 120

viii

ABSTRACT

In contrast to conventional wisdom, bacteria are capable of forming highly organized organelles in the cytoplasm to optimize metabolic processes. Salmonella enterica can degrade the carbon source 1,2-propanediol (1,2-PD) within a proteinaceous organelle named the 1,2-PD utilization microcompartment (Pdu MCP). Prior enzymatic assays indicated that a phosphotransacylase (PTAC) was required for 1,2-PD degradation but the enzyme involved was not identified. Here we showed that the PduL enzyme is an evolutionarily distinct PTAC used for

1,2-PD degradation and that it is associated with the Pdu MCP where it plays a role in coenzyme

A homeostasis. A series of genetic and growth studies showed that the house-keeping phosphotransacetylase Pta was not required for 1,2-PD metabolism in Salmonella while pduL mutants were impaired for aerobic and anaerobic growth on 1,2-PD. High-pressure liquid chromatography (HPLC) and enzymatic assays determined that PduL catalyzes the

2- interconversion of propionyl-CoA and propionyl-PO4 . Biochemical studies indicated that PduL has kinetic properties appropriate to a role in 1,2-PD degradation. Bioinformatics analysis showed that the PduL enzyme doesn’t share sequence homology with known PTAC enzymes, indicating that it represents a new enzyme class.

Recent studies suggested that the cofactors required for MCP-associated metabolism, Ado-

+ B12 and NAD and HS-CoA are internally recycled within MCPs. We performed in vivo as well as in vitro studies to show that PduL is a component of the Pdu MCP and that it is used to regenerate coenzyme A inside the Pdu MCP. Western blots showed enrichment of PduL and also a PduL-green fluorescent protein (GFP) fusion in purified MCPs compared to crude cell extracts, indicating that PduL is an MCP component. Tests of genetic effects of breaking MCPs on PduL suggested that PduL is required for proper functions of intact MCPs. Further genetic studies

showed that PduL in MCPs can’t utilize acetyl-CoA in the cytosol to support Salmonella growth, and PduL recycling coenzyme A in the Pdu MCP is indicated. In addition, our studies also identified an N-terminal targeting sequence on PduL which mediates its encapsulation into the

Pdu MCP. This short targeting sequence could also mediate the encapsulation of GFP. Together with recent publications, these studies indicate that cofactor recycling is a general mechanism used by MCPs of various types to optimize metabolic efficiency. The targeting studies advance our knowledge on spatial organization of enzymes into MCPs.

We also initiated studies on the higher order structure of the Pdu MCP by exploring the binding interactions among varied Pdu proteins. These studies identified two possible binding interactions: PduL with PduBB’ and PduU with PduV. These results were consistent with an interactome network formulated using computational models. A better understanding of architecture of the Pdu MCP provide a basis for further study on its mechanisms of assembly.

CHAPTER 1. GENERAL INTRODUCTION

1.1. Salmonella and Salmonella Infection

Salmonella is facultative intracellular pathogen and leads predominantly a host-associated lifestyle. It is usually isolated from contaminated food or water sources, which act as the bacteria reservoirs. Salmonella can be divided into two groups, nontyphoidal and typhoidal serovars.

Nontyphoidal serovars usually cause self-limiting gastrointestinal diseases. Most cases of invasive nontyphoidal Salmonella infections (iNTS) are caused by S. typhimurium and S.

Enteritidis. Typhoidal serovars S. typhi and S. paratyphi A cause more severe typhoid fever and systemic infections. According to CDC (centers for disease control and prevention), Salmonella is one of the leading causes of food-borne illness. Every year, Salmonella causes one million illnesses in the United States, with 19,000 hospitalizations and 380 deaths. Infection with

Salmonella becomes a global economic and health burden.

Salmonella utilizes sophisticated strategy to infect human beings. Ingestion of food containing a high concentration of Salmonella is usually required to cause infection in healthy adults. Infection may occur due to the ingestion of a small amount of the bacteria for infants and young children or inhalation of Salmonella-containing dust in infants. Salmonella has evolved to tolerate acid environment to certain extent and a portion of the bacterial cells survive the gastric acid and manage to reach the intestine, where they start to move towards the intestinal epithelial cells using flagella. Then Salmonella docks to the host cells and injects effector proteins including SipA, SopA, SopB, SopD, and SopE2 via the Salmonella pathogenicity island-1 (SPI-

1)-encoded type III secretion system (TTSS) which are required for invasion and enteropathogenesis (1). Salmonella can survive in phagocytic and non-phagocytic cells, which requires a second TTSS encoded by the Salmonella pathogenicity island 2 (SPI-2) (2).

In humans, non-typhoidal Salmonella (NTS) serovars induce a strong Th1 response with high levels of IFN-γ, TNF-α, IL-18, IL-12, IL-15 and IL-10 detected in serum. Several chemokines induced upon NTS infection lead to the recruitment and activation of macrophages and dendritic cells as well as a significant influx of neutrophils into the intestinal lumen. While typhoidal serovars usually are not associated with a strong neutrophil infiltration or acute diarrhea. The immune response to infection with typhoidal serovars is complicated and involves both humoral and cell-mediated immune responses (3). Clinical studies showed a significant CD4 and CD8 T cell response to specific S. typhi antigens during typhoid fever, with elevated levels of IFN-γ,

TNF-α, IL-6, IL-8 and IL-10, indicating that the human immune response to S. typhi infection is predominantly Th1-associated.

S. typhimurium and S. typhi share about 96% DNA sequence homology while inducing very different clinical manifestations and immune responses in humans. About 200 genes in S. typhi are inactivated while most of their homologs in S. typhimurium are intact. Many of these inactivated genes in S. typhi genome are involved in motility and chemotaxis or encode for TTSS effectors, fimbriae, or adhesins that play a role in Salmonella pathogenicity (4). It is possible that differences in virulence and colonization factor composition affect interactions between the host and the pathogens and disease outcome in humans. It was proposed that the polysaccharide capsular antigen Vi in S. typhi enables resistance to phagocytosis and complement killing. Vi may also mask host access to pattern recognition molecules, causing less production of IL-8, a neutrophil chemoattractant, which limits neutrophil infiltration and thereby reduces intestinal inflammation (5, 6).

As part of the host innate immunity, the inflammatory process, to some extent, prevents the spread of Salmonella infection. Salmonella is prone to spread more efficiently and cause

systemic infection in the absence of a neutrophilic response both in mice and cattles. S. typhi which causes systemic infections doesn’t elicit a significant intestinal neutrophilic response (7, 8,

9). However, Salmonella evolved to take advantage of the host intestinal inflammatory response.

A small fraction of the Salmonella population actively invades the intestinal mucosa and triggers a massive acute inflammatory response. The resulting acute neutrophil infiltration effectively restrict the infection mostly to the enteric sites to prevent systemic dissemination of this pathogen, while at the same time also creates an intestinal environment which favors the remaining large fraction of Salmonella population which stays in the intestinal lumen for effective transmission to the next hosts (10).

In a healthy individual, the microbiota in the intestinal system prevents translocation of pathogenic bacteria to the mesenteric lymph node and hence prevent an undesirable immune response. Earlier studies demonstrated that disruption of the intestinal microbiota due to antibiotic treatment promotes translocation of S. typhimurium by phagocytes to mesenteric lymph node and results in increased susceptibility to Salmonella infection (11). Under certain circumstances, carbohydrates metabolized by commensal microbiota provide energy source for

Salmonella. Bacteroides thetaiotaomicron encodes sialidase which releases sialic acid from glycoconjugates, but doesn’t have enzymes to utilize sialic acid as a food source. The free sialic acid can feed S. typhimurium which lacks sialidase (12). S. typhimurium can also utilize fucose and rhamnose as well as their fermentation product 1,2-propanediol which are readily present in the gut and can’t be used by even the closely related enteric bacterial species E. coli. Besides the enzymatic machinery to degrade the above carbon molecules, Salmonella can use tetrothionate generated by host during intestinal inflammation as an alternative respiratory electron acceptor to

actually keep the degradation processes running in the microaerobic/anaerobic gut environment

(see section 1.1.7).

Several other mechanisms by which Salmonella takes advantage of intestinal inflammation have emerged recently as well. Lipocalin-2, a host antimicrobial peptide, is generated during inflammation in response to IL-17 and IL-22. It binds enterobactin, a siderophore produced by enteric bacteria and inhibits iron acquisition by bacterial organisms. Salmonella produces salmochelin, another siderophore which is not bound by lipocalin-2, and shows competitive advantage over other intestinal bacteria (13). Besides, iron deprivation in the inflamed intestine induces expression of colicin Ib by Salmonella, which is a bacteriocin against other

Enterobacteriaceae and provides additional competitive advantage (14). In the inflamed gut, calprotectin produced by neutrophils damages bacterial growth by sequestering zinc. However,

Salmonella overcomes this host self-protective mechanism by expressing a zinc transporter with high affinity called ZnuABC (14).

It is estimated that over 20% of malignancies worldwide are attributed to infectious agents

(15). It has been shown that bacteria can produce potent carcinogens including nitrite and suitable amines. A cohort study showed that chronic carriers of S. typhi/paratyphi have greatly increased risks of cancers of the biliary tract and hepatobiliary carcinoma, and also with malignancy in the colorectum, pancreas and lung (16, 17). Salmonella infection may increase risk for the high prevalence disease IBS (irritable bowel syndrome) by eight times (18).

Antibiotic treatment may prolong the duration of excretion of NTS and therefore is only recommended for infants, elderly, and immunocompromised individuals. Whereas, typhoid fever is always immediately treated with antibiotics. However, multidrug-resistance is an increasing problem in S. enterica serovars. Resistance to multiple antibiotics is especially common in S.

typhimurium and is linked to more severe disease outcome (19). New resistant strains of S. enterica are emerging worldwide. Currently, three types of vaccines against S. typhi are commercially available, but there is no licensed vaccine against S. paratyphi A. Vaccines against

NTS serovars typhimurium are only effective in poultry. There are no vaccines for NTS in humans or other animals such as cattles or pigs. These significant limitations in treatment and prevention strategies created urgent needs for the complete understanding of Salmonella pathobiology. The ultimate goal is to identify potential targets for future therapeutic and vaccine development.

1.2. Bacterial Microcompartment (MCP)

1.2.1. Discovery

Prokaryotic cells were previously thought to have no intracellular compartmentalization. In

1969, “crystalline bodies” were first discovered under electron microscope in blue-green algae

(20). The next year, polyhedral structured bodies were reported by Shively et al. (21) in seven

Thiobacillus species. They were later named carboxysomes, the first identified bacterial microcompartments (MCPs). In the late 1990s, Salmonella enterica was found to form a polyhedral organelle during growth on 1,2-propanediol (1,2-PD), which was named 1,2-PD utilization MCP (Pdu MCP) (22). Later studies identified ethanolamine utilization microcompartment (Eut MCP), another type of MCPs in S. enterica which was associated with ethanolamine utilization. Since their discoveries, MCPs have been called polyhedral bodies, enterosomes and sometimes metabolosomes. The name bacterial microcompartments will be used throughout the rest of this thesis.

Figure 1.1. Electron micrograph of MCPs. a the carboxysomes of H. neapolitanus; b the organelles formed during growth of S. enterica on 1,2-PD. Triangles point to the MCPs (23).

The bacterial MCPs that have been well studies are similar in structure and architecture.

They are polyhedral in shape and about 100-150 nm in cross section as shown in Figure 1.1.

They consist of a 3-4 nm multiple protein shell encapsulating a series of enzymes which function together in a specific metabolic pathway. The presence of three MCPs, carboxysomes, Pdu

MCPs, and Eut MCPs have been experimentally confirmed and their structures and functions have been investigated. All identified MCPs have protein shells composed primarily of bacterial microcompartment (BMC) domain proteins while each specific type of MCPs encapsulate different metabolic processes and have related functions. To date, MCPs have only been found in bacteria, no MCPs have been reported in eukaryotic cells or archaea. MCPs are purely protein complexes, and there is no evidence for lipid components.

1.2.2. Distribution and Diversity

Comparative genomics approaches identified BMC-domain containing proteins in genomes of 358 bacterial species across 11 different bacterial phyla, indicating about 14.6% of bacteria may form different types of MCPs with diverse functionalities (24, 25). The major proposed functions for MCPs are to retain toxic or volatile intermediates and increase metabolic efficiency

by substrate channeling. The function of each distinct MCP type is related to a specific metabolism process.

1.2.2.1.Carboxysomes

First discovered 45 years ago, carboxysomes have been extensively investigated. They were found in bacteria which use Calvin cycle for carbon fixation including probably all cyanobacteria, for example, Phormidium uncinatum, Nostoc punctiforme, Synechococcus elongates, and many chemoautotrophs, especially sulphur and nitrifying bacteria like Thiobacillus, Halothiobacillus, and Thiomonas (26-28). There are two types of carboxysomes defined by organization of related genes and by protein sequence comparisons. α-type carboxysomes are present in α-cyanobacteria

(cyanobacteria which express form 1A ribulose bisphosphate carboxylase/oxygenase (RuBisCO)) and many chemoautotrophic proteobacteria (23, 29). The composition of α-type carboxysomes was investigated in representative species facultative chemoautotroph Halothiobacillus neapolitanus. β-type carboxysomes are found in β-cyanobacteria which are cyanobacteria producing form 1B RuBisCO and represented by Synechocystis spp. PCC 6803 as shown in

Figure 1.2. Both α- and β-carboxysomes are composed of RuBisCO and carbonic anhydrase (CA) encapsulated within a protein shell made of homologous polypeptides (30, 31). In α- carboxysomes, CsoS1 proteins (homologous proteins CsoS1A-C) are the major shell components and they count for around 17% of total weight of carboxysomes. In β-carboxysomes, CcmK1-4 and CcmO are the homologues of Csos1 and they contribute to the formation of β-carboxysome shell.

1.2.2.2. Pdu MCP

In 1994, a protein containing a BMC domain was found to be involved in vitamin B12- dependent 1,2-PD utilization in Salmonella enterica serovar Typhimurium (referred to hereafter

as S. enterica) (33). Four years later, Pdu MCPs were observed in S. enterica and K. oxytoca by electron microscopy as shown in Figure 1.3 (34). 1,2-PD can be utilized as a sole carbon and energy source by several enteric bacteria including Salmonella, Listeria, Klebsiella, Shigella and

Yersinia. Pdu MCPs are around 120-160 nm in diameter, appear polyhedral in shape but slightly larger and more irregular compared to carboxysomes (35-37). Pdu MCP has a mass of about 600

MDa and consists of 18,000 individual polypeptides of 16-18 different types (37). Pdu MCPs are the most complicated MCPs elucidated so far at the molecular level in that more enzymes are encapsulated in the lumen of MCPs, the shell is made of more different types of polypeptides, the shape is less geometrically regular and the size is slightly larger (vs 80-140 nm for carboxysome shell) (22, 34, 38). Besides, 1,2-PD metabolism requires multiple enzymatic cofactors adenosyl-cobalamin (AdoCbl), coenzyme A and Nicotinamide adenine dinucleotide

(NAD+) and also a route for them to be taken up by the Pdu MCPs.

Figure 1.2. Thin-section electron micrograph of A a dividing cell of the cyanobacterium Synechocystis spp. PCC6803 (scale bar, 200 nm); B enlargement of a single carboxysome (scale bar, 50 nm) (32).

Figure 1.3. Electron micrograph of Pdu MCPs of S. enterica. Left panel a representative image of wild-type S. enterica (39); right panel purified Pdu MCPs from S. enterica (40).

1.2.2.3. Eut MCP

Back in 1999, while Pdu MCP-related studies were ongoing, Kofoid et al. found that besides the capability of forming Pdu MCPs, S. enterica can grow on ethanolamine as the only carbon, nitrogen and energy source (41). When observed by electron microscopy, Eut MCPs from K. oxytoca show similar size and appearance to Pdu MCPs during growth on ethanolamine (34).

Many bacteria classes including Gammaproteobacteria like Erwinia, Salmonella, Escherichia,

Klebsiella and Pseudomonas, Betaproteobacteria like Achromobacter, Firmicutes such as

Clostridium, Actinobacteria such as Arthrobacter, Mycobacterium and Corynebacterium, as well as Enterococcus and Flavobacterium, can use ethanolamine as a sole source of carbon and/or nitrogen (42, 43).

1.2.2.4.Glycyl Radical-Based Propanediol Utilization Microcompartment (Grp MCP)

Compared to the well-known carboxysomes, Pdu MCPs and Eut MCPs, the Grp type MCP encoding operon was relatively new. The grp operon encodes four core enzymes including a

glycyl radical diol dehydratase (DDH), glycyl radical activase, aldehyde dehydrogenase and a chaperone protein (44, 45). Further evidence indicated that 1,2-PD is the primary substrate in this system. The enzyme glycyl radical diol dehydratase has since been referred as B12-independent

1,2-PD dehydratase. This enzyme catalyzes a similar reaction as DDH in the Pdu MCPs and produces propionaldehyde. The proposed function for Grp MCPs is to mediate 1,2-PD degradation in a manner similar to that of 1,2-PD metabolism associated with Pdu MCPs. The major difference that distinguishes Grp MCP from Pdu MCP is that the first reaction which generates aldehyde is carried out by a B12-independent glycyl radical enzyme rather than the B12- dependent DDH. Also, the glycyl radical enzyme is sensitive to oxygen due to the presence of a glycyl radical in its activated state, which indicates that the Grp MCPs may be most important in strict anaerobic conditions (44). Whereas Pdu MCP-encapsulated 1,2-PD degradation could occur in both aerobic and anaerobic environments.

1.2.2.5.Other Prospective MCPs

Recently, Clostridium phytofermentans was found to contain three loci encoding MCP- related genes in the genome. Petit et al. (46) reported genomic, transcriptional and physiological evidence which indicates that the MCP encoding locus is involved in the metabolism of fucose and rhamnose and converts both deoxyhexose sugars into a variety of fermentation products such as propanol, propionate, ethanol, lactate and acetate through the pathways partially enclosed in a

MCP. Polyhedral structures resembling classical MCPs were visible by transmission electron microscopy of the cells growing on fucose. A model (46) proposed that fucose and rhamnose are taken up by C. phytofermentans through ABC transporters into the cell where they are converted to lactaldehyde and dihydroxyacetone phosphate (DHAP) by an aldolase encoded within the

MCP locus. DHAP can feed into glycolytic pathway and generate fermentation products

including ethanol, acetate, lactate and formate, while lactaldehyde may be converted to 1,2-PD by a PD-oxidoreductase. 1,2-PD could be used to generate propanol or propionate through MCP- associated 1,2-PD degradation pathway. Comparative genomic analysis suggested that the proposed pathway might be present in other Clostridium species as well (46).

Another unique MCP operon is present in four Mycobacterium species and contains two

BMC shell proteins and a minor vertex shell protein genes as well as four functional proteins or enzymes. Two of the enzymes have been shown to be involved in metabolizing amino alcohols

(44).

1.2.3. Genetics

Diverse MCPs share genomic characteristics. All MCP-associated proteins are encoded within operons consisting of multiple copies of paralogous genes encoding the shell proteins along with the genes for the associated enzymes. This unique genomic signature was used to predict the presence of multiple types of MCPs across a broad range of bacterial phyla.

1.2.3.1.Carboxysome Encoding Gene Clusters

Carboxysomes can be divided into α- and β- types which differ in their protein composition and the organization of their respective genes. α-carboxysomes belong to α-cyanobacteria which have a single operon to include type 1A RuBisCO and their carboxysome genes, whereas organisms with β- types have 1B RuBisCO and their carboxysome genes arranged in multiple gene clusters (47). The two type carboxysomes have homologous but specific shell proteins which are not present in the other type of carboxysomes. The main protein constituent of the α- carboxysome shell from H. neapolitanus is CsoS1A which shares significant sequence similarity with CcmK and CcmO of β-carboxysomes (47, 48). The csoS2 and csoS3 genes are exclusively

found only in α-carboxysomes while the ccmM and ccmN genes only occur in β-carboxysomes.

The genomic organization of the two types of carboxysomes are listed in Fig. 1.4.

Figure 1.4. Genomic organization of carboxysomes. a composition of α-carboxysomes from H. neapolitanus. Paralogous genes have same colors. The SDS gel in the center showed proteins identified in purified carboxysomes. The main constituents of carboxysomes are RuBisCO (green) and the shell proteins CsoS1A, CsoS1B and CsoS1C (blue); b carboxysome gene cluster from β-carboxysomes. CcaA is colored orange to indicate its status as a β-type carbonic anhydrase. CcmM is colored in two parts to indicate homology to ɣ-type carbonic anhydrase at its N-terminus and homology to RuBisCO small subunit at the C-terminus. White- colored genes don’t show homology within the bacterial MCP clusters that are shown (47).

1.2.3.2.The pdu Operon

The genes specifically required for 1,2-PD utilization by S. enterica form a single contiguous cluster, the pdu locus, which is contiguous to the cobalamin B12 synthesizing gene clusters (cob operon) at minute 44 of the genome (22). The pdu operon includes 23 genes pduABCDEFGHJKLMNOPQSTUVWX and pocR as shown in Figure 1.5 (22, 49). Based on experimental and bioinformatics analyses, seven pdu genes pduCDEPQLW are thought to encode five enzymes needed for the 1,2-PD degradative pathway which are diol dehydratase

(PduCDE, DDH), propionaldehyde dehydrogenase (PduP), phosphotransacylase (PduL), acetate kinase (PduW), and alcohol dehydrogenase (PduQ) (50-53); pduF encodes a hydrophobic protein with 65% sequence identity to the GlpF protein of E. coli which facilitates the diffusion of glycerol across the cytoplasmic membrane (33). Based on the homology between PduF and GlpF protein as well as the chemical similarity of 1,2-PD and glycerol, PduF may function as a 1,2-PD diffusion facilitation protein (33). The pocR gene is located between the pdu and cob operons

(the cob operon contains genes needed for de novo B12 synthesis, see section 1.2.1). C-terminus of PocR shares 55% similarity with AraC family of transcriptional regulatory proteins, and both

PocR and AraC contain a DNA-binding helix-turn-helix motif. Genetic study indicated that pocR codes for a positive transcriptional regulator of both the pdu and cob operons (33, 54); pduGH code for the large (PduG) and small (PduH) subunits of a putative DDH-reactivating factor and are probably involved in DDH reactivation (22); pduSO-encoded cobalamin reductase (PduS) and adenosyltransferase (PduO) are used for the conversion of inactive vitamin B12 hydroxycobalamin (OH-Cbl) to active coenzyme B12 (AdoCbl) (55, 56); seven genes pduABB’JKNTU encode shell components (22, 34, 40, 49), and pduM has been shown to be

another structural gene (39); pduX is involved in cobalamin synthetic pathway (57). And finally pduV has unknown functions.

Figure 1.5. The pdu operon of S. enterica. The genes involved in 1,2-PD degradation cluster at a single locus with shell protein encoding genes interspersed with enzyme encoding genes (58).

Analysis of purified Pdu MCPs showed that the MCPs constitute around 10% of total cell proteins which matches with the abundant occupancy of these structures under electron microscopy. Two-dimensional electrophoresis followed by densitometry determined that diol dehydratase PduCDE is the most abundant Pdu enzyme with the approximate molar ratio of

C:D:E:P:G:H:O as 8:4:8:4:2:2:1 (49). PduABB’J are major shell components and PduKTU are minor shell proteins with the approximate molar ratios as A:B:B’:J:K:T:U=10:6:7:15:1:1:2.

Deletion of pduA, pduBB’, pduJ or pduN gene significantly impaired MCP formation and caused toxic levels of propionaldehyde accumulated in the cell, indicating damaged integrity of the

MCP shell (59). Deletion of pduU or pduT had no phenotype under the conditions tested.

1.2.3.3. The eut Operon

Ethanolamine metabolism in S. enterica is encoded by the 17-gene cluster eut operon as shown in Figure 1.6 (41). The eut operon located at 50 min on the S. enterica chromosome (60).

The eut operon codes for five structural proteins including four BMC proteins EutK, EutL, EutM,

EutS and a pentameric protein EutN, four enzymes EutB, EutC, EutE, EutD which break down ethanolamine, two enzymes EutA and EutT for enzyme reactivation and cofactor recovery, an ethanolamine transport protein EutH, and a positive regulatory protein EutR for induction of eut operon transcription (41, 61-64). Besides, three proteins expressed from the eut operon including a chaperone protein EutJ and two other proteins EutPQ have unknown functions. Induction of eut operon depends on the presence of transcriptional activator EutR, ethanolamine and adenosylcobalamin.

Figure 1.6. The ethanolamine utilization (eut) operon of S. enterica (58).

1.2.4. Functional Models

As a subcellular organelle, the unique feature of MCPs is the use of a dense protein shell instead of lipid bilayer membrane. The unique MCP shell plays a crucial role in the functions of diverse MCPs (47, 58, 65, 66). First of all, the shell functions as a diffusion barrier, which protect volatile intermediates from being lost through diffusion, while at the same time preventing contact-mediated toxicity caused to bacterial genomic DNA and cytoplasm proteins by the toxic intermediates (aldehyde in Pdu MCP and Eut MCP). Also, the shell shields the interior metabolic enzymes from reacting with reactive or competing molecules. Secondly, intensive structural studies indicated shell proteins bear central pores in their structures which may selectively transport substrates, products and even large cofactors involved in specific encapsulated pathways. Thirdly, several shell proteins have been found to interact with and encapsulate the core enzymes into the lumen of the specific MCP during MCP assembly and also possibly regulate the spatial organization of the enzymes for their optimal functions. Molecular transport through the shell will be discussed in this section. Interaction between shell proteins and core enzymes will be illustrated in chapter 4. Besides sharing the common features among different types of MCPs, the specific function of each MCP is defined by the metabolic processes which is involved in.

1.2.4.1.Carboxysomes-Carbon Fixation

Autotrophic prokaryotic cells accumulate inorganic carbon and improve carbon fixation through carbon-concentrating mechanism (CCM) as shown in Figure 1.7. During the first stage

- of CCM, bicarbonate (HCO3 ) is transported into the cell by transmembrane pumps and is concentrated inside the cell. The second stage of CCM starts with the translocation of bicarbonate into carboxysome. There carbonic anhydrase (CA) converts bicarbonate to CO2

which is incorporated into ribulose-1,5-bisphosphate (RuBP) by RuBisCO to form two molecules of 3-phosphoglycerate (3-PGA) which is then released from carboxysomes. RuBisCO is a kinetically inefficient enzyme. It catalyzes the competing fixation of both O2 (known as photorespiration) and CO2, it has relatively low Kcat value to substrate CO2, and it must bind a nonsubstrate CO2 at the activation site for the optimal activity (37, 67). Carboxysomes contribute to carbon fixation by retaining high concentration of CO2 in the vicinity to RuBisCO in a confined space to optimize the function of RuBisCO and hence minimize photorespiration. Since most of the oceanic microorganisms (cyanobacteria) carry out CO2 fixation in MCPs, it was estimated that >25% of global CO2 fixation occurs within carboxysomes (68, 69).

Figure 1.7. Carbon concentration and the carboxysome. The carboxysome is involved in the second part of the CCM. The two enzymes CA and RuBisCO in the carboxysome and their reactions are shown in reactions 1 and 2. Phosphoglycolate is generated from the side reaction where O2 is the substrate for RuBP instead of CO2 (47).

1.2.4.2.Pdu MCP-Cobalamin B12-Dependent 1,2-PD Metabolism

The current model for 1,2-PD metabolism associated with Pdu MCPs proposes that, by retaining this metabolic pathway in MCPs, bacteria are able to utilize 1,2-PD as carbon and energy source while avoiding cellular toxicity caused by aldehyde accumulation, preventing diffusive carbon loss and increasing metabolism efficiency by substrate channeling.

A working model was developed over time as pdu operon was identified in S. enterica. Like the gene clusters for α-carboxysomes, shell protein encoding genes were found interspersed with enzyme encoding genes, which provided a first clue that 1,2-PD metabolism pathway is associated with MCPs (33). As propionaldehyde was identified as an intermediate during 1,2-PD degradation, it was suggested that MCPs may form to sequester the toxic aldehyde (70). Later a comprehensive genetic study found that deletion of major Pdu shell components PduA, PduBB’,

PduJ and PduN significantly damaged MCP formation and resulted in accumulation of aldehyde and cellular toxicity during growth on 1,2-PD, which is consistent with the proposed function for

Pdu MCP (59). The propionaldehyde toxicity was represented by growth arrest for about 20 hours and then growth of S. enterica resumed, which was interpreted as the aldehyde intermediate accumulated until 1,2-PD in the medium was depleted, then the aldehyde level dropped due to the consumption by PduP and PduQ (59). Thus, it was reasonable to propose that

MCPs retain the toxic intermediate aldehyde while allowing Salmonella to utilize 1,2-PD as carbon and energy source. Recent structural studies on MCP shell proteins suggested the shell is composed of molecular layers formed from BMC-domain proteins with pentameric non-BMC proteins closing the vertices of the MCPs and giving rise to the icosahedral structures. Further

+ studies suggested vitamin B12 and NAD , the required cofactors for enzymes PduCDE and PduP,

are recycled inside the Pdu MCP. With all of the studies described above, the current model for

Pdu MCP emerged.

According to this model as shown in Figure 1.8, 1,2-PD is taken up by S. enterica from the growth medium probably by the diffusion facilitator PduF and then is transported into the lumen of Pdu MCP by an unclear mechanism. Recent crystallographic and genetic studies suggested it is possible that 1,2-PD may be transported through central pores on hexameric shell component

PduA or pseudohexameric shell protein PduB (36, 71-73). The first enzymatic reaction in the lumen is the conversion of 1,2-PD into propionaldehyde by AdoCbl-dependent diol dehydratase

PduCDE. Propionaldehyde is oxidized by CoA-dependent propionaldehyde dehydrogenase PduP to produce propionyl-CoA or alternatively reduced by alcohol dehydrogenase PduQ to generate

1-propanol. By forming and excreting propanol, the cells can keep internal redox balance (74).

Propionyl-CoA can feed into methylcitrate pathway or be consumed by phosphotransacylase

2- 2- PduL to produce propionyl-PO4 (according to our study in chapter 2). Propionyl-PO4 exits

MCP and is converted to propionate by propionate kinase PduW.

1,2-PD degradation generates one ATP by substrate level phosphorylation catalyzed by

PduW, one electron sink (propionaldehyde), and a three-carbon intermediate propionyl-CoA which may feed into the central metabolism via methylcitrate pathway (75, 76). Under anaerobic conditions, the methyl citrate pathway is inoperative due to the absence of an exogenous electron acceptor. Since both propanol and propionate are excreted, anaerobic 1,2-PD metabolism produces an electron sink and a source of ATP but can’t generate cell carbon (75, 77).

Recent studies propose that cofactor AdoCbl for DDH, cofactors NAD+ and HS-CoA for

PduP may be recycled internally inside the Pdu MCP by cobalamin reductase PduS as well as adenosyltransferase PduO (see section 1.2.5), 1-propanol dehydrogenase PduQ and

phosphotransacylase PduL (see chapter 3), respectively. In a recent study, Cheng et al. (52) showed that PduQ functions to help regenerate electron carriers and retain redox balance during

1,2-PD fermentation by converting a portion of the propionaldehyde to 1-propanol and NADH to

NAD+ to avoid depleting NAD+ within the Pdu MCP lumen. Extensive structural studies suggested that substrate, cofactors and products of 1,2-PD metabolism might be transported through the pores in the shell proteins (see section 1.1.5). Further studies are needed for a thorough understanding of the molecular transport mechanisms of MCPs.

Figure 1.8. Model for 1,2-PD degradation by S. enterica. 1,2-PD is transported through the shell and is degraded within the Pdu MCP. The MCP protein shell functions as a diffusion barrier to retain propionaldehyde to eliminate toxicity and carbon loss. Propionaldehyde is converted to propionyl-CoA which either feeds into methyl citrate pathway or is used to produce propionate (58).

1.2.4.3. Eut MCP-Ethanolamine Degradation

Ethanolamine degradation in Salmonella is carried out analogously to Pdu MCP-associated

1,2-PD degradation except ethanolamine is utilized instead of 1,2-PD and acetaldehyde is the toxic intermediate generated during the process rather than propionaldehyde. As shown in Figure

1.9, the first step is carried out by vitamin B12-dependent EutBC to convert ethanolamine to ammonia and acetaldehyde which is further metabolized to alcohol and CoA-esters within Eut

MCP. By retaining acetaldehyde within Eut MCP eliminates the damage to cellular DNA and also prevent carbon loss by diffusion of aldehyde out of MCP and across cell membrane.

Figure 1.9. Model for ethanolamine catabolism and Eut MCP function. The outer double rectangle represents the cell membrane, the dashed line confines the Eut MCP. Two large cofactors HS-CoA and NAD+ are recycled within the MCPs by enzyme EutE, EutD and EutG. This model proposes that acetaldehyde is retained inside the Eut MCP (white arrow) while substrate ethanolamine as well as intermediate ethanol and acetate can diffuse freely through the MCP shell (black arrows). According to this model, enzyme EutBC is located on the outside surface of the MCP and converts ethanolamine to acetaldehyde, which is released into the lumen of MCP and retained there (78).

1.2.5. BMC Domain and MCP Shell

MCPs are subcellular compartments with a specific function in bacteria cell, and the protein shell acts as a selective barrier, which fits the conventional definition of an organelle, except that

MCPs are exclusively proteinaceous bodies instead of membrane bound organelles.

Many bacterial MCP shell proteins share a conserved sequence referred to as bacterial microcompartment (BMC) domain (Pf00936). All genomes associated with MCPs usually have multiple paralogous copies of BMC shell proteins encoded together. Several thousand copies of these shell proteins of different types constitute a single microcompartment shell. Functionally diverse MCPs are all assembled from this protein family. To date, the BMC domain containing

MCP shell proteins have been found in only bacteria, no BMC proteins were found in archaea or eukaryotes. The only exception is that carboxysome shell proteins were once reported in the amoeboid Paulinella chromatophora (79).

English et al. (48) first reported the sequence of a major carboxysome shell peptide encoding gene csoS1 from proteobacteria T. neapolitanus, the organism from which carboxysomes were first purified in 1973 (30). The sequence shares homology with several ORFs from cyanobacteria

Synechococcus. All the other types of MCPs were discovered basically by identifying genes for

BMC domain-containing shell protein clustered with genes encoding metabolic enzymes during genomic sequence analysis followed by biophysical confirmation of observed polyhedral bodies by electron microscopy. Since the first crystal structure of a BMC domain protein was published in 2005 (35), crystallography has been an important tool for investigating assembling mechanisms in MCPs.

Early structural studies focused on shell proteins which are about 100 amino acids long and contain a single BMC domain. In 2005, the first crystal structures of BMC domain-containing

shell proteins CcmK2 and CcmK4 of β-carboxysomes were determined (35). N-terminal 80 amino acids of the BMC domain adopt an α/β-fold, six identical monomers self-assemble to form a wedge-shaped symmetric hexamer with a narrow pore down the center of the hexameric structure. The following key shell components from different types of MCPs also apply the symmetric hexameric structures and also have the central pores but with different sizes and charge properties due to the residues in the pore area which may be relate to their different molecular transport specificities on the shell (36, 80). These unique disc-shaped symmetric hexamers serve as the main building blocks for the facets of the MCP shell. According to

Crowley et al. (36), Pdu MCP key shell protein PduA forms a symmetric homohexamer and retains a disk-like hexagon, which is consistent to the majority of the BMC shell protein hexamers from other types of MCPs. Unlike other shell proteins that form flat hexameric structures, EutS shows bent hexamers which presumably form the edges where two facets of the

BMC shell meet (24, 66).

More recently, structure studies revealed two minor carboxysome shell components CcmL from Syn. 6803 and OrfA in H. neapolitanus assembling to form symmetric pentamers instead of hexamers (38). These pentameric proteins don’t contain BMC domains. Their subunits are packed tightly around the axis of symmetry with narrow pores around 3-3.5 Å in the center. Only

60 copies of the CcmL or OrfA are present among four to five thousand hexameric shell subunits.

Electron microscopy studies confirmed that carboxysomes are approximately icosahedral in shape (81, 82). Usually the architecture of large icosahedral structures shows a combination of hexamers and pentamers with hexamers constituting flat hexagonal sheet and pentamers generating curvature on the sheets and occupying the vertices of an icosahedron (83, 84).

Homologues to pentameric CcmL and OrfA were also found in other types of MCPs such as

EutN, PduN, and GrpN. Crystal structures of GrpN confirmed that it is pentameric (85). The original structural study indicated that PduN formed a hexamer, but further study demonstrated that EutN is a pentamer (85, 86). The name bacterial microcompartment vertex (BMV) was proposed to emphasize the special role of pentameric vertical proteins (85). Based on the resolved structures of hexameric and pentameric components of BMC-domain shell proteins, it was proposed, as shown in Figure 1.10, that the several thousand of BMC proteins self-assemble into cyclic hexameric units packed in a tile-like molecular layers forming the flat facets; minor

BMV proteins assemble into pentagonal units close the vertices of the MCPs giving rise to roughly icosahedral structures.

The structural studies suggested that the MCP shell doesn’t just function as a segregating device, but is also involved in controlling metabolic flow (54). The selectivity features of MCPs are opposite to those of lipid membranes in that MCP protein shell favors passage of polar instead of nonpolar molecules. Compared to carboxysomes, Pdu MCP and Eut MCP have special needs to transport large metabolic cofactors for 1,2-PD and ethanolamine degradation

(nicotinamide, cobalamin, coenzyme A and phosphorylated ATP derivatives). Current theories based on structural and genetic studies propose that the cofactors could be either moved across the MCP shell or regenerated internally in the MCPs, and both scenarios may partially contribute the supply of the bulky cofactor molecules. Structural and genomic studies indicated that double-

BMC-domain protein encoding genes are present in both pdu and eut operons. These proteins usually have open and closed conformations with large gated pores which may help transport bulky cofactors through the shell (66, 87). Systematic experimental studies are needed to confirm the involvement of gated pores in transport of cofactors.

Figure 1.10. A model for assembly of bacterial microcompartments. BMC-domain proteins (top left) form hexamers or pseudohexameric trimers which pack into a tile-like molecular layer (top right) as the majority of the shell. BMV proteins (bottom left) occupy the vertices of the polyhedron. Together they form a closed icosahedron (bottom right). BMC proteins have central pores possibly for molecular transport. (65).

Carboxysome shell proteins CcmK2 and CcmK4 have pores about 7 Å and 4 Å in diameter, respectively. Since each of the six subunits contribute a positively charged amino acid residue

Lys36 or Arg38 to the pore, a large net positive electrostatic potential is present in the pore region, which may attract negatively charged metabolites, particularly substrate bicarbonate, to the pore

(35). While the hydrophobic CO2 intermediate will not have the same electrostatic attraction to the pore as bicarbonate and hence is retained in the lumen of carboxysomes. At the same time the neutral O2 which competes with CO2 for binding RuBisCO doesn’t show any advantage of

passing through the positively charged pore area hence can’t get into carboxysomes. The PduA central pores with about 10 Å in diameter are mostly polar due to the presence of hydrogen bond acceptors in the center of the pores. This favors transport of the substrate 1,2-PD compared to the intermediate propionaldehyde (36, 73). X-ray diffraction data showed a glycerol molecule which is a 1,2-PD analog is found in the central pore of one PduA point mutant, indicating the possible function of PduA is to move 1,2-PD into the lumen of the Pdu MCP (71). Based on the fact that

PduJ, another Pdu BMC shell protein, has 86% sequence identity with PduA and identical pore- lining residues, it is possible that PduJ has similar structure and pores as PduA. Pang et al. (72) investigated another major MCP shell component PduB in Lactobacillus reuteri and found three glycerol molecules present in each pore within the trimeric structure of PduB. The amino acid residues lining in the pore area have affinity for glycerol and may function as a channel for substrate glycerol, whereas the hydrogen bonds at the glycerol-binding sites may also prevent efflux of hydrophobic intermediate aldehyde from the MCPs (72).

In 2009, the structures of the shell protein with tandem BMC-domain was first solved for

CsoS1D from α-carboxysome (87). Tandem BMC domain proteins contain two BMC domains duplicated in conjunction with each other to form trimers with a total of six BMC domains in a pseudohexameric symmetric arrangement, which are structurally similar to hexagonal shape of single BMC domain shell proteins. CsoS1D has a conserved Arg side chain which goes through conformational changes and causes the pores to either close or open up to 14 angstrom in diameter. Later structural studies suggested the presence of gated pores within another tandem

BMC protein EutL (88). EutL has three pores per pseudohexameric unit which may allow increased flux across the shell compared to one pore per unit area. Based on these observations, a gated opening in response to specific conditions or signals was suggested. For Pdu, Eut and Grp

systems, transport of large cofactors might explain the existence of the large gated pores. In carboxysomes, the gated pores might allow transport of substrate ribulose bisphosphate.

Tandem BMC shell protein PduT and some Grp MCP proteins bear an iron-sulfur cluster in the center of the pore region (36, 44). In the case of PduT, each subunit from the homotrimer contributes a single cysteine ligand to the 4Fe-4S cluster, which calls for the 4th ligand to an iron atom. This specific feature suggested that PduT may contribute to the regeneration of encapsulated redox cofactors by moving electrons across the shell. Alternatively, PduT may transport intact [4Fe-4S] clusters into the MCPs to replace the damaged [4Fe-4S] clusters in the lumen enzyme PduS (36).

Most recently, it has been found that point mutations replacing conserved edge lysine residues with alanine on PduA, PduJ or PduB are severely impaired for edge contacts with neighboring shell components, which cause unstable MCP sheet or blocking closing of the shell due to weak and nonspecific edge contact with other BMC domain proteins (73). Specifically,

K26, N29 and R79 on PduA are key residues which form hydrogen bonds within same hexamer and also with neighboring hexameric structures. This interaction hub strengthens the edges between the hexamers and thus stabilize the MCP shell. K26 and R79 are conserved across several BMC domain proteins among different MCPs, hence this could be a general mechanism for stabilizing MCP architecture.

1.2.6. Stability

Sinha et al. (73) initiated the investigation of the phenotypic effects of structural mutations of

Pdu MCP by identifying key surface residues in one major Pdu MCP protein PduA that are crucial to stabilize the MCP shell. Shortly after that, Kim et al. (89) tested the robustness of Pdu

MCP shell with regard to time, pH and temperature, and concluded that Pdu MCPs retain

integrity in various physiological relevant conditions which makes MCPs suitable for a diversity of applications esp. in vivo applications. MCPs are stable for at least 6 days at either 4°C or room temperature in vivo and remain stable in vitro for a longer period of time. MCPs display normal morphology between pH 6 and pH 10, although TEM images showed aggregates at pH 9 and above. Buffers with pH range below 5 or above 11 caused deterioration of MCP structure and reduced size. Also purified MCPs can tolerate temperatures up to 60°C without visible structural degradation under EM.

1.2.7. Association of MCPs with Salmonella Pathogenesis

1,2-PD is a main fermentation product of rhamnose and fucose (90, 91). L-rhamnose is a common component of plants cell walls and the outer cell membrane of certain Mycobacteria species (92). L-fucose is also present in the glycoconjugates of intestinal epithelial cells (50).

Hence, 1,2-PD is readily available in the anaerobic environment in the human and animal gut as well as aquatic sediments. Ethanolamine is present in the gastrointestinal tract of mammals as a product of degradation of phosphatidylethanolamine which is an essential component of membrane in both prokaryotic and eukaryotic cells and is constantly shed together with intestinal epithelial cells into the lumen of the gut (93).

1,2-PD and ethanolamine are present in the intestinal system but don’t not serve a role as carbon sources without a terminal electron acceptor (94). In the normal gut environment, anaerobic metabolism of microbiota constantly produce large amounts of hydrogen sulfide (H2S) which binds to mitochondrial respiratory enzymes and prevents oxidative phosphorylation during cellular respiration and hence causes cellular anoxia. Large intestinal mucosa convert H2S to

2- thiosulfate (S2O3 ) to prevent the toxicity. S. enterica, when introduced into the gut, employs two type III secretion systems T3SS-1 and T3SS-2 to effectively elicit acute intestinal

inflammation, penetrate the intestinal epithelial cells and survive in tissue macrophages (95).

During acute inflammation, neutrophils migrate from blood vessels to the site of infection via chemotaxis and release nitric oxide radicals (NO) and reactive oxygen species, the process known as host respiratory burst, which results in oxidation of thiosulfate to tetrothionate. Genetic studies indicated that tetrothionate can serve as an alternative electron acceptor to support anaerobic respiration hence promote anaerobic growth on ethanolamine or 1,2-PD (75).

The ability of S. enterica to take advantage of host-generated 1,2-PD as well as ethanolamine clearly provides growth advantages for Salmonella in an inflamed gut. By using tetrothionate as alternative electron acceptor, Salmonella can utilize 1,2-PD and ethanolamine as the only carbon, nitrogen and energy source, which may be one of the virulence strategies that S. enterica utilizes to exploit inflammation to overcome colonization resistance (protection by high density of commensal microbiota inhabiting the intestine and shielding from infection) as well as competing with other enteropathogenic bacteria. By using in vivo expression technology (IVET),

Conner et al. (96) pinpointed that genomic region containing pdu and cob operons are specifically induced during infection. Competitive index assay (CI) showed that removal of the pdu region conferred virulence defect in systemic survival, which indicated that the capability of using alternative carbon sources in the gut contribute to Salmonella virulence (96, 97). The findings that the pdu operon is induced in host cells may provide a starting point for drug and vaccine development against certain enteropathogenic bacterial infections. Further structural studies are needed to identify exposed peptide regions on BMC shell proteins.

One comparative genomic study identified a 55-kb gene locus in another enteric pathogen

Listeria monocytogenes genome which is highly homologous to cob-pdu regulon and eut operon in S. enterica suggested a role of this locus in anaerobic growth and perhaps also in virulence of

L. monocytogenes in the host (98). Indeed, by analyzing the transcriptional profile of intracellular

L. monocytogenes and validating the biological relevance of the intracellular gene expression profile, Joseph et al. (99) showed that, instead of competing with the host cells for their dominant carbon and nitrogen sources, L. monocytogenes can utilize alternative carbon and nitrogen source ethanolamine during replication in epithelial cells, which allows the longer survival in the host.

Thus, MCP-dependent 1,2-PD and ethanolamine utilizations contribute to pathogenesis and dissemination of enteric pathogens containing pdu and/or eut operons.

1.2.8. Applications

1.2.8.1.Bioreactor

Recently studies have been testing the feasibility of encapsulate metabolic pathways for producing high-value compounds within MCPs. It has been shown that coordinated expression of shell proteins PduA-B-J-K-N-U in E. coli generated empty MCPs and it is possible to target proteins to outside (PduV) or inside (PduC, PduD) of MCPs (100). Choudhary et al. (101) successfully expressed Salmonella enterica Eut MCP shell proteins in E. coli resulting in the formation of polyhedral protein shells. They also found one single shell protein EutS sufficiently formed the MCPs. By tagging a pyruvate decarboxylase and an alcohol dehydrogenase with

PduP N-terminal targeting sequence, Lawrence et al. (102) was able to efficiently generate ethanol from pyruvate and showed that the strains with modified empty MCPs produce alcohol with higher yield. Heterologous GFP proteins and β-galactosidase were successfully encapsulated into recombinant Eut MCPs by N-terminal 19 amino acids of EutC (101). Enzyme

RuBisCO from heterologous species was also been engineered into carboxysomes (103). These studies provide a proof-of-concept for engineering of MCPs for biosynthesis and biocatalysis.

1.2.8.2.Drug Delivery

Many common chemotherapeutic agents are small hydrophobic molecules and the use of potentially toxic organic solvents are required for effective delivery (104). Moreover, the use of therapeutic proteins and peptides is prevented by the fast elimination from the circulation, enzymatic degradation and accumulation in normal organs and tissues (105). Current nanoscale drug delivery systems include liposomes, polymer particles and virus like particles (VLPs) (106).

Liposomes and polymers may cause reduced activities of the peptides/enzymes due to chemical linking and liposomes sometimes exhibit drug leaking issues. VLPs have advantages compared to the lipid vesicles and polymers but have limited capacity (hundreds of proteins) and well- documented immunogenicity (107).

The main function of MCPs is to retain small hydrophobic and toxic aldehyde within the shell, which makes MCPs suitable for delivering the chemotherapeutic drugs as well as peptide- based anticancer drugs. Compared to the above mentioned delivering systems, the use of MCPs as drug carriers may results in higher active moiety: carrier ratio, better protection against degradation of drug activity without affecting normal cells because of their large internal volume and the demonstrated capability of keeping the activities of interior enzymes.

Two recent studies identified phenotypic effects of structural mutations of Pdu MCP and investigated the stability of Pdu MCPs over time, at elevated temperatures as well as in a diverse range of pH (73, 89). MCPs were found to retain the integrity in various physiological relevant conditions in regard to temperature and pH range which make MCPs great candidate for drug delivery (89). The robustness of MCPs helps prevent unspecific release of toxic or chemotherapeutic drugs. More importantly the breaking of the MCPs can be controlled by changing pH value of the buffers, meaning it is possible to regulate the release of the

encapsulated molecules, which opens the possibility of applying MCP in medical field, especially for drug delivery.

1.2.8.3.Reuterin Production in Probiotics

Probiotic bacterium Lactobacillus reuteri produces reuterin by fermentation of glycerol.

Reuterin is an equilibrium of 3-hydroxypropioaldehyde (3-HPA), HPA-hydrate and HPA-dimer.

It has a broad spectrum of antimicrobial activity against viruses, prokaryotes (Gram+ and Gram- bacteria) as well as eukaryotes (fungi and parasites) and is also described as anti-cancer agent

(108). Sriramulu et al. (109) reported that maximal reuterin production by L. reuteri was achieved with preincubation with 1,2-PD together with glycerol compared to glycerol alone.

Further evidence indicated that Pdu MCP-associated diol dehydratase is responsible to produce

3-HPA from glycerol. Probiotic bacteria are known to beneficially affect the host by inhibit pathogenic bacteria due to the excretion of reuterin while not interfere with normal microbiota in the gut. One problem of producing reuterin using probiotic bacteria is the low yield. To utilize

MCPs as a reuterin bioreactor may increase the production of this molecule and may have further application in pharmaceutical industry.

1.2.9. MCP Shell vs. Viral Capsid

When carboxysomes were first discovered by electron microscopy, they were thought to be viral particles, and attempts were made to induce a viral lytic response without success (110).

Through years of studies on MCPs, now it is clear that MCPs have complete different contents and functions compared to viral capsids. Still, structural studies drew an architectural parallel between bacterial MCPs and viral particles esp. the ultrastructure and the underlying design principles. Both MCP shell and capsid have symmetrical structural featuring icosahedral shape.

In MCPs, the conserved BMC domain shell proteins form cyclic hexamers, which bind side by

side to form layers representing the flat facets of the icosahedral shell (32). Certain minor shell components tend to form pentamers which serve as vertex to close the flat sheets and finally for a closed shell. The fact that different shell proteins may serve different roles in the architecture of the MCP shell resembles the structures of many viral capsids which also features combination of hexamers and pentamers, which applies to certain virus structure architecture as well.

So far there is no evident sequence resemblance or structural similarities between MCP shell proteins and viral capsid protein components, partly because there are relatively small number of viral capsid proteins have their structures solved by crystallization. The structural similarity between bacterial MCPs and virus may reflect geometric discipline of self-assembling macromolecular protein complex (24).

Crystal structures on MCP shell proteins proposed possibilities that hexameric central pores as well as the gaps between hexamers of BMC shell proteins. Similarly, certain viruses contains interstitial spaces between the capsid protein subunits which may also mediate the flow of metabolites. Yeast L-A viral capsid particle has 18 angstrom diameter openings at the icosahedral five-fold axes which provides a portal for the entry of nucleotide triphosphates and exit of the viral mRNA (111). The crystal structure of bluetongue virus icosahedral protein capsid disclosed that the pores on the 5-fold icosahedral axis allow the egress of mRNA through the interaction of RNA backbone phosphates with the amino acid side chains that line the pore

(112). Also a different binding site on the capsid provides a selective channel for input and output of substrates and byproducts. However, bluetongue virus differs from MCPs shells in that encountering the physiological environment rich in Mg2+ and other ions results in the overall expansion of the capsid esp. around the 5-fold axes which was not observed in MCPs. The structural studies also showed that interactions between the C-terminal tails of the BMC domain

proteins could influence microcompartment domain assembly in the same way that the flexible termini of certain viral capsid proteins often participate as switches for distinct types of interactions in the mature viral capsid (35).

Cowpea Chlorotic Mottle Virus (CCMV) capsid can dissemble and reassemble depending on pH (113). There has been one report that showed that Pdu MCPs are sensitive to pH and the stability requires pH between 6 and 10 (89). Whether Pdu MCPs also fall apart and reform according to pH is unclear and future studies are needed to clarify it.

1.3. Vitamin B12

Vitamin B12, also known as cobalamin, is a water-soluble vitamin with a crucial physiological role in cellular metabolism. Vitamin B12 is the largest and most structurally complicated vitamin and is synthesized only in bacteria and archaea. Adenosylcobalamin

(AdoCbl) and methylcobalamin (CH3Cbl) are two cofactors for a variety of enzymes found in bacteria and higher animals. Animals assimilate cobalamin through diet, while fungi and plants don’t utilize B12.

1.3.1. Control of the cob/pdu Regulon

The cob operon is located at minute 44 of the genetic map of S. enterica genome, which is adjacent to the pdu operon (see section 1.1.3.2). Enzymes encoded from cob operon are used for

B12 synthesis. The two operons are induced by 1,2-PD using a single regulatory protein PocR

(77). Two global regulatory systems Crp/Cya and ArcA/ArcB regulate inducibility of both cob and pdu operons in that both operons are activated aerobically and anaerobically by Crp protein and anaerobically by ArcA protein (77, 114). During anaerobic respiration of a poor carbon source, the Crp and ArcA proteins function additively and produce the maximum inducibility.

The control of the cob/pdu regulon relies on five promoters which are all located in the central

region between the two operons and four of them are activated by PocR. As shown in Figure

1.11, the cob and pdu operons are transcribed divergently from promoters Ppdu and Pcob, respectively. Ppdu and Pcob are activated when both PocR and 1,2-PD are present. Global control of the regulon is carried out by regulating PocR protein level. Transcription of the pocR gene is controlled by Ppoc which is regulated only by Crp/cAMuP, P2 which is autoregulated by PocR and requires ArcA and P1 whose regulation may involves Fnr or Crp in addition to PocR.

Figure 1.11. Control of the cob/pdu regulon in S. enterica. Boxes indicate structural genes. Black arrows designate transcripts. Gray arrows indicate regulatory effects. Dashed gray arrows indicate that higher level of PocR may be required to activate the promoter (77).

The regulon is proposed to have three states (77). During aerobic growth on glucose, the regulon is in the off state, all promoters are at their lowest level and PocR is expressed by the basal levels of Ppoc, P1 and P2. During aerobic growth on a poor carbon source and/or anaerobic growth, cells switch to a standby state. At this state, cells grow under global conditions appropriate for induction while no 1,2-PD is present, PocR expression increases, but the regulon remains uninduced. The presence of 1,2-PD is sensed by PocR resulting in the induction of both

P1 and P2, which increase PocR level and pushes the system to the on state. Thus, the high level of PocR/1,2-PD induces the expression of Ppdu and Pcob.

There are 30 genes used for cobalamin synthesis (25 constitute the cob operon) and 22 genes used for 1,2-PD metabolism (pduX from the pdu operon used for cobalamin synthesis), which together count for >1% of Salmonella genome (53 of 4747 genes in S. enterica LT2) (115).

Maintenance of such large collection of genes requires a strong selective force during evolution.

The co-regulation of the cob operon and pdu operon suggested that B12-dependent utilization of

1,2-PD could be a main reason for B12 synthesis in Salmonella spp (77).

1.3.2. Structure

Cobalamin has three parts: a central ring, a nucleotide loop, and a variable moiety, for example, a cyano group (-CN), a hydroxyl group (-OH), a methyl group (-CH3) or a 5’- deoxyadenosyl group, which yields the four B12 forms cyanocobalamin, hydroxocobalamin, methylcobalamin and adenosylcobalamin, respectively. The structure of cobalamin is based on a central corrin ring, which is structurally and biosynthetically similar to the porphyrin ring found in heme, chlorophyll, and a cytochrome. The central ring of B12 differs from other related rings by the lack of a carbon bridge between the A and D porphyrins, the ring oxidation state, the distribution of ring decorations, and by the unique central metal ion cobalt. Four of the six coordination sites are occupied by the corrin ring, while cobalt’s lower (Coα) axial ligand is N-7 of dimethylbenzimidazole (DMB) which is covalently bonded to the corrin ring as part of a nucleotide loop (58) (Fig. 1.12). Ado-B12 also has a 5’ deoxyadenosyl moiety acting as its upper

(Coβ) axial ligand, and the 5’ carbon of this group is joined by a covalent bond to the cobalt within the corrin ring. In the case of methylcobalamin, the deoxyadenosyl moiety is replaced by a methyl group (77).

1.3.3. Biosynthesis

Only bacteria and archaea contain the enzymes required for de novo synthesis of the central corrin ring. Certain bacterial species synthesize the corrinoid ring under aerobic conditions, others generate the corrin ring anaerobically. The two pathways may use nonhomologous enzymes to produce analogous intermediates during the biosynthesis. The first part of the process is identical between the aerobic and anaerobic routes, then at the later stage, the aerobic pathway incorporates oxygen as a prerequisite for the ring formation, while anaerobic route takes advantages of the chelated cobalt ion for the ring contraction (Fig. 1.13).

Figure 1.12. Structure of cobalamin. The central carbon atoms are numbered, and peripheral amidated carboxyl groups are lettered. The R group stands for -CN, -CH3, and adenosyl group (115).

Four major stages are required to synthesize cobalamin from the five-carbon precursor 5- aminolaevulinic acid (ALA). The first stage is the biosynthesis of uroporphyrinogen III (Uro III) from ALA. Uro III is a precursor of corrins as well as heme, siroheme, chlorophylls and factor

F430. Its biosynthesis from ALA involves the utilization of the enzymes ALA dehydratase

(HemB), porphobilinogen deaminase (HemC) and uro’gen III synthase (HemD). The second stage entails the conversion of Uro III to cobinamide by a series of reactions including extensive methylation of the porphyrin ring, amidation of carboxyl groups, removal of a ring carbon, addition of the R group and the aminopropanol side chain. In the aerobic pathway, this stage results in the precorrin 6, while the anaerobic pathway is distinguished by the early incorporation of cobalt and results in cobalt-precorrin 6 rather than precorrin 6. On stage II, DMB moiety is synthesized possibly from flavin molecules. Stage III involves the covalent linkage of cobinamide, DMB, and a phosphoribosyl moiety derived from an NAD precursor.

Figure 1.13. Biosynthesis of precorrin 6 through aerobic pathway (light arrows) and cobalt-precorrin 6 by anaerobic pathway (bold arrows) during B12 synthesis (116).

In mammals, exogenous cobalamin precursors cyanocobalamin (CNCbl) and hydroxycobalamin (OH-Cbl) are obtained through the diet and converted into cofactors AdoCbl and CH3Cbl. As shown in Figure 1.14, assimilation of CNCbl is carried out by the conversion of

CNCbl to cob(II)alamin by decyanation, further to cob(I)alamin by reduction, and finally to

AdoCbl by transfer of a 5’-deoxyadenosyl group from ATP. The enzymes involved in this process are β-ligand transferase, cob(III)alamin reductase, cob(II)alamin reductase and adenosyltransferase, respectively (56, 63, 117-123). Assimilation of OH-Cbl occurs in a similar pathway except that the first step is the reduction of OH-Cbl to cob(II)alamin by cobalamin reductase or the reducing environment of the cell (124, 125).

Figure 1.14. Proposed pathway for the conversion of CNCbl into AdoCbl (originally from (126), modified in (74))

1.3.4. Cobalamin-Dependent Reactions

AdoCbl and CH3Cbl act as cofactors for three types of enzymes: isomerases, methyltransferases, and dehalogenases (127). In humans, AdoCbl-dependent methylmalonyl-

CoA mutase MUT and CH3Cbl-dependent methionine synthase MTR play a key role for the formation of blood and in the normal functions of central nervous system.

1.3.4.1. AdoCbl-Dependent Isomerases

In bacteria, isomerases are the largest family of cobalamin-dependent enzymes which are important in fermentation pathways. These enzymes catalyze concomitant exchange of hydrogen atoms as well as a second substituent which can be a carbon atom, an oxygen atom of an alcohol, or a nitrogen atom of an amine. The isomerases in this family include methylmalonyl-CoA mutase, glutamate mutase, methyleneglutarate mutase, isobutyryl-CoA mutase, lysine 5,6 aminomutase, diol dehydratase and ribonucleotide reductase, etc. (127).

In enteric bacteria, the primary role of the cofactor B12 may be to support fermentation of small molecules by generating both an oxidizable molecule and an electron sink for balancing redox reactions. This role has been observed in Ado-B12-dependent degradation of ethanolamine,

1,2-PD and glycerol by ethanolamine ammonia lyase, propanediol dehydratases and glycerol dehydratase, respectively. The intermediate aldehyde can be oxidized to provide ATP and the oxidation reactions can be balanced by reducing part of aldehyde to an alcohol which is later excreted outside the cell. Ethanolamine is present in nature as components of lipids, phosphatidyl ethanolamine and phosphatidyl choline. Ethanolamine ammonia lyase produces ethanol and acetaldehyde which can be utilized to provide carbon and energy through acetyl-CoA in a Eut

MCP-dependent manner. Similarly, 1,2-PD is converted by dehydratase to propionaldehyde which is used as carbon and energy source by the function of Pdu MCPs. In nonenteric bacteria,

AdoCbl-dependent amino mutases catalyze similar reactions that support fermentation of glutamic acid, lysine, leucine, and ornithine.

The only AdoCbl-dependent isomerase that is also found in human beings is methylmalonyl-

CoA mutase MUT. The MUT function is necessary for proper myelin synthesis and is important in proper function of the central nervous system.

1.3.4.2. CH3Cbl-Dependent Methyltransferases

CH3Cbl-dependent methyltransferases play a crucial role in amino acid metabolism in both humans and bacteria. They are also important in carbon metabolism and CO2 fixation in anaerobic microbes (127). The enzymes catalyze the transfer of a methyl group from a methyl donor to a methyl group acceptor. Cofactor CH3Cbl is cleaved heterolytically and acts as a methyl group carrier. Methionine synthetase transfers a methyl group from methyl- tetrahydrofolate to homocysteine as the last step in methionine synthesis. Salmonella and E. coli both express a B12-dependent methionine synthase MetH and a B12-independent MetE which catalyze the same reaction. MetH is preferred when CH3Cbl is available, while MetE is induced to respond to homocysteine accumulation when MetH is inactive (128). The only CH3Cbl- dependent enzyme in human beings is methionine synthase MTR, which is important in blood formation.

1.3.4.3. B12-Dependent Reductive Dehalogenases

Bacteria containing B12-dependent reductive dehalogenases provide benefits for the environment by detoxification of aromatic and aliphatic chlorinated organics, many of which, for example, chlorinated phenols, chlorinated ethane, and polychlorinated biphenyls (PCBs), are on the top list of EPA (Environmental Protection Agency) priority pollutants (127). Even though both anaerobic and aerobic microbes can carry out dehalogenation, anaerobic bacteria are more efficient in removing halogen atoms from polyhalogenated molecules (129). Anaerobic organisms with the capability of reductive dehalogenation include Desulfomonile, Dehalobacter, and Desulfitobacterium (130). These organisms have different preference for the chlorine substituent they can remove (131, 132).

1.3.5. A Cobalamin Recycling System in the Pdu MCP

As mentioned above, the first step of 1,2-PD degradation in the lumen of Pdu MCP is the conversion of 1,2-PD into propionaldehyde catalyzed by AdoCbl-dependent diol dehydratase

(DDH). The adenosyl-group of AdoCbl is unstable and constantly lost and replaced by a hydroxyl group, resulting in inactivation of DDH by tight binding of inactive OH-Cbl to the active site (133, 134). DDH reactivase PduGH releases OH-Cbl and apo-DDH. Next OH-Cbl goes through reduction catalyzed by reductase PduS to cob(I)alamin and then adenosylation by adenosyltransferase PduO to active AdoCbl. AdoCbl automatically associates with apo-DDH to form active holoenzyme.

Figure 1.15. The proposed model for B12 recycling inside the Pdu MCP in S. enterica. Recycling of the inactive OH-Cbl-DDH into active AdoCbl is carried out by diol dehydratase reactivase (PduGH), cobalamin reductase (PduS) and adenosyltransferase (PduO). The process results in reactivation of AdoCbl and DDH (55).

All of the enzymes (PduGH, PduS and PduO) involved in B12 recycling have been established as components of the Pdu MCP, which suggested that cobalamin recycling occurs completely within the lumen of the MCPs as shown in Figure 1.15 (49, 55). The recycling of

AdoCbl is required for the proper activity of PduCDE and hence is essential for of 1,2-PD metabolism.

1.3.6. B12-Related Diseases

In humans, tetrahydrofolate (THF) plays a vital role in DNA synthesis and reduced THF levels caused by B12 deficiency results in loss of functionality of MTR enzyme which causes rapid turnover of red blood cells and intestinal epithelial cells and causes megaloblastic anemia and Pernicious anemia. Also the accumulation of homocysteine in the blood and urine may cause long term damage to arteries and clotting with increased risk of stroke and heart attack.

Myelin is a dielectric material which forms the myelin sheath around the axon of a neuron.

The main function of myelin sheath is to increase the speed of impulses propagation and prevent the electrical current from leaving the axon which are essential for the proper functions of the nervous system. B12 deficiency abolishes the function of human MUT, which causes elevated level of methylmalonic acid (MMA), a myelin destabilizer. Increased MMA level prevents normal fatty acid synthesis, and MMA may be assimilated into fatty acid rather than the normal malonic acid. Myelin with MMA incorporated becomes fragile resulting in demyelination, which is associated with degeneration of central nervous system and spinal cord, but the mechanisms involved are unclear (135).

Damaged MTR reactions due to B12 deficiency also have neurological effects. MTR reaction produces methionine which is required to generate S-adenosyl-methionine (SAM) is necessary for the methylation of myelin sheath phospholipids. In addition, SAM is involved in production

of neurotransmitters and hormones including epinephrine, norepinephrine and dopamine, and also in brain metabolism. Inadequate levels of SAM thus may cause depression as well as neurodegeneration (136).

1.4. Coenzyme A (HS-CoA)

1.4.1. Discovery and Structure

HS-CoA is an essential intermediate in many biosynthetic and energy-yielding metabolic pathways and also a crucial regulator of several key metabolic reactions. Every genome sequenced so far encode enzymes that utilize coenzyme A as a substrate. Coenzyme A was discovered by Fritz Lipmann accidently when he studied acetylation on the amino group of the drug sulfonamide in 1940s. He named it coenzyme A (CoA) where “A” stood for “activation of acetate”. Lipmann shared the Nobel Prize in Physiology or Medicine in 1953 with Hans Adolf

Krebs “for his discovery of co-enzyme A and its importance for intermediary metabolism”

(Krebs won the Prize “for his discovery of the citric acid cycle”). Later, Lipmann showed that

CoA consists of adenosine 5’-phosphate pantothenic acid and a sulfhydryl moiety (137, 138).

Figure 1.16. The structure of coenzyme A. A 2D structure cited from http://pubchem.ncbi.nlm.nih.gov/; B 3D stick model cited from Wikimedia Commons.

The structure of coenzyme A, as shown in Figure 1.16, was defined in the early 1950s. It is a coenzyme containing among its constituents pantothenic acid and a terminal thiol group that forms high-energy thioester linkages with various acids.

1.4.2. Biosynthesis

Biosynthesis of coenzyme A is a highly conserved process in all living organisms. CoA is synthesized from pantothenate and cysteine in a five-step process which consumes four ATP molecules and the pathway intermediates are common in prokaryotic and eukaryotic cells.

Pantothenate is commonly known as vitamin B5, which can be synthesized de novo in plants, fungi and most bacterial species, but is an essential nutrient for animals and certain microbes.

Vitamin B5 is usually present in sufficient quantities in the diet and is also produced by bacteria in human intestines. Pantothenate is taken into cells through specific carrier pantothenate permease (PanF) in E. coli, a hydrogen symporter in yeast or a sodium-dependent multivitamin transporter in the intestinal epithelium of mammals (139). In eukaryotic cells, the biosynthesis process starts with phosphorylation of pantothenate to 4'-phosphopantothenate by pantothenate kinase (PanK), which is rate-limiting step in most organisms (140, 141). 4’-P-pantothenate is condensed with cysteine consuming one ATP to generate 4’-P-pantothenoylcysteine by 4′- phosphopantothenoylcysteine synthase (PPCS). 4′-phosphopantothenoylcysteine decarboxylase

(PPCDC) catalyzes the decarboxylation of 4′-phosphopantothenoylcysteine to yield 4′- phosphopantetheine. Next, 4′-phosphopantetheine adenosyltransferase (PPAT) catalyzes the second rate-limiting reaction in CoA biosynthesis--transfer of AMP to form dephospho-CoA, which is finally phosphorylated to produce CoA by dephospho-CoA kinase (DPCK). A couple of studies pointed out the specific interaction between PanK1β and Coenzyme A synthase (CoAsy), the first and last enzymes of the CoA biosynthetic pathway, and showed CoAsy might be present

in a complex (142, 143). Although no publication to date demonstrated the existence of a CoA biosynthetic complex in mammals.

Figure 1.17. Schematic diagram of a universal pathway for CoA biosynthesis and its key players in bacteria, yeast and mammals (139).

In bacteria, the above five enzymes are instead called CoaA, CoaBC (bifunctional enzyme),

CoaD and CoaE respectively (141, 144, 145). To date, at least three types of CoaA/PanK have been identified. Type I is expressed in E. coli and is efficiently inhibited by CoA. Type II is present in S. aureus and is not inhibited by CoA and derivatives at all. The more recently identified type III enzyme is the only known PanK found especially in pathogenic bacterial species. PanK-III adopts a different structural fold compared to that of E. coli CoaA, and hence is resistant to CoA feedback inhibition (144). Exploration of inhibitors targeting PanK-III may

result in development of novel antibiotics (146). These bacterial CoA synthetic enzymes are usually present in quaternary structure: CoaA is a homodimer, CoaBC is a homododecamer,

CoaD and CoaE are trimers. Unlike in the yeast, where the enzymes for CoA biosynthesis are grouped together, bacterial Coa enzymes don’t associate in multienzyme complexes.

Intensive investigation has been focusing on subcellular localization of CoA biosynthetic enzymes, esp. the rate limiting PanK and CoAsy, in mammalian cells. PanK1α isoform is exclusively localized in the nucleus while PanK1β shows cytosolic localization. Human PanK2 possesses an N-terminal regulatory region which shows functional nuclear localization as well as export signals and is observed in the nucleus and the mitochondria (147).

1.4.3. CoA and its Thioester Derivatives

The unique structure of CoA relates to its function as a major acyl group carrier and carbonyl-activating group, the reactions generate various metabolically active thioester derivatives, including acetyl-CoA, malonyl-CoA and 3-hydroxy-3-methylglutaryl-CoA (HMG-

CoA) (Fig. 1.18). The levels of CoA and its derivatives in mammalian cells and tissues are tightly regulated by various extracellular stimuli, including nutrients, hormones and cellular metabolites. The ratio between CoA and its thioester derivatives is important for cellular homoeostasis.

1.4.4 Regulation of CoA Biosynthesis

The levels of CoA and its thioester derivatives are tightly regulated at various levels and in different cellular compartments by extracellular and intracellular signals. The regulatory mechanisms may include gene expression of the biosynthetic enzymes; posttranslational modifications; compartmentalization; metabolic flux of CoA/CoA thioester; and CoA degradation. Fasting, glucagon and glucocorticoids increase the level of CoA (149, 150), while

insulin, glucose, fatty acids and pyruvate cause decrease of CoA level (151, 152). Pathological conditions, such as diabetes, Reye’s syndrome, cancer, vitamin B12 deficiency and cardiac hypertrophy as well as treatment with hypolipidaemic drugs have been reported to change intracellular coenzyme A level (149, 153-155).

Figure 1.18. Chemical structures of CoA thioester derivatives (148).

1.4.4.1 Feedback Regulation

Feedback regulation of PanK is carried out by CoA and CoA thioesters and controls the overall availability of CoA in response to a cell’s metabolic state. In E. coli, culturing condition with different carbon sources basically decide the total abundance of CoA pool as well as the relative distribution of CoA species (156). Growth on glucose generates 15 times more pantothenate than the amount utilized for CoA synthesis, indicating that CoaA enzymatic

reaction is rate limiting for conversion from pantothenate to CoA (157). The CoaA activity is reduced in vivo with she shift from glucose to acetate, reflecting that CoaA is less sensitive to inhibition by acetyl-CoA (the major CoA species growing on glucose) compared to the inhibition by free CoA (the main CoA species growing on acetate). In bacteria, this lower sensitivity to acetyl-CoA allows the fasting expansion of the CoA pool to meet the metabolic requirements of fast growing cells on glucose. The eukaryotic PanK is also under the feedback regulation by the

CoA pool. The mouse PanK1 isoforms (mPanK1α and mPanK1β) are differentially regulated by

CoA species. The human PanK2 is localized in mitochondria and is very sensitive to inhibition from long-chain acyl-CoA, acetyl-CoA and malonyl-CoA, suggesting its enzymatic activity is diminished in most metabolic circumstances. All the other PanK enzymes are present in cytosol and hence are not sensitive to the levels and changes of the CoA pool.

Changes in the most abundant CoA species and ATP concentration work cooperatively modulate the rate of CoA biosynthesis (141). Mammalian CoAsy is present in complex with proteins involved in diverse signaling pathways, such as p85α regulatory subunit of phosphoinositide 3-kinase (PI3K) (158), Src homology 2 domain-containing protein tyrosine phosphatase (Shp2PTP) and tyrosine kinase Src (158), and enhancer of mRNA-decapping protein 4 (EDC4) (159). Tyrosine phosphorylation of CoAsy by Src tyrosine kinase family members is required for its interaction with PI3K. Moreover, EDC4 was shown to inhibit the dephospho-CoA kinase activity of CoAsy in vitro (159). The mechanisms behind these interactions are not very clear, their roles in regulation of formation of a possible CoA biosynthetic complex can’t be ruled out.

1.4.4.2 Regulation by Compartmentalization

In mammalian cells, CoA level varies with different compartments. CoA concentrations are estimated to be 0.7 mM and 2.2-5.0 mM in peroxisomes and mitochondria, respectively. The high level of mitochondrial CoA is used as a cofactor in the TCA cycle and fatty acid β- oxidation, and the concentrations of CoA pool regulate the rate of these processes. The level of free CoA determines the speed of oxidation-dependent energy production in both peroxisomes and mitochondria. Whereas CoA levels as low as 0.02 to 0.14 mM have been observed in cytosols of different tissues and organs, where it is utilized for lipid synthesis, fatty acid oxidation, protein modification and also membrane trafficking. The endoplasmic reticulum is another compartment for CoA. Acetyl-CoA is transported into the lumen of this organelle for formation of O-acetylated gangliosides and other sugar components. How CoA generated from these processes returns to the cytoplasm is unknown.

1.4.4.3. Various Size and Composition in CoA/CoA Thioester Pool in Bacteria

In bacteria E. coli, growing in different carbon sources lead to various size and composition in CoA pool. The CoA level is as high as 400 uM when E. coli grows on glucose and acetyl-CoA is the major derivative (156). While growing in amino acids mixture derived from hydrolyzed casein, the pool is only around 100 uM. This comparison indicates that coenzyme is important for the generation of amino acids from glucose through the Krebs cycle. Acetyl-CoA is the dominant component of the CoA pool (79.8%) during exponential growth in glucose minimal medium (160). Total amount of acetate in the cell controls the levels of CoA in bacteria. When the number of acetyl-CoA decreases significantly in an acetate auxotroph of E. coli, the amount of free CoA drops accordingly, and the excessive CoA is converted to 4′-phosphopantetheine and pumped out of the cell (161). While the increase of intracellular CoA by overexpressing CoaA

enzyme in E. coli causes increased carbon flux to acetate production (162). In S. aureus, CoA level accumulates to millimolar levels due to the facts that the CoaA enzymatic activity from this bacterium is not feedback-regulated by CoA pool. It actually meets the physiological needs of S. aureus since it lacks glutathione and hence depends on CoA/CoA disulfide reductase to keep the redox balance and protect from the oxidative damage (163).

1.4.4.4. Regulation by Gene Expression

In most bacterial species, coaA gene expression level controls the upper limit for the intracellular CoA concentration. The coaA promoter doesn’t share homology with consensus promoter sequence in E. coli, and the coaA sequence contains codons with low frequency of usage, which result in lower abundance of CoaA protein relative to the average E. coli proteins

(164). Even though CoaA was manipulated to be overexpressed from a multicopy plasmid in E. coli, feedback inhibition of the enzyme limit the increase of the CoA level (164). On the other hand, in higher animals, PanK expression is modulated in a long-term response to diet and diseases, while the mechanisms were unclear (141).

1.4.4.5. Regulation by CoA Degradation

CoA can be degraded by dephosphorylation to dephospho-CoA, cleavage of the phosphodiester bond to release 4'-phosphopantothenate, or direct transfer of the 4'- phosphopantetheine moiety on CoA to other carrier proteins such as acyl carrier protein (ACP) in bacteria and fatty acid synthase (FAS) in eukaryotic cells. ACP is the essential acyl group carrier which interacts with enzymes in fatty acid synthesis and FAS is a multi-enzyme protein that generate long-chain saturated fatty acid from acetyl-CoA and malonyl-CoA. In bacteria, another regulatory mechanism is mediated by a phosphodiesterase. Fast decrease of acetate supply stimulates the response from the CoA pool and results in reduction of acetyl-CoA level, increase

of free CoA transiently and finally CoA degradation and transported outside of the cells (161).

Nudix hydrolase has been identified to contribute in the above reactions (165).

1.4.5. Functions of CoA and CoA Derivatives

CoA is an essential metabolic cofactor for all living organisms. CoA basically functions as an acyl group carrier and carbonyl-activating group contributing to central metabolism including citric acid cycle and fatty acid synthesis. Around 4% of enzymes use it as an obligate cofactor.

It’s involved in more than 100 metabolic reactions (166, 167). CoA and its derivatives are involved in the catabolism of proteins, carbohydrates and lipids through metabolic reactions which releasing the energy from food in the form of acetyl-CoA.

1.4.5.1. The Citric Acid Cycle

Acetyl-CoA is the starting point for the citric acid cycle. The primary sources of acetyl-CoA are decarboxylation of pyruvate which comes from the breakdown of sugars and the oxidation of fatty acids. The citric acid cycle begins with the transfer of the acetyl group from acetyl-CoA to oxaloacetate to form six-carbon citrate. The carbons from acetyl-CoA become part of the oxaloacetate carbon backbone after the first TCA cycle. After several turns of TCA cycle, acetyl-

CoA-donated carbons are released as CO2. The TCA cycle utilizes acetyl-CoA derived from carbohydrate, fats and proteins to generate chemical energy in the form of adenosine triphosphate (ATP).

1.4.5.2. Synthesis of Fatty Acids

Acetyl-CoA and malonyl-CoA are used to generate palmitate by fatty acid synthase I (FAS I) in animals and yeast, and by FAS II in bacteria, plants, fungi, parasites and mitochondria. In the presence of NADPH, long-chain saturated fatty acids are produced. Desaturation of fatty acids in bacteria occurs in the anaerobic conditions and is dependent on β-hydroxydecanoyl-ACP

dehydrase FabA and β-ketoacyl-ACP synthase FabB in E. coli or their homologues in other bacteria to add the double bond before elongation by using the fatty acid synthesis machinery.

Aerobic desaturation is utilized by all eukaryotic and some prokaryotic cells and it uses desaturases to produce unsaturated fatty acid from full-length saturated fatty acid. Unsaturated fatty acid is essential component of prokaryotic and eukaryotic cell membrane, signal molecules,

Fatty acid synthesis is an important part of the lipogenesis process. Lipogenesis and glycolysis together contribute to create fats from sugar molecules in blood stream in living organisms.

1.4.5.3. Synthesis of Cholesterol and Ketone Bodies

Two acetyl-CoA molecules can be condensed to generate acetoacetyl-CoA, which is hydrated together with another acetyl-CoA molecule to form HMG-CoA. HMG-CoA feeds into mevalonate pathway/HMG-CoA reductase pathway to produce dimethylallyl pyrophosphate

(DMAPP) and isopentenyl pyrophosphate (IPP). The mevalonate pathway is present in all higher eukaryotes and many bacteria as well. This pathway is used to produce cholesterol and ketone bodies. Cholesterol has diverse functions in cell membrane maintenance, intracellular transport, cellular signaling and nerve conduction within the cell membrane. Cholesterol is also the precursor for the biosynthesis of bile, vitamin D and steroid hormones and their derivatives.

Ketone bodies are three water-soluble molecules acetone, acetoacetic acid and beta- hydroxybutyric acid. They are generated by the liver from fatty acids during fasting or carbohydrate restriction conditions, taken up by cells and converted back into acetyl-CoA to feed into TCA cycle for energy production. In the brain, ketone bodies are also used to produce long chain fatty acids since long chain fatty acids can’t pass the blood-brain barrier.

1.4.5.4. Balancing Carbohydrate Metabolism and Fat Metabolism

In animals and human beings, acyl-CoA thioesters are crucial to the balance between carbohydrate metabolism and fat metabolism. Under normal circumstances, acetyl-CoA generated from fatty acid metabolism feeds into the TCA cycle and is used for the cell’s energy supply. When circulating fatty acids accumulate to high levels, the generation of acetyl-CoA from fat catabolism surpasses the energy requirement of the cell. Then ketone bodies are produced from acetyl-CoA in the liver and released into blood. Acetyl-CoA and malonyl-CoA are precursors from which fatty acid is generated by fatty acid synthases. Fatty acid synthesis is an important part of the lipogenesis. One major role of fatty acid in animal metabolism is energy production by ATP synthesis; fatty acids also contribute to energy storage, phospholipid membrane formation and signaling pathways.

1.4.5.5. Protein Acetylation

Cell signaling networks are regulated by reversible protein post-translational modifications.

Reversible acetylation of ε-amino group on lysine was discovered on histones 50 years ago (168).

This transfer of an acetyl group from acetyl-CoA to the ε-amino group neutralizes lysine’s positive charge which is a crucial step in gene regulation in the nucleus. Studies showed that histone acetylation is critical in regulating global chromatin architecture and gene transcription.

It is also involved in DNA replication, repair and heterochromatin formation (169). In bacteria, acetylation serves as a central regulator that coordinate metabolism in response to changes in different carbon sources which are concurrent with changes in cell growth as well as metabolic flux (135). In S. enterica, 191 proteins were identified to be lysine-acetylated and about 90% enzymes of central metabolism were acetylated (170). In E. coli the number reached 349 (171).

Metabolic enzymes are common targets for acetylation. The balance between

acetylation/deacetylation is affected by switching carbon sources (e.g., from glucose to citrate), growth phase, and availability of acetylphosphate (116-118).

In 1985, tubulin was described as the first acetylated cytoplasmic protein (86). Tubulin is the principal element of microtubules which are major components of cytoskeleton and internal structures of cilia and flagella. Proteomic studies identified hundreds of proteins which are acetylated on one or multiple lysine residues. The reversible lysine acetylation in many cellular and even viral proteins have been found to be involved in critical signaling processes controlling and coordinating cytoskeleton dynamics, intracellular trafficking, vesicle fusion, metabolism, and stress response (91, 104-107).

1.4.6. HS-CoA-Related Diseases

CoA/CoA thioesters have important signaling functions, which may tightly regulate the function of ATP during injury (84). Imbalance of CoA and its thioester levels may cause significant inhibition of platelet aggregation. Malfunction of CoA molecules may also contribute to cardiovascular disorders through vasoconstriction (83). Increased Long chain acyl-CoA level has been observed in obese or Type 2 diabetic patients possibly due to the induction of rapid and potent opening of ATP-sensitive K+ channel causing glucose-stimulated insulin secretion (85).

Neurodegeneration with brain iron accumulation (NBIA) is a group of disorders characterized by iron accumulation in the brain. In 2001, it was discovered that pank2, the gene encoding CoA biosynthetic enzymes PANK, causes a subform of NBIA featured by dystonia, dysarthria and intellectual disabilities. After that, the disease was named PANK-associated neurodegeneration (PKAN). The influence of CoA level on protein acetylation and autophagy, the link between reduced histone and tubulin acetylation and neurodegeneration, and CoA as an

anti-apoptotic signaling molecule may all contribute to the pathogenesis of PANK and NBIA

(172).

1.5. Protein Targeting Mechanisms

In the crowded living cell environment, various mechanisms are applied to control and regulate targeting of the proteins for their proper functions at the right time and in the right subcellular location. Some proteins contain intrinsic signal sequences which are recognized by chaperones or bind transport proteins and are targeted to cellular membranes or secreted outside the cell. Some proteins are involved in protein-protein interactions which allow assembly and regulation of protein networks. Gene clustering and gene fusion are commonly employed for convenient targeting and interaction of functional related proteins. Cellular compartmentalization and enzyme encapsulation are utilized for high efficiency of metabolic processes.

1.5.1. Target Recognition of the PDZ Domains

PDZ domain represents one of the most commonly found protein-protein interaction domains and is widely distributed in organisms from bacteria to plants and vertebrates. It is named after the three unrelated proteins, mammalian postsynaptic density protein PSD95, Drosophila disc large tumor suppressor DLG1, and mammalian tight junction protein zonular occludens-1 ZO1, from which it was first identified (173). Since most PDZ proteins have multiple copies of PDZ domains, they establish interactions with multiple proteins which is crucial for their functions in coordinating signaling complex formation and scaffolding for protein-protein interactions (174-

178). In bacteria, PDZ domains are involved in diverse functions as membrane integrity, sporulation, lipid biosynthesis, pheromone production as well as pathogenesis.

The basic structure of most identified PDZ domains is similar. PDZ domain is usually 80-90 amino acid residues in length and contains 2 α-helices and 5-6 β-strands (174). This domain

contains a hydrophobic pocket formed by the β1-β2 loop and side chains from strand β2 and helix α2, which is highly conserved among PDZ family members (179). The interaction pocket on PDZ domain allows binding of target proteins with consensus sequence patterns. To date, four types of PDZ motifs which bind to PDZ domains have been identified, which are class I (S/T-x-

Φ), class II (Φ-x-Φ), class III (Ψ-x-Φ) and class IV (D-x-V), where x is any amino acid, Φ is a hydrophobic residue (V, I, L, A, G,W, C, M, F) and Ψ is a basic, hydrophilic residue (H, R, K)

(180). These motifs are usually 4 amino acids long, but recent studies indicated longer sequences

(up to C-terminal 12-mer on rabies virus glycoprotein and 13C-ter on phosphatase and tensin homolog) might be involved in binding stability (181, 182).

More recent studies on PDZ domains indicated that the binding specificity of PDZ domain not only involves binding to the carboxyl-terminus of various proteins, it also achieves greater versatility by PDZ-PDZ domain dimerization in a “head-to-tail” mode (Fig. 1.19). For example, the neuronal nitric oxide synthase (nNOS) PDZ domain contains a 30 amino acid C-terminal extension that forms a β-hairpin structure (β-finger) which docks in the peptide-binding sequence of PSD-95 PDZ domain, mimicking a free C-terminus (183). Alternatively, two PDZ domains from some protein like Na+/H+ exchanger regulatory factor (NHERF) or Glutamate receptor- interacting protein (GRIP) engage in homodimerization, which leaves the peptide-binding pocket oriented in the opposite direction for dual binding and allows the PDZ domains simultaneously interact with multiple proteins to form a network (172, 184). By using gel-filtration and surface plasmon resonance (SPR), Zimmermann et al. (185) demonstrated the existence of direct interaction between an abundant plasma membrane lipid component phosphatidylinositol 4,5- bisphosphate (PIP2) and a subset of PDZ domains. It was shown that this interaction is required

for plasma membrane targeting. Similar binding mechanisms were suggested and a systematic analysis will be critical to understanding the novel interactions.

1.5.2. Protein Targeting to the Sec Complex

The Sec translocon is a well-studied protein transport system which is conserved in the cytoplasmic membrane of bacteria and archaea (187). The two main functions of the Sec translocon are to transport secretory protein across the inner membrane and to integrate membrane proteins into the inner membrane. Targeting of many proteins begins with the ribosome during translation. The Sec transport system cooperates with cytosolic protein partners including the signal recognition particle (SRP) and the SecA/SecB system to target inner membrane proteins co-translationally and secretory proteins post-translationally, respectively

(188).

1.5.2.1. Ribosome as Docking Site for Targeting

Secretory proteins transported through the Sec complex usually possess an N-terminal signal sequence which consist of three parts—a positively charged N-terminal region followed by a hydrophobic region and a polar C-terminal region. The signal sequences have conserved architecture but various primary sequences and length (189). There is usually a cleavage site for signal peptidases on C-terminus of the signal sequences. Most membrane proteins use their highly hydrophobic N-terminal short sequences (referred to as a signal anchor sequence) as signal for recognition instead (190, 191).

Proteins are synthesized at about 10-20 amino acids per second and within several seconds after translation is initiated, the N-terminal signal sequences will be exposed to the outside at the ribosomal exit tunnel (192). To avoid undesired interactions of these hydrophobic signal sequences, the ribosome engages specific molecules for early substrate recognition. Studies

indicated that the starting signal for downstream targeting could be the organization of a nascent polypeptide into secondary structures within the ribosomal tunnel (193, 194). Recently, it was shown that the structural differences of the nascent chain in the ribosome tunnel differentially affect the recruitment of SRP or Trigger Factor (TF) (194, 195). Both SRP and TF contact the conserved ribosomal protein L23 which is located close to the tunnel exit. L23 appears to act as a major docking site for interactions between the transport molecules and the newly synthesized polypeptides (196, 197).

Figure 1.19. Possible PDZ interaction modes. PDZ domain proteins are involved in at least four different types of interaction: recognition of C-terminal motifs in peptides, recognition of internal motifs in proteins, PDZ- PDZ dimerization, and recognition of lipids (186).

1.5.2.2. SRP-Dependent Co-Translational Targeting of Inner Membrane Proteins

In E. coli, the SRP pathway is predominantly engaged in targeting of membrane proteins

(198, 199). SRP consists of protein Fifty-Four Homologue (Ffh, named due to its homology to the eukaryotic SRP54 subunit) and 4.5S RNA, both required for targeting. Ffh possesses a methionine-rich M-domain for signal sequence binding and an NG-domain with GTPase activity for binding of the membrane-bound SRP receptor (SR). The highly hydrophobic M-domain with high methionine content is thought to provide structural plasticity to accommodate signal sequences of different lengths and compositions (200). The 4.5S RNA and the Ffh M-domain form a binding pocket for signal sequences on newly synthesized peptides. The RNA is also required for the formation of stable interaction complex between Ffh and SR by modulating GTP hydrolysis (201).

During the SRP cycle, E. coli SRP rapidly scans ribosomes and its binding to the ribosomal tunnel exit is stabilized when a signal anchor sequence is present (202). Then the complex of the signal sequence binding to the M-domain is targeted to the membrane where SRP and NG- domains on SR build a pseudo-homodimer (203). GTP binding stabilizes the interaction between

SPR and SR which initiates the transfer of signal sequence from SRP to the Sec complex. The

Sec translocon controls the conformational changes in the SRP-SR complex and activates GTP hydrolysis which releases SRP back to the cytosol to start a new cycle (204, 205).

The core Sec translocon consists of a heterotrimeric protein complex SecYEG in bacteria.

SecYE are essential for protein transport while SecG is not (187). SecY channel contains 10 α- helical transmembrane (TM) domains which are divided into two halves connected by a periplasmic loop, together they form the aqueous protein transport domain. SecE is located at the back of SecY and functions to stabilize the two halves of SecY. The translocon assumes a closed

conformation in the resting state with the periplasmic side of the SecY channel blocked by a short α-helix. The translocation of preprotein displaces the plug α-helix and results in the open state. The SecYEG translocon opens laterally into the lipid bilayer to facilitate insertion of the newly synthesized membrane protein into the IM.

1.5.2.3. SecA/SecB Pathway for Secretory Protein Targeting

Nascent secretory polypeptides are escorted by cytosolic chaperones, including trigger factor and tetrameric SecB to the membrane. SecB shows a high affinity for ATPase SecA and the binding of SecB to the C-terminus of SecA potentially transfer the preprotein to SecA, which binds to signal sequences on the polypeptides and drives translocation through the SecYEG channel by ATP hydrolysis (187, 206, 207). SecB was thought to dissociate from the complex when SecA binds to ATP (208). Classically it was suggested the binding between SecA and the signal sequence occurs after the preprotein is targeted to the membrane, though some data indicate that SecA can bind to ribosome-nascent peptides, meaning it can also act co- translationally (209-211). After the preprotein is targeted to the translocon, the signal sequence is removed by the membrane-embedded signal peptidases and then degraded by membrane bound signal peptide peptidases.

1.5.3. Protein Encapsulation into Encapsulin-a Bacterial Nanocompartment

Encapsulin is a small protein-based organelle and is present in many bacterial species and a species of archaea C. methanoregula. Encapsulin is constructed from 60 copies of identical monomers which self-assemble into an icosahedral structure with an inner diameter of 20 nm and an exterior diameter of 24 nm (212). Several studies reported that encapsulin can function as a compartment to encapsulate dye decolorizing peroxidase (DyP) and ferritin like protein (Flp), both are involved in oxidative stress responses (212, 213).

DyP proteins are highly conservative across microbial species. Dyp homologues from encapsulin-associated genomes share a conserved C-terminal extension. The C-terminal extension also occurs in Flp homologues when associated with encapsulin. These highly conserved C-terminal extensions have varied lengths (30-40 amino acids for DyP homologues) with a high percentage of alanine, glycine and proline residues followed by a conserved anchor sequence (212). This specific sequence interacts with a hydrophobic binding pocket in the interior side of the protein cage and the binding interaction is proposed to direct the native DyP enzymes inside into the B. linens encapsulin (212). DyP from B. linens is assembled as a trimer of dimers inside the internal cavity of encapsulin. While Flp in T. maritima forms a pentamer of dimers arranged in a ring-like structure. Since the encapsulated Dyp and Flp are oligomers, they are likely to have multiple binding contacts with the interior of the encapsulin cage. These multiple interactions between the encapsulated enzymes and encapsulin may contribute to the higher binding avidity and specificity of the encapsulation. This docking sequence on C-terminus of encapsulin monomers have been shown to encapsulate teal fluorescent protein (TFP) into the encapsulin cage of B. linens (214).

Like ferritin, Flp also contains ferroxidase active site. The packaged Flp enzymes can oxidize the reactive ferrous ions which are retained in the encapsulin. The encapsulin could keep significant amount of iron due to its relative large inner volume compared to the size of ferritin

(220 Å vs 80 Å) (212). The presence of Flp in anaerobic bacteria may protect these organisms from accidental exposure to oxygen and from the adverse effects of metabolically produced peroxide (215, 216). In pathogenic bacteria like Mycobacteria, the encapsulation of peroxidase within encapsulin provides resistance to the killing by oxidative burst in the host and promotes survival in the host monocytes (217). Besides segregating the intermediates generated during

enzymatic reactions, encapsulin may also protect the packaged enzymes. Encapsulins in T. maritima and B. linens have high structural stability at various temperatures and pH conditions, which may significantly increase the lifetime of the interior enzymes.

1.5.4. Vault Ribonucleoprotein Particles

In mammals, certain molecular compartments are built exclusively from protein components.

One example is the vault particle which is named for its morphology with multiple arches similar to cathedral ceilings. Vault ribonucleoprotein particles (RNP) are the largest cytoplasmic ribonucleoprotein structures with dimensions of 725 x 410 x 410 Å3 and are found in the cytoplasm of most eukaryotic cells (218). Vault consist of a thin (≈20 Å) hollow protein shell made of 96 copies of major vault protein (MVP) which encapsulates 2-4 copies of telomerase associated protein I (TEP1), around 12 copies of poly (ADP-ribose)-polymerase (VPARP), and

8-16 copies of a small untranslated RNA. Vaults have been implicated in multidrug resistance, cellular signaling, and innate immunity, however, the relevant mechanisms remain elusive (219-

224).

The encapsulation of the VPARP into vault is mediated by the interactions between the minimal size interaction (M-INT) domain on the C-terminus of VPARP and the N-terminus of shell protein MVP (225). Electron microscopy studies showed that M-INT domain binds to the inside of recombinant vaults at both above and below the waist of the particle (226). It was demonstrated that the 162 amino acid INT domain on VPARP (VRARP residue 1563-1724) could be used to target heterologous proteins GFP and luciferase to the vault cavity (227). The

INT could be reduced to 147 amino acids (VRARP residue 1563-1709) without notable defect in

MVP binding (228). The possible INT binding site was mapped by NMR to MVP residues 113-

221, which are the third and fourth repeat domains of MVP (229). The NMR analysis fits well with the binding site determined by an independent cryo-EM difference mapping.

A yeast two-hybrid assay indicated that TEP1 didn’t interact with MVP monomers (230).

Another study carried out co-infection of baculoviruses expressing TEP1 and MVP produced vault-like particles and showed that TEP1 and MVP were co-purified (237). These studies suggested that TEP1 may be able to interact with intact vault particles but not the MVP monomers. Later, a series of experiments using a yeast three-hybrid system with a series of truncated TEP1 followed by co-infection of Sf9 insect cell cultures and purification of vault particles showed that the p80 region of TEP1 (amino acids 201-911) contains vault-interaction domains (231).

1.6. Research Overview

Bacterial MCPs are widely distributed among diverse bacterial phyla and contribute to the optimization of metabolic pathways in bacteria. Over the years crystal studies of MCP shell proteins suggested a route of molecular transport through the narrow pores in the shell. While several recent studies provided evidence that large cofactors might be recycled inside of MCPs.

Another fundamental question about MCPs is how enzymes are packaged into the lumen of

MCPs. This work is to trying to tackle these questions by identifying a specific lumen enzyme

PduL and investigating its specific role in cofactor recycling and the targeting mechanism which directs the enzyme into MCPs.

The first part of this thesis characterized PduL as a new class of PTAC enzyme by using biochemical and bacterial genetic methods. Genomic analysis and sequence alignment indicated

PduL doesn’t have homology with known PTACs and is distinct in evolution. PduL was

identified to be involved in 1,2-PD metabolism by genetic tests. The substrate and product of

PduL reaction were determined by HPLC, and the kinetic parameters were determined.

The second part of this study investigated the specific role of PduL associated with Pdu MCP and determined the short N-terminal targeting sequence on PduL. MCP purification followed by

Western blots showed PduL is a MCP component. Genetic studies indicated that PduL located inside MCP has limited access to acetyl-CoA molecules in the cytosol, which caused growth defect correctable by breaking MCPs. These data suggested an internal recycling of coenzyme A in Pdu MCPs. Growth studies and immunoblots using pduL deletion mutants lacking N-terminus together with GFP fusion tests indicated PduL contains a short terminal targeting sequence.

The third part of this work explored protein-protein network within Pdu MCP. In vivo two- hybrid tests plus in vitro affinity pull-downs followed by Western blotting suggested possible protein-protein interactions of L/BB’ and U/V. Our experiment results were consistent with PPIs predictions by use of computational models from our collaborating group.

Consistent with recent relevant research results, our studies in this thesis would expand the understanding of MCP functions and provide a reference for the future work.

1.7. Thesis Organization

This dissertation consists of five chapters. Chapter one contains an introductory review of the history, structure, function, application of bacterial microcompartments and their contribution to enteric bacteria pathogenesis and the comparison between bacterial MCPs and viral capsid as well as coenzyme A including its biosynthesis, functions and relatedness to diseases. Chapter two is a paper which was published in the Journal of Bacteriology. Chapter 3 is a paper to be submitted to a journal. Chapter 4 includes preliminary results of certain protein-protein interactions of 1,2-propanediol utilization microcompartments, partial contents of this chapter

has been incorporated into a paper and submitted to the journal PLoS Computational Biology.

Chapter 5 is a general concluding chapter, where the results and conclusions from the two papers and chapter 4 are summarized. All of the papers were written by me with the editorial guidance and assistance by my major professor, Dr. Thomas A. Bobik.

CHAPTER 2. PDUL IS AN EVOLUTIONARILY DISTINCT PHOSPHOTRANSACYLASE INVOLVED IN B12-DEPENDENT 1,2-PROPANEDIOL DEGRADATION BY SALMONELLA ENTERICA SEROVAR TYPHIMURIUM LT2

This chapter, in part or in full, is a reprint of the material as it appears in the Journal of Bacteriology (volume 189, pp. 1589-1596, 2007).

2.1. Abstract

Salmonella enterica degrades 1,2-propanediol (1,2-PD) in a coenzyme B12-dependent manner. Prior enzymatic assays of crude cell extracts indicated that a phosphotransacylase

(PTAC) was needed for this process, but the enzyme involved was not identified. Here we show that the pduL gene encodes an evolutionarily distinct PTAC used for 1,2-PD degradation.

Growth tests showed that pduL mutants were unable to ferment 1,2-PD and were also impaired for aerobic growth on this compound. Enzyme assays showed that cell extracts from a pduL mutant lacked measurable PTAC activity in a background that also carried a pta mutation (the pta gene was previously shown to encode a PTAC enzyme). Ectopic expression of pduL corrected the growth defects of pta mutant. PduL fused to 8 C-terminal histidine residues (PduL-

-1 -1 His8) was purified and its kinetic constants determined: Vmax = 51.7  7.6 µmol min mg ; and

2- 2- Km for propionyl-PO4 and acetyl-PO4 = 0.61 and 0.97 mM, respectively. Sequence analyses showed that PduL is unrelated in amino acid sequence to known PTAC enzymes and that PduL homologues are distributed among at least 49 bacterial species, but are absent from the Archaea and Eukarya.

2.2. Introduction

A number of bacterial genera including Salmonella, Klebsiella, Shigella, Yersinia, Listeria,

Lactobacillus and Lactococcus include members that grow on 1,2-propanediol (1,2-PD) in a coenzyme B12-dependent fashion. 1,2-PD is a major product of the fermentation of rhamnose and fucose which are common sugars in plant cell walls, bacterial exopolysaccharides and the glycoconjugates of intestinal epithelial cells. Accordingly, the ability to degrade 1,2-PD is thought to provide a selective advantage in anaerobic environments such as the large intestines of higher animals, sediments and the depths of soils. Recent studies with Salmonella enterica have

shown that 1,2-PD degradation is one of the most complex metabolic processes known (22, 49,

53, 70, 119, 126). The degradation of this small molecule requires catabolic enzymes, a system for recycling inactive cobalamins to coenzyme B12, and an unusual polyhedral body that is composed of metabolic enzymes encased within a multi-protein shell.

A pathway for 1,2-PD degradation by S. enterica has been proposed based on enzymatic studies of crude cell extracts and genetic analyses (90, 232). In the first step of the pathway, 1,2-

PD in converted to propionaldehyde via coenzyme B12-dependent diol dehydratase (233).

Propionaldehyde is then converted to 1-propanol and propionic acid by a reaction series thought to involve 1-propanol dehydrogenase, coenzyme A (CoA)-dependent propionaldehyde dehydrogenase, phosphotransacylase (PTAC), and propionate kinase (90, 224). This pathway generates one ATP, an electron sink, and a 3-carbon intermediate (propionyl-CoA), which feeds into central metabolism via the methylcitrate pathway (76).

The genes specifically required for 1,2-PD utilization (pdu) by S. enterica form a single contiguous cluster, the pdu locus. DNA sequence analyses indicate that this locus includes 23 genes (22, 33, 51). Based on experimental and/or bioinformatics analyses, six pdu genes are thought to encode enzymes needed for the 1,2-PD degradative pathway (22); two are involved in transport and regulation (33, 54); two are probably involved in diol dehydratase reactivation (22); two are used for the conversion of vitamin B12 to coenzyme B12 (119, 126); four are of unknown function; and seven share similarity to genes involved in the formation of carboxysomes, a polyhedral body found in certain cyanobacteria and chemoautotrophs (22, 33). Although the methylcitrate pathway is required for growth of S. enterica on 1,2-PD a sole carbon source, the genes for this pathway map outside the pdu locus (76).

1,2-PD 1,2-PD 1-propanol AdoCbl PduCDE PduQ propionaldehyde NAD NAD NADH PduP ATP NADH ADP METHYLCITRATE propionyl-CoA 2- propionate PATHWAY propionyl-PO4 PduL PduW

Figure 2.1. Model for 1,2-propanediol degradation by S. enterica. 1,2-propanediol (1,2-PD), coenzyme B12

(AdoCbl), coenzyme B12-dependent diol dehydratase (PduCDE), propionaldehyde dehydrogenase (PduP), phosphotransacylase (PduL), propionate kinase (PduW), 1-propanol dehydrogenase (PduQ). The first two steps of 1,2-PD degradation (conversion of 1,2-PD to propionyl-CoA) are proposed to occur within a microcompartment. The dashed line indicates the shell of the MCP which is composed of up to seven different 2- polypeptides. ATP is generated during the conversion of propionyl-PO4 to propionate. Given a suitable terminal electron acceptor such as O2, propionyl-CoA is degraded via the methylcitrate pathway where biosynthetic precursors and additional energy are produced. Two additional enzymes associated with the polyhedral bodies (but not shown in the figure) are a putative diol dehydratase reactivase (PduGH) and an

ATP:cob(I)alamin adenosyltransferase (PduO) involved in B12 recycling.

It was recently shown that S. enterica forms polyhedral bodies during growth on 1,2-PD (22,

34, 70). These bodies are extremely large macromolecular complexes that are 100-150 nm in cross-section and consist of metabolic enzymes encased within a protein shell. Purification of these structures showed that they are composed of at least 14 different polypeptides

(PduABB'CDEGHJKOPTU) (49). This includes 4 enzymes and as many as seven shell proteins all of which are encoded by the pdu locus. The first two steps of 1,2-PD degradation are thought to occur in the lumen of the PhBs and the remaining steps in the cytoplasm of the cell (Fig. 2.1)

(49). The function of PhBs is uncertain. Prior studies suggest they act as microcompartments that sequester propionaldehyde (an intermediate of 1,2-PD degradation) in order to minimize toxicity and/or prevent carbon loss (49, 70, 234, 235). Interestingly, recent genomic analyses tentatively indicate seven functionally distinct polyhedral bodies distributed among over 40 genera of

bacteria (23). Thus, polyhedral bodies may be a general mechanism of metabolic organization in the Bacteria.

As described above, prior enzymatic analysis of crude cell extracts indicated that a PTAC is used for 1,2-PD degradation (90, 232). However, the enzyme involved was not identified. Here, we show that the pduL gene encodes a conserved, evolutionarily distinct PTAC that functions in

1,2-PD degradation by S. enterica.

2.3. Materials and Methods

2- Chemicals and reagents. Antibiotics, vitamin B12 and acetyl-PO4 were from Sigma

2- Chemical Company (St. Louis, MO). Propionyl-PO4 was synthesized as described (236).

Isopropyl--D-thiogalactopyranoside (IPTG) was from Diagnostic Chemicals Limited

(Charlottesville PEI, Canada). Restriction enzymes and T4 DNA ligase were from New England

Biolabs (Beverly, MA). Coomassie Brilliant Blue R-250, EDTA, ethidium bromide, 2- mercaptoethanol, and SDS were from Bio-Rad (Hercules, CA). Other chemicals were from

Fisher Scientific (Pittsburgh, PA).

Bacterial strains, media, and growth conditions. The bacterial strains used in this study are listed in Table 2.1. The rich medium used was Luria-Bertani/Lennox (LB) medium (Difco,

Detroit, MI) (237). The minimal media used was no-carbon-E (NCE) medium containing supplements indicated in the text and figure legends (238, 239). MacConkey/1,2-PD indicator

1 medium contained MacConkey agar base (Difco), 1% 1,2-PD and 200 ng ml- of vitamin B12.

General molecular methods. Agarose gel electrophoresis was performed as described previously (240). Plasmid DNA was purified by the alkaline lysis procedure (240) or by using

Qiagen products (Qiagen, Chatsworth, CA) according to the manufacturer's instructions.

Following restriction digestion or PCR amplification, DNA was purified using Promega Wizard

PCR Preps (Madison, WI) or Qiagen gel extraction kits. Restriction digests were carried out using standard protocols (240). For ligation of DNA fragments, T4 DNA ligase was used according to the manufacturer's directions. Electroporation was carried out as previously described (22).

Table 2.1. Bacterial strains used in this study.

Strain Genotype Source aS. enterica serovar Typhimurium LT2 BE47 thr-480::Tn10dCam BE188 ∆pduL670 T.A.Bobik lab collection BE281 pta-406::Tn10 T.A.Bobik lab collection BE284 pduL670 /pLAC22-pduL (AmpR) T.A.Bobik lab collection BE285 pduL670 /pLAC22-no insert (AmpR) T.A.Bobik lab collection BE286 pLAC22-pduL (AmpR) T.A.Bobik lab collection BE287 pLAC22-no insert (AmpR) T.A.Bobik lab collection BE291 ∆pduL670, pta209::Tn10 T.A.Bobik lab collection BE527 pta209::Tn10 T.A.Bobik lab collection BE529 pta-209::Tn10 /pLAC22-no insert (AmpR) T.A.Bobik lab collection BE530 pta-209::Tn10 /pLAC22-pduL (AmpR) T.A.Bobik lab collection BE548 pta-209::Tn10 /pBE522-no insert (KanR) T.A.Bobik lab collection BE549 pta-209::Tn10 /pBE522-pta (KanR) T.A.Bobik lab collection - - - + r E. coli BL21(DE3)RIL (E. coli B) F ompT hsdS (rB mB ) dcm Tet gal 1 (DE3) endA Hte (argU ileY leuW Camr) BE119 BL21 DE3 RIL/ bpTA925-no insert (Kanr) T.A.Bobik lab collection BE554 BL21 DE3 RIL/ pTA925-pduL-His8 This study S17.1pir recA(RP4-2-Tc::Mu)pir T.A.Bobik lab collection aformerly S. typhimurium LT2 b r T7 expression vector containing a minimal clone for production of PduL-His8, Kan

Protein methods. Polyacrylamide gel electrophoresis (PAGE) was performed using Bio-Rad

Redigels and Bio-Rad Mini-Protean II electrophoresis cells according to the manufacturer’s instructions. Following gel electrophoresis, Coomassie Brilliant Blue R-250 was used to stain proteins. The protein concentration of solutions was determined using Bio-Rad Protein Assay

Reagent (Bio-Rad).

P22 transduction. Transductional crosses were performed as described using P22 HT105/1 int-210 (241), a mutant phage that has high transducing ability (242). Transductants were tested for phage contamination and sensitivity by streaking on green plates against P22 H5.

Construction of plasmids for production of PduL and PduL-His8. PCR was used to amplify the pduL coding sequence from template pMGS2 (70). The primers used for amplification were

5’-GCCGCCAGATCTATGGATAAAGAGCTTCTGCAATCA-3’ and 5’-

GCCGCCAAGCTTATTATCGCGGGCCTACCAGCCG-3’. These PCR primers introduced

BglII and HindIII restriction sites that were used for cloning into vector pLAC22 (243).

Following ligation, clones were introduced into E coli DH5α by electroporation and transformants were selected by plating on LB agar supplemented with 100 g ml-1 ampicillin

(Amp) (119). Pure cultures were prepared from selected transformants. The presence of insert

DNA was verified by restriction analysis or PCR, and the DNA sequence of selected pduL clones was determined. Clones having the expected DNA sequence were used for further study.

A similar procedure was used to clone PduL fused to eight C-terminal histidine residues

(PduL-His8) with the following differences. The primers used for PCR amplification were 5'-

GCCGCCAGATCTATGGATAAAGAGCTTCTGCAATCA-3' and 5'-

GCCGCCAAGCTTATTAATGATGATGATGATGATGATGATGTCGCGGGCCTACCAGCC

G-3. The PCR product was ligated to T7 expression vector pTA925 (119). Transformation was done by electroporation and LB plates supplemented with 25 µg ml-1 kanamycin (Kan) were used to select for transformants.

Growth of the PduL-His8 production strain. A T7 expression plasmid (pTA925) was used for production of PduL-His8. The host used for protein expression was E. coli BL21DE3 RIL

(Stratagene). This strain expresses T7 RNA polymerase as well as rare tRNAs for arginine,

isoleucine and leucine. Expression and control strains (BE554 and BE119) were grown in 400 ml

LB broth containing 25 g/ml Kan and 10 µg/ml chloramphenicol. Cultures were incubated at 15

C with shaking at 275 rpm in a 1 liter baffled Erlenmeyer flask. Cells were grown to an optical density of 0.6-0.8 at 600 nm and protein expression was induced by the addition of 1 mM IPTG.

Cells were incubated for an additional 16 hours, and harvested by centrifugation at 8,000 x g for

10 minutes and at 4 C using a Beckman JA-10 rotor and Avanti J-25 centrifuge.

Purification of PduL-His8. The PduL-His8 production strain (BE554) was grown as described above. One gram of cells (wet weight) was suspended in 3 ml of buffer containing 50 mM Tris-HCl, pH 7.2, 200 mM (NH4)2SO4 and 0.4 mM AEBSF. Cells were broken using a

French pressure cell at 20,000 psi, and the resulting cell extract was centrifuged at 20,000 x g using a Beckman JA-17 rotor. The supernatant fraction was filtered through a 0.45 µm syringe filter. A chromatography column containing 1 ml of Ni-NTA resin (Qiagen) was equilibrated with 5 ml of buffer A: 50 mM Tri-HCl, pH 7.7, 25 mM KCl, 300 mM NaCl, 200 mM (NH4)2SO4

10 mM imidazole and 5 mM 2-mercaptoethanol. Filtered cell extract (2.5 ml, about 50 mg protein) was applied to the column. The column was washed with 15 ml of buffer A supplemented with 10% glycerol and 80 mM imidazole. The column was eluted with 10 ml of buffer A (in two of 5 ml steps) supplemented with 10% glycerol and 400 mM imidazole.

Preparation of cell extracts of S. enterica. Cells were grown under conditions that induce the pdu operon (54). Cell paste was suspended in 25 mM Tris-HCl, pH 7.2 containing 0.4 mM protease inhibitor 4-(2-Aminoethyl)benzenesulfonylfluoride.HCl (AEBSF) (3 ml of buffer per gram cells wet weight). Cells were broken using a French pressure cell (Thermo Electron Corp.,

Waltham, MA) at 20,000 psi.

2- PTAC assays. PTAC assays measured the conversion of acyl-PO4 + HS-CoA to acyl-CoA, and were performed as described (244). Standard assays contained 50 mM Tris-HCl, pH 7.2, 20

2- mM KCl, 0.2 mM HS-CoA, and 1 mM acyl-PO4 in a total volume of 1 ml. For kinetic studies, the concentrations of certain assay components were varied as indicated in the text. Activity was

-1 -1 determined by following absorbance at 232 nm over time and by using E232 = 5.5 mM cm for calculations.

Construction of a nonpolar PduL deletion. Bases 22 to 612 of the pduL coding sequence were deleted via a PCR-based method (245). The deletion was designed to leave all predicted translational start and stop signals of pdu genes intact. The following primers were used for PCR amplification of the flanking regions of the pduL gene: primer 1, 5’-

GCTCTAGAGCCGAAATCAGCCTAATCGATGGCG-3’; primer 2, 5’-

CGTTCATCGCGGGCCTACCAGCCGATCCATTACGCTTCACCTCGC-3’; primer 3, 5’-

CGGCTGGTAGGCCCGCGATGAACG-3’; and primer 4, 5’-

CGAGCTCGCCAGATGCATGATTTACTC-3’. Primers 1 and 2 were used to amplify a 498 base-pair region upstream of the pduL gene, and primers 3 and 4 were used to amplify 527 bases downstream of pduL gene. The upstream and downstream amplification products were purified then fused by a PCR reaction that included 1 ng/µl of each product and primers 1 and 4. The fused product was digested with XbaI and SacI (these sites were designed into primers 1 and 4, respectively), and ligated to suicide vector pCVD442 that had been similarly digested. The ligation mixture was used to transform E. coli S17.1 by electroporation and transformants were selected on LB medium supplemented with Amp (100g/ml). Six transformants were screened by restriction analysis and all released an insert of the expected size (1058 bp). One of these transformants was used to introduce the pduL deletion into the S. enterica chromosome using the

procedure of Miller and Mekelanos (245) with the following modification. For the conjugation step, strain BE47 was used as the recipient and exconjugants were selected by plating on LB agar supplemented with Amp (100 g/ml) and chloramphenicol (20 g/ml). Deletion of the pduL coding sequence was verified by PCR using chromosomal DNA as a template. Lastly, the thr-

480 dCAM insertion used for selection of exconjugants was "crossed-off" by P22 transduction using a phage lysate prepared with the wild-type strain and by selecting for prototrophy on NCE glucose minimal medium.

Aerobic growth curves. Growth media are described in the figure legends. For strains carrying pLAC22, media were also supplemented with 100 g/ml Amp and 0.2 mM IPTG. To prepare the inoculum, LB cultures (2 ml) were incubated overnight at 37 C, and then cells were collected by centrifugation and resuspended in growth curve medium. Media were inoculated to a density of 0.15 AU and growth was followed by measuring optical density at 600 nm using a

BioTek Synergy microplate reader as follows: 48-well flat bottom plates (Falcon); 0.5 ml of growth medium per well, 37 C; and shaking set a level 4. A relatively low volume of growth medium (0.5 ml) in a 48-well microplate was necessary for adequate aeration and controls showed that growth under these conditions was similar to growth in shake flasks (data not shown).

Anaerobic growth curves. A procedure similar to that for aerobic growth curves was used to measure the fermentation of 1,2-PD, but with the following differences. Cultures were inoculated to an initial density of 0.1 AU. Microplates were sealed with polyolefin membranes

(Fisher Scientific) inside an anaerobic growth chamber (Coy Laboratory Products, Grass Lake,

MI). To monitor whether anaerobic conditions were maintained within the microplate, NCE glycerol minimal medium was added to several wells and inoculated with wild-type S. enterica.

S. enterica cannot ferment glycerol; therefore, growth on glycerol would indicate the presence of oxygen. For the experiments described in this study, no growth was observed in wells containing glycerol minimal medium indicating that the headspace contained minimal amounts of oxygen

(data not shown).

DNA sequencing and analysis. DNA sequencing was carried out at the Iowa State

University (ISU) DNA Facility using Applied Biosystems Inc. automated sequencing equipment.

The template for DNA sequencing was plasmid DNA purified using Qiagen 100 tips or Qiagen mini-prep kits. BLAST software was used for sequence similarity searching (246).

High-pressure liquid chromatography (HPLC). A Microsorb C18 column (150 x 4.6 mm) was used with a Varian ProStar system that included a model 230 solvent delivery module, a model 430 autosampler and a model 325 UV-Vis detector (Varian, Palo Alto, CA). Buffer A and

B contained 10 mM NH4 formate pH 4.6 and 10% or 90% methanol, respectively. The flow rate was 1 ml min-1 and the buffer composition was varied as follows: (minute:%B), 0:0, 5:0, 17:100,

22:100, 23:0, 28:0.

HPLC electrospray ionization mass spectrometry (HPLC-ESI-MS). An Agilent ion trap model 1100 mass spectrometer (Agilent, Palo Alto, CA) was operated in the positive mode. The ion source parameters were optimized for the formation of [M + H]+ ions with source temperature of 310 C, capillary voltage of 3.2 kV, and cone voltage of 25 V. Nitrogen was used as the nebulizing gas and as the drying gas at flow rates of 15 and 400 L h-1, respectively.

2.4. Results

The pta gene is nonessential for 1,2-PD degradation. Prior biochemical studies indicated that PTAC was needed for 1,2-PD degradation by S. enterica (90, 232). In the proposed biochemical pathway, this enzyme catalyzed the conversion of propionyl-CoA to propionyl-

phosphate (transfer of a 3-carbon acyl-group) (Fig. 2.1). An enzyme with homology to known

PTAC enzymes is not encoded by the pdu locus (22). However, S. enterica produces a phosphotransacetylase (Pta) encoded by the pta gene which maps outside the pdu operon at centisome 50 (247). Pta is used for growth on acetate and inositol and plays a key role in the

2- interconversion of acetyl-CoA and acetyl-PO4 (transfer of a 2-carbon acyl-group) (247). In vitro, Pta not only mediates the interconversion of acetyl-CoA and acetyl-phosphate, it also

2- catalyzes the interconversion propionyl-CoA and propionyl-PO4 (247, 248). This raised the question of whether pta plays a role in 1,2-PD degradation. Two independent pta mutants

(BE281 and BE527) were examined for growth on 1,2-PD. Aerobically, both grew slightly faster than wild-type S. enterica. The doubling times for the wild-type were typically 7.5-8.5 h and those for the pta mutants were 6-7 h. Both pta mutants were reconstructed via P22 transduction to assure that they were isogenic with the wild-type and both still exhibited a small increase in growth rate on 1,2-PD (not shown). These results confirmed prior studies which showed that pta is nonessential for aerobic growth of S. enterica on 1,2-PD minimal medium (249).

Similarly, pta was unnecessary for the fermentation of 1,2-PD. Strains with pta mutations and wild-type S. enterica both had doubling times of 5-6 h during 1,2-PD fermentation.

Phenotypic test showed that the pta mutants used in the above studies were impaired for grow on acetate or inositol minimal medium and lacked detectable pta activity in cell extracts as is expected for pta mutants (247). In addition, the growth defects on acetate and inositol were corrected by a pta minimal clone and the chromosomal location of both pta insertions was verified by PCR (data not shown). Hence, we conclude that the pta gene is nonessential for B12- dependent 1,2-PD degradation under either aerobic or anaerobic growth conditions.

Strains with pduL mutations produce less propionic acid on MacConkey/1,2- propanediol indicator medium. The finding that the pta gene was unnecessary for 1,2-PD degradation suggested that S. enterica expresses an additional PTAC that is sufficient to mediate this process. As part of ongoing studies of 1,2-PD degradation by S. enterica, we constructed a series of precise pdu deletion mutations using PCR-based methods (245, 250). Tests with

MacConkey/1,2-PD medium (251) indicated that three independent pduL deletion mutants each produced less propionic acid from 1,2-PD than did wild-type S. enterica. The enzymes that convert 1,2-PD to propionic acid were proposed to include coenzyme B12-dependent diol dehydratase, propionaldehyde dehydrogenase, propionate kinase and PTAC (Fig. 2.1) (90, 232).

Of these, only the gene for the PTAC was unidentified (51, 53, 249). Hence, these results tentatively suggested that the pduL gene encodes a PTAC enzyme.

Enzyme assays indicate that pduL encodes a PTAC enzyme. Wild-type S. enterica, as well as pduL and pta null mutants were grown under conditions that induce the pdu operon (54).

Cell extracts were prepared and PTAC activity was measured using an assay that follows the conversion of acyl-phosphate + HS-CoA to acyl-CoA and inorganic phosphate (the reverse

2- 2- reaction with respect to 1,2-PD degradation). Both propionyl-PO4 and acetyl-PO4 were used as substrates. Cell extract from the wild-type strain had 8.9 µmol min-1 mg-1 PTAC activity with

2- propionyl-PO4 (Table 2.2). Extracts from pta or pduL mutants had partial activity (6.0 and 2.6

µmol min-1 mg-1, respectively) (Table 2.2). PTAC activity was undetectable in cell extracts from the double mutant (pta pduL). The simplest interpretation of these results is that pduL and pta each encode PTAC enzymes which was previously shown for pta (247). Furthermore, the pta pduL double mutant lacked detectable PTAC activity demonstrating that the activity in Pta-

PduL+ cell extracts required PduL.

Table 2.2. Phosphotransacylase activity in cell extracts from selected strains of S. enterica. aPTAC activity (µmol min-1 mg-1) 2- 2- b 2- Relevant genotype propionyl-PO4 acetyl-PO4 propionyl-PO4 preference wild type (PduL+ Pta+) 8.9 3.5 2.5 BE188 (PduL- Pta+) 2.6 3.2 0.8 BE527 (PduL+ Pta-) 6.0 0.4 15.0 BE291 (PduL- Pta-) ND ND -

a 2- 2- PTAC assays contained .2mM HS-CoA and 1mM propionyl-PO4 or acetyl-PO4 and assay buffer. Activity was determined by following the absorbance of reactions at 232 nm. b 2- 2- 2- Propionyl- PO4 preference = (activity with propionyl-PO4 )/(activity with acetyl-PO4 ) cND none detected (the lower detection limit of the assay is estimated to be 0.03umol min-1 mg-1). Cells were grown on NCE succinate minimal medium supplemented with 1,2-PD to ensure induction of the pdu operon.

Enzyme assays also showed that the PTAC in Pta- PduL+ cell extracts was 15-fold more

2- 2- -1 -1 active with propionyl-PO4 compared to acetyl-PO4 (6 µmol min mg compared to 0.4 µmol min-1 mg-1) (Table 2.2). This is consistent with a role for PduL in 1,2-PD degradation. On the other hand, the PTAC activity in Pta+ PduL- cell extracts was 1.6-fold more active with acetyl-

2- PO4 consistent with the role of Pta in acetyl-group metabolism (247).

For the above enzyme assays, controls showed that detectable PTAC activity required acyl- phosphate and HS-CoA, and in all cases PTAC activity was linear with enzyme concentration

(data not shown).

PduL mutants are impaired for aerobic growth on 1,2-PD. Compared to wild-type S. enterica, a pduL mutant (BE188) was impaired for aerobic growth on 1,2-PD (Fig. 2.2A). The doubling time for S. enterica was 7.9  0.5 h and that of a pduL mutant was 12.6  0.6 h. Further impairment of growth was not seen in the pduL pta double mutant even though this mutant lacked measurable PTAC activity (data not shown). These findings are consistent with the idea that PduL is a PTAC involved in 1,2-PD degradation. Aerobically, growth on 1,2-PD is expected

to proceed in the absence of PTAC since propionyl-CoA (which is formed prior to the PTAC reaction) can be metabolized via the methylcitrate pathway (Fig. 2.1) (76). The observed growth impairment of the pduL mutant was likely due to reduced ATP synthesis resulting from a block in the oxidative branch of the 1,2-PD degradative pathway although other explanations cannot be ruled out.

pduL mutants are unable to obtain energy from 1,2-PD under fermentative conditions.

In the absence of an exogenous electron acceptor, the methylcitrate pathway is inoperative, and S. enterica is unable to grow on 1,2-PD as a sole carbon source (48). Under these conditions, S. enterica converts 1,2-PD to 1-propanol and propionate (Fig. 2.1). This stimulates growth on minimal medium supplemented with a small amount of yeast extract by providing a source of

ATP (48). The production of ATP via the conversion of 1,2-PD to propionate is expected to require PTAC (Fig 2.1). Therefore, we tested the effect of a pduL mutation on this process. The doubling times of wild-type S. enterica in the presence and absence of 1,2-PD under fermentative conditions were 5.5  0.35 h and 10.7  0.29 h, respectively (Fig. 2.2B); the fermentation of 1,2-PD stimulated the growth rate about 2-fold which is consistent with previous studies (48). The doubling times of the pduL mutant (BE188) under fermentative conditions in the presence and absence of 1,2-PD were 13.1  1.0 and 13.3  0.95 h, respectively. This indicated that energy production via 1,2-PD fermentation was completely eliminated by the pduL mutation. This is consistent with a role for PduL in 1,2-PD degradation.

The pduL mutant used in the above studies contained a wild-type pta gene indicating that Pta did not substitute for PduL under the conditions used to measure the 1,2-PD fermentation.

Furthermore, a pta null mutation (alone or in combination with a pduL null mutation) did not

affect the fermentation of 1,2-PD (data not shown). These results indicate that the Pta enzyme is insufficient to support the fermentation of 1,2-PD and that PduL is required.

Figure 2.2. Growth a pduL null mutant on 1,2-PD. A. Aerobic growth of wild-type S. enterica and a pduL mutant (BE188) on 1,2-PD minimal medium. B. Fermentation of 1,2-PD. BE188 carries a precise deletion of the pduL gene made by a PCR-based method. The observed growth defects of the BE188 were corrected by a pduL minimal clone. Each growth curve was performed 3 times in quadruplicate. Representative curves are shown. Log phase doubling times with standard deviations are given in the text. The aerobic growth medium

-1 was NCE supplemented with 52 mM 1,2-PD, 200 ng ml vitamin B12 and 0.3 mM valine, isoleucine, leucine and threonine. The fermentation medium was NCE supplemented with 52 mM 1,2-PD, 200 ng ml-1 vitamin

B12 and 0.2% yeast extract. The inoculation and incubation procedures are described in materials and methods.

The observed phenotypes of a pduL mutant are complemented by a pduL minimal clone.

Growth tests showed that the aerobic and anaerobic growth phenotypes described above for a pduL mutant (BE188) were fully corrected by expression of a pduL minimal clone using a tightly regulated expression vector, pLAC22. Under aerobic conditions, the doubling times BE287

(wild-type S. enterica/pLAC22-no insert) and BE286 (wild-type S. enterica/pLAC22-pduL) were

5.4  0.12 h compared to 6.0  0.32 h, respectively. The doubling times for strains for BE285

(pduL/pLAC22-no insert) and BE284 (pduL/pLAC22-pduL) were 8.2  0.71 and 5.5  0.25 h, respectively. Under anaerobic conditions, the doubling times S. enterica/pLAC22-no insert, S. enterica/pLAC22-pduL, pduL/pLAC22-no insert and pduL/pLAC22-pduL were 5.0  0.22, 5.2 

0.37, 11.1  0.19, and 5.3  0.40 respectively. Thus, under aerobic and anaerobic conditions, a pduL minimal clone complemented the growth defect to the pduL mutation. On the other hand, vector without insert did not correct the observed growth defects of the pduL mutant. These results show that observed phenotypes of pduL mutant (BE188) resulted from the deletion of the pduL gene, but not from polarity or a mutation acquired during strain construction.

0.7

0.5 - Pta /pduL

600 control

0.3 OD

- 0.1 Pta /no insert 0 0 5 10 15 20 Time (h)

Figure 2.3. Ectopic expression of a pduL minimal clone corrects the growth defect of a pta mutant on acetate minimal medium. S. enterica harboring expression vector pLAC22 without insert (control), a pta mutant carrying pLAC22 without insert (Pta-/no insert), a pta mutant harboring a pduL minimal clone in pLAC22 (Pta-/pduL). The relevant strain numbers are BE287, BE529 and BE530 (Table 2.1). Controls showed that pLAC22 with or without the pduL insert did not observably affect the growth of the wild-type strain on acetate minimal (data not shown). In addition, the growth defect of the pta mutant was fully corrected by a pta minimal clone showing that the effects of this mutation were not due to polarity (data not shown). The growth medium was NCE supplemented with 0.2% Na acetate, and 0.2 mM IPTG. Each growth study was performed twice in quadruplicate. Representative curves are shown. Log phase doubling times with standard deviations are given in the text.

PduL substitutes for Pta in vivo during acetate utilization. Growth tests were performed to determine whether PduL could substitute for Pta in vivo. Prior studies showed that pta mutants grow very slowly on minimal medium with >30 mM acetate (247). Here we show that ectopic expression of PduL corrects this defect (Fig. 2.3). The doubling times of the Pta+ strain and the pta mutant on acetate minimal medium were 4.4  0.41 h and 28.6  1.4 h, respectively.

The doubling time of a pta mutant producing PduL from expression vector pLAC22 was 3.7 

0.1 h which was slightly faster than the control (4.4  0.41 h). Hence, ectopic expression of pduL fully corrected the growth defect of the pta mutant on acetate minimal medium. Since Pta is a well-studied PTAC enzyme, this finding indicates that PduL also has PTAC activity.

Propionyl-CoA is a product of the PduL reaction. The finding that HS-CoA was required for PduL activity (see above) indicated that an acyl-CoA was formed in assay mixtures.

Moreover, the PTAC assay used to measure PduL activity followed absorbance at 232 nm which monitors the formation of thioester bonds (252). To confirm that an acyl-CoA was formed in

PduL enzyme assays, HPLC and HPLC-mass spectrometry were performed. Assays containing

- + - - 2- cell extract from BE527 (Pta PduL ) or BE291 (Pta PduL ), HS-CoA, propionyl-PO4 and standard components; therefore, propionyl-CoA was the expected acyl-CoA product. Reverse- phase HPLC with UV detection at 260 nm was used to identify CoA compounds present in reaction mixtures. A single major compound with a retention time of 10.9 min was detected when Pta- PduL+ cell extracts were used. This compound co-eluted with authentic propionyl-

CoA following co-injection and produced a mass spectrum characteristic of propionyl-CoA via

HPLC-ESI-MS (Fig. 2.4). The major peaks for authentic propionyl-CoA and the PduL reaction product that eluted at 10.9 min were at m/z = 824.1 or 824.2, respectively. These peaks correspond to [M + H]+ for propionyl-CoA. In contrast, propionyl-CoA was undetectable by

HPLC in assay mixtures containing cell extract from BE291 (Pta- PduL-) and the major HPLC peak observed corresponded to HS-CoA (retention time = 4.5 min). Our interpretation of the above results is that propionyl-CoA is a product of the PduL reaction.

8.0 A 824.1

6.0

4.0

894.6 2.0 862.1

844.6 6 746.6 792.7 765.6 813.1 878.4 0.0 750 800 850 900

2.0 824.2

intensity x 10 x intensity B

1.5

1.0

0.5 851.2 772.5 744.1 862.1 891.1 0.0 750 800 850 900 m/z

Figure 2.4. Propionyl-CoA is produced by the PduL reaction. A. HPLC-ESI-MS of propionyl-CoA standard. B. HPLC-ESI-MS of a PduL assay after 10 min. incubation at 30 C. The assay initially contained 10 µg cell - + 2- extract from BE527 (Pta PduL ), 0.4 mM HS-CoA, 1 mM propionyl-PO4 and buffer.

Production of PduL-His8 protein. E. coli strain BE554 was constructed to produce high levels of recombinant PduL fused to 8 C-terminal histidine residues (PduL-His8). Protein expression by BE 554 and a control strain (BE119) were examined by SDS-PAGE (not shown) and enzyme assay. Relatively high amounts of a protein with a molecular mass near 27 kDa were present in the inclusion body fraction from the pduL expression strain, which is near the

predicted molecular mass for PduL-His8 (24 kDa). However, a modest amount of soluble protein near this mass was produced.

To better assess the production of PduL-His8, cell extracts were tested for PTAC activity.

Soluble extracts from the PduL production strain typically contained about 50 mol min-1 mg-1 activity whereas the control strain produced only 2 mol min-1 mg-1 PTAC activity. Hence, a reasonable amount of PduL-His8 appeared to be soluble and active under the production conditions used. On the other hand, although the inclusion body fraction from the expression strain contained high amounts of PduL-His8 protein, the PTAC activity of this fraction was typically about 2-fold lower than that of the soluble fraction indicating that most of the PduL in this fraction was inactive. No PTAC activity was detected in the inclusion body fraction from the

2- control strain. The PTAC assay used measured the conversion of propionyl-PO4 and HS-CoA

2- to propionyl-CoA and HPO4 . No activity could be detected when enzyme, HS-CoA or

2- propionyl-PO4 was omitted from the assay mixture. In each case, PTAC activity was linear with enzyme concentration (data not shown).

a Table 2.3. Kinetic constants for purified recombinant PduL-His8 .

Substrate Km Vmax Kcat Kcat/Km -1 -1 -1 (mM) (µmol min-1 mg-1) (sec ) (mM sec ) 2- propionyl-PO4 0.61  0.06 51.7  7.6 20.7 33.9 2- acetyl-PO4 0.97  0.26 13.4  1.8 5.4 5.6 aPTAC assays were performed by following the absorbance of reaction mixtures at 232 nm as described in 2- 2- material and methods. The kinetic constants for propionyl-PO4 and acetyl-PO4 were determined using an excess of HS-CoA (0.4 mM).

Purification and kinetic characterization of PduL-His8. PduL-His8 was purified from cell extracts of strain BE554 by nickel-affinity chromatography (Fig. 2.5). Purified PduL-His8

appeared homogenous following SDS-PAGE and staining Coomassie blue (Fig. 2.5). Kinetic

2- analysis showed that purified PduL-His8 was 3.8-fold more active with propionyl-PO4

2- -1 -1 compared to acetyl-PO4 (Vmax = 51.7 and 13.5 µmol min mg , respectively) (Table 2.3). It

2- 2- also had a lower Km for propionyl-PO4 than for acetyl-PO4 (0.61  0.06 versus 0.97  0.26

2- 2- mM). No activity was detected when PduL-His8, HS-CoA or propionyl-PO4 (or acetyl-PO4 ) was omitted from the assay mixture. In each case, the activity of PduL-His8 activity was linear with enzyme concentration (data not shown).

1 2 3 4 kDa

PduL-his8

Figure 2.5. SDS-PAGE analysis of PduL-His8 purification. Lane 1, molecular mass markers; lanes 2, soluble crude extracts from the PduL-His8 production strain (BE554) (10 µg protein); lanes 3, fraction obtained from the first elution step using buffer with 400 mM imidazole (2 µg protein); lane 4, fraction obtained from the second elution step using buffer with 400 mM imidazole (2 µg protein). A 12% acrylamide gel was stained with Coomassie Brilliant Blue R250.

-1 -1 For PduL-His8, crude cell extracts typically had PTAC activity of about 50 µmol min mg .

Nickel-affinity chromatography yielded purified enzyme with specific activities of about 45-55

-1 -1 µmol min mg . Visual inspection of SDS-PAGE gels indicated that PduL-His8 comprised

about 2-5% of the total cell protein (Fig. 2.5); hence, these results suggest that PduL-His8 lost activity during the course of purification. Indeed, PduL-His8 was unstable in the purified form and lost about 50% of its activity 48 h after purification. The reported kinetic constants are based on assays done as soon as possible after purification. As the activity of PduL-His8 decreased, its

2- Km for HS-CoA and propionyl-PO4 remained unchanged within experimental error. Its activity

2- 2- with propionyl-PO4 and acetyl-PO4 decreased in direct proportion. We infer that the Km values reported here are representative, but the Vmax values are probably an underestimate.

2.5. Discussion

Here we presented genetic and biochemical evidence that the pduL gene encodes a PTAC involved in 1,2-PD degradation. Growth tests showed that a pduL mutant was impaired for 1,2-

PD degradation under aerobic and anaerobic conditions indicating a role for pduL in 1,2-PD degradation. Growth studies also showed that ectopic expression of PduL corrected the growth defect of a pta mutant on acetate minimal medium indicating that PduL has phosphotransacetylase activity in vivo. Enzyme assays of crude cell extracts demonstrated that pduL was necessary for the production of detectable PTAC activity in a genetic background that contained a pta mutation. HPLC-ESI-MS showed that propionyl-CoA was produced from

2- propionyl-PO4 and HS-CoA in PduL enzyme assays. PduL-His8 was purified and found to have

-1 -1 2- a specific activity of 57.4 µmol min mg with propionyl-PO4 as substrate indicating the PduL is sufficient for PTAC activity. Thus, genetic and biochemical studies indicate that PduL is a

PTAC involved in 1,2-PD degradation.

Blast and PSI-Blast analyses (5 iterations) showed that PduL lacks significant similarity to known PTAC enzymes. Thus, PduL is evolutionarily distinct. It either evolved independently of known PTAC enzymes or it became highly divergent from them over time. Sequence analyses

also showed that PduL homologues (expect <7x10-30) are found in 49/337 (14.5%) complete microbial genomes present in GenBank. No homologues of PduL were found among the Archaea or Eukarya. Hence, PduL is apparently a conserved protein specific to the Bacteria. In addition,

Blast analyses identified 4 PduL homologues that are fused to sequences with homology to acetate kinases. These may be dual function enzymes that convert acyl-CoA to its corresponding organic acid with the synthesis of ATP.

Including PduL, there are three known classes of phosphotransacylases. The S. enterica PduL,

Pta, and EutD enzymes are representatives of these classes. PduL is comprised of 210 amino acids while Pta and EutD are 714 and 338 amino acids in length, respectively (234). EutD and the C-terminal region of Pta are homologous. PduL lacks significant similarity to EutD and Pta.

The N-terminal region of Pta contains BioD-like and "DRTGG" domains. These two domains are absent from EutD and PduL. The physiological role of these domains is uncertain. In enterica bacteria, acetyl-group metabolism and is linked to two-component signal transduction systems and DNA damage and repair in a complex manner (254-256). The N-terminal domains of Pta may be needed to accommodate expanded physiological roles.

Kinetic studies were performed with purified PduL-His8 (Table 2.3). The Vmax of PduL with

2- -1 -1 propionyl-PO4 as substrate was 51.7  7.6 µmol min mg . Previously purified Pta enzymes were reported to have specific activities between 32 and 9006 µmol min-1 mg-1 when an acyl-

2- PO4 was used as substrate (257). The activity of PduL-His8 is at the low end of this range, but may be an underestimate due to its instability (see results section).

2- The Km values of PduL-His8 for HS-CoA and propionyl-PO4 were determined to be 0.61 

0.06 and 0.032  0.006 mM, respectively. Reported Km values of Pta homologues for HS-CoA

2- range from 0.03 to 1.7 mM and for acetyl-PO4 range from 0.024 to 4.7 mM (257). Thus, Km

2- values of PduL for HS-CoA and propionyl-PO4 appear to be within a physiologically meaningful range. Kinetic studies found that Kcat/Km of Pdu-His8 was about 6-fold higher for

2- 2- 2- propionyl-PO4 compared to acetyl-PO4 . The selectivity of PduL for propionyl-PO4 is consistent with a role in 1,2-PD degradation. Furthermore, its relatively low activity with acetyl-

2- PO4 may be important to minimize perturbation of acetyl-group metabolism (254, 256).

Two findings in this report were surprising to us: 1) the observation that pduL mutants were significantly impaired for growth on 1,2-PD minimal medium (Fig. 2.2) the finding that a pduL pta double mutant was not further impaired for growth on 1,2-PD. Because Pta is known to be

2- active with propionyl-PO4 , we expected that Pta would partly substitute for PduL during growth of S. enterica on 1,2-PD. However, in the studies reported here, it did not. A simple explanation is that PduL is more efficient at propionyl-group metabolism in vivo. Two explanations we think are more interesting are differential regulation of pta and pduL, as well as the possibility that

PduL is specifically adapted to function in concert with the microcompartment involved in 1,2-

PD degradation.

2.6. Acknowledgements

We thank the ISU DNA Sequencing and Synthesis Facility for assistance with DNA analyses.

We thank Ann Perera and the ISU Metabolomics Laboratory for help with mass spectrometry.

We thank the ISU Office of Biotechnology for financial support.

CHAPTER 3. PDUL RECYCLES COENZYME A IN PDU MICROCOMPARTMENT IN SALMONELLA ENTERICA

This chapter, in part or in full, contains the manuscript to be submitted to a journal.

3.1. Abstract

Salmonella enterica is capable of growing on 1,2-propanediol, whose degradation is carried out in propanediol utilization microcompartments (Pdu MCPs). Our previous study showed PduL is a new class phosphotransacylase involved in MCP-associated 1,2-propanediol (1,2-PD) metabolism. Here, we demonstrated by Western blots that PduL is a component of Pdu MCP.

Genetic studies showed PduL in MCPs can’t utilize acetyl-CoA in the cytoplasm to support

Salmonella growth, and PduL recycling coenzyme A in the Pdu MCPs is indicated. Growth tests as well as protein immunoblots also provided evidence that N-terminal region of PduL functions as a targeting sequence for MCP encapsulation. These data, together with recent publications, indicate that cofactor recycling might be a general mechanism utilized among MCPs to optimize metabolic efficiency. The targeting study further expands our knowledge of enzyme encapsulation during MCP assembly. Complete understanding of MCPs might provide a basis for the development of nanocontainers for use in biotechnology.

3.2. Introduction

Microcompartments (MCPs) are a family of subcellular proteinaceous organelles utilized by

14.6% of bacteria to optimize metabolic pathways with toxic or volatile intermediates involved

(24, 25, 65). These polyhedral protein complexes are 100-150 nm in size and consist of a multi- protein shell sequestering enzymes which work sequentially in a specific metabolic pathway.

Gene expression analysis and bioinformatics studies predicted the presence of at least seven types of MCPs with various functions (23, 65). The carboxysome is the first identified MCP which functions in retaining CO2 to increase carbon fixation efficiency (258). The two other

MCPs that have been investigated so far are named 1,2-propanediol utilization microcompartment (Pdu MCP) and ethanolamine utilization microcompartment (Eut MCP).

They are used for metabolizing a three-carbon molecule 1,2-PD and a two-carbon compound ethanolamine, respectively while sequestering the toxic metabolite aldehyde generated during the degradation processes (23, 57). By retaining these metabolic pathways in MCPs, bacteria are able to utilize 1,2-PD and ethanolamine as carbon and energy source while avoiding cellular toxicity caused by aldehyde accumulation and increasing metabolism efficiency by substrate channeling. 1,2-PD and ethanolamine are present in the human gut, the ability of S. enterica of growing on both of them contribute to Salmonella pathogenesis (75, 94, 95, 97, 259).

In Salmonella enterica, degradation of 1,2-PD associated with Pdu MCPs is a coenzyme B12- dependent process which involves multiple enzymes and cofactors. Nine different polypeptides

PduABB’JKMNTU self-assemble to form a polyhedral shell that encapsulates key enzymes for metabolizing 1,2-PD (39, 49). 1,2-PD is taken into MCP by unknown mechanisms which might involve the selective pores on the shell (36, 58, 73). The first two reactions for 1,2-PD degradation are carried out in the lumen of MCP (Fig. 3.1). It starts with the conversion of 1,2-

PD to propionaldehyde by Ado-B12-dependent diol dehydratase PduCDE (22). Next, another major enzyme propionaldehyde dehydrogenase PduP turns propionaldehyde into propionyl-CoA in the MCPs (22, 49). Our previous model claimed that the following rest steps of 1,2-PD degradation happen outside of MCPs. Propionyl-CoA is either utilized by phosphotransacylase

2- PduL to produce propionyl-PO4 , which is converted to final product propionate by the reaction of propionate kinase PduW or feeds into methylcitrate pathway (50). Alcohol dehydrogenase

PduQ could alternatively convert propionaldehyde to 1-propanol. Recently Cheng et al. (52) showed that PduQ was also associated with Pdu MCPs and it catalyzed propionaldehyde reduction inside MCPs to regenerate NAD+, the cofactor required for PduP reaction. Besides the enzymes and shell components, Ado-B12, is an active form of coenzyme B12 and is required for

the reaction catalyzed by diol dehydratase PduCDE, another two cofactors NAD+ and HS-CoA are required for the function of PduP.

PduW propionate ADP ATP

Figure 3.1. A model for 1,2-PD metabolism associated with Pdu MCPs and the role of PduL. The Pdu MCP consists of a series of enzymes encapsulated within a multi-protein shell. The enzymes are used for degrading 1,2-PD. The shell is composed of 9 polypeptides PduABB’JKMNTU. The enzymes encapsulated are four 1,2- PD degrading enzymes dial dehydratase DDH (PduCDE), propionaldehyde (PPA) dehydrogenase (PduP), 1- propanol dehydrogenase (PduQ) and phosphotransacylase (PduL). The proper function of DDH requires + vitamin B12 in its active form Adenosyl-cobalamin (Ado B12). Both NAD and HS-CoA is a required coenzyme for PduP. The proposed function of Pdu MCP is to retain aldehyde to avoid the toxicity to genomic DNA and to channel the intermediate to the downstream enzymes to increase the efficiency of the metabolic pathway. PduQ regenerate NAD+ inside of MCPs (52). PduL could recycle propionyl-CoA (prp-CoA) to regenerate HS-CoA which is a necessary coenzyme for the function of PduP (this study).

The shells of MCPs function as a barricade to keep volatile substrates or toxic metabolites inside MCPs and segregate them from cytoplasm. While at the same time the metabolic processes in MCPs require input of substrates as well as cofactors and output of products.

Intensive studies have been focusing on solving how metabolites cross the shell. Analysis of crystal structures of MCP shell proteins indicated that major shell components of different types of MCPs contain small pores in the center of their disc-shaped homohexamers, which are the basic building blocks of the shell (35, 36). Caboxysome shells tend to have positive electrostatic potential around the central pores which may bring in negatively charged substrate bicarbonate

(35). One abundant Pdu MCP shell protein PduA has been reported to contain hydrogen bond acceptors in the central pores which may facilitate the transport of the polar substrate 1,2-PD (36,

73). Compared to the size of the substrates, cofactors NAD+ and HS-CoA are much bigger and hence harder to get through the shell. Recent studies revealed that tandem BMC-domain containing shell proteins CsoS1D and CcmP have conserved glutamate and arginine residues which adopt different conformations and causes the pores to either close or open (up to 14 angstrom in diameter for CsoS1D) (87, 260). Yet so far these theories are based on analysis on protein structures and no direct experimental evidence has shown that bulky cofactors are transported through gated pores. Recently, Cheng et al. (52) showed that alcohol dehydrogenase

PduQ regenerated NAD+ within the Pdu MCPs in Salmonella and thus provided the first evidence for cofactor internal recycling within bacterial MCPs. Soon after that, Huseby et al. (78) used genetic methods to demonstrate that NAD+ and coenzyme A were recycled and retained within the Eut MCPs by two enzymes EutD and EutG during ethanolamine catabolism. These two studies open up the possibilities that MCPs are capable of reusing cofactors independent of outside input.

One of the fundamental questions about MCP functions is that how the enzymes are encapsulated into MCPs during their assembly. Sequence alignment identified short N-terminal extra sequence on MCP-associated enzymes PduP, PduD and EutC compared to their

homologues which don’t have MCP associated genomic context (101, 261, 262). Following this prediction, Fan et al. (262) reported that N-terminal 18 amino acids on PduP sufficiently package the enzyme and several foreign protein tags into the lumen of MCP. Later N-terminal extension of the medium subunit of diol dehydrase was also found to package the whole enzyme into Pdu

MCPs (261). Further in vivo targeting study proved that interaction between the targeting sequence on PduP and shell protein PduA is critical for encapsulation of PduP into MCPs (263).

In our previous model, PduL was proposed to be outside of MCPs. In this study, we demonstrated that PduL is a component of Pdu MCPs. We then investigated the specific functions of PduL related to MCPs and presented evidence that it recycles coenzyme A inside

Pdu MCPs. These results support the earlier studies of cofactor recycling and suggest that internal recycling could be an alternative mechanism for MCPs to acquire cofactors. We also showed that N-terminal region of PduL has encapsulation effects. This finding, together with recent targeting studies, indicate that short targeting sequences use a general mechanism to direct protein encapsulation in MCPs.

3.3. Materials and Methods

Chemicals and reagents. Antibiotics, CN-Cbl, DNase I were from Sigma-Aldrich (St. Louis,

MO). Coenzyme A was from MP Biomedicals (Santa Ana, CA). Isopropyl b-D-1- thiogalactopyranoside (IPTG) was from Diagnostic Chemicals Limited (Charlottesville PEI,

Canada). KOD DNA polymerase was from Novagen (Cambridge, MA). Restriction enzymes and

T4 DNA ligase were from New England Biolabs (Beverly, MA). Bacterial protein extraction reagent (B-PER II) was from Pierce (Rockford, IL). The cOmplete protease inhibitor was from

Roche Diagnostics Corporation (Indianapolis, IN). Other chemicals were from Fisher Scientific

(Pittsburgh, PA).

Bacterial strains and growth conditions. The bacterial strains used in this study are derivatives of Salmonella enterica serovar Typhimurium LT2 (Table 3.1.) The rich media used were LB/Lennox medium (Difco, Detroit, MI) (237). The minimal medium used was no-carbon-

E (NCE) medium (239).

Table 3.1. Bacterial strains used in this study.

Strain Genotype Source S. enterica serovar Typhimurium LT2 BE188 ∆pduL670 T.A.Bobik lab collection BE291 ∆pduL670, pta209::Tn10 T.A.Bobik lab collection BE527 pta209::Tn10 T.A.Bobik lab collection BE647 ∆pduABB’ T.A.Bobik lab collection BE717 pta209::Tn10∆pduABB’ T.A.Bobik lab collection BE881 ∆pduABB’∆pduL670 This study BE790 ∆pduL670, pta209::Tn10/Plac22 T.A.Bobik lab collection BE791 ∆pduL670, pta209::Tn10/Plac22-pduL T.A.Bobik lab collection BE1112 ∆pduL670, pta209::Tn10/Plac22- PduL∆1-5 This study BE1114 ∆pduL670, pta209::Tn10/Plac22- PduL∆1-10 This study BE1766 ∆pduL670, ∆pduABB’, pta209::Tn10 This study BE2021 ∆pduL670, pta209::Tn10 /Plac22- eGFP This study BE2022 ∆pduL670, pta209::Tn10 /Plac22- pduL1-20-eGFP This study BE2025 ∆pduL670, pta209::Tn10 /Plac22-pduL-linker-eGFP This study

MCP purification, SDS-PAGE and Western blots. Pdu MCPs were purified as previously described (39) except carbenicillin and IPTG were added to induce protein expression from a vector pLAC22. Protein assay was carried out by use of protein assay reagent (Bio-rad, Hercules,

CA) with 1 mg/ml bovine serum albumin (BSA) as a standard. SDS-PAGE was performed using

4–20% Mini-PROTEAN TGX gels from Bio-Rad followed by staining using Bio-Safe

Coomassie Stain (Bio-Rad). For Western blots, the nitrocellulose membranes were probed using mouse monoclonal anti-GFP (Fisher Scientific) or rabbit polyclonal anti-PduL (Genscript,

Piscataway, NJ) with 0.5 ug/ml in TBST buffer and goat anti-mouse or anti-rabbit IgG HRP

(horseradish peroxidase) conjugate at 40 ng/ml in TBST buffer (Santa Cruz Biotechnology,

Santa Cruz, CA). Signal development was carried out by using SuperSignal West Pico

Chemiluminescent Substrate (Pierce, Rockford, IL) according to the manufacturer’s instructions.

The result was analyzed by ChemiDoc XRS+ Imager (Bio-Rad).

Growth studies. Growth studies were carried out using Synergy HT microplate reader

(BioTek, Winooski, VT) as previously described (50). Each growth curve was a representative of three independent experiments in triplicate. Doubling times were calculated from semilog plots with the formula doubling time = 0.693/(2.303 X slope of the linear region of the plot).

3.4. Results

Sequence analysis. Prior studies showed that short N-terminal are used to target proteins to the lumen of bacterial MCPs (261, 262). Sequence alignment of 8 representative PduL homologues from different bacterial organisms showed that N-terminal sequence extensions are present in four of the eight homologues whose genomic context suggests the potential formation of MCPs in the organism (Fig. 3.2A). Prediction of secondary structure of PduL was carried out using programs PredictProtein (Fig. 3.2B). 14 residues on the N-terminus were predicted to have propensity for an α-helix formation, which was consistent with the recent reports that PduP and encapsulated proteins from more than one type of MCPs have terminal targeting sequence which was prone to form α-helices (262, 263). This findings suggested that PduL might be a component of the Pdu MCP even though it was not identified by prior proteomics analyses purified MCPs

(49).

PduL is a component of Pdu MCP. To investigate the cellular location of PduL, MCPs purified from wild-type S. enterica as well as a pduL deletion mutant BE291 were analyzed by

SDS-PAGE and Western blots. By Western blot (Fig. 3.3), a strong band near 24 kD was

detected in MCPs purified from Salmonella using anti-PduL sera, while a weaker band at same size was readily detected in crude extracts. No bands at this size were detected in either whole cell extracts or MCPs purified from a pduL deletion strain BE291. These results indicated that

PduL is a component of Pdu MCP.

Figure 3.2. Sequence analysis and structure prediction of PduL N-terminus. Panel A. Sequence alignment of the N-terminal region of 8 representative PduL homologues from different bacterial organisms. Terminal sequence extensions are present in four of the eight homologues whose genomic context suggests the potential formation of MCPs in the organism. PduL of S. enterica belongs to PduL superfamily (cl05584). The sequence contains two tandem domains. Several members in this superfamily catalyze interconversion of acetyl-CoA and acetyl-phosphate, and propionyl-CoA and propionyl-phosphate. The National Center for Biotechnology Information gene accession identifications, from the top to the bottom, are 496567468, 485596631, 515719363, 239826378, 5069455, 16503278, 375257112, and 332161170. Panel B, C and D. Secondary structure prediction of N-terminal region of PduL by software B: Psipred; C: JPRED; D: PredictProtein.

100

A LT2 ∆pduL B kDa M 1 2

75 50 37

25 PduL 20 1 2 3 4

Figure 3.3. PduL is a component of the Pdu MCP. A. Western blot for PduL with anti-PduL. Lane 1 and 2, 10 μg whole-cell extract and 10 μg purified MCPs from wild-type strain LT2; lane 3 and 4, 10 μg whole-cell extract and 10 μg purified MCPs from a pduL deletion mutant (BE188). B. 10-20% SDS-PAGE gel stained with Bio-Safe Coomassie. Lane 1, molecular mass markers; lane 2, 10 μg Pdu MCPs purified from LT2; lane 3, 10 μg Pdu MCPs purified from BE188.

A specific band with molecular mass close to PduL was not identified by SDS-PAGE. Most likely, PduL was obscured by major components of Pdu MCPs PduB’ and PduD, which had similar molecular mass as PduL (24.1 kD for PduB', 24.2 kD for PduD vs 23.1 kD for PduL).

PduL is capable of packing heterologous GFP into the Pdu MCPs. To further investigate the cellular location of PduL, we tested whether PduL could mediate the encapsulation of eGFP into Pdu MCPs. A PduL-eGFP fusion protein was expressed from plasmid pLAC22 in a pduL deletion mutant. PduL fused to eGFP was readily detected in purified MCPs as a strong band near 50 kD by anti-GFP monoclonal antibody (Fig. 3.4). eGFP was not detected in MCPs purified from wild type Salmonella or a ∆pduL mutant expressing native eGFP. The fusion proteins appeared in the whole-cell extract fraction, indicating the normal expression. Higher

101

amount of the fusion protein present in eMCPs versus the whole-cell lysate fraction was further evidence that PduL is targeted to MCPs.

MCPs Whole cell extracts

50kD 37kD

25kD 20kD

Lane 1 2 3 4 5 6 7

Figure 3.4. PduL-eGFP fusion protein was detected in Pdu MCPs. Lane 1 and 6: Western standards; lane2-4: 10ug Pdu MCPs purified from: a wild type S. enterica LT2, ∆pduL/pLac22-eGFP (BE2021), and ∆pduL/pLac22-pduL1-20-eGFP (BE2025); lane 5-7: 10ug whole-cell extracts of LT2, BE2021 and BE2025.

N-terminal region of PduL is required for targeting into MCPs. To test for an N-terminal targeting sequence on PduL, we expressed truncated pduL lacking N-terminal 5, 10 and 20 amino acids from vector pLAC22 in a pduL deletion background. To test the cellular location of the truncated PduL proteins, MCPs were purified from the mutants and analyzed by Western

Blots for PduL. Compared to wild type, significantly less PduL was detected in purified MCPs from PduL∆1-5 and PduL∆1-10 (Fig. 3.5). Comparable amount of proteins in the whole-cell extracts for the mutants and control strain indicating proper expression of the mutants. We also examined targeting of PduL ∆1-20; however none could be detected by Western blot suggesting this mutant protein was unstable. As an additional control we did an in vivo test for PduL enzyme activity as

102

described (50). This test showed that PduL∆1-5 and PduL∆1-10 were active enzymes but that

PduL∆1-20 was inactive enzymatically or unstable.

M MCPs Whole-cell lysates

25 kD 20 kD

Lane 1 2 3 4 5 6 7

Figure 3.5. N-terminal deletion affects encapsulation of PduL into Pdu MCPs. Lane 1: Molecular mass marker; lane 2-4: 10, 25 and 7.5 ug purified Pdu MCPs from control strain ∆pduL/pLac22-pduL (BE791), ∆pduL/pLac22-pduL∆1-5 (BE1112) and ∆pduL/pLac22-pduL∆1-10 (BE1114); lane 5-7: 10, 25 and 7.5 ug whole- cell lysates from BE791, BE1112 and BE1114.

PduL N-terminal 20 amino acids can target eGFP to the Pdu MCP. To determine whether the N-terminal extension of PduL is sufficient for encapsulating proteins into the Pdu

MCP, we fused PduL N-terminal 20 amino acids to eGFP and looked for the fusion proteins in the purified MCP by Western blots with anti-GFP. PduL1-20-eGFP was produced using the IPTG- inducible promoter on plasmid pLAC22 in a background carrying a pduL deletion mutation.

Western blots indicated that PduL1-20-eGFP (~28 kD) co-purified with Pdu MCPs while eGFP did not (Fig. 3.6). At the same IPTG level, similar amounts of eGFP and PduL1-20-eGFP were detected in cell lysates, indicating normal production and folding of both proteins. At lower

IPTG levels (0.02 mM), Western blots of purified MCPs showed a band for PduL1-20-eGFP but not for eGFP. Low levels of both PduL1-20-eGFP and eGFP were present in cell lysates. These

103

results indicated that a short N-terminal 20 amino acid sequence from PduL is sufficient for packaging proteins into Pdu MCPs.

MCPs Cell lysates

Lane 1 2 3 4 Figure 3.6. PduL N-terminal 20 amino acids showed targeting effect. Lane 1, 2: 15ug MCPs purified from: eGFP-expressing strain BE2021 and PduL1-20-eGFP expressing strain BE2022, respectively. Lane 3, 4: 15ug whole-cell extracts from: eGFP-expressing strain BE2021 and PduL1-20-eGFP expressing strain BE2022, respectively. 0.2mM IPTG induction was used for all samples.

Genetic effects of preventing MCP formation on the growth of a pduL deletion mutant.

Prior studies showed that pduL deletion mutant had a reduced growth rate on 1,2-PD even

2- though Salmonella expresses a housekeeping Pta enzyme which is active with propionyl-PO4

2- (the primary substrate of Pta is acetyl-PO4 ) (50). Previously, it was proposed that PduL provides additional enzyme activity needed to support growth on 1,2-PD (50). However, an alternative explanation is that PduL might have an MCP-specific function such as internal HS-

CoA recycling (Fig. 3.1). To test for an MCP-specific function for PduL, we looked at the effect of breaking the Pdu MCP genetically (with a pduABB’ deletion which was shown to prevent formation of the Pdu MCP (70)) on the phenotype of a pduL deletion mutant. Results showed

104

that in a MCP+ background, a pduL deletion mutant grew slower than the wild type strain LT2 as was previously reported. In a MCP- background, the ∆pduL mutant actually grew faster than

PduL+ (doubling time of 3.43 ± .1 h vs 4.16 ± .12 h) (Fig. 3.7).

1.2

0.8 O D 6 0.6 0 0 0.4 LT2 ∆pduAB 0.2 ∆pduL ∆pduAB ∆pduL 0 0 5 10 15 20 25 30 35 40 45 TIME(H)

Figure 3.7. Genetic effects of preventing MCP formation on the growth of a pduL deletion mutant. Wild type LT2, a ∆pduL mutant (BE188), a ∆pduABB’ strain with broken MCP phenotype (BE647), and a ∆pduL ∆pduABB’ triple mutant (BE881) were grown on 1,2-PD minimal medium supplemented with 100 nM CN-Cbl

(cyanocobalamin, vitamin B12).

This suggested that PduL is necessary for the proper functioning of an intact Pdu MCP and might have a role in coenzyme A recycling (Fig. 3.1). This also argues against the idea that the role of PduL is to provide more activity. We noticed that PduL- MCP- grew faster than the wild type strain (PduL+ MCP+) (doubling time of 3.43 ± .1 h vs 4.85 ± .3 h). One possible

105

interpretation is that the PduL- strain accumulated propionyl-CoA to higher levels increasing flux through the methylcitrate pathway (Fig. 3.1) (264).

Slow growth of a pta mutant on acetate minimal medium was partially corrected by mutations that broke the MCPs. Next we tested whether the Pdu MCP restarts access of

2- acetyl-CoA to PduL. Salmonella can growth on acetate by converting it first to acetyl-PO4 and then to acetyl-CoA by the ordered action of two house-keeping enzymes acetate kinase (AckA) and Pta (247). Prior studies showed a pta mutant grows poorly on acetate but this phenotype is

2- corrected by production of PduL from a plasmid (PduL also converted acetyl-PO4 to acetyl-

2- CoA) (50). To test whether acetyl-PO4 and acetyl-CoA can move freely across the MCP shell we tested whether PduL encapsulated within the Pdu MCP could correct the growth defect of a

Pta mutant on acetate. Results show that induction of the pdu operon (which leads to MCP formation) only slightly increased growth of a Pta strain on acetate (doubling time = 7.03 ± .05 h)

(Fig. 3.8). This indicated that PduL could not effectively replace the function of Pta when encapsulated within the Pdu MCPs. However, when the pdu operon was induced in a background that cannot form the MCP shell (∆pduAB) much better growth of a Pta strain on acetate was observed (3.46 ± .07 h) (Fig. 3.8). The growth difference of PduL+ Pta- MCP+ strain and PduL+

- - 2- Pta MCP suggest that the shell of the Pdu MCP restricts the movement acetyl-PO4 and/or acetyl-CoA and impairs the ability of PduL to correct that Pta defect. Restriction of metabolite movement by the shell of the Pdu MCP is consistent with the need for internal recycling of CoA as shown in Figure 3.1. For the experiments described above, expression of the pdu operon was induced by supplementation of the growth medium with 1,2-PD. No 1,2-PD metabolism occurred under these conditions because coenzyme B12 (a required cofactor for 1,2-PD degradation) is not present.

106

0.7

0.6

0.5

O 0.4 D LT2 6 ∆pta 0 0.3 0 ∆pta+PD ∆pta ∆AB 0.2 ∆pta ∆AB+PD ∆pta ∆L 0.1 ∆pta ∆L+PD ∆pta ∆L ∆AB 0 0 5 10 15 20 25 30 Time (h)

Figure 3.8. Slow growth of a pta mutant on acetate minimal medium was partially corrected by mutations that broke the MCPs. LT2 and the Mutants were grown in acetate minimal medium with or without 1,2-PD induction.

N-terminal deletion of pduL showed targeting defect. Based on the above finding that

2- acetyl-PO4 /acetyl-CoA do not pass freely through the shell of the Pdu MCP, we developed a growth test for PduL targeting. As mentioned above, the pduL gene expression was induced by addition of 1,2-PD to the acetate growth medium. Mutations that prevent encapsulation of PduL within the Pdu MCP should enhance growth on acetate in a Pta- background. Tests showed that both PduL∆1-5 and PduL∆1-10 grew faster than the Pta- control with 1,2-PD supplementation, indicating more PduL enzyme was present in the cytosol (Fig. 3.9). This finding further supported the western blots shown above which indicated PduL has an N-terminal targeting sequence. The comparison between the control and each mutant was calculated for both doubling time and cell density (Table. 3.2). Both PduL∆1-5 and PduL∆1-10 grew faster than the control. It is

107

possible that the partial deletions resulted in either higher phosphotransacylase activity or more efficient protein expression and folding. Here the growth test was done at 30 °C instead of 37 °C due to the concern that, at higher temperature, the fast growth of the mutants might exceed the

MCP assembling rate which would artificially result in more significant results. This growth test further quantitatively showed that deletion of N-terminal 10 amino acids had more severe effect on targeting compared to N5 deletion (50% of control strain vs 78% of control strain, table 3.2).

A B

Figure 3.9. N-terminal deletion of pduL showed targeting defect. The growth of the control strain with wild type PduL (BE791), a PduL∆1-5 mutant (BE1112, panel A), and a PduL∆1-10 mutant (BE1114, panel B) in acetate minimal medium were adjusted to the same level, then the growth of BE1112 (panel A) and BE1114 (panel B) were compared to BE791 in acetate medium supplemented with 1,2-PD, respectively. All strains contained a pduL and pta double mutation.

3.5. Discussion

In our previous study, Havemann and Bobik (49) identified 15 polypeptides associated with

Pdu MCPs by using of two-dimensional electrophoresis followed by N-terminal sequencing and protein mass fingerprinting. PduL was not found in this process. We thought PduL might not be

108

detected on the acrylamide gel due to its low abundance in the purified MCPs and its high theoretical pI (9.09) which might cause the migration of PduL off the gel during isoelectric focusing (49). In this study, we provided evidence that PduL is a component of Pdu MCP.

Western blots clearly showed the presence and enrichment of PduL in purified MCPs. PduL fused to eGFP was readily detected in purified MCPs, which further suggested that PduL is associated with Pdu MCPs. Although we didn’t visualize PduL on SDS-PAGE. We also measured the phosphotransacylase activity and didn’t observe higher enzymatic activity in purified MCPs versus that of whole-cell lysate, which is possibly due to the unstable nature of

PduL and also the interference of enzymatic activity by nonionic detergent B-per II used for

MCP purification (50, 265).

Table 3.2. N-terminal deletion of pduL showed targeting defect on acetate/PD growth assay.

a b Strain IPTG conc: mM % of cell density (OD600) % of doubling time (h) BE1112 .01 77.7 ± 2.5 80 ± 5 .02 73.7 ± 6.5 73.3 ± 14.3 BE1114 .003 59 ± 9 49 ± 10 .005 57 ± 11 46 ± 9

Each growth curve was repeated three times in triplicate. The numbers are mean value ± 1 standard deviation. a% of cell density = (difference in cell density of the mutant growing in acetate medium and acetate/PD medium)/(difference in cell density of strain BE791 at the time point when BE791 reached maximum optical density measured at a wavelength of 600 nm). b% of doubling time = (difference in the doubling time of a mutant growing in acetate medium and acetate/PD medium)/(difference in the doubling time of control strain BE791 growing in acetate medium and acetate/PD medium)

We also investigated the reasons that PduL is associated with the Pdu MCP. Our recent study indicated that PduQ functions by recycling NAD+ internally within the Pdu MCPs (52). Cofactor

109

NAD+ can’t diffuse freely through the shell, thus regeneration of NAD+ provides a possible way for the MCP to acquire cofactors. This proposed theory was further supported by another study showing that Eut MCP retains a private pool of HS-CoA (78). Considering that large coenzyme

A may also be reused in Pdu MCPs, we carried out growth study to test the effects of breaking

MCPs on PduL. The results showed reduced growth of a pduL mutant with intact MCPs on 1,2-

PD compared to wild type Salmonella and faster growth of PduL- versus PduL+ with broken

MCPs, indicating that PduL is required for an intact Pdu MCP to function properly. Further investigation of growth on acetate showed that PduL only slightly corrected the growth defect of a Pta- strain, suggesting PduL within the Pdu MCP could barely replace the function of Pta. After the MCPs were broken, PduL functioned more efficiently in the cytoplasm and partially corrected the growth of the Pta- strain. These results indicated that the access to substrate acetyl-

CoA is restricted by the shell, and coenzyme A internal recycling inside the Pdu MCP is strongly suggested. Based on our results and the aforementioned findings, we propose that internal recycling of both HS-CoA and NAD+ helps Pdu MCPs function as a self-supporting metabolism system and contribute to optimal growth of S. enterica on 1,2-PD. Our proposed theory didn’t rule out the possibility that CoA might also be transported through the shell. In fact, we observed very slow growth of Pta- PduL+ MCP+ on acetate, indicating that limited but not eliminated amount of acetyl-CoA manage to trespass the MCP shell.

Function of MCP requires lumen enzymes are encapsulated into the protein shell. Fan et al.

(261, 262) showed that the short N-terminal sequences of PduD and PduP efficiently package core enzymes PduCDE and PduP into the MCPs, respectively. Their bioinformatics analysis also predicted N-terminal targeting sequences for EutC and EutG, lumen proteins associated with Eut

MCPs. Later studies by Choudhary et al. (101) demonstrated that N-terminal 19 amino acids of

110

EutC is sufficient to target eGFP and β-galactosidase into native and recombinant Eut MCPs.

Our sequence alignment and structure predictions suggested that PduL homologues from MCP- associated bacterial genomes feature a short N-terminal sequence containing alternative charged and hydrophobic residues with potential of forming an α-helix, which is conservative among

MCP-encapsulated proteins. Western blots showed that PduL lacking N-terminal residues appeared at significantly lower levels in purified MCPs, indicating targeting is affected by the N- terminal deletion. We also detected the presence of PduL1-20-eGFP in purified MCPs, which suggested that N-terminal 20 amino acids of PduL is efficient in packaging the enzyme into

MCPs. These results, in conjunction with the above studies, expands our understanding of how

MCP-associated enzymes and shell components form the highly organized protein structures.

Self-assemble and compartmentalization significantly increase metabolic efficiency by sequestering toxic and volatile intermediates, and promoting substrate channeling, which make

MCPs great candidates as bioreactors. This study, together with other targeting and recycling explorations, promotes the discovery of short targeting sequences and a thorough understanding of molecular transport mechanisms of MCPs. The acquired knowledge makes it possible for genetic and biochemical engineering of self-assembling bioreactors with desired metabolic enzymes encapsulated into the MCP shell (266). Successful encapsulation of functional heterologous enzymes into reconstructed MCPs have been observed (101, 187). As a proof of concept, an ethanol bioreactor was generated in vivo by packaging ethanol producing enzymes with a targeting sequence into recombinant Pdu MCPs (102). A recent study showed that Pdu

MCPs retain integrity over time and at various temperatures, and breaking of MCPs could be controlled by changing pH (89). This further indicates the robustness of MCPs and provides a

111

possible drug releasing mechanism. To utilize MCPs for delivery of chemotherapeutic drugs might protect normal tissues without compromising enzymatic efficiency of therapeutic peptides.

3.6. Acknowledgements

This work was supported by grant MCB-0956451 from NSF. We thank the ISU DNA

Sequencing and Synthesis Facility for assistance with DNA analyses. We thank Dr. Chiranjit

Chowdhury from Bobik lab for the help with sequence alignment. We thank Dr. Shamistha Sinha and Brent Lehman for providing the plasmid pTA925-pduBB’102/207A.

112

CHAPTER 4. PROTEIN-PROTEIN INTERACTIONS IN THE PDU MCP

Part of this work has been incorporated into a paper and submitted to the journal PLoS Computational Biology.

113

4.1. Abstract

Studies on diverse MCPs indicated that protein-protein interactions (PPIs) between the lumen enzymes and the shell components are crucial in assembly of MCPs and in the organization of these protein complexes. Our targeting studies (see chapter 3) indicated that lumen enzyme PduL, similar to major lumen enzymes PduP and PduD, also has a short N-terminal targeting sequence, while the shell protein which interacts with PduL to mediate the encapsulation was not identified.

Here we provided preliminary evidence that PduL binds to PduBB’. In vivo bacterial two-hybrid screening test indicated the possible binding interaction between PduL and PduB’. Further investigation using in vitro nickel-affinity pull-down tests followed by Western blotting provided another evidence of L-BB’ binding. Our data is consistent with the Pdu interactome prediction by Jorda et al. (unpublished data) using computer based docking simulations. Our two-hybrid system also suggested a positive interaction between PduV, a protein with unknown functions, and shell protein PduU, which was also predicted within the PPI network. Prediction by interactome model followed by experimental confirmation may significantly improve the accuracy and efficiency in discovering PPIs which underlie MCP functions.

4.2. Introduction

Various mechanisms are utilized for targeting proteins into organelles or certain parts within eukaryotic and prokaryotic cells. As more studies revealed the encapsulation of enzymes into

MCPs and the molecular transport through the shell to meet the needs of the enzymatic functions, a picture emerges that fully functional MCPs are highly organized through the network of interactions between proteins associated with MCPs. Recent studies showed that some interior enzymes in Pdu MCPs contain special short N-terminal extensions which are necessary and sufficient for targeting into the Pdu MCPs (101, 261, 262). Alignment of targeting sequences

114

from various enzymes showed little sequence similarity. But the targeting sequence generally have propensity of forming an amphipathic α-helix. Further studies indicated that these α-helices might interact and bind to certain BMC domain proteins on MCP shells. Subsequent studies by

Fan et al. (263) carried out in vitro binding studies, in vivo targeting studies together with structural model analysis which provided evidence that targeting sequence on N-terminal region of enzyme PduP binds to C-terminus of shell proteins PduA and PduJ to mediates the encapsulation of PduP into the lumen of MCPs.

Several studies have successfully utilized short N-terminal targeting sequences to direct green fluorescence proteins and β-galactosidase into MCPs. Another group reported the solution structure of the N-terminal 18-amino-acid peptide of PduP from Citrobacter freundii and confirmed that the peptide has a helical conformation (102). They further identified the binding interaction between the P18 peptide with another shell protein PduK. These results indicated that the processes controlling the interactions among MCP-associated proteins may vary between different MCP types across bacterial species. In the case of Eut MCPs, an N-terminal targeting sequence on EutC is crucial for this enzyme to be packaged into Eut MCP (101). Also EutC preferentially bind shell protein EutS, a shell protein with the capability of forming a properly delimited MCP on its own (101). These results indicated the targeting of EutC into Eut MCPs may be mediated by the interaction between EutC and EutS (101).

Besides the targeting related interactions between encapsulated enzymes and MCP shell proteins, protein-protein interactions associated with MCPs might have broader significance.

CcmN is an essential carboxysomal component. According to Kinney et al. (267), its N-terminal domain binds the encapsulated CcmM which interacts with RuBisCO and CA. Its C-terminal

115

peptide, which binds major shell component CcmK2, is crucial for carboxysome formation, suggesting that CcmM may function as a scaffolding protein during MCP assembly.

Packaging of metabolic enzymes into MCPs via interactions between shell components and targeting sequences on lumen enzymes may represent a general mechanism across the MCP systems. But to this date, details about atomic interactions between shell protein and lumen enzymes were all based on mutagenesis studies and computer modeling. By far only one study extensively investigated and determined the mechanism by which the MCP targeting sequence functions (263). Comprehensive investigative system is needed to fill in this knowledge gap and provide directions for the potential applications of MCPs in biotechnology.

In this chapter, we performed in vivo two-hybrid tests together with in vitro nickel affinity pull-down experiments and identified more possible interacting proteins PduL and PduBB’ as well as PduU and PduV. Our experimental data was consistent with MCP interactome model predicted by Jorda et al. (unpublished data) using a coevolution-based machine learning algorithm. This work may help direct future work on protein-protein interactions in MCPs.

4.3. Material and Methods

Bacterial strains and growth conditions. Bacterial strains used in this chapter are derivatives of E. coli and are listed in Table 4.1. The minimal medium used for two-hybrid assays were M9 media with Histidine-dropout supplement (BD/clontech), 5mM 3-amino-1,2,4- triazole (3-AT) and streptomycin was also added into the selective minimal medium.

Two-hybrid assay. The BacterioMatch II Two-Hybrid System (Agilent technologies) was used according to the manufacturer's instructions with the following modification: co- transformation was carried out by using 30 ng each of the bait and target vector. To construct the needed plasmids, each pdu protein encoding DNA sequences were amplified by PCR and then

116

restricted and ligated into pBT for expression as fusions with the λcI protein, and into pTRG for expression as fusions with the RNAPα protein. Ligation reactions were used to transform E. coli

XL1-Blue MRF’. Plasmid DNA was purified using a Qiagen mini prep kit, and all clones were verified by DNA sequencing. The bacterial strains used for preparation of recombinant plasmids for two-are listed in Table 4.1. Self-activation by each recombinant bait and target was tested before the two-hybrid interaction assays to determine if the bait or target was capable of activating the reporter cassette on its own. Determination of protein-protein interaction was carried out by co-transforming BacterioMatch II validation reporter competent cells using recombinant bait and target.

Table 4.1. Bacterial strains used in this study.

Strain Genotype Source E. coli XL1 Blue MRF’ BE768 pTRG-pduU This study BE797 pBT-LGF2 T.A. Bobik laboratory collection BE798 pTRG-Gal11P T.A. Bobik laboratory collection BE984 pBT-pduU This study BE991 pBT-pduL This study BE998 pBT-pduV This study BE1004 pTRG-pduB’ This study BE1013 pTRG-pduV This study BE1017 pBT-pduB’ This study BE1034 pTRG-pduL This study E. coli BL21(DE3) Rosetta 2 BE1795 pCDF-pduLHis6-pduB’ This study BE1804 pCDF-pduLHis6-pduBB’ This study BE1816 pCDF-pduL-His6pduBB’ This study BE1827 pCDF-pduL-His6pduB’ This study BE1858 pCDF-pduLHis6-no insert at MCS2 This study BE2000 pCDF-no insert at MCS1- This study pduBB’102/207AHis6 BE2001 pCDF-no insert at MCS1- This study His6pduBB’102/207A BE2002 pCDF-pduL-no insert at MCS2 This study BE2003 pCDF-pduL-pduBB’102/207AHis6 This study BE2004 pCDF-pduL-His6pduBB’102/207A This study BE2024 pCDF-no insert at MCS1- This study His6pduBB’

117

Protein purification and nickel-affinity pull-down assay. pCDFDuet-1 (Novagen) was used for cloning and the constructs were transformed into E. coli BL21 (DE3) Rosetta 2 cells

(Stratagene). The transformants were grown in Luria–Bertani medium. When the OD600 reached early log phase, protein production was induced with 0.5 mM isopropyl-β-D-1- thiogalactopyranoside (IPTG) at 16 °C for 16 h. The cells were collected and suspended in 50 mM 2-(Cyclohexylamino) ethanesulfonic acid (CHES) (pH 7.4), 300 mM NaCl, 10 mM imidazole and broken with a French press (Thermo Electron) operated at 20,000 psi. The cell lysate was clarified at 35,000 × g for 20 min at 4 °C. The supernatant was loaded onto a column containing 4 mL Ni-NTA Superflow resin (Qiagen) pre-equilibrated with 50 mM CHES (pH

7.5), 300 mM NaCl, 10 mM imidazole. The column was washed with 80 mM imidazole, and then eluted with 1 mL 50 mM CHES (pH 8.5), 300 mM NaCl, and 300 mM imidazole. All elution fractions were analyzed by SDS-PAGE to test for coelution of the His-tagged and untagged protein.

4.4. Results

BacterioMatch II two-hybrid screening for binding partners for PduL. To screen for the possible binding partners for PduL, we carried out a BacterioMatch II two-hybrid test using

E.coli as the host strain. Using this system, we identified several proteins including PduB’, PduD,

PduN and PduW as candidates for further investigating their possible interactions with PduL. In this system, a reporter strain is co-transformed with appropriate target and bait fusion proteins. A protein-protein interaction between the target and bait activates the HIS3 gene allowing growth in the presence of 3-amino-1,2,4-triazole (3-AT), a competitive inhibitor of the His3 enzyme. If a large number of colonies is obtained following co-transformation, an interaction between the target and bait proteins is indicated. We fused pduL into the bait vector pBT and introduced each

118

of the shell protein PduABB’JKNTU into the target plasmid pTRG. The reciprocal clones were also constructed. For each protein pair, the number of colonies obtained following co- transformation was comparable to that of a positive control with bait and prey proteins (LGF2 and Gal11P) that are known to strongly interact (Table 4.2). The cfu number on selective medium was significantly higher compared to the respective self-test (negative control) but lower than the positive control, suggesting medium or weak interactions. Although other possibilities can’t be ruled out, for example, the interacting protein pair needs other proteins or molecules for optimal binding. We found pBT-pduL construct produced much less colonies during self-test and the interacting tests, indicating more severe cellular toxicity caused to the host cells.

Table 4.2. Two hybrid assay to test the interaction with PduL

Pairs tested NSa Sb pBT-LGF2, pTRG-Gal11P 1000 ± 87 1000 ± 72 pBT, pTRG w/ pduL TNTCc 9 ± 5 pduB’ TNTC 35 ± 5 pduB 39 ± 9 1 ± 0 pTRG, pBT w/ pduL 75 ± 21 6 ± 1 pduB 19 ± 10 0 ± 0 pduB’ 895 ± 65 7 ± 3 pduW TNTC 14 ± 4 pduN 1000 ± 50 45 ± 6 pBT-pduB’, pTRG-pduL 960 ± 40 490 ± 55 pBT-pduW, pTRG-pduL TNTC 123 ± 49 pBT-pduD, pTRG-pduL TNTC 450 ± 26 pBT-pduN, pTRG-pduL TNTC 240 ± 61 pBT-pduL, pTRG-pduB’ 1 ± 0 0 ± 0 pBT-pduL, pTRG-pduB 14 ± 13 0.3 ± 0.6 Different combinations of bait (pBT) and prey (pTRG) genotypes are tested. The LGF2 and Gal11p pair serves as a positive control. Tests with each pair were repeated three times and the numbers shown are the mean ± 1 standard deviation. acfu obtained on nonselective (NS) screening medium plates (no 3-AT) with 1:100 diluted cells. bcfu obtained on selective (S) screening medium (3-AT) plates. ccfu too numerous to count.

119

A B kDa 1 2 3 4 5 250 KDa 1 2 3 4 5 6 250 150 150 100 100 75 75 50 50 37 37 25 PduB PduB’ 20 PduB102/207A 25 15 20 PduL PduB’ PduL 15

Construct hisBB’ L-hisBB’ Construct hisBB’ L-hisBB’ BB’his L-BB’his

Fraction elution Fraction whole-cell lysate soluble elution

Figure 4.1. SDS-PAGE analysis of nickel-affinity pull-downs showed binding interactions between PduL and

PduBB’. Panel A (using pduBB’ double mutants with higher solubility). Lane 1, -His6pduBB’102/207A

(BE2001); lane 2, pduL-His6pduBB’102/207A (BE2004); lane 3, molecular markers; lane 4, - pduBB’102/207AHis6 (BE2000); lane 5, pduBB’102/207AHis6 (BE2003). Panel B (using w.t. pduBB’). Lane 1, molecular markers; lane 2 and 3, whole-cell extract and soluble fraction from strain BE2024 (His6pduBB’); lane 4 and 5, whole-cell extract, soluble fraction and eluted fraction from strain BE1816 (pduL-His6pduBB’).

PduL interacts with PduBB’. To further identify the protein-protein interactions suggested by the two-hybrid system, we performed nickel affinity copurification followed by Western blots.

The results indicated a binding interaction between PduL and PduB or/and PduB’. Due to the presence of a restriction cutting site in pduL sequence, we chose to introduce pduL into MCS1 of

Duet vector. The test of L-BB’ resulted in a weak band with the size of PduL on the SDS-PAGE.

The control strain with only BB’ cloned into MCS2 site on Duet didn’t produce enough protein in the eluted samples for a conclusive result. We next tested the pair of L and edge mutant of

PduBB’ which allows much better solubility of the shell protein. We got the positive results by using BB’ mutant with six histidine residues tagged to both N-terminus and C-terminus (Fig. 4.1).

Next, we did Western blots to test the presence of PduL and PduBB’ in the elution fraction using

120

both anti-PduL and anti-PduB antibodies. The results indicated PduL coeluted with PduBB’ (Fig.

4.2).

1 2 3 4 Anti-PduL

Anti-PduBB’

Sample L-hisBB’ -hisBB’ L-BB’his -BB’his

Figure 4.2. Western blots supported that PduL coeluted with PduBB’. Lane 1-4: elution fraction of strains expressing pduL-His6pduBB’102/207A (BE2004), -His6pduBB’102/207A (BE2001), pduL- pduBB’102/207AHis6 (BE2003), and -pduBB’102/207AHis6 (BE2000), respectively.

Table 4.3. Two hybrid assay to test the interaction between PduU and PduV.

Pairs tested NSa Sb pBT-LGF2, pTRG-Gal11P 1000 ± 87 1000 ± 72 pBT-pduU, pTRG 11 ± 3 0 ± 0 pBT-pduV, pTRG 19 ± 10 0 ± 0 pBT, pTRG-pduU 167 ± 21 0 ± 0 pBT, pTRG-pduV 740 ± 75 7 ± 3 pBT-pduU, pTRG-pduV 1000 ± 66 600 ± 12 pBT-pduV, pTRG-pduU 840 ± 79 TNTCc acfu obtained on nonselective (NS) screening medium plates (no 3-AT) with 1:100 diluted cells. bcfu obtained on selective (S) screening medium (3-AT) plates. ccfu too numerous to count.

121

A binding interaction between PduU and PduV. The BacterioMatch II Two-Hybrid system was used to test for interactions between the PduU and PduV proteins. Results showed that PduU and PduV interacted in reciprocal tests, that is, when used as bait and prey, or prey and bait (Table 4.3). Negative controls showed the PduU or PduV alone did interact spuriously to confer 3-AT resistance (Table 4.3). The positive results of the UV pairs were confirmed by testing the expression of a second reporter gene aadA which confers streptomycin resistance in co-transformed E. coli (data not shown). Thus, overall, these results support an interaction between PduU and PduV.

4.5. Discussion

The nickel affinity pull-down study of PduL and PduBB’ provided valuable information on their possible interactions. A small histidine tag on N- and C- terminal region both produced positive results, which might indicate that a small six histidine tag doesn’t interfere with binding interaction, or PduL doesn’t bind to termini on PduBB’, which is consistent with the recent docking simulation of interaction between L and BB’ which predicted PduL might bind to the middle part on PduBB’.

We observed that some bacterial two-hybrid data were inconsistent with the co-pulldown results. According to the pull-downs, PduL co-eluted with PduBB’, while two-hybrid only suggested L/B’ but not L/B binding. One possible explanation is that certain reconstructed bait and/or target plasmids used in the two-hybrid system became toxic when growing in the host, resulting in low number of colonies even on nonselective medium plates and difficulty interpreting the data.

122

This study promotes our understanding of higher level organization in MCP systems.

Computer based prediction followed by experimental confirmation may provide more efficient model system for further MCP-associated protein-protein interaction studies.

4.6. Acknowledgements

This work was supported by NSF grant MCB-0956451 to TAB. We thank Julien Jorda from

University of California, Los Angles for providing the docking prediction data.

123

CHAPTER 5. SUMMARY AND CONCLUSIONS

This dissertation focuses on biochemical and genetic analysis of the distinct phosphotransacylase PduL, its coenzyme A recycling functions specifically associated with Pdu

MCPs as well as protein-protein interactions in Pdu MCPs.

Chapter 2 is about the identification of a new class phosphotransacylase PduL involved in

1,2-PD metabolism associated with Pdu MCPs. preliminary data indicated the presence of a phosphotransacylase in this process, but the exact enzyme involved was not initially found. Here, we showed that the house-keeping PTAC pta is not involved in 1,2-PD metabolism. Next we conducted a series of biochemical and genetic studies and established the role of PduL as a phosphotransacylase in 1,2-PD utilization in Salmonella. Further bioinformatics analysis showed that PduL lacks sequence similarity with currently known phosphotransacylases, indicating that

PduL is an evolutionarily distinct PTAC found in bacteria. Also PduL is a much smaller enzyme compared to other classes of PTACs in Salmonella. Further studies on the origin of PduL may disclose unique events during evolution.

Chapter 3 aims to determine the Pdu MCP-specific role of PduL and interpret how PduL is targeted into MCPs. Prior studies showed that another PTAC enzyme EutD is recycling HS-CoA internally within the Eut MCP, and an alcohol dehydrogenase PduQ regenerates cofactor NAD+ inside the Pdu MCP. PduL, in this work, was demonstrated to be an MCP component, might also function in cofactor recycling. Therefore, we performed a series of genetic and growth studies which showed that PduL is required for the proper function of intact MCPs, and furthermore,

PduL in MCPs can’t utilize acetyl-CoA in the cytoplasm to support Salmonella growth. Together, coenzyme A internal recycling by PduL within the Pdu MCP is indicated. Sequence alignment and structure prediction indicated PduL has a targeting sequence, protein immunoblots provided

124

evidence that N-terminal region of PduL functions as a targeting sequence for MCP encapsulation.

Chapter 4 explores the protein-protein interactions in the Pdu MCPs. Here we used the bacterial two-hybrid screen tests and nickel-affinity pull-down experiments followed by immunoblots to show that PduL may have binding interaction with main shell protein PduBB’, which might mediate the encapsulation of PduL into the Pdu MCP. Our results also indicated that PduV, a protein with unknown functions, interacts with shell protein PduU. Protein-protein interactions among the MCP-associated proteins are crucial for assembly and spatial organization of MCPs.

Through the above series of studies in this dissertation, we drew the following conclusions.

1. The PduL enzyme was established to function as a phosphotransacylase in 1,2-PD

utilization. PduL represents a new class of phosphotransacylase.

2. The primary role of PduL specifically associated with MCPs is to recycle HS-CoA within

the Pdu MCP, which provides a way for MCPs to acquire and replenish cofactors. This

enhances our knowledge level of the working mechanisms of MCPs.

3. A short N-terminal targeting sequence of PduL is responsible for its encapsulation into

the Pdu MCP. Targeting sequence-mediated interactions between lumen enzymes and

shell proteins direct protein encapsulation into MCPs. These mechanisms advance our

understanding of the self-assembly and the higher-level organization of bacterial MCPs.

4. A complete understanding of the mechanisms of the functions of MCPs may allow their

potential applications as bioreactors for industrial chemical production, biochemical

pathway optimization, and as drug delivery systems for therapeutic applications.

125

REFERENCES

1. Zhang S, Santos RL, Tsolis RM, Stender S, Hardt WD, Bäumler AJ, Adams LG. 2002. The Salmonella enterica serotype Typhimurium effector proteins SipA, SopA, SopB, SopD, and SopE2 act in concert to induce diarrhea in calves. Infect Immun. 70: 3843–55.

2. Hansen-Wester I, Hense lM. 2001. Salmonella pathogenicity islands encoding typeIII secretion systems. Microbes Infect. 3: 549–59.

3. Sztein, M.B. 2007. Cell-mediated immunity and antibody responses elicited by attenuated Salmonella enterica serovar Typhi strains used as live oral vaccines in humans. Clin. Infect. Dis. 45(Suppl.1): S15–S19.

4. McClelland, M., Sanderson, K.E., Clifton, S.W., Latreille, P., Porwollik, S., Sabo, A., Meyer R, Bieri T, Ozersky P, McLellan M, Harkins CR, Wang C, Nguyen C, Berghoff A, Elliott G, Kohlberg S, Strong C, Du F, Carter J, Kremizki C, Layman D, Leonard S, Sun H, Fulton L, Nash W, Miner T, Minx P, Delehaunty K, Fronick C, Magrini V, Nhan M, Warren W, Florea L, Spieth J, Wilson RK. 2004. Comparison of genome degradation in Paratyphi A and Typhi, human-restricted serovars of Salmonella enterica that causetyphoid. Nat. Genet. 36: 1268–74.

5. Robbins, J.D., and Robbins, J.B. 1984. Reexamination of the protective role of the capsular polysaccharide (Vi antigen) of Salmonella Typhi. J. Infect.Dis. 150: 436–49.

6. Raffatellu, M., Chessa, D., Wilson, R.P., Dusold, R., Rubino, S., and Baumler, A. J. 2005. The Vi capsular antigen of Salmonella enterica serotypeTyphireduces Toll-likereceptor-dependent interleukin-8 expression in the intestinal mucosa. Infect. Immun. 73: 3367–74.

7. Cheminay C, Chakravortty D, Hense lM. 2004. Role of neutrophils in murine salmonellosis. InfectImmun. 72: 468–77.

8. Nunes JS, Lawhon SD, Rossetti CA, Khare S, Figueiredo JF, Gull T, Burghardt RC, Bäumler AJ, Tsolis RM, Andrews-Polymenis HL, Adams LG. 2010. Morphologic and cytokine profile characterization of Salmonella enterica serovar Typhimurium infection in calves with bovine leukocyte adhesion deficiency. Vet Pathol. 47: 322–33.

9. Sprinz H, Gangarosa EJ, Williams M, Hornick RB, Woodward TE. 1966. Histopathology of the upper small intestines in typhoid fever. Biopsy study of experimental disease in man. Am J Dig Dis. 11: 615–24.

10. Santos RL, Bevins CL, Adams LG, Raffatellu M, Tükel Ç, Tsolis RM, Bäumler AJ. 2009. Life in the inflamed intestine, Salmonella style. Trends Microbiol. 17: 498–506.

11. DiehlGE, Longman RS, Zhang JX, BreartB, Galan C, Cuesta A, Schwab SR, Littman DR. 2013. Microbiota restricts trafficking of bacteria to mesenteric lymph nodes by CX(3)CR1(hi) cells. Nature. 494: 116–20.

12. Ng KM, Ferreyra JA, Higginbottom SK, Lynch JB, Kashyap PC, Gopinath S, Naidu N, Choudhury B, Weimer BC, Monack DM, Sonnenburg JL. 2013. Microbiota-liberated host sugars facilitate post- antibiotic expansion of enteric pathogens. Nature. 502: 96–9.

13. Raffatellu M, George MD, Akiyama Y, Hornsby MJ, Nuccio SP, Paixão TA, Butler BP, Chu H, Santos RL, Berger T, Mak TW, Tsolis RM, Bevins CL, Solnick JV, Dandekar S, Bäumler AJ. 2009. Lipocalin-2 resistance confers an advantage to Salmonella enterica serotype Typhimurium for growth and survival in the inflamed intestine. Cell Host Microbe. 5: 476–86.

14. Nedialkova LP, Denzler R, Koeppel MB, Diehl M, Ring D, Wille T, Gerlach RG, Stecher B. 2014. Inflammation fuels colicin Ib-dependent competition of Salmonella serovar Typhimurium and E. coli in enterobacterial blooms. PloS Pathog. 10: e1003844.

126

15. Bouvard V, Baan R, Straif K, Grosse Y, Secretan B, El Ghissassi F, Benbrahim-Tallaa L, Guha N, Freeman C, Galichet L, Cogliano V. 2009. A review of human carcinogens--Part B: biological agents. WHO International Agency for Research on Cancer Monograph Working Group. Lancet Oncol. 10: 321-22.

16. Nagaraja V, Eslick GD. 2014. Systematic review with meta-analysis: the relationship between chronic Salmonella typhi carrier status and gall-bladder cancer. Aliment Pharmacol Ther. 39: 745-50.

17. Caygill CP, Braddick M, Hill MJ, Knowles RL, Sharp JC. 1995. The association between typhoid carriage, typhoid infection and subsequent cancer at a number of sites. Eur J Cancer Prev. 4: 187-93.

18. Catanzaro R, Occhipinti S, Calabrese F, Anzalone MG, Milazzo M, Italia A, Marotta F. 2014. Irritable bowel syndrome: new findings in pathophysiological and therapeutic field. Minerva Gastroenterol Dietol. 60: 151-63.

19. Krueger, A.L., Greene, S.A., Barzilay, E.J., Henao, O., Vugia, D., Hanna, S., Meyer S, Smith K, Pecic G, Hoefer D, Griffin PM. 2014. Clinical outcomes of nalidixic acid, ceftriaxone, and multi drug-resistant nontyphoidal Salmonella infections compared with pansusceptible infections in FoodNet sites, 2006–2008. Foodborne Pathog. Dis. 11: 335–41.

20. Gantt E, Conti SF. 1969. Ultrastructure of blue-green algae. J Bacteriol. 97: 1486-93.

21. Shively JM, Decker GL, Greenawalt JW. 1970. Comparative ultrastructure of the thiobacilli. J Bacteriol. 101: 618-27.

22. Bobik TA, Havemann GD, Busch RJ, Williams DS, Aldrich HC. 1999. The propanediol utilization (pdu) operon of Salmonella enterica serovar Typhimurium LT2 includes genes necessary for formation of polyhedral organelles involved in coenzyme B12-dependent 1, 2-propanediol degradation. J Bacteriol. 181: 5967-75.

23. Bobik TA. 2006. Polyhedral organelles compartmenting bacterial metabolic processes. Appl Microbiol Biotechnol. 70: 517-25.

24. Kerfeld CA, Heinhorst S, Cannon GC. 2010. Bacterial microcompartments. Annu Rev Microbiol. 64: 391-408.

25. Abdul-Rahman F, Petit E, Blanchard JL. 2013. The distribution of polyhedral bacterial microcompartments suggests frequent horizontal transfer and operon reassembly. J. Phylogenet. Evol. Biol. 1: 4.

26. Shively JM, Bryant DA, Fuller RC, Konopka AE, Stevens SE Jr, Strohl WR. 1988. Functional inclusions in prokaryotic cells. Int. Rev. Cytol. 113: 35–100.

27. Codd, G. A. & Marsden, W. J. N. 1984. The carboxysomes (polyhedral bodies) of autotrophic prokaryotes. Biol. Rev. 59: 389–422.

28. Heinhorst, S., Cannon, G. C. & Shively, J. M. 2006. In Complex Intracellular Structures in Prokaryotes (ed. Shively, J. M.) Springer-Verlag, Berlin. 141–65.

29. MR, Hanson D, Price GD. 2002. Evolution and diversity of CO2 concentrating mechanisms in cyanobacteria. Funct Plant Biol. 29: 161-73.

30. Shively JM, Ball F, Brown DH, Saunders RE. 1973. Functional organelles in prokaryotes: polyhedral inclusions (carboxysomes) of Thiobacillus neapolitanus. Science. 182: 584–6.

31. So AK, Espie GS, Williams EB, Shively JM, Heinhorst S, Cannon GC. 2004. A novel evolutionary lineage of carbonic anhydrase (epsilon class) is a component of the carboxysome shell. J Bacteriol. 186: 623–30.

127

32. Beeby M, Bobik TA, Yeates TO. 2009. Exploiting genomic patterns to discover new supramolecular protein assemblies. Protein Sci. 18: 69-79.

33. Chen P, Andersson D, Roth JR. 1994. The control region of the pdu/cob regulon in Salmonella typhimurium. J Bacteriol. 176: 5474–82.

34. J M Shively, C E Bradburne, H C Aldrich, T A Bobik, J L Mehlman, S Jin, S H Baker. 1998. Sequence homologs of the carboxysomal polypeptide CsoS1 of the thiobacilli are present in cyanobacteria and enteric bacteria that form carboxysomes - polyhedral bodies. Canadian Journal of Botany. 76: 906-16.

35. Kerfeld CA, Sawaya MR, Tanaka S, Nguyen CV, Phillips M, Beeby M, Yeates TO. 2005. Protein structures forming the shell of primitive bacterial organelles. Science. 309: 936-8.

36. Crowley CS, Cascio D, Sawaya MR, Kopstein JS, Bobik TA, Yeates TO. 2010. Structural insight into the mechanisms of transport across the Salmonella enterica Pdu microcompartment shell. J Biol Chem. 285: 37838-46.

37. Cheng S, Liu Y, Crowley CS, Yeates TO, Bobik TA. 2008. Bacterial microcompartments: their properties and paradoxes. Bioessays. 30: 1084-95.

38. Tanaka S, Kerfeld CA, Sawaya MR, Cai F, Heinhorst S, Cannon GC, Yeates TO. 2008. Atomic-level models of the bacterial carboxysome shell. Science. 319: 1083-6.

39. Sinha S, Cheng S, Fan C, Bobik TA. 2012. The PduM protein is a structural component of the microcompartments involved in coenzyme B12-dependent 1,2-propanediol degradation by Salmonella enterica. J Bacteriol. 194: 1912–8.

40. Crowley CS, Sawaya MR, Bobik TA, Yeates TO. 2008. Structure of the PduU shell protein from the Pdu microcompartment of Salmonella. Structure. 16: 1324-32.

41. Kofoid, E., C. Rappleye, I. Stojiljkovic, and J. Roth. 1999. The 17-gene ethanolamine (eut) operon of Salmonella typhimurium encodes five homologues of carboxysome shell proteins. J. Bacteriol. 181: 5317– 29.

42. Blackwell CM, Scarlett FA, Turner JM. 1976. Ethanolamine catabolism by bacteria, including Escherichia coli. Biochem Soc Trans. 4: 495–7.

43. Del Papa MF, Perego M. 2008. Ethanolamine activates a sensor histidine kinase regulating its utilization in Enterococcus faecalis. J Bacteriol. 190: 7147-56.

44. Jorda J, Lopez D, Wheatley NM, Yeates TO. 2013. Using comparative genomics to uncover new kinds of protein-based metabolic organelles in bacteria. Protein Sci. 22: 179-95.

45. Scott KP, Martin JC, Campbell G, Mayer CD, Flint HJ. 2006. Whole genome transcription profiling reveals genes up-regulated by growth on fucose in the human gut bacterium “Roseburia inulinivorans.” J. Bacteriol. 188: 4340–9.

46. Petit E, LaTouf WG, Coppi MV, Warnick TA, Currie D, Romashko I, Deshpande S, Haas K, Alvelo- Maurosa JG, Wardman C, Schnell DJ, Leschine SB, Blanchard JL. 2013. Involvement of a bacterial microcompartment in the metabolism of fucose and rhamnose by Clostridium phytofermentans. PLoS One. 8: e54337.

47. Yeates TO, Kerfeld CA, Heinhorst S, Cannon GC, Shively JM. 2008. Protein-based organelles in bacteria: carboxysomes and related microcompartments. Nat Rev Microbiol. 6: 681-91.

48. English RS, Lorbach SC, Qin X, Shively JM. 1994. Isolation and characterization of a carboxysome shell gene from Thiobacillus neapolitanus. Mol Microbiol. 12: 647-54.

128

49. Havemann GD, Bobik TA. 2003. Protein content of polyhedral organelles involved in coenzyme B12- dependent degradation of 1,2-propanediol in Salmonella enterica serovar Typhimurium LT2. J Bacteriol. 185: 5086-95.

50. Liu Y, Leal NA, Sampson EM, Johnson CL, Havemann GD, Bobik TA. 2007. PduL is an evolutionarily distinct phosphotransacylase involved in B12-dependent 1,2-propanediol degradation by Salmonella enterica serovar typhimurium LT2. J Bacteriol. 189: 1589-96.

51. Bobik TA, Xu Y, Jeter RM, Otto KE, Roth JR. 1997. Propanediol utilization genes (pdu) of Salmonella typhimurium: three genes for the propanediol dehydratase. J Bacteriol. 179: 6633-9.

52. Cheng S, Fan C, Sinha S, Bobik TA. 2012. The PduQ enzyme is an alcohol dehydrogenase used to recycle NAD+ internally within the Pdu microcompartment of Salmonella enterica. PLoS One. 7: e47144.

53. Leal NA, Havemann GD, Bobik TA. 2003. PduP is a coenzyme A-acylating propionaldehyde dehydrogenase associated with the polyhedral bodies involved in B12-dependent 1,2-propanediol degradation by Salmonella enterica serovar Typhimurium LT2. Arch Microbiol. 180: 353–61.

54. Bobik TA, Ailion M, Roth JR. 1992. A single regulatory gene integrates control of vitamin B12 synthesis and propanediol degradation. J. Bacteriol. 174: 2253–66.

55. Cheng S, Bobik TA. 2010. Characterization of the PduS cobalamin reductase of Salmonella enterica and its role in the Pdu microcompartment. J Bacteriol. 192: 5071–80.

56. Johnson CL, Buszko ML, Bobik TA. 2004. Purification and initial characterization of the Salmonella enterica PduO ATP:Cob(I)alamin adenosyltransferase. J Bacteriol. 186: 7881–7.

57. Fan C, Bobik TA. 2008. The PduX enzyme of Salmonella enterica is an L-threonine kinase used for coenzyme B12 synthesis. J Biol Chem. 283: 11322-9.

58. Chowdhury C, Sinha S, Chun S, Yeates TO, Bobik TA. 2014. Diverse Bacterial Microcompartment Organelles. Microbiol Mol Biol Rev. 78: 438-68.

59. Cheng S, Sinha S, Fan C, Liu Y, Bobik TA. 2011. Genetic analysis of the protein shell of the microcompartments involved in coenzyme B12-dependent 1,2-propanediol degradation by Salmonella. J Bacteriol. 193: 1385–92.

60. Roof, D. M., and J. R. Roth. 1989. Functions required for vitamin B12-dependent ethanolamine utilization in Salmonella typhimurium. J. Bacteriol. 171: 3316–23.

61. Stojiljkovic I, Baeumler AJ, Heffron F. 1995. Ethanolamine utilization in Salmonella typhimurium: nucleotide sequence, protein expression, and mutational analysis of the cchA cchB eutE eutJ eutG eutH gene cluster. J. Bacteriol. 177: 1357–66.

62. Mori K, Bando R, Hieda N, Toraya T. 2004. Identification of a reactivating factor for adenosylcobalamin-dependent ethanolamine ammonia lyase. J. Bacteriol. 186: 6845–54.

63. Buan NR, Escalante-Semerena JC. 2006. Purification and initial biochemical characterization of ATP:cob(I)alamin adenosyltransferase (EutT) enzyme of Salmonella enterica. J. Biol. Chem. 281: 16971–7.

64. Penrod JT, Mace CC, Roth JR. 2004. A pH-sensitive function and phenotype: evidence that EutH facilitates diffusion of uncharged ethanolarnine in Salmonella enterica. J. Bacteriol. 186: 6885–90.

65. Yeates TO, Jorda J, Bobik TA. 2013. The shells of BMC-type microcompartment organelles in bacteria. J. Mol Microbiol Biotechnol. 23: 290-9.

129

66. Tanaka S, Sawaya MR, Yeates TO. 2010. Structure and mechanisms of a protein-based organelle in Escherichia coli. Science. 327: 81-4.

67. Cannon GC, Bradburne CE, Aldrich HC, Baker SH, Heinhorst S, Shively JM. 2001. Microcompartments in prokaryotes: carboxysomes and related polyhedra. Appl Environ Microbiol. 67: 5351-61.

68. Field, C. B., Behrenfeld, M. J., Randerson, J. T. & Falkowski, P. 1998. Primary production of the biosphere: integrating terrestrial and oceanic components. Science. 281: 237–40.

69. Rae BD, Long BM, Badger MR, Price GD. 2013. Functions, compositions, and evolution of the two types of carboxysomes: polyhedral microcompartments that facilitate CO2 fixation in cyanobacteria and some proteobacteria. Microbiol. Mol. Biol. Rev. 77: 357–79.

70. Havemann GD, Sampson EM, Bobik TA. 2002. PduA is a shell protein of polyhedral organelles involved in coenzyme B12-dependent degradation of 1,2-propanediol in Salmonella enterica serovar typhimurium LT2. J Bacteriol. 184: 1253-61.

71. Pang A, Frank S, Brown I, Warren MJ, Pickersgill RW. 2014. Structural insights into higher order assembly and function of the bacterial microcompartment protein PduA. J Biol Chem. 289: 22377-84.

72. Pang A, Liang M, Prentice MB, Pickersgill RW. 2012. Substrate channels revealed in the trimeric Lactobacillus reuteri bacterial microcompartment shell protein PduB. Acta Crystallogr D Biol Crystallogr. 68: 1642-52.

73. Sinha S, Cheng S, Sung YW, McNamara DE, Sawaya MR, Yeates TO, Bobik TA. 2014. Alanine scanning mutagenesis identifies an asparagine-arginine-lysine triad essential to assembly of the shell of the Pdu microcompartment. J Mol Biol. 426: 2328-45.

+ 74. Cheng, SQ. 2010. Recycling of vitamin B12 and NAD within the Pdu microcompartment. PhD dissertation, Iowa State University, Ames: ProQuest/UMI. (Publication No. AAT 3438684.)

75. Price-Carter M, Tingey J, Bobik TA, Roth JR. 2001. The alternative electron acceptor tetrathionate supports B12-dependent anaerobic growth of Salmonella enterica serovar typhimurium on ethanolamine or 1,2-propanediol. J Bacteriol. 183: 2463-75.

76. Horswill, A., and J. Escalante-Semerena. 1997. Propionate catabolism in Salmonella typhimurium LT2: two divergently transcribed units comprise the prp locus at 8.5 centisomes, prpR encodes a member of the sigma-54 family of activators, and the prpBCDE genes constitute an operon. J. Bacteriol. 179: 928–40.

77. Roth JR, Lawrence JG, Bobik TA. 1996. Cobalamin (coenzyme B12): synthesis and biological significance. Annu. Rev. Microbiol. 50: 137–81.

78. Huseby DL, Roth JR. 2013. Evidence that a metabolic microcompartment contains and recycles private cofactor pools. J Bacteriol. 195: 2864-79.

79. Marin B, Nowack EC, Glöckner G, Melkonian M. 2007. The ancestor of the Paulinella chromatophore obtained a carboxysomal operon by horizontal gene transfer from a Nitrococcus-like gamma- proteobacterium. BMC Evol Biol. 7: 85.

80. Tsai Y, Sawaya MR, Cannon GC, Cai F, Williams EB, Heinhorst S, Kerfeld CA, Yeates TO. 2007. Structural analysis of CsoS1A and the protein shell of the Halothiobacillus neapolitanus carboxysome. PLoS Biol. 5: e144.

81. M. F. Schmid., Paredes AM, Khant HA, Soyer F, Aldrich HC, Chiu W, Shively JM. 2006. Structure of Halothiobacillus neapolitanus carboxysomes by cryo-electron tomography. J. Mol. Biol. 364: 526-35.

130

82. C. V. Iancu., Ding HJ, Morris DM, Dias DP, Gonzales AD, Martino A, Jensen GJ. 2007. The structure of isolated Synechococcus strain WH8102 carboxysomes as revealed by electron cryotomography. J. Mol. Biol. 372: 764-73.

83. J. E. Johnson, J. A. Speir. 1997. Quasi-equivalent viruses: a paradigm for protein assemblies. J. Mol. Biol. 269: 665-75.

84. D. L. Caspar, A. Klug. 1962. Physical principles in the construction of regular viruses. Cold Spring Harbor Symp. Quant. Biol. 27: 1-24.

85. Wheatley NM, Gidaniyan SD, Liu Y, Cascio D, Yeates TO. 2013. Bacterial microcompartment shells of diverse functional types possess pentameric vertex proteins. Protein Sci. 22: 660–5.

86. Forouhar F, Kuzin A, Seetharaman J, Lee I, Zhou W, Abashidze M, Chen Y, Yong W, Janjua H, Fang Y, Wang D, Cunningham K, Xiao R, Acton TB, Pichersky E, Klessig DF, Porter CW, Montelione GT, Tong L. 2007. Functional insights from structural genomics. J Struct Funct Genomics. 8: 37-44.

87. Klein MG, Zwart P, Bagby SC, Cai F, Chisholm SW, Heinhorst S, Cannon GC, Kerfeld CA. 2009. Identification and structural analysis of a novel carboxysome shell protein with implications for metabolite transport. J. Mol. Biol. 392: 319–33.

88. Sagermann M, Ohtaki A, Nikolakakis K. 2009. Crystal structure of the EutL shell protein of the ethanolamine ammonia lyase microcompartment. Proc. Natl. Acad. Sci. USA. 106: 8883–7.

89. Kim EY, Slininger MF, Tullman-Ercek D. 2014. The effects of time, temperature, and pH on the stability of PDU bacterial microcompartments. Protein Sci. 23: 1434-41.

90. Obradors, N., Badia, J., Baldoma, L. and Aguilar, J. 1988. Anaerobic metabolism of the L-rhamnose fermentation product 1,2-propanediol in Salmonella typhimurium. J. Bacteriol. 170: 2159-62.

91. Badia, J., J. Ros, and J. Aguilar. 1985. Fermentation mechanisms of fucose and rhamnose in Salmonella typhimurium and Klebsiella pneumoniae. J. Bacteriol. 161: 435-7.

92. Golan, David E., ed. 2004. "Chapter 35 - Pharmacology of the Bacterial Cell Wall". Principles of Pharmacology: The Pathophysiologic Basis of Drug Therapy. 2nd ed. Armen H. Tashjian Jr., Ehrin J. Armstrong, Joshua N. Galanter, April Wang Armstrong, Ramy A. Arnaout, Harris S. Rose. Lippincott Williams and Wilkins.

93. Randle, C. L., P. W. Albro, and J. C. Dittmer. 1969. The phosphoglyceride composition of gramnegative bacteria and the changes in composition during growth. Biochim. Biophys. Acta. 187: 214–20.

94. Thiennimitr P, Winter SE, Winter MG, Xavier MN, Tolstikov V, Huseby DL, Sterzenbach T, Tsolis RM, Roth JR, Bäumler AJ. 2011. Intestinal inflammation allows Salmonella to use ethanolamine to compete with the microbiota. Proc Natl Acad Sci U S A. 108: 17480-5.

95. Winter SE, Thiennimitr P, Winter MG, Butler BP, Huseby DL, Crawford RW, Russell JM, Bevins CL, Adams LG, Tsolis RM, Roth JR, Bäumler AJ. 2010. Gut inflammation provides a respiratory electron acceptor for Salmonella. Nature. 467: 426-9.

96. Conner CP, Heithoff DM, Julio SM, Sinsheimer RL, Mahan MJ. 1998. Differential patterns of acquired virulence genes distinguish Salmonella strains. Proc Natl Acad Sci U S A. 95: 4641-5.

97. Heithoff DM, Conner CP, Hentschel U, Govantes F, Hanna PC, Mahan MJ. 1999. Coordinate intracellular expression of Salmonella genes induced during infection. J Bacteriol. 181: 799-807.

131

98. Buchrieser C, Rusniok C, Kunst F, Cossart P, Glaser P. 2003. Comparison of the genome sequences of Listeria monocytogenes and Listeria innocua: clues for evolution and pathogenicity. FEMS Immunol. Med. Microbiol. 35: 207–13.

99. Joseph B, Przybilla K, Stühler C, Schauer K, Slaghuis J, Fuchs TM, Goebel W. 2006. Identification of Listeria monocytogenes genes contributing to intracellular replication by expression profiling and mutant screening. J Bacteriol. 188: 556-68.

100. Parsons JB, Frank S, Bhella D, Liang M, Prentice MB, Mulvihill DP, Warren MJ. 2010. Synthesis of empty bacterial microcompartments, directed organelle protein incorporation, and evidence of filament- associated organelle movement. Mol Cell. 38: 305-15.

101. Choudhary S, Quin MB, Sanders MA, Johnson ET, Schmidt-Dannert C. 2012. Engineered protein nano-compartments for targeted enzyme localization. PLoS One.7: e33342.

102. Lawrence AD, Frank S, Newnham S, Lee MJ, Brown IR, Xue WF, Rowe ML, Mulvihill DP, Prentice MB, Howard MJ, Warren MJ. 2014. Solution structure of a bacterial microcompartmen PduB, targeting peptide and its application in the construction of an ethanol bioreactor. ACS Synth Biol. 3: 454-65.

103. Menon BB, Dou Z, Heinhorst S, Shively JM, Cannon GC. 2008. Halothiobacillus neapolitanus carboxysomes sequester heterologous and chimeric RubisCO species. PLoS One. 3: e3570.

104. Sparreboom, A.; Barker, S. D.; Verweij, J. 2005. Paclitaxel Repackaged in an Albumin-Stabilized Nanoparticle: Handy or Just a Dandy? J. Clin. Oncol. 23: 7765–7.

105. Torchilin, V. P.; Lukyanov, A. N. 2003. Peptide and Protein Drug Delivery to and into Tumors: Challenges and Solutions. Drug Discovery Today. 8: 259–66.

106. Allen, T. M.; Cullis, P. R. 2004. Drug Delivery Systems: Entering the Mainstream. Science. 303:1818–22.

107. Mastrobattista, E.; van der Aa, M. A. E. M.; Hennink, W. E.; Crommelin, D. J. A. 2006. Artificial Viruses: A Nanotechnological Approach to Gene Delivery. Nat. Rev. Drug Discovery. 5: 115–21.

108. Dobrogosz WJ, Casas IA, Pagano GA, Sjöberg B-M, Talarico TL, Karlsson M. 1989. Lactobacillus reuteri and the enteric microbiota. In: Grubb R, Midtvedt T, Norin E (eds). The regulatory and protective role of the normal microflora. MacMillan, London, pp 283–92.

109. Sriramulu DD, Liang M, Hernandez-Romero D, Raux-Deery E, Lünsdorf H, Parsons JB, Warren MJ, Prentice MB. 2008. Lactobacillus reuteri DSM 20016 produces cobalamin-dependent diol dehydratase in metabolosomes and metabolizes 1,2-propanediol by disproportionation. J Bacteriol. 190: 4559-67.

110. J. M. Shively, R. S. English. 1991. The carboxysome, a prokaryotic organelle: a mini-review Can. J. Bot. 69: 957.

111. Naitow H, Tang J, Canady M, Wickner RB, Johnson JE. 2002. L-A virus at 3.4 A resolution reveals particle architecture and mRNA decapping mechanism. Nat Struct Biol. 9: 725-8.

112. Diprose JM, Burroughs JN, Sutton GC, Goldsmith A, Gouet P, Malby R, Overton I, Ziéntara S, Mertens PP, Stuart DI, Grimes JM. 2001. Translocation portals for the substrates and products of a viral transcription complex: the bluetongue virus core. EMBO J. 20: 7229-39.

113. Comellas-Aragonès M, Engelkamp H, Claessen VI, Sommerdijk NA, Rowan AE, Christianen PC, Maan JC, Verduin BJ, Cornelissen JJ, Nolte RJ. 2007. A virus-based single-enzyme nanoreactor. Nat Nanotechnol. 2: 635-9.

132

114. Ailion M, Bobik TA, Roth JR. 1993. Two global regulatory systems (Crp and Arc) control the cobalamin/propanediol regulon of Salmonella typhimurium. J. Bacteriol. 175: 7200–8.

115. Roth JR, Lawrence JG, Rubenfield M, Kieffer-Higgins S, Church GM. 1993. Characterization of the cobalamin (vitamin B12) biosynthetic genes of Salmonella typhimurium. J Bacteriol. 175: 3303-16.

116. Scott, A. I., and C. A. Roessner. 2002. Biosynthesis of cobalamin (vitamin B12). Biochem. Soc. Trans. 30: 613-20.

117. Kim, J., C. Gherasim, and R. Banerjee. 2008. Decyanation of vitamin B12 by a trafficking chaperone. Proc. Natl. Acad. Sci. USA. 105: 14551-4.

118. Watanabe, F., Y. Oki, Y. Nakano, and S. Kitaoka. 1988. Occurrence and characterization of cyanocobalamin reductase (NADPH; CN-eliminating) involved in decyanation of cyanocobalamin in Euglena gracilis. J. Nutr. Sci. Vitaminol. (Tokyo) 34: 1-10.

119. Johnson, C. L. V., E. Pechonick, S. D. Park, G. D. Havemann, N. A. Leal, and T. A. Bobik. 2001. Functional genomic, biochemical, and genetic characterization of the Salmonella pduO gene, an ATP:cob(I)alamin adenosyltransferase gene. J. Bacteriol. 183: 1577-84.

120. Leal, N. A., S. D. Park, P. E. Kima, and T. A. Bobik. 2003. Identification of the human and bovine ATP:Cob(I)alamin adenosyltransferase cDNAs based on complementation of a bacterial mutant. J. Biol. Chem. 278: 9227-34.

121. St Maurice, M., P. E. Mera, M. P. Taranto, F. Sesma, J. C. Escalante-Semerena, and I. Rayment. 2007. Structural characterization of the active site of the PduO-type ATP:Co(I)rrinoid adenosyltransferase from Lactobacillus reuteri. J. Biol. Chem. 282: 2596-605.

122. Zhang, J., C. M. Dobson, X. Wu, J. Lerner-Ellis, D. S. Rosenblatt, and R. A. Gravel. 2006. Impact of cblB mutations on the function of ATP:cob(I)alamin adenosyltransferase in disorders of vitamin B12 metabolism. Mol. Genet. Metab. 87: 315-22.

123. Fan C, Bobik TA. 2008. Functional characterization and mutation analysis of human ATP:Cob(I)alamin adenosyltransferase. Biochemistry. 47: 2806–13.

124. Fonseca, M. V., and J. C. Escalante-Semerena. 2000. Reduction of Cob(III)alamin to Cob(II)alamin in Salmonella enterica serovar Typhimurium LT2. J. Bacteriol. 182: 4304-9.

125. Watanabe, F., Y. Oki, Y. Nakano, and S. Kitaoka. 1987. Purification and characterization of aquacobalamin reductase (NADPH) from Euglena gracilis. J. Biol. Chem. 262: 11514-8.

126. Sampson, E. M., C. L. V. Johnson, and T. A. Bobik. 2005. Biochemical evidence that the pduS gene encodes a bifunctional cobalamin reductase. Microbiology. 151: 1169-77.

127. Banerjee R, Ragsdale SW. 2003. The many faces of vitamin B12: catalysis by cobalamin-dependent enzymes. Annu Rev Biochem. 72: 209-47.

128. Cauthen, S. E., Foster, M. A., and Woods, D. D. 1966. Methionine synthesis by extracts of Salmonella typhimurium. Biochem. J. 98: 630-5.

129. El Fantroussi S, Naveau H, Agathos SN. 1998. Anaerobic dechlorinating bacteria. Biotechnol. Prog. 14: 167–88.

130. Linkfield TG, Tiedje JM. 1990. Characterization of the requirements and substrates for reductive dehalogenation by strain DCB-1. J. Ind. Microbiol. 5: 9–15.

133

131. Krasotkina J, Walters T, Maruya KA, Ragsdale SW. 2001. Characterization of the B12- and iron-sulfur- containing reductive dehalogenase from Desulfitobacterium chlororespirans. J. Biol. Chem. 276: 40991–7.

132. Wiegel J, Zhang X, Wu Q. 1999. Anaerobic dehalogenation of hydroxylated polychlorinated biphenyls by Desulfitobacterium dehalogenans. Appl. Environ. Microbiol. 65: 2217–21.

133. Toraya, T. 2003. Radical catalysis in coenzyme B12-dependent isomerization (eliminating) reactions. Chem. Rev. 103: 2095-128.

134. Toraya, T., and K. Mori. 1999. A reactivating factor for coenzyme B12-dependent diol dehydratase. J. Biol. Chem. 274: 3372–7.

135. Naidich MJ, Ho SU. 2005. Case 87: Subacute combined degeneration. Radiology. 237: 101–5.

136. Lee S, Lemere CA, Frost JL, Shea TB. 2012. Dietary supplementation with S-adenosyl methionine delayed amyloid-β and tau pathology in 3xTg-AD mice. J Alzheimers Dis. 28: 423-31.

137. Nicole Kresge, Robert D. Simoni and Robert L. Hill. 2005. Fritz Lipmann and the Discovery of Coenzyme A. J. Biol. Chem. 280: e18.

138. Baddiley J, Thain EM, Novelli GD, Lipmann F. 1953. Structure of coenzyme A. Nature. 171: 76.

139. Martinez DL, Tsuchiya Y, Gout I. 2014. Coenzyme A biosynthetic machinery in mammalian cells. Biochem Soc Trans. 42: 1112-7.

140. Jackowski S, Alix J-H. 1990. Cloning, sequence, and expression of the pantothenate permease (panF) gene of Escherichia coli. J Bacteriol. 172: 3842–8.

141. Leonardi, R., Zhang, Y.-M., Rock, C.O. and Jackowski, S. 2005. Coenzyme A: back in action. Prog. Lipid Res. 44: 125-53.

142. Nemazanyy, I., Panasyuk, G., Zhyvoloup, A., Panayotou, G., Gout, I.T. and Filonenko, V. 2004. Specific interaction between S6K1 and CoA synthase: a potential link between the mTOR/S6K pathway, CoA biosynthesis and energy metabolism. FEBS Lett. 578: 357–62.

143. So¨ derberg, O., Leuchowius, K.-J., Gullberg, M., Jarvius, M., Weibrecht, I., Larsson, L.-G. and Landegren, U. 2008. Characterizing proteins and their interactions in cells and tissues using the in situ proximity ligation assay. Methods. 45: 227–32.

144. Yang, K., Eyobo, Y., Brand, L.A., Martynowski, D., Tomchick, D., Strauss, E. and Zhang, H. 2006. Crystal structure of a type III pantothenate kinase: insight into the mechanism of an essential coenzyme A biosynthetic enzyme universally distributed in bacteria. J. Bacteriol. 188: 5532-40.

145. Bucovaz, E.T., Macleod, R.M., Morrison, J.C. and Whybrew, W.D. 1997. The coenzyme A- synthesizing protein complex and its proposed role in CoA biosynthesis in bakers’ yeast. Biochimie. 79: 787-98.

146. Choudhry, A. E., T. L. Mandichak, J. P. Broskey, R. W. Egolf, C. Kinsland, T. P. Begley, M. A. Seefeld, T. W. Ku, J. R. Brown, M. Zalacain, and K. Ratnam. 2003. Inhibitors of pantothenate kinase: novel antibiotics for staphylococcal infections. Antimicrob. Agents Chemother. 47: 2051-5.

147. Alfonso-Pecchio, A., Garcia, M., Leonardi, R. and Jackowski, S. 2012. Compartmentalization of mammalian pantothenate kinases. PLoS ONE. 7: e49509.

148. Davaapil H, Tsuchiya Y, Gout I. 2014. Signalling functions of coenzyme A and its derivatives in mammalian cells. Biochem Soc Trans. 42: 1056-62.

134

149. Reibel DK, Wyse BW, Berkich DA, Palko WM, Neely JR. 1981. Effects of diabetes and fasting on pantothenic acid metabolism in rats. Am J Physiol. 240: e597-601.

150. Smith, C.M. and Savage, C.R. 1980. Regulation of coenzyme A biosynthesis by glucagon and glucocorticoid in adult rat liver parenchymal cells. Biochem. J. 188: 175–84.

151. Robishaw, J.D., Berkich, D. and Neely, J.R. 1982. Rate-limiting step and control of coenzyme A synthesis in cardiac muscle. J. Biol. Chem. 257: 10967–72.

152. Berge, R.K., Hosøy, L.H. and Farstad, M.N. 1984. Influence of dietary status on liver palmitoyl-CoA hydrolase, peroxisomal enzymes, CoASH and long-chain acyl-CoA in rats. Int. J. Biochem. 16: 403–10.

153. Brass, E.P., Tahiliani, A.G., Allen, R.H. and Stabler, S.P. 1990. Coenzyme A metabolism in vitamin B- 12-deficient rats. J. Nutr. 120: 290–7.

154. Corkey, B.E., Hale, D.E., Glennon, M.C., Kelley, R.I., Coates, P.M., Kilpatrick, L. and Stanley, C.A. 1988. Relationship between unusual hepatic acyl coenzyme A profiles and the pathogenesis of Reye syndrome. J. Clin. Invest. 82: 782–8.

155. Reibel, D.K., Uboh, C.E. and Kent, R.L. 1983. Altered coenzyme A and carnitine metabolism in pressure-overload hypertrophied hearts. Am. J. Physiol. 244: H839–H843.

156. Vallari DS, Jackowski S, Rock CO. 1987. Regulation of pantothenate kinase by coenzyme A and its thioesters. J Biol Chem. 262: 2468-71.

157. Jackowski S, Rock CO. 1981. Regulation of coenzyme A biosynthesis. J Bacteriol. 148: 926–32.

158. Breus, O., Panasyuk, G., Gout, I.T., Filonenko, V. and Nemazanyy, I. 2010. CoA Synthase is phosphorylated on tyrosines in mammalian cells, interacts with and is dephosphorylated by Shp2PTP. Mol. Cell. Biochem. 335: 195–202.

159. Gudkova, D., Panasyuk, G., Nemazanyy, I., Zhyvoloup, A., Monteil, P., Filonenko, V. and Gout, I. 2012. EDC4 interacts with and regulates the dephospho-CoA kinase activity of CoA synthase. FEBS Lett. 586: 3590–5.

160. Jackowski S, Rock CO. 1986. Consequences of reduced intracellular coenzyme A content in Escherichia coli. J Bacteriol. 166: 866-71.

161. Vallari DS, Jackowski S. 1988. Biosynthesis and degradation both contribute to the regulation of coenzyme A content in Escherichia coli. J Bacteriol. 170: 3961-6.

162. Vadali RV, Bennett GN, San KY. 2004. Cofactor engineering of intracellular CoA/acetyl-CoA and its effect on metabolic flux redistribution in Escherichia coli. Metab Eng. 6: 133-9.

163. Leonardi R, Chohnan S, Zhang YM, Virga KG, Lee RE, Rock CO, Jackowski S. 2005. A pantothenate kinase from Staphylococcus aureus refractory to feedback regulation by coenzyme A. J Biol Chem. 280: 3314-22.

164. Song WJ, Jackowski S. 1992. Cloning, sequencing, and expression of the pantothenate kinase (coaA) gene of Escherichia coli. J Bacteriol. 174: 6411-7.

165. Kang LW, Gabelli SB, Bianchet MA, Xu WL, Bessman MJ, Amzel LM. 2003. Structure of a coenzyme A pyrophosphatase from Deinococcus radiodurans: a member of the Nudix family. J Bacteriol. 185: 4110- 8.

166. Lipmann F, Kaplan NO, Novelli GD. 1947. Chemistry and distribution of the coenzyme for acetylation (coenzyme A). Fed Proc. 6: 272.

135

167. Begley TP, Kinsland C, Strauss E. 2001. The biosynthesis of coenzyme A in bacteria. Vitam Horm. 61: 157-71.

168. Allfrey, V. G., Faulkner, R. & Mirsky, A. E. 1964. Acetylation and methylation of histones and their possible role in the regulation of RNA synthesis. Proc. Natl Acad. Sci. USA. 51: 786–94.

169. Kurdistani SK, Grunstein M. 2003. Histone acetylation and deacetylation in yeast. Nat Rev Mol Cell Biol. 4: 276-84.

170. Wang Q, Zhang Y, Yang C, Xiong H, Lin Y, Yao J, Li H, Xie L, Zhao W, Yao Y, Ning ZB, Zeng R, Xiong Y, Guan KL, Zhao S, Zhao GP. 2010. Acetylation of metabolic enzymes coordinates carbon source utilization and metabolic flux. Science. 327: 1004-7.

171. Jaouen, G., ed. 2006. Bioorganometallics: Biomolecules, Labeling, Medicine. Weinheim: Wiley-VCH.

172. S. Maudsley, A. M. Zamah, N. Rahman, J. T. Blitzer, L. M. Luttrell, R. J. Lefkowitz, R. A. Hall. 2000. Platelet-derived growth factor receptor association with Na+/H+ exchanger regulatory factor potentiates receptor activity. Mol. Cell. Biol. 20: 8352-63.

173. Kennedy MB. 1995. Origin of PDZ (DHR, GLGF) domains. Trends Biochem Sci. 20: 350.

174. Fanning, A.S.; Anderson, J.M. 1999. Protein modules as organizers of membrane structure. Curr. Opin. Cell. Biol. 11: 432-9.

175. Kim, E.; Sheng, M. 2004. PDZ domain proteins of synapses. Nat. Rev. Neurosci. 10: 771-81.

176. Scott, J.D.; Pawson, T. 2009. Cell signaling in space and time: where proteins come together and when they're apart. Science. 326: 1220-4.

177. Pawson, T.; Kofler, M. 2009. Kinome signaling through regulated protein– protein interactions in normal and cancer cells. Curr. Opin. Cell. Biol. 21: 147-53.

178. Harris, B.Z.; Lim, W. A. 2001. Mechanism and role of PDZ domains in signaling complex assembly. J. Cell Sci. 114: 3219-31.

179. Morais Cabral JH, Petosa C, Sutcliffe MJ, Raza S, Byron O, Poy F, Marfatia SM, Chishti AH, Liddington RC. 1996. Crystal structure of a PDZ domain Nature. 382: 649-52.

180. Songyang, Z.; Fanning, A.S.; Fu, C.; Xu, J.; Marfatia, S.M.; Chishti, A.H.; Crompton, A.; Chan, A.C.; Anderson, J.M.; Cantley, L.C. 1997. Recognition of unique carboxyl-terminal motifs by distinct PDZ domains. Science. 275: 73-7.

181. Babault, N.; Cordier, F.; Lafage, M.; Cockburn, J.; Haouz, A.; Prehaud, C.; Rey, F.A.; Delepierre, M.; Buc, H.; Lafon, M.; Wolff, N. 2011. Peptides targeting the PDZ domain of PTPN4 are efficient inducers of glioblastoma cell death. Structure. 19: 1518-24.

182. Terrien, E.; Chaffotte, A.; Lafage, M.; Khan, Z.; Préhaud, C.; Cordier, F.; Simenel, C.; Delepierre, M.; Buc, H.; Lafon, M.; Wolff, N. 2012. Interference with the PTEN-MAST2 interaction by a viral protein leads to cellular relocalization of PTEN. Sci. Signal. 5: ra58.

183. H. Tochio, Y. K. Mok, Q. Zhang, H. M. Kan, D. S. Bredt, M. Zhang. 2000. Formation of NOS/PSD-95 PDZ dimer requires a preformed β-finger structure from the nNOS PDZ domain. J. Mol. Biol. 303: 359-70.

184. Y. J. Im, S. H. Park, S. H. Rho, J. H. Lee, G. B. Kang, M. Sheng, E. Kim, S. H. Eom. 2002. Crystal structure of GRIP1 PDZ6-peptide complex reveals the structural basis for class II PDZ target recognition and PDZ domain-mediated multimerization. J. Biol. Chem. 278: 8501-7.

136

185. P. Zimmermann, K. Meerschaert, G. Reekmans, I. Leenaerts, J. V. Small, J. Vandekerckhove, G. David, J. Gettemans. 2002. PIP2-PDZ domain binding controls the association of syntenin with the plasma membrane. Mol. Cell. 9: 1215-25.

186. Nourry C, Grant SG, Borg JP. 2003. PDZ domain proteins: plug and play! Sci STKE. 179: re7.

187. Kudva R, Denks K, Kuhn P, Vogt A, Muller M, Koch HG, Mu¨ller M. 2013. Protein translocation across the inner membrane of Gramnegative bacteria: The Sec and Tat dependent protein transport pathways. Res Microbiol. 164: 505–34.

188. Koch, H.G., Moser, M., Mu¨ller, M. 2003. Signal recognition particle dependent protein targeting, universal to all kingdoms of life. Rev. Physiol. Biochem. Pharmacol. 146: 55-94.

189. Von Heijne G. 1986. A new method for predicting signal sequence cleavage sites. Nucleic Acids Res. 14: 4683–90.

190. Von Heijne G. 1984. Analysis of the distribution of charged residues in the N-terminal region of signal sequences: Implications for protein export in prokaryotic and eukaryotic cells. EMBO J. 3: 2315–8.

191. Von Heijne G. 1990. The signal peptide. J Membr Biol. 115: 195–201.

192. Talkad, V., Schneider, E., Kennell, D. 1976. Evidence for variable rates of ribosome movement in Escherichia coli. J. Mol. Biol. 104: 299-303.

193. Woolhead, C.A., McCormick, P.J., Johnson, A.E. 2004. Nascent membrane and secretory proteins differ in FRET-detected folding far inside the ribosome and in their exposure to ribosomal proteins. Cell. 116: 725-36.

194. Peterson, J.H., Woolhead, C.A., Bernstein, H.D. 2010. The conformation of a nascent polypeptide inside the ribosome tunnel affects protein targeting and protein folding. Mol. Microbiol. 78: 203-17.

195. Lin, K.F., Sun, C.S., Huang, Y.C., Chan, S.I., Koubek, J., Wu, T.H., Huang, J.J. 2012. Cotranslational protein folding within the ribosome tunnel influences trigger-factor recruitment. Biophys. J. 102: 2818-27.

196. Schlu¨nzen, F., Wilson, D.N., Tian, P., Harms, J.M., McInnes, S.J., Hansen, H.A., Albrecht, R., Buerger, J., Wilbanks, S.M., Fucini, P. 2005. The binding mode of the trigger factor on the ribosome: implications for protein folding and SRP interaction. Structure. 13: 1685-94.

197. Baram, D., Pyetan, E., Sittner, A., Auerbach-Nevo, T., Bashan, A., Yonath, A. 2005. Structure of trigger factor binding domain in biologically homologous complex with eubacterial ribosome reveals its chaperone action. Proc. Natl. Acad. Sci. U S A. 102: 12017-22.

198. De Gier J-WL, Mansournia P, Valent QA, Phillips GJ, Luirink J, von Heijne G. 1996. Assembly of a cytoplasmic membrane protein in Escherichia coli is dependent on the signal recognition particle. FEBS Lett. 399: 307–9.

199. Macfarlane J, Muller M. 1995. The functional integration of a polytopic membrane protein of Escherichia coli is dependent on the bacterial signal-recognition particle. Eur J Biochem. 233:766–71.

200. Keenan RJ, Freymann DM, Walter P, Stroud RM. 1998. Crystal structure of the signal sequence binding subunit of the signal recognition particle. Cell. 94:181–91.

201. Ataide SF, Schmitz N, Shen K, Ke A, Shan SO, Doudna JA, Ban N. 2011. The crystal structure of the signal recognition particle in complex with its receptor. Science. 331:881–6.

202. Holtkamp W, Lee S, Bornemann T, Senyushkina T, Rodnina MV, Wintermeyer W. 2012. Dynamic switch of the signal recognition particle from scanning to targeting. Nat Struct Mol Biol. 19: 1332–7.

137

203. Egea PF, Shan SO, Napetschnig J, Savage DF, Walter P, Stroud RM. 2004. Substrate twinning activates the signal recognition particle and its receptor. Nature. 427: 215–21.

204. Akopian D, Dalal K, Shen K, Duong F, Shan SO. 2013. SecYEG activates GTPases to drive the completion of cotranslational protein targeting. J Cell Biol. 200: 397–405.

205. Akopian D, Shen K, Zhang X, Shan SO. 2013. Signal recognition particle: An essential protein-targeting machine. Annu Rev Biochem. 82: 693–721.

206. Karamyshev AL, Johnson AE. 2005. Selective SecA association with signal sequences in ribosome- bound nascent chains: A potential role for SecA in ribosome targeting to the bacterial membrane. J Biol Chem. 280: 37930–40.

207. Hartl, F.U., Lecker, S., Schiebel, E., Hendrick, J.P., Wickner, W. 1990. The binding cascade of SecB to SecA to SecY/E mediates preprotein targeting to the E. coli plasma membrane. Cell. 63: 269-79.

208. Fekkes, P., van der Does, C., Driessen, A.J. 1997. The molecular chaperone SecB is released from the carboxy-terminus of SecA during initiation of precursor protein translocation. EMBO J. 16: 6105-13.

209. Fekkes P, De Wit JG, Van Der Wolk JPW, Kimsey HH, Kumamoto CA, Driessen AJM. 1998. Preprotein transfer to the Escherichia coli translocase requires the co-operative binding of SecB and the signal sequence to SecA. Mol Microbiol. 29: 1179–90.

210. Randall LL, Topping TB, Smith VF, Diamond DL, Hardy SJ. 1998. SecB: A chaperone from Escherichia coli. Methods Enzymol. 290: 444–59.

211. Huber D, Rajagopalan N, Preissler S, Rocco MA, Merz F, Kramer G, Bukau B. 2011. SecA interacts with ribosomes in order to facilitate posttranslational translocation in bacteria. Mol Cell. 41: 343–53.

212. Sutter, M.; Boehringer, D.; Gutmann, S.; Gunther, S.; Prangishvili, D.; Loessner, M. J.; Stetter, K. O.; Weber-Ban, E.; Ban, N. 2008. Structural basis of enzyme encapsulation into a bacterial nanocompartment. Nat. Struct. Mol. Biol. 15: 939−47.

213. Rahmanpour, R.; Bugg, T. D. H. 2013. Assembly in vitro of Rhodococcus jostii RHA1 encapsulin and peroxidase DypB to form a nanocompartment. FEBS J. 280: 2097−104.

214. Rurup, W. F.; Snijder, J.; Koay, M. S. T.; Heck, A. J. R.; Cornelissen, J. J. L. M. 2014. Self-sorting of foreign proteins in a bacterial nanocompartment. J. Am. Chem. Soc. 136: 3828−32.

215. Harrison, P.M. & Arosio, P. 1996. The ferritins: molecular properties, iron storage function and cellular regulation. Biochim. Biophys. Acta. 1275: 161–203.

216. Rocha, E.R., Andrews, S.C., Keen, J.N. & Brock, J.H. 1992. Isolation of a ferritin from Bacteroides fragilis. FEMS Microbiol. Lett. 95: 207–12.

217. Manca, C., Paul, S., Barry, C.E. III, Freedman, V.H. & Kaplan, G. 1999. Mycobacterium tuberculosis catalase and peroxidase activities and resistance to oxidative killing in human monocytes in vitro. Infect. Immun. 67: 74–9.

218. Mikyas,Y., Makabi,M., Raval-Fernandes,S., Harrington,L., Kickhoefer,V.A., Rome,L.H. and Stewart,P.L. 2004. Cryoelectron microscopy imaging of recombinant and tissue derived vaults: localization of the MVP N termini and VPARP. J. Mol. Biol. 344: 91–105.

219. Steiner E, Holzmann K, Pirker C, Elbling L, Micksche M, Sutterluty H, Berger W. 2006. The major vault protein is responsive to and interferes with interferon-{gamma}-mediated STAT1 signals. J Cell Sci. 119: 459–69.

138

220. Kolli S, Zito CI, Mossink MH, Wiemer EA, Bennett AM. 2004. The Major Vault Protein Is a Novel Substrate for the Tyrosine Phosphatase SHP-2 and Scaffold Protein in Epidermal Growth Factor Signaling. J Biol Chem. 279: 29374–85.

221. Kim E, Lee S, Mian MF, Yun SU, Song M, Yi KS, Ryu SH, Suh PG. 2006. Crosstalk between Src and major vault protein in epidermal growth factor-dependent cell signalling. Febs J. 273: 793–804.

222. Eichenmuller B, Kedersha N, Solovyeva E, Everley P, Lang J, Himes RH, Suprenant KA. 2003. Vaults bind directly to microtubules via their caps and not their barrels. Cell Motil Cytoskeleton. 56: 225– 36.

223. Gopinath, S.C., Matsugami, A., Katahira, M. and Kumar, P.K. 2005. Human vault-associated noncoding RNAs bind to mitoxantrone, a chemotherapeutic compound. Nucleic Acids Res. 33: 4874–81.

224. Kowalski, M.P., Dubouix-Bourandy, A., Bajmoczi, M., Golan, D.E., Zaidi, T., Coutinho-Sledge, Y.S., Gygi, M.P., Gygi, S.P., Wiemer, E.A.C. and Pier, G.B. 2007. Host resistance to lung infection mediated by major vault protein in epithelial cells. Science. 317: 130–2.

225. Kickhoefer, V. A., Siva, A. C., Kedersha, N. L., Inman, E. M., Ruland, C., Streuli, M. & Rome, L. H. 1999. The 193-kD vault protein, VPARP, is a novel poly(ADP-ribose) polymerase. J. Cell Biol. 146: 917– 28.

226. Casañas A, Guerra P, Fita I, Verdaguer N. 2012. Vault particles: a new generation of delivery nanodevices. Curr Opin Biotechnol. 23: 972-7.

227. Kickhoefer VA, Garcia Y, Mikyas Y, Johansson E, Zhou JC, Raval-Fernandes S, Minoofar P, Zink JI, Dunn B, Stewart PL, Rome LH. 2005. Engineering of vault nanocapsules with enzymatic and fluorescent properties. Proc Natl Acad Sci U S A. 102: 4348–52.

228. Goldsmith, L. E.; Pupols, M.; Kickhoefer, V. A.; Rome, L. H.; Monbouquette, H. G. 2009. Utilization of a protein “shuttle” to load vault nanocapsules with gold probes and proteins. ACS Nano. 3: 3175–83.

229. Kozlov, G.; Vavelyuk, O.; Minailiuc, O.; Banville, D.; Gehring, K.; Ekiel, I. 2006. Solution Structure of a Two-Repeat Fragment of Major Vault Protein. J. Mol. Biol. 356: 444–52.

230. Van Zon,A., Mossink,M.H., Schoester,M., Scheffer,G.L., Scheper,R.J., Sonneveld,P. and Wiemer,E.A. 2002. Structural domains of vault proteins: a role for the coiled coil domain in vault assembly. Biochem. Biophys. Res. Commun. 291: 535–41.

231. Poderycki MJ, Rome LH, Harrington L, Kickhoefer VA. 2005. The p80 homology region of TEP1 is sufficient for its association with the telomerase and vault RNAs, and the vault particle. Nucleic Acids Res. 33: 893–902.

232. Toraya, T., S. Honda, and S. Fukui. 1979. Fermentation of 1,2-propanediol and 1,2-ethanediol by some genera of Enterobacteriaceae, involving coenzyme B12-dependent diol dehydratase. J. Bacteriol. 139: 39– 47.

233. Abeles, R. H., and H. A. Lee. 1961. Intramolecular oxidation-reduction requiring a cobamide coenzyme. J. Biol. Chem. 236: 2347–50.

234. Brinsmade, S. R., T. Paldon, and J. C. Escalante-Semerena. 2005. Minimal functions and physiological conditions required for growth of Salmonella enterica on ethanolamine in the absence of the metabolosome. J. Bacteriol. 187: 8039–46.

235. Penrod, J. T., and J. R. Roth. 2006. Conserving a volatile metabolite: a role for carboxysome-like organelles in Salmonella enterica. J. Bacteriol. 188: 2865–74.

139

236. Paoli, P., P. Cirri, L. Camici, G. Manao, G. Cappugi, G. Moneti, G. Pieraccini, G. Camici, and G. Ramponi. 1997. Common-type acylphosphatase: steady-state kinetics and leaving-group dependence. Biochem. J. 327: 177–84.

237. Miller, J. H. 1972. Experiments in molecular genetics. Cold Spring Harbor Laboratory, Cold Spring Harbor, NY.

238. Berkowitz, D., J. M. Hushon, H. J. Whitfield, Jr., J. Roth, and B. N. Ames. 1968. Procedure for identifying nonsense mutations. J. Bacteriol. 96: 215–20.

239. Vogel, H. J., and D. M. Bonner. 1956. Acetylornithinase of Escherichia coli: partial purification and some properties. J. Biol. Chem. 218: 97–106.

240. Sambrook, J., E. F. Fritsch, and T. Maniatis. 1989. Molecular cloning: a laboratory manual, 2nd ed. Cold Spring Harbor Laboratory, Cold Spring Harbor, NY.

241. Davis, R. W., D. Botstein, and J. R. Roth. 1980. Advanced bacterial genetics: a manual for genetic engineering. Cold Spring Harbor Laboratory, Cold Spring Harbor, NY.

242. Schmieger, H. 1971. A method for detection of phage mutants with altered transducing ability. Mol. Gen. Genet. 110: 378–81.

243. Warren, J. W., J. R. Walker, J. R. Roth, and E. Altman. 2000. Construction and characterization of a highly regulable expression vector, pLAC11, and its multipurpose derivatives, pLAC22 and pLAC33. Plasmid. 44: 138–51.

244. Lundie, L. L., Jr., and J. G. Ferry. 1989. Activation of acetate by Methanosarcina thermophila. J. Biol. Chem. 264: 18392–6.

245. Miller, V. L., and J. J. Mekalanos. 1988. A novel suicide vector and its use in construction of insertion mutations: osmoregulation of outer membrane proteins and virulence determinants in Vibrio cholerae requires toxR. J. Bacteriol. 170: 2575–83.

246. Altschul, S. F., T. L. Madden, A. A. Schaffer, J. Zhang, Z. Zhang, W. Miller, and D. J. Lipman. 1997. Gapped BLAST and PSI-BLAST: a new generation of protein database search programs. Nucleic Acids Res. 25: 3389–402.

247. LeVine, S. M., F. Ardeshir, and G. F. Ames. 1980. Isolation and characterization of acetate kinase and phosphotransacetylase mutants of Escherichia coli and Salmonella typhimurium. J. Bacteriol. 143: 1081–5.

248. Shimizu, M., T. Suzuki, K. Y. Kameda, and Y. Abiko. 1969. Phosphotransacetylase of Escherichia coli B, purification and properties. Biochim. Biophys. Acta. 191: 550–8.

249. Palacios, S., V. J. Starai, and J. C. Escalante-Semerena. 2003. Propionyl coenzyme A is a common intermediate in the 1,2-propanediol and propionate catabolic pathways needed for expression of the prpBCDE operon during growth of Salmonella enterica on 1,2-propanediol. J. Bacteriol. 185: 2802–10.

250. Datsenko, K. A., and B. L. Wanner. 2000. One-step inactivation of chromosomal genes in Escherichia coli K-12 using PCR products. Proc. Natl. Acad. Sci. USA. 97: 6640–5.

251. Jeter, R. M. 1990. Cobalamin-dependent 1,2-propanediol utilization by Salmonella typhimurium. J. Gen. Microbiol. 136: 887–96.

252. Dawson, R. M. C., D. C. Elliott, W. H. Elliott, and K. M. Jones (ed.). 1969. Data for biochemical research, 2nd ed. Oxford University Press, Oxford, United Kingdom.

140

253. Brinsmade, S. R., and J. C. Escalante-Semerena. 2004. The eutD gene of Salmonella enterica encodes a protein with phosphotransacetylase enzyme activity. J. Bacteriol. 186: 1890–2.

254. McCleary, W. R., J. B. Stock, and A. J. Ninfa. 1993. Is acetyl phosphate a global signal in Escherichia coli? J. Bacteriol. 175: 2793–8.

255. Shi, I. Y., J. Stansbury, and A. Kuzminov. 2005. A defect in the acetyl coenzyme A7acetate pathway poisons recombinational repair-deficient mutants of Escherichia coli. J. Bacteriol. 187: 1266–75.

256. Wolfe, A. J., D. E. Chang, J. D. Walker, J. E. Seitz-Partridge, M. D. Vidaurri, C. F. Lange, B. M. Pruss, M. C. Henk, J. C. Larkin, and T. Conway. 2003. Evidence that acetyl phosphate functions as a global signal during biofilm development. Mol. Microbiol. 48: 977–88.

257. Schomburg, I., A. Chang, C. Ebeling, M. Gremse, C. Heldt, G. Huhn, and D. Schomburg. 2004. BRENDA, the enzyme database: updates and major new developments. Nucleic Acids Res. 32: D431–3.

258. Yeates TO, Thompson MC, Bobik TA. 2011. The protein shells of bacterial microcompartment organelles. Curr Opin Struct Biol. 21: 223-31.

259. Klumpp J, Fuchs TM. 2007. Identification of novel genes in genomic islands that contribute to Salmonella typhimurium replication in macrophages. Microbiology. 153: 1207-20.

260. Cai F, Sutter M, Cameron JC, Stanley DN, Kinney JN, Kerfeld CA. 2013. The structure of CcmP, a tandem bacterial microcompartment domain protein from the β-carboxysome, forms a subcompartment within a microcompartment. J Biol Chem. 288: 16055-63.

261. Fan C, Bobik TA. 2011. The N-terminal region of the medium subunit (PduD) packages adenosylcobalamin-dependent diol dehydratase (PduCDE) into the Pdu microcompartment. J Bacteriol. 193: 5623-8.

262. Fan C, Cheng S, Liu Y, Escobar CM, Crowley CS, Jefferson RE, Yeates TO, Bobik TA. 2010. Short N-terminal sequences package proteins into bacterial microcompartments. Proc Natl Acad Sci U S A. 107: 7509-14.

263. Fan C, Cheng S, Sinha S, Bobik TA. 2012. Interactions between the termini of lumen enzymes and shell proteins mediate enzyme encapsulation into bacterial microcompartments. Proc Natl Acad Sci U S A. 109: 14995-5000.

264. Horswill AR1, Escalante-Semerena JC. 1999. Salmonella typhimurium LT2 catabolizes propionate via the 2-methylcitric acid cycle. J Bacteriol. 181: 5615-23.

265. Vincenzini MT, Favilli F, Stio M, Vanni P, Treves C. 1985. Detergents as selective inhibitors and inactivators of enzymes. Physiol Chem Phys Med NMR. 17: 279-95.

266. Lee H, DeLoache WC, Dueber JE. 2012. Spatial organization of enzymes for metabolic engineering. Metab. Eng. 14: 242–51.

267. Kinney JN, Salmeen A, Cai F, Kerfeld CA. 2012. Elucidating essential role of conserved carboxysomal protein CcmN reveals common feature of bacterial microcompartment assembly. J Biol Chem. 287: 17729- 36.

141

ACKNOWLEDGEMENTS

I would like to express the deepest appreciation to my major professor Dr. Bobik, who has the intelligence and the spirits of a wonderful mentor. Since the day I began my study in his lab,

Dr. Bobik has always been there for me, educating me through experiments, encouraging me to think independently and critically, helping me with professional writing, and transforming me into a better person. Without his endless support and guidance this dissertation would not have been possible.

I would like to thank my co-major professor Dr. Qijing Zhang, my committee members Dr.

Marian Kohut, Dr. Alan Dispirito, and Dr. Michael Cho for their time, support and suggestions.

Their deep knowledge and expertise in their respective field of research helped me thinking about my projects from new perspectives.

It has been such a pleasure studying with my previous labmates: Dr. Shouqiang Cheng, Dr.

Chuanguang Fan, Dr. Sharmistha Sinha, Dr. Huilin Zhu, and Dr. David Gogerty. Especially deep thanks to Shouqiang and Chenguang for providing a family-like lab environment and unlimited help on and collaboration with my research. Many thanks to Sharmistha for the help and encouragement on my study. I would also like to mention my current lab members Christian

Bartholomay, Dr. Ryan Sturms, Dr. Chiranjit Chowdhury, Brent Lehman, Nick Streausin, and

Alexandra Volker. It has been a pleasure working with them. Thanks to Ryan and Chiranjit for the suggestions on my experiments.

I would also like to give special thanks to Megan O’Donnell from Iowa State University library for helping with obtaining permissions and rights for citing publications.

Finally and most importantly, I would like to thank my family: my mother, Xiuqing Guo; my father, Xikui Liu; my brother, Dayong Liu; and my fiancé, Bryan Gontarek for their endless love

142

and support. Over the years my family has been always extremely supportive. Without their encouragement and infinite love, I would never had traveled across the ocean and pursued my study here. Bryan has been so wonderful since the first day we met and has always been there with me, for the good and the bad. I feel so lucky to have him and his love in my life.

143

NOMENCLATURE

1,2-PD 1,2-Propanediol

3-AT 3-Amino-1,2,4-Triazole

3-HPA 3-Hydroxypropioaldehyde

3-PGA 3-Phosphoglycerate

ACP Acyl Carrier Protein

AdoCbl Adenosylcobalamin

AEBSF 4-(2-Aminoethyl) Benzenesulfonylfluoride.HCl

ALA 5-Aminolaevulinic Acid

ATP Adenosine Triphosphate

BMC Bacterial Microcompartment

BMV Bacterial Microcompartment Vertex

BSA Bovine Serum Albumin

CA Carbonic Anhydrase

Cbl Cobalamin

CCM Carbon-Concentrating Mechanism

CCMV Cowpea Chlorotic Mottle Virus

CDC Centers for Disease Control and Prevention

CFU Colony Forming Unit

CH3Cbl Methylcobalamin

CI Competitive Index

CNCbl Cyanocobalamin

CoA Coenzyme A

144

CoAsy Coenzyme A synthase

CI Competitive Index Assay

DDH Diol Dehydratase

DHAP Dihydroxyacetone Phosphate

DLG1 Drosophila Disc Large Tumor Suppressor

DMAPP Dimethylallyl Pyrophosphate

DMB Dimethylbenzimidazole

DPCK Dephospho-CoA Kinase

DyP Dye Decolorizing Peroxidase

EM Electron Microscopy

EPA Environmental Protection Agency

Eut MCP Ethanolamine Utilization Microcompartment

FAS Fatty Acid Synthase

Ffh Fifty-Four Homologue

Flp Ferritin Like Protein

GRIP Glutamate Receptor-Interacting Protein

Grp MCP Glycyl Radical-based Propanediol Utilization Microcompartment

HMG-CoA 3-Hydroxy-3-Methylglutaryl-CoA

HPLC High-Pressure Liquid Chromatography

HPLC-ESI-MS HPLC electrospray ionization mass spectrometry

IPP Isopentenyl Pyrophosphate

IPTG Isopropyl--D-Thiogalactopyranoside

IVET In Vivo Expression Technology

145

LB Luria-Bertani/Lennox

MCP Microcompartment

MMA Methylmalonic Acid

MVP Major Vault Protein

NAD+ Nicotinamide adenine dinucleotide

NBIA Neurodegeneration with Brain Iron Accumulation

NCE No-Carbon-E

NHERF Na+/H+ Exchanger Regulatory Factor nNOS Neuronal Nitric Oxide Synthase iNTS Invasive Non-Typhoidal Salmonella

NTS Non-Typhoidal Salmonella

OH-Cbl Hydroxycobalamin

PAGE Polyacrylamide Gel Electrophoresis

PanF Pantothenate Permease

PanK Pantothenate Kinase

PCB Polychlorinated Biphenyl

Pdu MCP Propanediol Utilization Microcompartment

PI3K Phosphoinositide 3-Kinase

PIP2 Phosphatidylinositol 4,5-Bisphosphate

PKAN PANK-Associated Neurodegeneration

PPAT Phosphopantetheine Adenylyltransferase

PPCDC Phosphopantothenoylcysteine Decarboxylase

PPCS Phosphopantothenoylcysteine Synthase

146

PPIs Protein-Protein Interactions

PSD95 Postsynaptic Density Protein

PTAC Phosphotransacylase

RNP Ribonucleoprotein

RuBisCO Ribulose Bisphosphate Carboxylase/Oxygenase

RuBP Ribulose-1,5-Bisphosphate

SAM S-Adenosyl-Methionine

Shp2PTP Src Homology 2 Domain-Containing Protein Tyrosine Phosphatase

SPR Surface Plasmon Resonance

SPI Salmonella Pathogenicity Island

SRP Signal Recognition Particle

SR Signal Recognition Particle Receptor

PCB Polychlorinated Biphenyl

PI3K Phosphoinositide 3-Kinase

PPA propionaldehyde

T3SS/TTSS Type III Secretion System

TEM Transmission Electron Microscopy

TEP1 Telomerase Associated Protein I

TF Trigger Factor

TFP Teal Fluorescent Protein

THF Tetrahydrofolate

TM Transmembrane

TNTC Too Numerous to Account

147

Uro III Uroporphyrinogen III

VLP Virus Like Particle

VPARP Vault Poly (ADP-Ribose)-Polymerase

ZO1 Zonular Occludens-1