University of Kentucky UKnowledge

Theses and Dissertations--Pharmacy College of Pharmacy

2021

INTERACTIONS OF POST-PKS ENZYMES OF THE MITHRAMYCIN BIOSYNTHETIC PATHWAY

Ryan Wheeler University of Kentucky, [email protected] Author ORCID Identifier: https://orcid.org/0000-0002-2466-4018 Digital Object Identifier: https://doi.org/10.13023/etd.2021.043

Right click to open a feedback form in a new tab to let us know how this document benefits ou.y

Recommended Citation Wheeler, Ryan, "INTERACTIONS OF POST-PKS ENZYMES OF THE MITHRAMYCIN BIOSYNTHETIC PATHWAY" (2021). Theses and Dissertations--Pharmacy. 123. https://uknowledge.uky.edu/pharmacy_etds/123

This Doctoral Dissertation is brought to you for free and open access by the College of Pharmacy at UKnowledge. It has been accepted for inclusion in Theses and Dissertations--Pharmacy by an authorized administrator of UKnowledge. For more information, please contact [email protected]. STUDENT AGREEMENT:

I represent that my thesis or dissertation and abstract are my original work. Proper attribution has been given to all outside sources. I understand that I am solely responsible for obtaining any needed copyright permissions. I have obtained needed written permission statement(s) from the owner(s) of each third-party copyrighted matter to be included in my work, allowing electronic distribution (if such use is not permitted by the fair use doctrine) which will be submitted to UKnowledge as Additional File.

I hereby grant to The University of Kentucky and its agents the irrevocable, non-exclusive, and royalty-free license to archive and make accessible my work in whole or in part in all forms of media, now or hereafter known. I agree that the document mentioned above may be made available immediately for worldwide access unless an embargo applies.

I retain all other ownership rights to the copyright of my work. I also retain the right to use in future works (such as articles or books) all or part of my work. I understand that I am free to register the copyright to my work.

REVIEW, APPROVAL AND ACCEPTANCE

The document mentioned above has been reviewed and accepted by the student’s advisor, on behalf of the advisory committee, and by the Director of Graduate Studies (DGS), on behalf of the program; we verify that this is the final, approved version of the student’s thesis including all changes required by the advisory committee. The undersigned agree to abide by the statements above.

Ryan Wheeler, Student

Dr. Jürgen Rohr, Major Professor

Dr. David Feola, Director of Graduate Studies INTERACTIONS OF POST-PKS ENZYMES OF THE MITHRAMYCIN

BIOSYNTHETIC PATHWAY

______

DISSERTATION ______

A dissertation submitted in partial fulfillment of the requirements for the degree of Doctor of Philosophy in the College of Pharmacy at the University of Kentucky

By

Ryan Wheeler

Lexington, Kentucky

Director: Dr. Jürgen Rohr, Professor of Pharmaceutical Sciences

Lexington, Kentucky

2021

Copyright © Ryan Wheeler 2021

https://orcid.org/0000-0002-2466-4018 ABSTRACT OF DISSERTATION

INTERACTIONS OF POST-PKS ENZYMES OF THE MITHRAMYCIN BIOSYNTHETIC PATHWAY Combinatorial biosynthesis is a powerful tool for generating new, more active drug analogues to combat disease. But in order for combinatorial biosynthesis to be employed to its full potential, a deep understanding of the enzymes that produce the parent molecule must be had. The goals of the work presented in this thesis are to characterize the reaction catalyzed by MtmW, the final enzyme in the mithramycin (MTM) biosynthetic pathway, and to discover the interaction between MtmW and MtmOIV. MtmW is an aldol-ketoreductase responsible for reducing the most distal carbonyl on the MTM pentyl side chain. It forms an octamer that has structural homology to the Kvβ subunits of the Kv1 potassium ion channel. While performing in vitro reactions with MtmW, we discovered that an inactive C-2 epimer, iso-mithramycin (iso-MTM), was produced. After verifying that iso-MTM was the correct product both in vitro and in vivo, we determined that iso-MTM isomerized spontaneously at a slow rate in the presence of Mg2+ and rapidly when passing through the S. argillaceus cell membrane. Finally, we showed that MTM DK, the product of MtmOIV, is the substrate for MtmW, and that MtmW could react with MTM DK without the presence of MtmOIV. There are, however, weaknesses to the combinatorial biosynthesis method including low yields and low substrate affinity. A contributing factor to these weaknesses, we propose, is loss of the native protein-protein interactions (PPIs) that are present in the biosynthetic pathways subjected to combinatorial biosynthesis. To demonstrate this, we have shown that a PPI exists between MtmOIV and MtmW, and we have characterized the nature of this interaction using 19F NMR and covalent labeling strategies combined with MS/MS. KEYWORDS : Natural products, Mithramycin, Protein-Protein interactions, Ketoreductase Ryan Wheeler 01-30-2021 INTERACTIONS OF POST-PKS ENZYMES OF THE MITHRAMYCIN

BIOSYNTHETIC PATHWAY

By

Ryan Wheeler

Dr. Jürgen Rohr Director of Dissertation

Dr. David Feola Director of Graduate Studies

01-30-2021

Table of Contents

List of Tables ...... vii

List of Figures ...... viii

List of Acronyms ...... xix

Background Information ...... 1

1.1 Mithramycin Structure and Activity ...... 1

1.1.1 MTM as an SP1 inhibitor ...... 2

1.1.2 MTM as an EWS-Fli1 inhibitor ...... 4

1.1.3 MTM and Hypercalcemia ...... 6

1.2 Mithramycin Biosynthesis...... 7

1.2.1 MTM PKS ...... 7

1.2.2 Deoxysugar Biosynthesis ...... 12

1.2.3 Methylation ...... 16

1.2.4 Glycosylation ...... 20

1.2.5 MtmOIV, MtmW ...... 29

1.2.6 Regulation and Resistance ...... 36

1.3 Specific Aims ...... 39

Discovery of a Cryptic Intermediate in Late Steps of Mithramycin Biosynthesis ...... 41

2.1 Introduction ...... 41

2.2 Experimental Design and Results ...... 41

iii 2.2.1 MtmW Crystallization ...... 41

2.2.2 MtmW Polydispersity ...... 50

2.2.3 MtmOIV and MtmW PPI ...... 52

2.2.4 MtmW in vitro Assay ...... 56

2.2.5 iso-MTM NMR ...... 58

2.2.6 iso-MTM Bioassay ...... 59

2.2.7 Macromolecular Crowding ...... 62

2.2.8 Total Cell Assay ...... 66

2.2.9 Extracellular Isomerase ...... 69

2.2.10 iso-MTM Mg2+ Assay ...... 72

2.2.11 Whole Cell Isomerization ...... 76

2.2.12 iso-MTM Cell Extraction ...... 77

2.2.13 MtmW Substrate ...... 79

2.3 Conclusions and Future Directions ...... 83

2.4 Materials and Methods ...... 90

Bacterial strains and plasmids...... 90

Premithramycin B (PreB) production and isolation...... 91

Enzyme production and isolation...... 91

In vitro reaction...... 92

EMSA assay...... 93

iv Formaldehyde Crosslinking...... 93

MgCl2 isomerization assay...... 93

Antibiotic assays ...... 93

Analytical ultracentrifugation...... 94

Crystallization...... 95

Total cell assay...... 96

MtmW Substrate Assay...... 96

Interactions and Channeling of the Substrates in the Mithramycin Post-PKS Pathway ... 98

3.1 Background ...... 98

3.2 Experimental Design and Results ...... 99

3.2.1 19F NMR ...... 99

3.2.2 Protein Footprinting ...... 105

3.2.3 Label Transfer ...... 113

3.3 Conclusions and Future Directions ...... 116

3.4 Materials and Methods ...... 122

Enzyme production and isolation...... 122

Enzyme Labeling...... 123

Digestion...... 124

19F NMR...... 125

Label Transfer...... 125

v

LC/MS/MS ...... 125

References ...... 127

Vita ...... 138

vi

List of Tables

Table 2.1 Crystallographic statistics for MtmW and its complexes with ligands….…………………………………………………………………………….…..43

Table 3.1 A summary of the EDC/GEE labeling results. The IonScore is the is the (10)-

Log10 of the probability that the observed match is a random event…………………………………………………………………..…………….…..109

Table 3.2 A summary of the label transfer results. Score is the measure of how well the

MS/MS data matches the sequence of the proteins that were queried. Coverage is the percentage of the amino acids in the protein that were identified. PSMs are the number of

MS/MS spectra that were matched to the protein……………………………………………………………...…………………...116

vii

List of Figures

Figure 1.1 Mithramycin structure ...... 2

Figure 1.2 MTM dimer (yellow) coordinated around Mg2+ (green) and bound to DNA

(grey). Disaccharide, trisaccharide, and pentyl side chain present as indicated...... 3

Figure 1.3 Fli1 DNA binding domain (Orange) binding to DNA (Magenta)

(GACCGGAAGTG) (Protein Data Bank accession code: 5JVT) ...... 5

Figure 1.4 EWS-Fli1 translocation event ...... 5

Figure 1.5 Claisen condensation begins with enolate formation and nucleophilic attack on the carbonyl. The reformation of the ketone causes -OR to become a leaving group ...... 7

Figure 1.6 A general type II PKS reaction using malonyl as a starter and extender unit malonyl is added to the phosphopantotheine coenzyme by malonylacyltransferase

(MAT). B. The structure of the KS (Blue ribbon) and CLF (Teal spheres) in actinorohdin

PKS (actI-ORF1,ORF2). The catalytic cysteine (yellow spheres) is on the KS subunit, while the pocket that determines the number of iterations is formed by the magenta spheres on the CLF subunit. (Protein Data Bank accession code: 1TQY)27 ...... 9

Figure 1.7 Early MTM biosynthesis steps ...... 10

Figure 1.8 A. Minimal type I PKS. Additional domains can be present in each module to tailor the acyl unit they add to the chain. B. A general type III PKS...... 12

Figure 1.9 Early deoxysugar biosynthesis...... 13

Figure 1.10 The production of TDP-4-keto-2,6-dideoxy-D-olivose, a common intermediate in the biosynthesis of MTM deoxysugars...... 14

viii

Figure 1.11 The late steps of MTM deoxysugar biosynthesis. Note that MtmC catalyzes both the reduction and methylation of TDP-4-keto-2,6-dideoxy-D-olivose into TDP-D- olivose and TDP-D-4-keto-mycarose respectively (Red box and red arrows)...... 15

Figure 1.12 Proposed TDP-D-4-keto-mycarose cyclization reaction...... 16

Figure 1.13 A. MtmMI O-methylation (red). B. MtmMII aromatic C-methylation (red).

Additional enzyme and intermediate names omitted for clarity...... 17

Figure 1.14 Select aureolic acid compounds. C-7 functional groups highlighted in light blue...... 18

Figure 1.15 A. The structure of urdamycin A B. Novel compounds from the expression of

UrdGT2 in S. argillaceus mutants. 1. 9-C-olivosylpremithramycinone 2. 4-C-O- demethyl-9-C-olivosylpremithramycinone 3. 9-C-mycarosylpremithramycinone 4. 4-C-

O-demethyl-9-C-mycarosylpremithramycinone ...... 19

Figure 1.16 A. Inverting glycosyltransferase (GT) mechanism. B. Proposed Retaining GT mechanisms. GTs in SN1, SNi, and orthogonal mechanisms participate by stabilizing the intermediate ...... 21

Figure 1.17 MTM trisaccharide chain biosynthesis ...... 23

Figure 1.18 Disaccharide biosynthesis ...... 24

Figure 1.19 A. Schematic of the determination of the order of disaccharide glycosyltransferase enzymes. B. Proposed effect of knocking out MtmOIV if the enzymes form a multienzyme complex. Dashed line indicates a minor product...... 25

Figure 1.20 Results of the heterologous expression of L-digitoxose biosynthesis genes into S. argillaceus. 1. Demycarosyl-3D-β-D-digitoxosyl MTM 2. Deoliosyl-3C-α-L- digitoxosyl-MTM 3. Deoliosyl-3C-β-D-mycarosyl-MTM ...... 27

ix

Figure 1.21 Chromomycin A3. The acyl groups are marked with green boxes ...... 28

Figure 1.22 Acetylated compounds generated in the S. griseus feeding experiments. Note that the parent compound was MTM SK in compound 7. 1. 4E-O-acetyl-mithramycin 2.

Demycarosyl-3D-β-D-digitoxosyl-3E-O acetyl-mithramycin 3. Demycarosyl-3D-β-D- digitoxosyl-4D-O acetyl-mithramycin 4. Demycarosyl-3D-β-D-digitoxosyl-4E-O acetyl- mithramycin 5. Demycarosyl-3D-β-D-digitoxosyl-3E,4D-O acetyl-mithramycin 6.

Demycarosyl-3D-β-D-digitoxosyl-3E,4E,4D-O acetyl-mithramycin 7. 4D-O-acetyl- mithramycin SK ...... 29

Figure 1.23 The reactions of MtmOIV and MtmW ...... 31

Figure 1.24 MTM DK shunt product formation. All reactions begin with MTM DK and occur on the C-3 side chain (red) ...... 32

Figure 1.25 Baeyer-Villiger oxidation of PreB. FAD (blue) is reduced to FADH which in turn captures molecular oxygen. PreB (black) is attacked and a criegee intermediate is formed. Rearrangement and dissociation lead to PreB-lactone and hydroxylated FAD, the latter is regenerated by dehydration...... 33

Figure 1.26 A general aldol ketoreductase catalyzed reduction ...... 34

Figure 1.27 A MTM SA and its two most active amino acid conjugates, MTM SA-Trp and MTM SA-Phe. B. The coupling reaction between MTM SA and the amino acids

(green box)...... 36

Figure 1.28 Regulation of the MTM pathway (grey). MtmR (green, arrows) acts as a transcription activator for the pathway but its expression is repressed by MtrY (blue, inverted arrows). However when bound to MTM (orange), MtrY ceases to act as a repressor and switches to an transcription activator (blue arrows) for the operon

x containing MtmR. This leads to increased expression of MTM. This figure was adapted from Florez et. al. 2015 ...... 38

Figure 1.29 A. A hypothetical quorum system. Quorum system (QS) molecules (orange) and the transcriptional activator (blue) are produced at a low basal rate. When the concentration of QS molecules is sufficient, they bind to the activator causing it to begin activating transcription of the gene cluster (grey), the QS molecules, and the activator itself...... 39

Figure 2.1 A monomer of MtmW. A view into the TIM barrel is shown, with the short N‐ terminal β-sheet capping the barrel. The N‐and C‐termini are labeled as “N” and “C”, respectively...... 42

Figure 2.2 Two orthogonal views of the MtmW octamer of the MtmW‐NADP+ complex.

A) The view into the inter-subunit channel. B) The view where the channel is oriented along the vertical axis. The bound NADP+ is shown by purple sticks...... 44

Figure 2.3 The crystal structure of MtmW‐NADP+‐PEG complex. NADP+ is shown by purple sticks, PEG (orange sticks) is well defined by the polder omit electron density map, contoured at 5.5σ (shown as a mesh). The residues in the putative substrate binding site are shown as sticks with the labels. The orientation of this MtmW monomer is the same as that of the similarly colored monomer in Figure 2.2a ...... 45

Figure 2.4 The sequence alignment of MtmW and other structurally characterized representatives from the AKR family. These enzymes are: AKR_P.sp.JS666: AKR from

Polaromonas sp. JS666, Kvb_rat: Kvb from Rattus norvegicus, STM2406: 3- hydroxybutanal reductase STM2406 from Salmonella enterica serovar Typhimurium,

Cbei_3974: AKR Cbei_3974 from Clostridium beijerinckii, YghZ_E.coli: YghZ from

xi

Escherichia coli, AKR_R.serpent: perakine reductase from Rauvolfia serpentina.

Identical residues are shown in red boxes and homologous residues are shown in red font.

The tetrad of active site residues conserved in the aldo-keto reductase (AKR) superfamily59 is shown by the red lollipops, and the functionally validated residues of the substrate binding site in the structurally similar STM240673 are shown by the green lollipops. The ketone oxygen of the substrate is predicted to abstract a proton from Tyr57 in the reduction mechanism.74...... 46

Figure 2.5 Views of the active sites of MtmW and STM2406. A. The active site of

MtmW. The protein is in a similar orientation to that in Figure 2.3. B. The active site of

STM2406 (PDB ID: 3ERP73). A cacodylate ion (yellow sticks), a 1,2-ethanediol (orange sticks) and a sodium ion (a green sphere) are bound in the substrate binding pocket of

STM2406, defined by the conserved NADPH site, the conserved catalytic Tyr66 and the residues whose mutations perturbed the substrate profile (Trp33, Asn65, Tyr100,

Asn169).73 C. The active site of STM2406 in the crystal structure of STM2406-NADP+ complex (PDB ID:5T7973). NADP+ is shown as purple sticks...... 47

Figure 2.6 A zoomed-in view of the active site of MtmW in complex with NADP+ and a

PEG region. The color scheme is the same as in Figure 2.5a ...... 49

Figure 2.7 A comparison of the structures of MtmW-NADP+ complex (this study) and

Kvβ in complex with the T1 domain (PDB ID: 1EXB15, 75). A. An octamer of MtmW is shown in the same view as in Figure 2.2b. B. A tetramer of Kvβ-T1-NADP+ complex.

The Kvβ tetramer is oriented similarly to the top tetrameric layer of MtmW in panel A.

The bound NADP+ is shown as purple sticks in both panels...... 50

xii

Figure 2.8 Three factors that can effect polydispersity. A. Substrate-mediated conformational changes. B. PPIs with other enzymes. C. The equilibrium between the enzyme’s product and substrate...... 52

Figure 2.9 The result of formaldehyde (FM) crosslinking between MtmOIV and MtmW.

A. 1. Ladder 2. MtmOIV + MtmW 3. MtmOIV +MtmW + FM 4. MtmW + FM 5.

MtmOIV + FM 6. MtmW + TMK + FM 7. MtmOIV + TMK + FM 8. TMK + FM B.

Close up of the top of wells 1-5. Note the aggregation present in 3 and 5 (Black arrows)

...... 53

Figure 2.10 A time course experiment with MtmOIV, MtmW and FM. 1. Ladder 2. 1 hr

3. 2 hr. 4. 3 hr 5. Overnight. Note the aggregation at the top of each well (black arrow) 54

Figure 2.11 A. A Coomassie blue stained 1.7 % agarose gel showing the EMSA with

MtmOIV (20 μM in monomers, 10 μM in dimers) and MtmW (20 μM in monomers, 2.5

μM in octamers). The smear above the MtmW band in lanes 3–6 indicated that MtmW was retarded by binding MtmOIV. B. A similar gel showing a titration of MtmW (10

μM) with MtmOIV. The fraction of free MtmW is shown below each lane...... 56

Figure 2.12 HPLC chromatograms of reactions catalyzed by MtmW and MtmOIV at pH

8.25. A) Pure PreB as standard; B) pure MTM as standard; C) PreB + MtmOIV + MtmW

+ FAD + NADPH; D) PreB + MtmOIV + FAD + NADPH; E) PreB +MtmOIV + boiled

MtmW + FAD + NADPH; F) PreB + MtmW + FAD + NADPH. Compounds 1, 2, and 3 are PreB, MTM, and iso-MTM, respectively...... 57

Figure 2.13 Distinction between iso-MTM and MTM through NMR...... 59

Figure 2.14 Bioactivity of iso-MTM. A. Disk diffusion assay. Zones of inhibition are marked with black circles, disks are marked with grey circles. B. The IC50 assay with

xiii

TC-32 cells. C. The MIC assay. All experiments were done in triplicate, and the data shown represents the average of the experiments...... 61

Figure 2.15 Fate of iso-MTM in MIC samples...... 62

Figure 2.16 Transition state-limited rate constant and Diffusion-limited rate constant. A.

At low concentrations of crowding agent (grey) the local concentration enzyme (red) increases. Addition of crowding agent causes the enzymes to “crowd together” increasing

PPIs (solid arrow). B. At high concentrations of crowding agent, diffusion and the rate at which the enzymes interact are reduced (dashed arrow). Addition of crowding agent decreases diffusion and decreases PPIs...... 63

Figure 2.17 The influence of crowding agent concentration on PPI and enzyme reaction rates. Addition of crowding agent increases interaction (orange dashed line) until its concentration begins to hamper diffusion (green dashed line). Overall rate of interaction and reaction is shown by the black line...... 64

Figure 2.18 The influence of crowding agent concentration on iso-MTM production.

Reactions were performed as in section 2.2.4 and analyzed by LC/MS. At all concentrations of PEG, iso-MTM (orange) is the major product...... 66

Figure 2.19 Two bifunctional enzymes in the MTM Pathway. A. MtmGIV (blue) adds the

C and E sugars to the trisaccharide chain B. MtmC (red) acts as a methyltransferase/ ketoreductase in deoxysugar biosynthesis...... 68

Figure 2.20 Select results from the total cell assay experiments. S. argillaceus ΔmtmGI and S. argillaceus ΔmtmGIII lysates were used as enzyme sources for the PreB to iso-

MTM reaction. In all experiments, iso-MTM was the major product...... 69

xiv

Figure 2.21 iso-MTM (blue) with S. argillaceus and S. aureus cell-free supernatant after two days incubation at 30 °C. In both cases the majority of iso-MTM epimerized into

MTM (green) ...... 71

Figure 2.22 Proposed mechanism for iso-MTM to MTM epimerization...... 72

Figure 2.23 LC/MS analysis of the isomerization of iso-MTM in the presence of 50 mM

Tris-HCl pH 8.0 and different concentrations of MgCl2 ...... 75

Figure 2.24 Extracts of a 24 hr culture of S. argillaceus. A. Supernatant butanol extract.

B. Cell pellet butanol extract...... 75

Figure 2.25 Proposed mechanism for iso-MTM to MTM epimerization. Updated to include the influence of Mg2+ ...... 76

Figure 2.26 Isomerization by intact cells...... 77

Figure 2.27 MTM is the only compound from the MTM pathway found in S. argillaceus

WT cultures. This suggests that there is a complex between the PKS (blue) enzymes and the post-PKS (green) enzymes. Intermediates such as premithramycin A2 are never seen, which implies that MTM intermediates are channeled through the entire MTM pathway because discrete enzymes would require pools of intermediates from which to draw. .... 79

Figure 2.28 A. The appearance of the m/z 1183 peak from the MtmOIV-MtmW in vitro reaction as the concentration of formic acid (FA) in the mobile phase was reduced. B.

The m/z 1183 peak is abolished when the buffer in the in vitro reaction is switched from

Tris to MOPS...... 81

Figure 2.29 The proposed mechanism of the formation of the MTM DK-Tris ketal adduct...... 82

xv

Figure 2.30 Isolated MTM DK-Tris adduct was used as a substrate for MtmW in a range of pHs. As the pH dropped, the ketal formation was reversed, allowing MtmW to use

MTM DK as a substrate, as shown by the formation of iso-MTM. MTM SA and other

MtmOIV shunt products (not shown) remained major products at every tested pH...... 82

Figure 2.31 A. As inactive iso-MTM passes through the channel formed by MtrAB, it is enzymatically isomerized to active MTM as a form of self-protection. B. However, active

MTM (orange) is required for the production of more MTM via interaction with MtrY.

This, it would seem, counters iso-MTM’s role as a self-defense mechanism.

Translocation of MTM from the environment into the cytosol is likely not mediated solely by MtrAB, as non-producing S. argillaceus strains become sensitive to MTM when mtrA is inactivated.65 ...... 85

Figure 2.32 A cell produces a signal molecule that induces gene set A. When that signal molecule is altered by environmental conditions, it induces gene set B. The environmental conditions relevant to iso-MTM and MTM according to our research are pH changes and transition elements. Figure adapted from Decho et. al.110 ...... 87

Figure 2.33 Proposed possible alternative ecological roles of iso-MTM and MTM

(orange). At low divalent metal concentrations (Mg2+shown here), making a lethal complex through chelation by MTM prevents its use from competing species (green). If iso-MTM rather than MTM is exported from the cell, its isomerization by divalent metal may be detected by S. argillaceus (blue) through its ability to bind MtrY. At high concentrations, MTM may protect S. argillaceus from divalent metal poisoning ...... 88

Figure 3.1 19F NMR of TFAA labeled MtmW with and without MtmOIV ...... 103

xvi

Figure 3.2 A. TFAA labeling reaction. B. Proposed mechanism of TFAA label disassociation...... 103

Figure 3.3 A-C. Three sides of the MtmOIV dimer showing the proposed cysteine mutations (magenta). * indicates nearest amino acid if the mutation fell on a flexible loop that was not included in the crystal structure. PreB is shown as green sticks. D. The 3- bromo-1,1,1-trifluoroacetone (BTFA) reaction with cysteine...... 104

Figure 3.4 PEG contamination in a DEPC labeled trypsin digest of MtmW. Red arrows highlight 44 Da intervals indicating PEG contamination...... 107

Figure 3.4 Labeling of MtmOIV. Red spheres are amino acids that are labeled only when

MtmOIV is alone. Yellow spheres are only labeled when MtmOIV is in contact with

MtmW. Green spheres are labeled in both conditions. Supposed PPI interface involving

D94 and D99 (black arrows) ...... 110

Figure 3.5 Labeling results that question the established MtmOIV dimer. E75 and E119

(green) appear to be blocked from labeling by the MtmOIV dimer interface. E123 (red) is only labeled when MtmOIV is alone in solution, and E425 and E431 (yellow) are labeled only when MtmOIV is in solution with MtmW. For colors see Figure 3.4 ...... 111

Figure 3.6 Labeling of MtmW. Red spheres are amino acids that are labeled only when

MtmW is alone. Green spheres are labeled in all conditions. A. Proposed PPI interface B.

Proposed PPI interface. Green amino acids in this area have lower IonScores when

MtmOIV is in solution...... 112

Figure 3.7 Two views of the MtmW octamer with one monomer removed for clarity. A.

The face of MtmW that we believe forms a PPI with MtrAB has no adjacent primary

xvii amines (magenta) B. The remainder of the primary amines are clustered around the active site defined here by the co-crystallized NADPH and polyethylene glycol (orange)...... 114

Figure 3.8 Three angles on the proposed PPI between MtmOIV (blue) and the MtmW octamer (grey). This Figure only shows a single MtmOIV for clarity. Red spheres are amino acids that are labeled only when MtmOIV or MtmW are alone. Yellow spheres are only labeled when MtmOIV is with MtmW. Green spheres are labeled in all conditions.

Note: This Figure was docked by hand, the only data that were considered were the proximity of the active sites, and the results from the label transfer and protein footprinting experiments...... 118

Figure 3.9 Detail on the active sites of MtmOIV and MtmW in the proposed PPI model.

MTM DK travels from the active site of MtmOIV (blue) to the active site of MtmW

(grey). The magenta sticks are co-crystallized PreB with MtmOIV and co-crystallized

NADPH and PEG with MtmW...... 119

Figure 3.10 The Kv1 potassium ion channel (PMDB structure: 2A79). The Kvβ subunit

(green) is structurally homologous to MtmW. The remainder of the Kv1 ion channel

(purple) may also be homologous to the MTM pathway enzymes which are listed on the right. The proposed MtmOIV binding site on the left (red box) would fit into this system in a manner where is is accessible to MtrA, which is in agreement with our label transfer data ...... 120

Figure 3.11 A schematic of the proposed MTM trisaccharide synthesis complex...... 121

Figure 3.12 A schematic of the proposed MTM complex involving MtmGI-MtrB. MtmW is shown as a tetramer for clarity, we currently hypothesize that MtmW remains an octamer in vivo...... 122

xviii

List of Acronyms

7-DMTM 7-demethylmithramycin ACP Acyl carrier protein ARK Aldo‐keto reductases AT Acetyl transfer AUC Analytical ultracentrifugation BD 2,3-butanedione BTFA 3-bromo-1,1,1-trifluoroacetone BVMO Bayer-Villiger monooxygenase CLF Chain length factor CMA3 Chromomycin A3 DEBS 6-deoxyerythronolide B synthase DEPC Diethylpyrocarbonate DH Dehydratase EDC N-(3-(dimethylamino)propyl)-N′-ethylcarbodiimide hydrochloride EIC Extracted ion chromatogram ER Enoylreductase EWS-Fli1 Ewing sarcoma breakpoint region 1-Friend leukemia integration 1 transcription factor FA Formic acid FAS Fatty acid Synthase FPA Fuculose 1-phosphate aldolase G3PD Glyceraldehyde-3-phosphate dehydrogenase GEE Glycine ethyl ester GT Glycosyltransferase HDX Hydrogen deuterium exchange KR Ketoreductase KS Ketoacyl synthase KV1 Shaker family of voltage‐dependent potassium channels Kvβ Kv1 β subunit LB Lysogeny broth MAT Malonylacyltransferase MMC Macromolecular crowding MTM Mithramycin MTM DK Mithramycin (di-keto C-3 side chain) MTM SA Mithramycin (short acid C-3 side chain) MTM SDK Mithramycin (short di-keto C-3 side chain)

xix

MTM SK Mithramycin (short keto C-3 side chain) NHS N-hydroxysuccinimide NOE Nuclear Overhauser Enhancement PDB Protein Data Bank PEG Polyethylene glycol PKS Polyketide synthase PPI Protein-protein interaction PPIase Peptidyl prolyl cis, trans-isomerase PreB Premithramycin B QS Quorum system Sulfo-SBED Sulfo-N-hydroxysuccinimidyl-2-(6-[biotinamido]-2-(p-azido benzamido)-hexanoamido SV Sedimentation velocity TDP Thiamine diphosphate TE Thioesterase TFAA Trifluoroacetic anhydride TMK Thymidylate kinase TNBSA 2,4,6-trinitrobenzene sulfonic acid TUNEL Terminal deoxynucleotidyl transferase-mediated dUTP nick end labeling

xx

Background Information

1.1 Mithramycin Structure and Activity

Mithramycin (MTM) is a type II polyketide produced by Streptomyces argillaceus. MTM has antimicrobial activity against Gram positive bacteria, which suggests that its purpose is to suppress competition from other bacteria in the soil, the environment where S. argillaceus is found. MTM has previously been used in the treatment of hypercalcemia, testicular cancer, and Ewing’s sarcoma.1-3 Unfortunately, the prescription of MTM was halted due to hepatic, gastrointestinal, bone marrow, and renal toxicities.4 In 2011, the National Cancer Institute performed a high-throughput screening for activity against the TC-32 cell line, which carries the EWS-Fli1 fusion protein that is likely responsible for Ewing’s sarcoma. Of the 50,000 compounds screened, MTM was the most active, reducing the transcription of the luciferase reporter at concentrations as low as 10 nM.5, 6 Because of its toxicity and its potential for inhibiting the transcription process leading to Ewing’s sarcoma, MTM is a prime lead candidate to treat Ewing’s sarcoma, and a broad derivatization program evolved, initially focusing on combinatorial biosynthesis, then using medicinal chemistry derivatization strategies.

MTM is an aureolic acid antibiotic. It consists of an aromatic, tricyclic aglycon, a

C-3 pentyl side chain, and two oligosaccharide chains, a C-2-O attached trisaccharide which is formed by β(1-3) linked D-deoxyhexoses, an oliose, an olivose, and a mycarose, and the C-6 attached disaccharide, which is made up of two β(1-3) connected olivoses.

MTM also has C-7 C-methyl and C-1’ O-methyl groups. All of the keto groups remaining from the polyketide biosynthesis are reduced, save for the C-1 and C-2’

1 carbonyls, while the C-8, C-9, C-3’, and C-4’ carbonyls were reduced to hydroxyl groups

(Figure 1.1).7

Figure 1.1 Mithramycin structure

1.1.1 MTM as an SP1 inhibitor

MTM binds to regions of GC rich DNA in a non-intercalative fashion. Its pentyl side chain and disaccharide interact with the DNA backbone and the aglycone and trisaccharide chain fit into the minor groove. In binding DNA, MTM forms a dimer coordinated around a divalent metal ion such as Mg2+ and DNA recognizing a minimum

5'-X(G/C)(G/C)-3' sequence.8 The divalent metal is bound to the O-1 carbonyl and the O-

9 enolate anion of the aglycone on each of the MTM monomers. NMR experiments on the binding of the MTM dimer with Mg2+ on GC rich DNA show that the C-9 hydroxyl groups form hydrogen bonds with amine group of guanidine in the recognition sequence which gives MTM some of its binding specificity (Figure 1.2).8

2 Figure 1.2 MTM dimer (yellow) coordinated around Mg2+ (green) and bound to DNA

(grey). Disaccharide, trisaccharide, and pentyl side chain present as indicated.

A result of the binding of MTM to DNA in human cells is the inhibition of the transcription factor SP1. SP1 governs a large number of genes, including regulating cyclins, and pro–oncogenes like CLDN4 and MDM2, which are implicated in cancer genesis or progression. Using a zinc finger motif, SP1 has a recognition sequence of 5'-

(G/T)GGGCGG(G/A)(G/A)(C/T)-3' on DNA9, a sequence which shares several features with the 5'-X(G/C)(G/C)-3' minimum recognition sequence of MTM. For this reason,

MTM is a potent competitive inhibitor of SP1 binding. In an RT-PCR assay, 0.5 μM of demycarosyl-3D-β-D-digitoxosyl-mithramycin SK (EC-8042), an MTM analogue obtained through combinatorial biosynthesis, was sufficient to inhibit the C-MYC and

XIAP, two downstream targets of SP1, to less than 10% of their wild type expression.10

The SP1 inhibition by MTM has led to its use as a treatment for testicular carcinoma, and

3 has also had success in inhibiting other cancers, such as colon cancer and pancreatic cancer when used alone or in conjunction with other drugs.11, 12

1.1.2 MTM as an EWS-Fli1 inhibitor

In addition to blocking the SP1 transcription factor, MTM also has activity toward

Ewing’s sarcoma through an alternate mechanism of action. Ewing’s sarcoma is the second most common bone cancer. It affects 1-3 million people per year in western countries, and most of the patients are children. It has a 50% 5 year survival rate which drops below 30% for 10 years.13 Ewing’s sarcoma is a cancer caused by an aberrant translocation between chromosomes 11 and 22, which forms the fusion protein, EWS-

Fli1 (Ewing sarcoma breakpoint region 1-Friend leukemia integration 1 transcription factor) (Figure 1.3). In the newly generated fusion protein, EWS is the N-terminal portion and it forms a transactivation domain, a domain that increases gene expression. EWS acts as a transactivator through its association with other transcription factors, such as TFIID, as well as with mRNA and RNA splicing enzymes. On the C-terminal side, Fli1 is a transcription factor that regulates several genes, many of which are pro-oncogenes that are expressed during development (Figure 1.4). For example, Fli1 can repress the transcription retinoblastoma protein, a tumor suppressor that inhibits a cell from entering the S stage of the cell cycle from G1. When fused, EWS-Fli1 is more active than Fli1 alone. This leads to abnormal gene expression and the Ewing’s sarcoma phenotype.13, 14

4 Figure 1.3 Fli1 DNA binding domain (Orange) binding to DNA (Magenta)

(GACCGGAAGTG) (Protein Data Bank accession code: 5JVT)

Figure 1.4 EWS-Fli1 translocation event

5 MTM binds to a GGAA repeat sequence that also makes up the recognition sequence for EWS-Fli1. In binding to this sequence, MTM also alters EWS-Fli1 binding and preventing transcription. In 2011, MTM was the best of 50,000 compounds screened against the Ewing’s sarcoma cell line TC-32 in a high throughput assay, with a mean IC50 of 15 nM. Further, the MTM analogue EC-8042, had comparable activity to MTM in regards to EWS-Fli1 suppression but was 130 and 32 fold less toxic, respectively, depending on whether the method of administration was intravenous or intraperitoneal.5,

15-17

1.1.3 MTM and Hypercalcemia

MTM also has been used in the treatment of hypercalcemia, particularly in association with Paget's disease. 18 Hypercalcemia is a condition where there is an unusually high concentration of calcium in the blood serum, which can lead to cardiac arrest or coma. It has been shown that hypercalcemia is a symptom of cancer in 5-30% of patients depending on cancer type. Hypercalcemia in cancer is typically caused by overexpression of parathyroid hormone, a hormone that normally maintains calcium homeostasis by stimulating osteoclast activity.18, 19 A study by Perlia et. al. involving 32 patients with hypercalcemia brought upon by cancer, found that a 25 µg/kg dose of MTM significantly reduced calcium serum levels, and repeated doses stabilized levels to the normal range. The dose given in this study was a degree of magnitude lower than that given in cancer treatment, indicating an alternative mechanism than simple tumor suppression.20 A study of MTM on the treatment of hypercalcemia in dogs showed similar results. Not only did MTM lower serum calcium levels, it also inhibited osteoclastic activity and, increased parathyroid hormone levels.21 While the exact

6 mechanism of action of MTM in the treatment of hypercalcemia is unknown, given its mechanism of action in SP1 and EWS-Fli1 inhibition it is likely due to an alteration of transcription levels in osteoclasts.20, 21

1.2 Mithramycin Biosynthesis

1.2.1 MTM PKS

Polyketide biosynthesis is analogous to fatty acid biosynthesis in primary metabolism as both catalyze the polymerization of acyl groups. The acyl groups, commonly acetate and malonate are first activated by coenzyme A. The activated acyl groups are joined by a series of Claisen reactions resulting a poly-β-keto ester. During the

Claisen reactions, malonyl-CoA is decarboxylated so that a single acetate unit is added to the polymer, the additional CO2 only being necessary to increase the nucleophilicity of the α-carbon of malonyl-CoA and aid the Claisen reaction (Figure 1.5). In fatty acid synthesis, carried out by fatty acid synthase (FAS), each carbonyl is reduced as the polymer is synthesized. However in polyketide synthesis this is not the case. The fate of the poly-β-keto ester in polyketide synthesis is determined by the type of polyketide synthase (PKS) in question.22, 23

Figure 1.5 Claisen condensation begins with enolate formation and nucleophilic attack on the carbonyl. The reformation of the ketone causes -OR to become a leaving group

7 Mithramycin is a type II polyketide, which means that it is produced in an iterative fashion by a PKS that contains an KSα and a KSβ subunit, and an acyl carrier protein (ACP), which in contrast to type I PKSs remains attached throughout the elongation. First, the phosphopantotheine functional group of ACP is primed with the acyl starter unit. In MTM biosynthesis, this is acetate, but several other acyl groups can be used. For example, malonamyl is the starter unit for tetracycline biosynthesis, while doxorubicin starts with a propionyl unit.24 To this tethered starter unit, 9 acetate extender units are added by Claisen condensation, and the poly-β-keto ester of MTM is formed.

The KSα subunit, MtmP in the MTM pathway, catalyzes the addition of the extender units, typically using malonyl-CoA as a substrate, which decarboxylates during the reaction facilitating the condensation while adding a net total of an acetate subunit

25 (Figure 1.6a). The length of the MTM poly-β-keto ester is determined by the KSβ subunit, also known as the chain length factor (CLF) and was named MtmK in the MTM biosynthetic pathway. The CLF determines the length of the polyketide through the size and geometry of the cavity where the polyketide is formed (Figure 1.6b). The growing acyl chain is modified by a ketoreductase (MtmTI), cyclase, and aromatase (MtmQ) and an oxygenase (MtmOII) to form the product, 4-demethylpremithramycinone (Figure

1.7).25, 26

8 Figure 1.6 A general type II PKS reaction using malonyl as a starter and extender unit malonyl is added to the phosphopantotheine coenzyme by malonylacyltransferase

(MAT). B. The structure of the KS (Blue ribbon) and CLF (Teal spheres) in actinorohdin PKS (actI-ORF1,ORF2). The catalytic cysteine (yellow spheres) is on the

KS subunit, while the pocket that determines the number of iterations is formed by the magenta spheres on the CLF subunit. (Protein Data Bank accession code: 1TQY)27

9 Figure 1.7 Early MTM biosynthesis steps

In contrast to type II PKSs, type I PKSs consist of discrete modules. An acetyl transfer (AT) domain loads the acyl group onto the ACP domain on a serine-attached phosphopantotheine, and the ketoacyl synthase (KS) catalyzes the Claisen reaction that elongates the acyl polymer. This is a single, minimal module. The module can also include ketoreductase (KR), dehydratase (DH), or enoylreductase (ER) domains to introduce variety to the molecule. The growing chain is passed from module to module, until it is released from the PKS by a thioesterase (TE) domain in the final module

(Figure 1.8a). For example, 6-deoxyerythronolide B synthase (DEBS) produces the

10 precursor to erythromycin and is a representative of type I PKSs. It contains six modules each adding methylmalonyl-CoA CoA. Modules 1 and 2 each have KR domains that reduce the carbonyl on the incoming acyl group, module 3 lacks a KR domain leaving the carbonyl intact. Module 4 contains the KR domain as well as DH and ER domains which results in the complete reduction of the added acyl group. Modules 5 and 6 are both minimal modules with the exception of the TE domain at the end of module 6 which cyclizes the product into the 14 member macrolactone 6-deoxyerythronolide B.28, 29 Type

III PKSs, differ from types I and II, in that they lack an ACP domain. They rely on acyl

CoA substrates directly and are commonly found as homodimers. As in the case of type

II PKSs, Type III PKS’s act iteratively, and typically produce small molecules compared to the other PKS types (Figure 1.8b). In addition to the canonical PKS types, there are hybrid systems such as type I-type II, and PKS-NRPS.30, 31

11 Figure 1.8 A. Minimal type I PKS. Additional domains can be present in each module to tailor the acyl unit they add to the chain. B. A general type III PKS.

1.2.2 Deoxysugar Biosynthesis

Deoxysugar moieties are a common feature in secondary metabolites and polyketides are no exception. The role of deoxysugars in polyketides is molecular recognition, increasing the binding and specificity to their target. Additionally, many glycosyltransferases, the enzymes responsible for adding the deoxysugars to secondary metabolites, are promiscuous. This makes deoxysugars and their associated glycosyltransferases a good tool for combinatorial biosynthesis, the derivatization of natural products using genetic manipulations.32 In general, deoxysugar biosynthesis begins with the addition of the nucleotide group thiamine diphosphate (TDP) to glucose-

12 1-phosphate. This sugar is then oxidized at its C-4 position and dehydrated at the C-2 and

C-6 positions (Figure 1.9). From here the biosynthesis branches wildly, with isomerizations, decarboxylations, transaminations, ketoreductions, and other reactions that give rise to over 90 different deoxysugars.23, 33-35

Figure 1.9 Early deoxysugar biosynthesis.

The role of deoxysugars in MTM is the recognition of GC-rich DNA. While we have no crystal structures of MTM and DNA, we do have structures of DNA with MTM analogues. These structures reveal that all the deoxysugars on the MTM analogue participate in either hydrophobic interactions or hydrogen bonding with DNA depending on the sugar moiety in question and the DNA sequence to which it is bound. Here I will describe the biosynthesis of the natural deoxysugars in MTM, and later the glycosyltransferases and associated combinatorial biochemistry associated with them.15, 36

In the course of the post-PKS tailoring steps, MTM is glycosylated five times with three different deoxysugars. The biosynthesis of these sugars begins with glucose-1- phosphate to which a TDP group is added by MtmD to produce TDP-D-glucose.

Following this, MtmE, a TDP-D-glucose 4,6 dehydratase, oxidizes the substrate and reduces NAD+ to form TDP-D-4-keto-6-deoxyglucose and NADH respectively (Figure

1.9).34 Next MtmV and an unknown enzyme act as a 2,3 dehydratase and C-3

13 ketoreductase, respectively, resulting in the loss of the C-2 hydroxy group (Figure 1.10).

From this step, the sugar biosynthetic paths diverge, MtmU reduces the C-4 keto group of the TDP-D-4-keto-2,6-deoxyglucose such that the newly formed C-4 hydroxyl is in the R stereochemistry giving TDP-D-oliose as a product. MtmC is unusual in that it is a bifunctional enzyme. It acts as a C-3 methyltransferase, which results in TDP-D-4-keto- mycarose, which is later reduced to TDP-D-mycarose by MtmTIII. MtmC also operates as a 4 ketoreductase, which reduces the C-4 hydroxyl in the S stereochemistry giving

TDP-D-olivose (Figure 1.11).34, 37, 38

Figure 1.10 The production of TDP-4-keto-2,6-dideoxy-D-olivose, a common intermediate in the biosynthesis of MTM deoxysugars.

14 Figure 1.11 The late steps of MTM deoxysugar biosynthesis. Note that MtmC catalyzes both the reduction and methylation of TDP-4-keto-2,6-dideoxy-D-olivose into TDP-D- olivose and TDP-D-4-keto-mycarose respectively (Red box and red arrows).

MtmC is also noteworthy because it processes a labile product in TDP-D-4-keto- mycarose. The addition of the methyl group at C-3 forces the hydroxyl group into the axial position, putting it in close proximity to the axial TDP group on C-1. The proximity of these two groups encourages a nucleophilic attack from the C-3 hydroxyl to the phosphate group of the TDP, leading to a cyclized shunt product that is unable to be used by MtmGIV as a substrate. For this reason, there is likely to be an interaction between

MtmC and MtmGIV, so that TDP-D-4-keto-mycarose can be channeled into the

MtmGIV active site in a way that preserves its structure (Figure 1.12).38, 39

15 Figure 1.12 Proposed TDP-D-4-keto-mycarose cyclization reaction.

1.2.3 Methylation

4-demethylpremithramycinone, the aglycone of MTM is methylated in two positions. First, it is O-methylated in on the C-4 hydroxyl group resulting premithramycinone. This is an essential step in MTM biosynthesis, as the unmethylated product is not be used as a substrate for other MTM enzymes in vivo. The second methylation is a C-methylation at the aromatic C-9 position of premithramycinone. The timing on this methylation is more ambiguous, in vivo knockout studies have shown that demethylated MTM intermediates can be isolated at any point following the first glycosylation in the trisaccharide synthesis, up to and including 7-demethylmithramycin

(7-DMTM), note that later biosynthetic steps alter the numbering of MTM intermediates

(Figure 1.13).40

16 Figure 1.13 A. MtmMI O-methylation (red). B. MtmMII aromatic C-methylation (red).

Additional enzyme and intermediate names omitted for clarity.

Because post-PKS steps following MtmMII methylation are promiscuous enough to proceed with an unmethylated product, it became a good target for combinatorial biosynthesis. Another member of the aureolic acid family is olivomycin A (Figure

1.14b), it has a similar aglycone to MTM (it lacks a C-7 methyl group), but has a different glycosylation pattern and lacks the C-7 methyl group, and has a greater chemotherapeutic index than MTM. Unfortunately 7-DMTM exhibits no bioactivity towards Micrococcus luteus even at concentrations as high as 200 μg/ml, the same study showed MTM started inhibiting growth at 5 μg/ml.41, 42 Further, while 7-DMTM accumulated in the nucleus of HeLa cells like MTM, it has as much as a 200-fold loss in affinity to GC DNA sequences compared to MTM. It was thought that the C-7 methyl

17 group was necessary for the proper positioning of the mycarose sugar on the trisaccharide chain in the minor grove of DNA. The loss of the C-7 methyl group proved to correlate with a loss of activity. Durhamycin and UCH9 are two members of the aureolic acid family with sec-butyl groups at this position, but they are less active as chemotherapeutic drugs than MTM (Figures 1.14c, 1.14d). The C-7 sec-butyl group appears to push aside the sugars from the minor grove of DNA, and thereby decreases the amount of perturbation in the DNA molecule. To my knowledge, there have been no attempts of introduce this group onto an MTM analogue.43

Figure 1.14 Select aureolic acid compounds. C-7 functional groups highlighted in light blue.

In addition to alkyl groups, aromatic C-glycosylations are possible in type II polyketides. Urdamycin A (Figure 11a), for example, has an oligosaccharide chain that begins with a C-glycosidically linked D-olivose. Urdamycin A has both antitumor and antibacterial properties. Expressing the urdamycin C-glycosyltransferase UrdGT2 in S.

18 argillaceus ΔmtmMII, ΔmtmGIV and S. argillaceus ΔmtmMII, ΔmtmGI-GIV resulted in four new, C-glycosylated compounds, 9-C-olivosylpremithramycinone, 9-C- mycarosylpremithramycinone, and C-4 demethyl versions of both. These novel compounds, however, showed little to no activity against the three tested cancer cell lines

(Figure 11b).44-46

Figure 1.15 A. The structure of urdamycin A B. Novel compounds from the expression of UrdGT2 in S. argillaceus mutants. 1. 9-C-olivosylpremithramycinone 2. 4-C-O- demethyl-9-C-olivosylpremithramycinone 3. 9-C-mycarosylpremithramycinone 4. 4-C-

O-demethyl-9-C-mycarosylpremithramycinone

19 1.2.4 Glycosylation

Glycosyltransferases catalyze the addition of sugar moieties to molecules. In secondary metabolism, the deoxysugars they transfer are responsible for molecular recognition and are often necessary for bioactivity. The deoxysugars that glycosyltransferases use as a substrate are usually activated by attachment of a nucleotide group, TDP in MTM biosynthesis, at the C-1 anomeric carbon.23 Glycosyltransferases attach the sugar to nucleophiles of secondary metabolites using either an inverting or a retaining mechanism. The inverting mechanism, of which all MTM glycosyltransferases use, uses an SN2 nucleophilic attack on the anomeric carbon (Figure 1.16a). The nucleotide acts as a leaving group. This gives the deoxysugar inverted stereochemistry in regards to the attachment of the nucleotide and the substrate. The retaining mechanism preserves the stereochemistry of the anomeric carbon. There are many theorized mechanisms for retaining glycosyltransferases, two consecutive SN2 reactions known as double displacement, SN1, SNi, and orthogonal mechanisms have been proposed (Figure

1.16b).47, 48

20 Figure 1.16 A. Inverting glycosyltransferase (GT) mechanism. B. Proposed Retaining

GT mechanisms. GTs in SN1, SNi, and orthogonal mechanisms participate by stabilizing the intermediate

The glycosyltransferases in MTM biosynthesis can be divided into two groups, those that are responsible for the trisaccharide chain and those responsible for the disaccharide chain. In the order of the biosynthesis, MtmGIV acts first adding a D- olivose to the 12a position of premithramycinone. MtmGIV is unusual in that it catalyzes the addition of the first and third deoxysugar of the trisaccharide chain, D-olivose and D- mycarose, respectively. However, unpublished research from our lab suggests that

21 MtmGIV transfers 4-keto-D-olivose in both positions after which MtmC and MtmTIII finalize the chemistry while the deoxysugars are part of an MTM intermediate. Acting in the step between the MtmGIV glycosylations, MtmGIII adds D-oliose in the second position of the trisaccharide chain. The activity of these MTM analogues would be low, as both the trisaccharide chain and pentyl side chain are vital for MTM to bind in the minor groove of DNA, and MTM intermediates will not proceed through further in the biosynthetic pathway to produce the pentyl side chain with an incomplete trisaccharide chain, according to gene inactivation experiments (Figure 1.17).39, 49

22 Figure 1.17 MTM trisaccharide chain biosynthesis

After the completion of the trisaccharide chain, MtmGI and MtmGII each add a D- olivose to form the disaccharide chain at the C-8 position of premithramycin A3 (Figure

1.18). Interestingly, inactivation of either MtmGI or MtmGII resulted in the loss of the entire disaccharide chain, and it was initially believed that the disaccharide was formed then transferred as a diolivoside to the MTM intermediate rather than two successive

23 glycosylations. This was later disproven by examining the products of the inactivation of

MtmOIV, the enzyme directly succeeding MtmGI and MtmGII in the MTM pathway.

Among the products of the MtmOIV inactivation was 3A‐deolivosylpremithramycin B, proving that the D-olivoses of the disaccharide were added sequentially. By feeding this intermediate to S. argillaceus mutants that had MtmGI or MtmGII inactivated, it was found that MtmGI added the first D-olivose, and MtmGII added the second (Figure

1.19a).50-52 The inactivity of MtmGI without MtmGII is the second example of a necessary PPI in the MTM biosynthetic pathway. The discovery of 3A‐ deolivosylpremithramycin B as a minor product in the ΔmtmOIV mutant also suggests the presence of an MtmGII, MtmOIV interaction, as the disruption of such an interaction would lead to intermediates leaking out of the enzyme complex into the cytosol (Figure

1.19b).53

Figure 1.18 Disaccharide biosynthesis

24 Figure 1.19 A. Schematic of the determination of the order of disaccharide glycosyltransferase enzymes. B. Proposed effect of knocking out MtmOIV if the enzymes form a multienzyme complex. Dashed line indicates a minor product.

As stated before, alternating the glycosylation pattern is a fruitful strategy of combinatorial biosynthesis. Nowhere is this more evident than the natural occurrence of other aureolic acids with different deoxysugars. For example, the trisaccharide chain of chromomycin A3 has two D-olivose sugars and an acetylated L-chromose in place of

MTM’s D-mycarose. Durhamycin has a tetradedeoxysugar chain containing D-olivose,

D-oliose, followed by two D-olivose sugars (Figure 1.14). All of these secondary

25 metabolites have been optimized for activity through natural selection, which encourages the use of combinatorial biosynthesis to generate novel compounds with altered saccharide patterns.53, 54

While there have been too many experiments using combinatorial biosynthesis targeting the glycosylation patterns in MTM to go over each in detail, I would like to draw attention to a few that have shown promise for clinical application or that provide insight into the techniques of combinatorial biosynthesis. Heterologous expression of L- digitoxose biosynthesis genes into S. argillaceus flooded the native glycosyltransferases with NDP-L-digitoxose substrate resulting in four new MTM analogues including demycarosyl-3D-β-D-digitoxosyl-MTM (Figure 1.20). Using a terminal deoxynucleotidyl transferase-mediated dUTP nick end labeling (TUNEL) assays all the compounds were better at inducing apoptosis than MTM when used in MDA-231 breast cancer cells, with demycarosyl-3D-β-D- digitoxosyl-MTM showing over 20-fold more activity.45, 55

26 Figure 1.20 Results of the heterologous expression of L-digitoxose biosynthesis genes into S. argillaceus. 1. Demycarosyl-3D-β-D-digitoxosyl MTM 2. Deoliosyl-3C-α-L- digitoxosyl-MTM 3. Deoliosyl-3C-β-D-mycarosyl-MTM

CmmA is the gene in the biosynthesis of chromomycin A3 (CMA3) that is responsible for acetylation of its L-chromose and D-oliose deoxysugar moieties (Figure

1.21). The acetylation of its deoxysugars is necessary for CMA3s activity, as deacetylated analogues had about 100-fold less activity when tested on a panel of cancer cells. For this reason, it was suspected that acetylation of MTM could yield analogues with greater activity. MTM, demycarosyl-3D-β-D- digitoxosyl-MTM, and several analogues generated in other studies were fed to Streptomyces griseus, the species that produces CMA3, then extracted and analyzed (Figure 1.22). Seven new, acetylated compounds were found, all of which had a GI50 of less than 0.1 mM against the cancer cell lines studied. Some were more active than their nonacetylated counterparts, especially when tested against pancreas tumor cells. The acetyl groups were subject to hydrolysis, however, limiting their clinical use.43, 56, 57

27 Figure 1.21 Chromomycin A3. The acyl groups are marked with green boxes

28 Figure 1.22 Acetylated compounds generated in the S. griseus feeding experiments.

Note that the parent compound was MTM SK in compound 7. 1. 4E-O-acetyl- mithramycin 2. Demycarosyl-3D-β-D-digitoxosyl-3E-O acetyl-mithramycin 3.

Demycarosyl-3D-β-D-digitoxosyl-4D-O acetyl-mithramycin 4. Demycarosyl-3D-β-D- digitoxosyl-4E-O acetyl-mithramycin 5. Demycarosyl-3D-β-D-digitoxosyl-3E,4D-O acetyl-mithramycin 6. Demycarosyl-3D-β-D-digitoxosyl-3E,4E,4D-O acetyl- mithramycin 7. 4D-O-acetyl-mithramycin SK

1.2.5 MtmOIV, MtmW

The final two enzymes in MTM biosynthesis are MtmOIV and MtmW, which catalyze the 4th ring opening using a Bayer-Villiger monooxygenase (BVMO)

29 mechanism, and reduce the distal ketone on the nascent pentyl side chain, respectively

(Figure 1.23). These enzymes are often grouped together due to their interdependence.

MtmOIV produces MTM DK, a labile intermediate that is quickly transformed into the shunt products MTM SA, MTM SK, and MTM SDK by nucleophilic attack by water on the β-ketones of the pentyl side chain (Figure 1.24). The protection of MTM DK to prevent the formation of these shunt products was the first piece of evidence that

MtmOIV and MtmW require a PPI to produce MTM.58

Baeyer–Villiger oxidation is rearrangement reaction resulting in the introduction of an oxygen adjacent to a keto group, thereby forming an ester or a lactone. The general mechanism of the enzyme catalyzing the Baeyer-Villiger oxidation begins with the keto oxygen of the substrate to be protonated and attacked by molecular oxygen. Then the keto group reforms and the more stably substituted group migrates leaving the new oxygen next to the ketone. In MtmOIV, the substrate PreB is oxygenated between C-1 and C-2 to give a lactone intermediate. This intermediate is hydrolyzed, opening the 4th ring of the tetracyclic premithramycin and leaving a carboxylic group on the newly numbered C-2 which in turn is removed by decarboxylation, leaving the trisaccharide chain, also attached to C-2, with S stereochemistry (Figure 1.25).23, 58

30 Figure 1.23 The reactions of MtmOIV and MtmW

31 Figure 1.24 MTM DK shunt product formation. All reactions begin with MTM DK and occur on the C-3 side chain (red)

32 Figure 1.25 Baeyer-Villiger oxidation of PreB. FAD (blue) is reduced to FADH which in turn captures molecular oxygen. PreB (black) is attacked and a criegee intermediate is formed. Rearrangement and dissociation lead to PreB-lactone and hydroxylated

FAD, the latter is regenerated by dehydration.

The aldol-ketoreductase reaction begins with an attack on the keto group of the substrate by the hydride group of NADPH. The electrons that were forming the double bond of the ketone, in turn, attack another proton supplied from a hydroxyl group of an amino acid of the enzyme, typically a tyrosine. In MtmW, this reaction reduces the most distal of the β-ketones on the pentyl side chain, thereby preventing the formation of shunt products. Aldol ketoreductases catalyze their reactions using a catalytic tetrad of

33 histidine, tyrosine, lysine, and aspartate. Tyrosine has an partial positive charge, which is due to the presence of the histidine. Histidine also donates a proton to tyrosine at the completion of the reaction. Lysine stabilizes the charge of tyrosine and facilitates the interaction between tyrosine and the substrate, while aspartate stabilizes the charge of lysine (Figure 1.26).59, 60

Figure 1.26 A general aldol ketoreductase catalyzed reduction

The MTM DK shunt products are clinically interesting, as MTM SK has comparable IC50 concentrations to MTM when used on a panel of ovarian cancer cell lines, and MTM SDK showed a 2-fold improvement in the same experiment. Both MTM

SK and MTM SDK were also better than MTM at inhibiting SP1 governed genes.61, 62 In a separate experiment, MTM SK was tested against 60 cancer cell lines and was an average of 9-fold more active than MTM, and was 1,500-fold less active against non- cancerous mouse 3T3 fibroblast cells. While MTM SK and MTM SDK had better therapeutic indices than MTM, MTM SA was relatively inactive. However, its shorter,

34 two carbon side chain terminated in a carboxylic acid group, allowed for the possibility of selective, semisynthetic modifications. The carboxylic acid group was reacted with different amino acids to form an amide linkage resulting, among others, in MTM SA- amino acid conjugants.63 Glycine, alanine, and valine all had comparable activity to

MTM SA, while tryptophan and phenylalanine conjugants had greater activity and specificity toward Ewing’s sarcoma (Figure 1.27).61-64

35 Figure 1.27 A MTM SA and its two most active amino acid conjugates, MTM SA-Trp and MTM SA-Phe. B. The coupling reaction between MTM SA and the amino acids

(green box).

1.2.6 Regulation and Resistance

Finally, the last step in the MTM biosynthetic pathway is its export from the cell.

This is accomplished by the ABC transporter MtrAB. MtrA is the ATP binding protein, which used the energy from hydrolyzing ATP to drive conformational changes in the transmembrane protein, MtrB, which allows for the export of MTM into the extracellular environment.65 An unusual fact of MTM resistance is that it seemingly relies on a single 36 mechanism. To our knowledge the only other species that contains a single resistance gene is Saccharopolyspora erythraea, the species that produces erythromycin. This suggests that it is likely that S. argillaceus contains an alternative mechanism for self- resistance to prevent poisoning its own DNA with its own antibiotic.65, 66

While not directly involved in MTM biosynthesis, regulation of the MTM pathway is important to consider in our studies. MtmR and MtrY are transcription factors, which regulate the production of MTM pathway enzymes, and by extension, the production of MTM. In the MTM pathway, MtrY is constitutively expressed and acts as a repressor to MtmR as well as the MTM resistance genes MtrA and MtrB. MtmR acts as a transcriptional activator for all seven operons in the MTM pathway and its inhibition keeps MTM production levels low. MtrY’s role as a repressor is reversed upon binding to

MTM, however, the MtrY-MTM complex ceases to inhibit MtmR, and acts as a transcriptional activator for MtmR, MtrA and MtrB.67 The phenomenon in which MTM acts as its own positive regulator is known as a feed-forward system (Figure 1.28). The feed-forward system is not uncommon in secondary metabolite biosynthesis, particularly in Streptomyces species. Other examples of feed-forward mechanisms in Streptomyces biosynthesis include the angucycline polyketide antibiotic auricin, jadomycin, and landomycin as well as the anthracycline daunorubicin.68-71 The feed-forward system also has similarities to quorum sensing systems, where secondary metabolites are produced at low, basal levels until their concentration becomes sufficient to promote large-scale production coupled with transcriptional alterations (Figure 1.29).67-72

37 Figure 1.28 Regulation of the MTM pathway (grey). MtmR (green, arrows) acts as a transcription activator for the pathway but its expression is repressed by MtrY (blue, inverted arrows). However when bound to MTM (orange), MtrY ceases to act as a repressor and switches to an transcription activator (blue arrows) for the operon containing MtmR. This leads to increased expression of MTM. This figure was adapted from Florez et. al. 2015

38 Figure 1.29 A. A hypothetical quorum system. Quorum system (QS) molecules (orange) and the transcriptional activator (blue) are produced at a low basal rate. When the concentration of QS molecules is sufficient, they bind to the activator causing it to begin activating transcription of the gene cluster (grey), the QS molecules, and the activator itself.

1.3 Specific Aims

In review, the biosynthesis of MTM has been well characterized and our understanding of the MTM biosynthetic pathway correlates with the increased opportunities for combinatorial biosynthesis, and by extension, better medicines. With this in mind, the specific aims of my research are as follows.

Aim 1: Increase our understanding of MtmW.

MtmW is the final enzyme in the MTM pathway and has so far only contributed to combinatorial biosynthesis through its absence. Also, the initial steps of our

39 investigations into MtmW revealed that it produced an unexpected epimer of MTM, iso-

MTM. Included in my first aim is to determine if iso-MTM had bioactivity, determine if iso-MTM was the natural product of MtmW or an artifact of the MtmW in vitro reaction, and to determine what, if any, mechanism of isomerization of iso-MTM into MTM.

Aim 2: Characterize the interaction between MtmOIV and MtmW.

There are three sets of enzymes that have putative PPIs in the MTM pathway,

MtmC and MtmGIV, MtmGI and MtmGII, and MtmOIV and MtmW. Of these, the pair that has been responsible for the most promising MTM analogues is MtmOIV and

MtmW. The evidence for the interaction between MtmOIV and MtmW includes (i) the apparent inactivity of MtmW in the absence of MtmOIV, (ii) that the product of MtmOIV is subject to degradation and therefore requires channeling to be used by MtmW, and (iii) intermediates and shunt products of the MTM pathway are not detected in S. argillaceus cultures, which suggests that enzymes in the pathway form a tightly controlled multi- enzyme complex.

40 Discovery of a Cryptic Intermediate in Late Steps of Mithramycin Biosynthesis

2.1 Introduction

MtmW is an enzyme that directly influences the bioactivity of MTM in the MTM biosynthetic pathway, and it has been previously exploited via combinatorial biosynthesis to produce safer and more active analogues. MtmW also has a compelling codependence on MtmOIV. For these reasons, it is an excellent subject for study of the field of natural products. As a brief overview of this chapter, MtmW produced an unexpected C-2 epimer of MTM that we named iso-MTM. We determined the structure of this epimer by NMR, and characterized its activity. We also performed experiments to determine whether iso-

MTM was the in vivo product of MtmW, or an artifact of the in vitro reaction in which it was discovered. We also performed experiments in attempt to explain why iso-MTM is the primary product in vitro, while it is never observed in S. argillaceus cell cultures.

Crystallization and polydispersity studies give us a clear picture of the structure of

MtmW and inform on its role in a putative multienzyme complex, which we believe the post-PKS enzymes in the MTM pathway organize themselves into for efficient biosynthesis of MTM.

2.2 Experimental Design and Results

2.2.1 MtmW Crystallization

We crystallized MtmW (ΔN3) in two crystal forms: the apo‐MtmW crystallized in one form (diffracting to 1.8 Å resolution) and its complex with NADP+ in the other (2.1

Å resolution), with a similar crystal packing (Table 2.1). We determined the crystal structures of these two states of MtmW. The structure of an MtmW monomer (Figure

2.1) consists of a common TIM barrel fold, which makes up most of the protein, a small

41 N‐terminal two‐stranded β‐sheet capping the TIM barrel and two extra helices outside of the barrel. Other short helices were also present, but their degree of disorder varied among monomers of the same crystal form and between the two forms. The apo‐MtmW crystals contained two MtmW monomers per asymmetric unit, with each monomer generating its own tetramer by crystal symmetry operations. In one monomer, two loops spanning residues 26–30 and 224–238 were not observed in the electron density map due to disorder. In the other monomer, in addition to two analogous loops (residues 24–30 and 218–247), a short loop spanning residues 55–56 was also disordered. The disordered loops are located on the same face of the protein (in the back of the structure in Figure

2.1).60

Figure 2.1 A monomer of MtmW. A view into the TIM barrel is shown, with the short

N‐terminal β-sheet capping the barrel. The N‐and C‐termini are labeled as “N” and

“C”, respectively.

42 Table 2.1 Crystallographic statistics for MtmW and its complexes with ligands

The structure shows that the monomers of MtmW form a well-packed homotetramer (colored in Figure 2.2a), in which both the monomer fold and the tetrameric assembly are highly conserved in STM2406 and similar aldo‐keto reductases

(AKR) as well as in Kvβ proteins. The tetramers in both crystal forms are loosely packed against each other (Figure 2.2b) so that their molecule, with the 4‐fold rotational symmetry axis running along the middle of the channel (Figure 2.2). The two helices lying outside of the TIM barrels are located on the periphery of the tetrameric scaffold.

The crystal structure of MtmW in complex with NADP+ showed that the cofactor was 43 bound on the outside of the channel to each monomer (Figure 2.2), where the access to the nicotinamide moiety of NADPH would occur between the tetramer subunits. The nicotinamide moiety of the bound NADP+ exits into a large cleft (Figure 2.2 and Figure

2.3), where substrate binding and catalysis must take place. The NADP+ binding interface is highly conserved among homologous proteins of known structure. As predicted from the sequence alignment, the catalytic tetrad universally conserved in the AKR is also present in MtmW at proper positions (Figure 2.4; conserved catalytic Tyr 57 is shown in

Figure 2.5).60

Figure 2.2 Two orthogonal views of the MtmW octamer of the MtmW‐

NADP+ complex. A) The view into the inter-subunit channel. B) The view where the channel is oriented along the vertical axis. The bound NADP+ is shown by purple sticks.

44 Figure 2.3 The crystal structure of MtmW‐NADP+‐PEG complex. NADP+ is shown by purple sticks, PEG (orange sticks) is well defined by the polder omit electron density map, contoured at 5.5σ (shown as a mesh). The residues in the putative substrate binding site are shown as sticks with the labels. The orientation of this MtmW monomer is the same as that of the similarly colored monomer in Figure 2.2a

45 Figure 2.4 The sequence alignment of MtmW and other structurally characterized representatives from the AKR family. These enzymes are: AKR_P.sp.JS666: AKR from

Polaromonas sp. JS666, Kvb_rat: Kvb from Rattus norvegicus, STM2406: 3- hydroxybutanal reductase STM2406 from Salmonella enterica serovar Typhimurium,

Cbei_3974: AKR Cbei_3974 from Clostridium beijerinckii, YghZ_E.coli: YghZ from

Escherichia coli, AKR_R.serpent: perakine reductase from Rauvolfia serpentina.

Identical residues are shown in red boxes and homologous residues are shown in red font. The tetrad of active site residues conserved in the aldo-keto reductase (AKR) superfamily59 is shown by the red lollipops, and the functionally validated residues of the substrate binding site in the structurally similar STM240673 are shown by the green lollipops. The ketone oxygen of the substrate is predicted to abstract a proton from

Tyr57 in the reduction mechanism.74

46 Figure 2.5 Views of the active sites of MtmW and STM2406. A. The active site of

MtmW. The protein is in a similar orientation to that in Figure 2.3. B. The active site of

STM2406 (PDB ID: 3ERP73). A cacodylate ion (yellow sticks), a 1,2-ethanediol

47 (orange sticks) and a sodium ion (a green sphere) are bound in the substrate binding pocket of STM2406, defined by the conserved NADPH site, the conserved catalytic

Tyr66 and the residues whose mutations perturbed the substrate profile (Trp33, Asn65,

Tyr100, Asn169).73 C. The active site of STM2406 in the crystal structure of STM2406-

NADP+ complex (PDB ID:5T7973). NADP+ is shown as purple sticks.

In addition to these crystals, extensive crystallization trials were carried out to obtain crystals of MtmW in complex with MTM and iso‐MTM. While we did not observe either of these ligands in the active site of MtmW, the crystals that were grown with MtmW pre-incubated with iso‐MTM and NADP+ yielded strong tubular electron density in the putative substrate binding cleft that was best modeled by a terminal region of a PEG molecule, in addition to a well resolved NADP+ molecule (Figure 2.3, 2.5a, and 2.6). The PEG is bound in the substrate binding pocket because of its proximity to the nicotinamide ring of NADP+ and Tyr 57. The structurally analogous substrate binding pocket STM2406 (Figures 2.5b, and 2.5c) was functionally validated by mutagenesis.73

The loops that were disordered in the apo‐ and NADP+‐bound crystal structures were all ordered in the MtmW‐NADP+‐PEG crystal structure. These conformational changes are likely coupled to MtmOIV binding, which we observed to be enhanced in the presence of

NADPH and PreB. The PEG is bound in a hydrophobic pocket, which is large enough to accommodate the tricyclic mithramycin core and the 3‐side chain. Some of the residues lining the pocket are located on the flexible loops: Leu 24, Tyr 26, Pro 27, Val 56, Tyr 57, others include Met 86, and Pro 87. The cleft defined by the nicotinamide ring of NADP+ contains charged residues Glu 308, Arg 125, Glu 155, consistent with the polar nature of the 3‐side chain of MTM.60

48

Figure 2.6 A zoomed-in view of the active site of MtmW in complex with NADP+ and a

PEG region. The color scheme is the same as in Figure 2.5a

In vertebrate Kv1 channels, each Kvβ subunit forms a complex with a T1 domain of the integral membrane β subunit responsible for the functional tetrameric assembly of the ion channel.75, 76 T1 is bound to Kvβ on the face of the tetramer analogous to the solvent exposed face of the MtmW octamer (Figure 2.7, top tetramer in Figure 2.2b). Kvβ is also an active NADPH‐dependent reductase whose dominant endogenous substrate is yet unknown, but the change in the redox state, even when induced by surrogate substrates, has been shown to regulate the channel.77 By the same logic, we propose that

49

MtmW, being the ultimate mithramycin biosynthetic enzyme, may be associated with a transmembrane channel to regulate the export of mithramycin.60 This work was performed by Drs. Oleg Tsodikov and Caixia Hu

Figure 2.7 A comparison of the structures of MtmW-NADP+ complex (this study) and

Kvβ in complex with the T1 domain (PDB ID: 1EXB15, 75). A. An octamer of MtmW is shown in the same view as in Figure 2.2b. B. A tetramer of Kvβ-T1-NADP+ complex.

The Kvβ tetramer is oriented similarly to the top tetrameric layer of MtmW in panel A.

The bound NADP+ is shown as purple sticks in both panels.

2.2.2 MtmW Polydispersity

The crystal structure of MtmW showed tetrameric polydispersity. To confirm this, we performed size exclusion chromatography and analytical ultracentrifugation. The results of the size exclusion chromatography suggested an octameric assembly and this result was confirmed by ultracentrifugation which showed a sedimentation coefficient of

10.8 S, which agreed with a model of MtmW octamer modeled as an oblate spheroid.

Structurally and biochemically characterized AKR superfamily proteins that are most

50 similar in sequence to MtmW include an octameric 3‐hydroxybutanal reductase

STM2406 from Salmonella enterica serovar Typhimurium (36 % sequence identity) and a tetrameric cytosolic β subunit (Kvβ) of the Shaker family of voltage‐dependent potassium channels (Kv1) in vertebrates (34 % sequence identity) (Figure 2.4, Figure

2.7).60 The ultracentrifugation was performed by Drs. Frank Herkules and Dmitri N.

Ivanov.

It's worth mentioning that polydispersity is highly contextual. Isopentenyl diphosphate isomerase in Sulfolobus shibatae forms an octamer until its substrate binds and causes a conformational change splitting the octamer into two tetramers (Figure

2.8a). Similarly, phosphofructokinase in Thermus thermophilus exists as an inactive dimer, which forms an active tetramer in the presence of other enzymes in the glycolysis pathway (Figure 2.8b). In the pentose phosphate pathway of Trypanosoma brucei, the equilibrium of 6-Phosphogluconate dehydrogenase shifts between favoring the dimeric form to favoring the tetrameric form depending on the concentration of product and substrate (Figure 2.8c). This is not to say that we believe that MtmW in the octameric form does not reflect its polydispersity in vivo, only that future investigations should account for the possibility of tetrameric or dimeric species when making predictions based on MtmW’s in vivo state.78-80

51 Figure 2.8 Three factors that can effect polydispersity. A. Substrate-mediated conformational changes. B. PPIs with other enzymes. C. The equilibrium between the enzyme’s product and substrate.

2.2.3 MtmOIV and MtmW PPI

To confirm the existence of our proposed PPI between MtmOIV and MtmW, an

EMSA assay and a formaldehyde crosslinking assay were performed. The pilot formaldehyde crosslinking assay was performed for two hours in 1% formaldehyde solution. While a new band did not show up at ~95 Kda, which would represent an

MtmOIV-MtmW crosslinking, there was a mass of enzyme in the crosslinked samples that did not enter the gel, suggesting crosslinking had taken place. Thymidylate kinase

(TMK) was chosen for a negative control, but exhibited a similar band that did not enter

52 the gel when subjected to formaldehyde by itself (Figure 2.9). As a follow up, a time course formaldehyde crosslinking assay was done with MtmOIV and MtmW for 1-16 hr.

The results from this assay showed both MtmOIV and MtmW were removed from the solution in a time dependent manner, but we were unable to ascertain whether we were capturing a PPI or simple aggregation caused by the formaldehyde (Figure 2.10).

Figure 2.9 The result of formaldehyde (FM) crosslinking between MtmOIV and MtmW.

A. 1. Ladder 2. MtmOIV + MtmW 3. MtmOIV +MtmW + FM 4. MtmW + FM 5.

MtmOIV + FM 6. MtmW + TMK + FM 7. MtmOIV + TMK + FM 8. TMK + FM B.

Close up of the top of wells 1-5. Note the aggregation present in 3 and 5 (Black arrows)

53 Figure 2.10 A time course experiment with MtmOIV, MtmW and FM. 1. Ladder 2. 1 hr

3. 2 hr. 4. 3 hr 5. Overnight. Note the aggregation at the top of each well (black arrow)

We tested whether MtmOIV and MtmW (each at 20 μM) form a complex by a electrophoretic mobility shift assay (EMSA) in the absence and in the presence of PreB and NADPH (a co‐substrate for both enzymes). (MtmOIV is strongly bound to FAD; therefore, addition of FAD was not necessary.) Besides the bands corresponding to

MtmOIV and MtmW, no new band was observed at any conditions, indicating that the two proteins did not form a long‐lived complex. However, a distinct and reproducible

54 smear of the band corresponding to MtmW was observed in the presence of the MtmOIV

(lanes 3–6), which migrated much more slowly. This effect was not observed in the absence of MtmOIV (lane 1), indicating that MtmW and MtmOIV formed a complex that dissociated during the initial stages of the electrophoresis. This smear was enhanced in the presence of either NADPH or PreB or both ligands (Figure 2.11a). The formation of

MtmOIV‐MtmW complex in a substrate‐dependent fashion is consistent with the above mechanistic models of the proper sequence of reaction occurring only when the product of MtmOIV can be handed over directly to MtmW. We also performed this assay as a titration, where the concentration of MtmOIV increased while the MtmW concentration was kept constant at 10 μm (Figure 2.11b). In agreement with the previous experiment, we observed gradual weakening of the intensity of the band corresponding to the free

MtmW upon titrating MtmOIV, indicating weak MtmOIV‐MtmW binding.60 The protein

EMSA was performed by Drs. Oleg Tsodikov and Caixia Hu.

55 Figure 2.11 A. A Coomassie blue stained 1.7 % agarose gel showing the EMSA with

MtmOIV (20 μM in monomers, 10 μM in dimers) and MtmW (20 μM in monomers, 2.5

μM in octamers). The smear above the MtmW band in lanes 3–6 indicated that MtmW was retarded by binding MtmOIV. B. A similar gel showing a titration of MtmW (10

μM) with MtmOIV. The fraction of free MtmW is shown below each lane.

2.2.4 MtmW in vitro Assay

With the knowledge that MtmOIV produces a labile intermediate, MTM DK, the in vitro assays of MtmW necessarily included MtmOIV as well as the cofactors NADPH and FAD, the NADPH being necessary for aldo ketoreductase and BVMO activity and the FAD necessary for BVMO activity. Surprisingly, in the presence of both enzymes,

56 only very small amounts of the expected MTM intermediates and shunt products of the reaction were observed, while a new HPLC peak emerged as major product in this assay.

The production of this new compound clearly required MtmW, since this product was not observed in the control assays carried out in the absence of MtmW or when MtmW was heat‐denatured. HR‐MS showed that this new compound has an identical molecular weight to that of MTM, hence we named this molecule iso‐MTM.

Figure 2.12 HPLC chromatograms of reactions catalyzed by MtmW and MtmOIV at pH 8.25. A) Pure PreB as standard; B) pure MTM as standard; C) PreB + MtmOIV +

MtmW + FAD + NADPH; D) PreB + MtmOIV + FAD + NADPH; E) PreB +MtmOIV

+ boiled MtmW + FAD + NADPH; F) PreB + MtmW + FAD + NADPH. Compounds

1, 2, and 3 are PreB, MTM, and iso-MTM, respectively.

57 2.2.5 iso-MTM NMR

To determine the nature of the stereochemical differences between MTM and iso-

MTM, iso-MTM was purified and analyzed by NMR. The NMR spectra revealed that iso‐MTM is a C-2 epimer of MTM. The NMR data for iso‐MTM (10), when compared with those for MTM,81 revealed a distinct difference in the coupling constant of C-2‐H in the 1H NMR spectrum. In MTM the C-2 proton resonates at 4.78 ppm with a coupling constant of 11.5 Hz (trans‐coupling), whereas in iso-MTM, a coupling constant of 6.7 Hz

(cis‐coupling) and a chemical shift of 4.40 ppm were found, indicating that iso-MTM was a C-2 epimer of MTM. These data also confirmed that MtmW reduced the C-4′ keto group, as seen in MTM (Figure 2.13).60 The NMR experiment was performed by Dr.

Prithiba Mitra.

58 Figure 2.13 Distinction between iso-MTM and MTM through NMR.

2.2.6 iso-MTM Bioassay

The activity of iso-MTM was assayed by disk diffusion, MIC, and IC50 on the

TC-32 Ewing’s sarcoma cell line. Salmonella enterica and Staphylococcus aureus were chosen to represent Gram-negative and Gram-positive bacteria, respectively. As expected, neither MTM nor iso-MTM was active against S. enterica, Gram-negative inhibition having not been previously reported. Unexpectedly, both MTM and iso-MTM were active against S. aureus, with iso-MTM producing a zone of inhibition of similar size to MTM (Figure 2.14a). An MIC assay was done using S. aureus to investigate

59 further. Iso-MTM had an MIC of 0.25 µg/mL versus 0.125 µg/mL for MTM, indicating iso-MTM had roughly half the bioactivity of MTM (Figure 2.14b). The IC50 assay using the EWS-Fli1 containing TC-32 cell line showed that iso‐MTM and MTM had an IC50 of

209 nm and 45 nm, respectively (Figure 2.14c). These results were unexpected because the trisaccharide chain of MTM compounds is vital for activity, as it sits in the minor grove of DNA and provides some of the molecule’s specificity for GC-rich DNA.

However, iso-MTM has R stereochemistry at the C-2 carbon where the trisaccharide attaches. This should cant the deoxysugars away from the DNA, restricting its ability to bind. On suspicion of iso-MTM’s apparent activity, I extracted the inhibiting MIC cultures and found that roughly half of the iso-MTM had isomerized into MTM (Figure

2.15). The IC50 assay was performed by Dr. Joe Eckenrode.

60 Figure 2.14 Bioactivity of iso-MTM. A. Disk diffusion assay. Zones of inhibition are marked with black circles, disks are marked with grey circles. B. The IC50 assay with

TC-32 cells. C. The MIC assay. All experiments were done in triplicate, and the data shown represents the average of the experiments.

61 Figure 2.15 Fate of iso-MTM in MIC samples.

2.2.7 Macromolecular Crowding

Macromolecular crowding (MMC) describes the native environment of an enzyme. The concept of MMC was developed by Minton & Wilf, when they reasoned that the crowding altered the polydispersity of glyceraldehyde-3-phosphate dehydrogenase (G3PD). G3PD forms monomers, dimers, and tetramers in solution and its activity increases in correlation with self-association. Because MMC increases the local concentration of monomers, it shifts the equilibrium of towards the more active tetrameric state. There are two factors to consider in oligomerization of enzymes into complexes. First, MMC increases the equilibrium constant for oligomerization. The volume displaced by the crowding agent increases the local concentration of enzyme and lowers the free energy required for activation for the complex, this is the transition state- limited rate (Figure 2.16a). Second, MMC lowers the diffusion coefficient, meaning the rate in which the monomers encounter each other is reduced, this is known as the

62 diffusion-limited rate (Figure 2.16b). Research shows that as the amount of crowding agent increases, the reaction rate moves from transition-state limited to diffusion limited.

(Figure 2.17) In practical terms, this means that there is an optimal concentration of macromolecules that will favor oligomerization.82, 83

Figure 2.16 Transition state-limited rate constant and Diffusion-limited rate constant.

A. At low concentrations of crowding agent (grey) the local concentration enzyme (red) increases. Addition of crowding agent causes the enzymes to “crowd together” increasing PPIs (solid arrow). B. At high concentrations of crowding agent, diffusion and the rate at which the enzymes interact are reduced (dashed arrow). Addition of crowding agent decreases diffusion and decreases PPIs.

63 Figure 2.17 The influence of crowding agent concentration on PPI and enzyme reaction rates. Addition of crowding agent increases interaction (orange dashed line) until its concentration begins to hamper diffusion (green dashed line). Overall rate of interaction and reaction is shown by the black line.

MMC can also affect the structure of proteins. Homouz et. al. showed that the

“football-shaped” protein VlsE, a virulence factor found in Borrelia burgdorferi, changed shape in a manner dependent on macromolecule concentration. VlsE showed an increase in alpha helix structure from 50% to 80% in the presence of 400 mg/mL Ficoll 70 using far-UV circular dichroism.83 This phenomenon was also observed by Stagg et. al. using apoflavodoxin from Desulfovibrio desulfuricans, where the alpha helix content rose 20% and the random coil content fell 10% upon the addition of 400 mg/ml Ficoll 70.86

Disordered regions of proteins tend to get compacted with increasing MMC. This compacting indirectly favors the more ordered types of secondary structure such as alpha

64 helix and beta sheet. Computer models in these studies show that the secondary and tertiary structures become more packed and crystalline in nature.83-87

The typical concentration of the intercellular environment is between 80-400 mg/mL and is composed of proteins, sugars, lipids, nucleic acids, and other large molecules. In contrast, the concentration of macromolecules in our MtmOIV-MtmW in vitro reaction is less than 2 mg/mL. While studies on MMC typically describe differences in the efficiency of a reaction, it was possible that it could have been the factor responsible for the formation of the creation of the MTM isomer in vivo. Three scenarios exist in regard to the previously mentioned research. The equilibrium of the MtmOIV dimer or the MtmW octamer may have been altered by performing the in vitro experiment in an uncrowded solution, the protein interactions between MtmOIV and

MtmW could have been altered, or the secondary structures of MtmOIV or MtmW many have become disordered in the uncrowded solution.

To test the effects of MMC on the MtmOIV MtmW reaction in regard to the unexpected formation of iso-MTM, the reaction was performed in the presence of 100,

200, and 300 mg/mL of PEG 8000. The concentrations were chosen based on the solubility of PEG 8000 and match the range of concentrations of macromolecules found in the intercellular environment. The reactions in PEG 8000 became less efficient as the concentration of crowding agent increased, as seen in the lower ratio of products to substrate. And they produced a similar ratio of iso-MTM to MTM; 70-85% of the products were iso-MTM. Overall, all the reactions with crowding agent produced a similar proportion of iso-MTM to MTM, leading me to conclude that loss of MMC was not responsible for the production of iso-MTM in my in vitro reaction (Figure 2.18).

65 Figure 2.18 The influence of crowding agent concentration on iso-MTM production.

Reactions were performed as in section 2.2.4 and analyzed by LC/MS. At all concentrations of PEG, iso-MTM (orange) is the major product.

2.2.8 Total Cell Assay

If iso-MTM is an intermediate in the MTM biosynthetic pathway, an isomerase in

S. argillaceus could have been responsible for the absence of iso-MTM in vivo. While a dedicated isomerase has not been detected in the MTM gene cluster, several enzymes in the MTM pathway are bifunctional, namely MtmC and MtmGIV which are a methyltransferase/ketoreductase, and a non-iterative glycosyltransferase that transfers 4-

66 keto mycarose and olivose to the C and E sugar positions of MTM, respectively (Figure

2.19). Additionally, there could be an isomerase outside the MTM gene cluster. For example, the elloramycin cluster in Streptomyces olivaceus lacks the genes for the synthesis of L-rhamnose, a deoxysugar incorporated in the final product. To detect the presence of an iso-MTM isomerase S. argillaceus, we reasoned that altering the PreB to iso-MTM reaction by omitting the purified MtmOIV and MtmW and using S. argillaceus cell lysate as the enzyme source, we could determine if iso-MTM was the natural product of the MTM biosynthetic pathway or an aberration caused by our out of context in vitro reaction.39, 88

67 Figure 2.19 Two bifunctional enzymes in the MTM Pathway. A. MtmGIV (blue) adds the C and E sugars to the trisaccharide chain B. MtmC (red) acts as a methyltransferase/ ketoreductase in deoxysugar biosynthesis.

Early trials of this experiment used WT S. argillaceus lysate as the enzyme source, but the results of these experiments were hampered by the native MTM in lysate.

To rectify this the mutant strains S. argillaceus ∆mtmGI and S. argillaceus ∆mtmGIII were used as the enzyme sources as they produce premithramycin A3 and premithramycin A1, respectively. The results of these reactions showed that iso-MTM was the main product by greater than 90%. The experiments were also repeated with the addition of ATP to rule out an ATP-dependent isomerase, and with the addition of purified MtmGIII in the ∆mtmGIII strain. Iso-MTM was also the main product of these additional experiments (Figure 2.20).

68 From these results we concluded that there was no iso-MTM isomerase present in the lysate of S. argillaceus. However, the loss of an intact membrane and its associated osmotic and electrochemical gradient could have altered the function of a transmembrane or membrane dependent enzyme that was responsible for the isomerization.89

Figure 2.20 Select results from the total cell assay experiments. S. argillaceus ΔmtmGI and S. argillaceus ΔmtmGIII lysates were used as enzyme sources for the PreB to iso-

MTM reaction. In all experiments, iso-MTM was the major product.

2.2.9 Extracellular Isomerase

The previous experiment investigated a potential isomerase or PPI binding partner to MtmOIV or MtmW in S. argillaceus that I suspected led to the discrepancy of the

69 products in the in vivo and in vitro reactions. This experiment did not take into account secreted, extracellular enzymes as the cells were pelleted and washed before lysis.

Extracellular protein transport is a common ability for all bacterial species. For example,

Helicobacter pylori secretes HP0175, a peptidyl prolyl cis, trans-isomerase (PPIase) as a virulence factor.90 PPIases isomerize proline peptide bonds from cis to trans and in the case of HP0175 it promotes autophagy in phagocytic cells and gastric epithelial cells through the conversion of the soluble human protein LC3I to its membrane bound form,

LC3II.90 Another example of an extracellular isomerase catalyzes the conversion of xylose into xylulose. Though typically an intercellular enzyme, actinomycetes species

NCL 82-5-1 secretes a soluble species of xylose isomerase when grown in xylan- containing media.90-94

According to these examples, it is possible that S. argillaceus secretes an extracellular iso-MTM isomerase. To test this, I incubated iso-MTM in a cell-free supernatant of S. argillaceus ∆mtmGIII that had been grown to stationary phase. The results of this experiment showed that iso-MTM slowly converted into MTM. Over the course of two days, over 80% of iso-MTM had been replaced by MTM. However, the conversion also occurred in a S. aureus cell-free supernatant, though to a lesser extent, leading me to believe that there was no extracellular isomerase as that would not have been present in the S. aureus supernatant. However, it did appear that there was some media component that was common to the R5A streptomyces media and the LB that was used to culture S. aureus was at least in part responsible for iso-MTM isomerization

(Figure 2.21).

70 Figure 2.21 iso-MTM (blue) with S. argillaceus and S. aureus cell-free supernatant after two days incubation at 30 °C. In both cases the majority of iso-MTM epimerized into MTM (green)

The mechanism we propose for iso-MTM isomerization, whether spontaneous or enzyme mediated, starts with the formation of an enolate at the C-2 position. The enolate results in planar geometry at the C-1 and C-2 carbons, and a loss of the R stereochemistry that defines iso-MTM. From this step, H+ attack on the si face results in MTM. The S stereochemistry at C-1 is energetically favorable due to the steric hindrance between the trisaccharide chain and the pentyl side chain at C-3. Equilibrium between the keto and enol form of MTM would still occur, but the MTM keto form should be vastly preferred after the initial enolization (Figure 2.22).

71 Figure 2.22 Proposed mechanism for iso-MTM to MTM epimerization.

2.2.10 iso-MTM Mg2+ Assay

As stated previously, the C-9 hydroxyl and the C-1 carbonyl of MTM coordinate with divalent metal ions to form a dimer. It is our hypothesis that the positive charge of the metal ion catalyzes the enolate formation in a pH dependent manner. The use of metal ions to stabilize enolates is not uncommon. Type II aldolases such as fuculose 1-

72 phosphate aldolase (FPA) use Zn2+ to catalyze their reactions. FPA to converts the keto group of dihydroxyacetone phosphate into the enolate form to facilitate its aldol addition onto the aldehyde group of lactaldehyde to form fuculose 1-phosphate.98 For an example of divalent metal mediated isomerases, we can look to the biosynthesis of the rare sugar psicose. Psicose is a C-3 epimer of fructose and is used as a sugar substitute in some processed foods. The enzyme D-psicose 3-epimerase isomerizes D-fructose into D- psicose in Agrobacterium tumefaciens and Clostridium scindens, and depends on the presence of Mn2+ ions for catalytic activity.100, 101 D-tagatose 3-epimerase is a similar enzyme to D-psicose 3-epimerase in that it isomerizes a sugar using a Mn2+. In the proposed mechanism for D-tagatose 3-epimerase, Mn2+ coordinates with the C-2 carbonyl and the C-3 hydroxyl groups of D-fructose and stabilizes the enolate form of the sugar. In the enolate form, hydride is allowed to attack the si face of the sugar resulting in the S C-

3 stereochemistry that differentiates psicose from the R C-3 stereochemistry that defines fructose.95-103

Metal-mediated isomerization reactions can also be catalyzed without the aid of enzymes like D-psicose 3-epimerase and D-psicose 3-epimerase. Both Ni2+ complexes and molybdic acid can catalyze a 1,2 C-shift of the C-3 carbons of sugar molecules resulting in epimerization. Although these reactions follow a fundamentally different mechanism than my proposed enzyme-free Mg2+ epimerization of iso-MTM, they show that metal ions in solution have the ability to catalyze chemical reactions of organic compounds, particularly those with carbonyl functional groups.104-107

To test if the presence of Mg2+ alone was responsible for iso-MTM isomerization,

I incubated iso-MTM in concentrations of Mg2+ up to 50 mM, the concentration found in

73 the R5A media with which I see no iso-MTM production. The results show that iso-

MTM is converted to MTM over days in a pH and Mg2+ concentration dependent manner when compared to controls without Mg2+. At pH 8, with Mg2+, 57.7% of iso-MTM had converted to MTM compared to 9.8% in the control (Figure 2.23). While these data show that non-enzymatic conversion of iso-MTM is mediated by Mg2+, the level of conversion was still lower than that of the supernatants of S. argillaceus and S. aureus (80% and

70%, respectively). As there is no iso-MTM present in S. argillaceus cultures (Figure

2.24), enzymes from cell lysate produce iso-MTM, and Mg2+ ions facilitate iso-MTM to

MTM conversion, a new hypothesis was formed. The product of MtmW is iso-MTM, which is isomerized upon passage through the cell membrane, via MtrAB, to MTM. Mg2+ or other divalent metal ion likely plays a catalytic role, as either added upon membrane translocation or as part of a catalytic motif in MtrAB (2.25).108

74 Figure 2.23 LC/MS analysis of the isomerization of iso-MTM in the presence of 50 mM

Tris-HCl pH 8.0 and different concentrations of MgCl2

Figure 2.24 Extracts of a 24 hr culture of S. argillaceus. A. Supernatant butanol extract. B. Cell pellet butanol extract.

75 Figure 2.25 Proposed mechanism for iso-MTM to MTM epimerization. Updated to include the influence of Mg2+

2.2.11 Whole Cell Isomerization

To test the new hypothesis, I incubated iso-MTM with S. argillaceus ∆mtmGIII.

The purpose of this experiment was to discover if alive, intact cells would push the iso-

MTM isomerization to completion. If the isomerization were mediated by environmental conditions alone, the results of this experiment would be similar to the Mg2+ assays or the extracellular isomerase experiments. The results showed near complete conversion of iso-

MTM to MTM by the second day. The iso-MTM peak was undetectable in the LC trace,

76 and only a small peak was observed by extracted ion chromatogram (EIC) (Figure 2.26).

These data show that iso-MTM conversion is more efficient with intact cells present than in supernatant alone, and showed results that nearly mirrored my observations of S. argillaceus WT cultures. The results also suggest that transport of iso-MTM across the cell membrane is necessary for complete isomerization.

Figure 2.26 Isomerization by intact cells.

2.2.12 iso-MTM Cell Extraction

Streptomyces coelicolor A3(2) produces, among many secondary metabolites, actinorhodin. Actinorhodin is only found intercellularly, though it is transformed into its lactone form, γ-actinorhodin upon export from the cell. To test if iso-MTM had accumulated in the cell in the same manner that actinorhodin did in S. coelicolor A3(2), 1 mL of culture was pelleted by centrifuge and washed 3 times with buffer, lysed by

77 sonication and the results extracted with twice with equal volumes of butanol. The extract was exchanged with methanol and analyzed by LC/MS. No iso-MTM was detected in this experiment, only MTM was found. The presence of MTM in the cells was expected, recall from chapter 1 that the MTM pathway is governed by a feed forward mechanism, and that MtrY is a transcription factor that requires MTM binding for activity. These results suggest that iso-MTM is produced by MtmW and exported directly from the cell.

It is likely that the putative MTM enzyme complex includes MtrAB, if we accept that the absence of iso-MTM in our experiment reflects the absence of iso-MTM in the cytosol

(Figure 2.27).109

78 Figure 2.27 MTM is the only compound from the MTM pathway found in S. argillaceus

WT cultures. This suggests that there is a complex between the PKS (blue) enzymes and the post-PKS (green) enzymes. Intermediates such as premithramycin A2 are never seen, which implies that MTM intermediates are channeled through the entire MTM pathway because discrete enzymes would require pools of intermediates from which to draw.

2.2.13 MtmW Substrate

Previously, we believed that MTM DK was the substrate for MtmW, and that

MtmW was inactive in the absence of MtmOIV. Studies with MTM DK were hampered by its rapid degradation into the shunt products MTM SK, MTM SDK, and MTM SA.58

79 In my attempts to isolate MTM DK, I discovered a new compound using extraction and

LC/MS at higher pHs with an mass to charge ratio of 1183 m/z (Figure 2.28). While attempts to characterize this compound using NMR failed, its presence was only observed in in vitro reactions using a Tris buffering system, leading us to speculate that it was a ketal adduct between Tris and either the C-2’ or C-4’ carbonyl of MTM DK

(Figure 2.29). Using this putative MTM DK-Tris adduct in a reaction with MtmW alone in decreasing pHs showed its conversion to iso-MTM and the MtmOIV shunt products

(Figure 2.30). This both supports the hypothesis that the newly isolated compound is a

MTM DK-Tris adduct, and counters the hypothesis that MtmW is inactive without the presence of MtmOIV. While this evidence would also seem to go against our conclusion from the EMSA experiments and our belief that MtmOIV and MtmW form a PPI, the majority of the products in these reactions were shunt products. In vitro reactions in

MOPS behave the same way as those in Tris, both MtmOIV and MtmW are required for iso-MTM production and MtmOIV alone generates only shunt products. MtmOIV shunt products are never found in S. argillaceus WT, suggesting that MTM DK channeling is still present in vitro and in vivo.

80 Figure 2.28 A. The appearance of the m/z 1183 peak from the MtmOIV-MtmW in vitro reaction as the concentration of formic acid (FA) in the mobile phase was reduced. B. The m/z 1183 peak is abolished when the buffer in the in vitro reaction is switched from Tris to MOPS.

81 Figure 2.29 The proposed mechanism of the formation of the MTM DK-Tris ketal adduct.

Figure 2.30 Isolated MTM DK-Tris adduct was used as a substrate for MtmW in a range of pHs. As the pH dropped, the ketal formation was reversed, allowing MtmW to use MTM DK as a substrate, as shown by the formation of iso-MTM. MTM SA and other MtmOIV shunt products (not shown) remained major products at every tested pH.

82 2.3 Conclusions and Future Directions

In this study, we have demonstrated that MtmW produces an epimer of MTM, iso-MTM, in its reaction. While iso-MTM is not present in cultures of S. argillaceus, it is the product of the combined enzymes in the MTM pathway, and is present a long as an intact cell membrane is missing. Macromolecular crowding was also investigated, and ruled out as a determinant in iso-MTM production. Our bioassay appears to show that iso-MTM has biological activity, however, we believe that this activity that was observed is only due only to its spontaneous transformation into MTM. This isomerization is rapid and complete when crossing the S. argillaceus cell membrane, but still appreciable under basic conditions and in the presence of divalent metal. And finally, the structure of

MtmW is similar to Kvβ, the cytosolic subunit of an ion channel protein. Based on these data, we propose that iso-MTM is the true product of MtmW both in vitro and in vivo, and that it isomerizes into MTM upon travel through MtrAB as it is exported from the cell.

The biological significance of iso-MTM production is unclear, but reasonable hypotheses can be formed. As an inactive form of MTM, iso-MTM provides protection from the producing strain should the putative MTM multienzyme complex be uncoupled, or the resistance enzymes, MtrAB, lose activity. As one of the few antibiotic producers with a single resistance mechanism apparent, it is likely that iso-MTM is an alternate form of self-protection. Production of an inactive precursor exists in the secondary metabolite schemes of other Streptomyces as well. Streptomyces antibioticus is a producer of oleandomycin, which exits the cell in an inactive, glycosylated form.

Produced alongside oleandomycin is OleR, an extracellular glycosidase that removes the

83 inactivating glucose to produce active oleandomycin a safe distance from the S. antibioticus. Similarly, Streptomyces griseus and Streptomyces bikiniensis produce streptomycin intermediates in their phosphorylated forms until just before export from the cell. Neither oleandomycin nor streptomycin are as naturally self-activating as iso-

MTM though, and recall from chapter 1, MTM’s presence in the cytosol of S. argillaceus is required for the production of MTM (Figure 2.31). So perhaps iso-MTM is a relic of and older function of MTM.

84 Figure 2.31 A. As inactive iso-MTM passes through the channel formed by MtrAB, it is enzymatically isomerized to active MTM as a form of self-protection. B. However, active MTM (orange) is required for the production of more MTM via interaction with

MtrY. This, it would seem, counters iso-MTM’s role as a self-defense mechanism.

Translocation of MTM from the environment into the cytosol is likely not mediated solely by MtrAB, as non-producing S. argillaceus strains become sensitive to MTM when mtrA is inactivated.65

Decho et. al. hypothesize that environmental conditions can alter quorum sensing molecules, giving the bacteria that reabsorb them information about their environment

(Figure 2.32).110 In the evolutionary past of MTM its conversion into iso-MTM, coupled with the feed-forward mechanism facilitated by MtrY, the MTM system may have served as a signal molecule that detected extracellular divalent metal concentrations. The

85 production of an antibiotic that also acted as a siderophore would make an efficient system for sequestering iron, magnesium, zinc, and other divalent metals, or alternatively, neutralize divalent metals when their concentrations reach toxic levels. MTM has a high

−8 −11 2 affinity toward divalent metals, ranging from a Kd of 1.14 × 10 to 5.03 × 10 M in

Mg2+ and Fe2+, respectively.111 In this model, addition of divalent metal would cause iso-

MTM to isomerize into MTM, activate MtrY and the associated feed-forward system, leading to full production of the MTM pathway. Fundamentally, the ability of iso-MTM to activate MtrY is important to answer this question. If iso-MTM were unable to initiate the feed forward system, it would indicate that iso-MTM may have a larger ecological role beyond self-protection. In comparison to iso-MTM, actinorhodin has also been shown to act as an intracellular signal, regulating the genes in its own biosynthetic pathway, has its production increased in S. coelicolor when cultured in divalent iron, and can chelate divalent iron. Iso-MTM may also exist at low concentrations just outside the cell early in the cell culture in a medium deficient in divalent metal. Early cultures in media with low Mg2+ were searched for traces of iso-MTM, but we were unable to detect it in any of these conditions. But these were in liquid media, which is not the natural environment for S. argillaceus, and technical limitations prevented the rapid examination of solid media cultures (Figure 2.33). This is, at the moment, only speculation and outside the scope of our current research. But it presents an interesting question for further study.112-116

86 Figure 2.32 A cell produces a signal molecule that induces gene set A. When that signal molecule is altered by environmental conditions, it induces gene set B. The environmental conditions relevant to iso-MTM and MTM according to our research are pH changes and transition elements. Figure adapted from Decho et. al.110

87 Figure 2.33 Proposed possible alternative ecological roles of iso-MTM and MTM

(orange). At low divalent metal concentrations (Mg2+shown here), making a lethal complex through chelation by MTM prevents its use from competing species (green). If iso-MTM rather than MTM is exported from the cell, its isomerization by divalent metal may be detected by S. argillaceus (blue) through its ability to bind MtrY. At high concentrations, MTM may protect S. argillaceus from divalent metal poisoning

MtmW has similarity to Kvβ, a subunit of the Shaker family of voltage‐dependent potassium channels, which form a PPI with the enzyme that forms the ion channel, we propose that a PPI exists between MtmW and MtrAB. Such a channel would also, the experiments suggest, catalyze the isomerization of iso-MTM to MTM. Analogy for this 88 model can be found in the actinorhodin biosynthesis in S. coelicolor. The diffusible, lactone form of actinorhodin, known as γ-actinorhodin, is only found after passage through an RND transporter and in conjunction with a basic extracellular environment. A similar transporter‐mediated catalysis may also be present in the MTM pathway, as iso‐

MTM has not been found in S. argillaceus cultures. Alternatively, as stated above, the local environment around the cell may be sufficient to force isomerization. We know that

Mg2+ is ubiquitous in the environment and essential for S. argillaceus growth. This is unlikely however, because early cultures of S. argillaceus show no iso-MTM, and our experiments show that media alone are insufficient for the complete isomerization of iso-

MTM at such timescales.

The interaction between MtmOIV and MtmW was another intriguing result from this research. A PPI between the enzymes was long suspected due to the instability of

MTM DK, the product of the MtmOIV reaction. The formaldehyde crosslinking experiment and the EMSA assay confirmed this interaction and showed that it was dependent on the binding of NADPH. This, along with the evidence for binding of

MtmGI and MtmGII, and the requirement for TDP-D-4-keto-mycarose to be channeled from MtmC to MtmGIV, suggest that the MTM post-PKS pathway is organized into one or more multienzyme complexes. There is a large amount of literature regarding multienzyme complexes in primary metabolism, PKS, and NRPS pathways, but there has been little research done on post-PKS multienzyme complexes. Aside from the MTM enzymes that require a PPI mentioned above, the landomycin, gaudimycin, and gilvocarcin pathways all contain codependent enzymes, which are indicative of PPIs if

89 not larger multienzyme complexes.117-119 Further study in this area would be beneficial to further combinatorial biosynthesis efforts, and should be investigated further.117-119

2.4 Materials and Methods

Bacterial strains and plasmids. Mithramycin producer Streptomyces argillaceus wild-type strain ATCC 12956 and premithramycin B producer S. argillaceus ΔmtmOIV mutant strain were stocked in the lab. Escherichia coli XL1 Blue (Stratagene) and SoluBL21

(Amsbio) were used for cloning and expression experiments, respectively. Vector pET28a(+) (Novagen) was used for expression in E. coli. PCR amplification and gene cloning PCR amplification of mtmW was carried out using Phusion High-Fidelity PCR

Master Mix (Thermo Scientific). The primers used for cloning of the gene for protein overexpression were MtmW_NcoI_for (5'-

AGCCATGGAGTTCCGAAGCCTTGGCCGAAGT-3') and MtmW_XhoI_rev (5'-

AACTCGAGCTCCTCGTCGTGCACGACGGAGTC-3'). The underlined letters represent restriction site NcoI at the start codon and XhoI located at the predicted stop codon. Cycling parameters started with 1 cycle at 98 °C for 30 s, followed by a touchdown PCR reaction with 20 cycles, with the first cycle at 98 °C for 10 s, 70 °C for

30 s and 72 °C for 30 s, and a decrease of 0.5 °C in each cycle afterwards. Then 10 cycles with an annealing temperature at 60 °C were used followed by a final elongation at 72 °C for 10 min and a final hold at 4 °C. The agarose gel electrophoresis was used for detection of DNA fragments. The PCR fragment was digested with NcoI and XhoI, and subsequently ligated into pET28a(+) cut by the same enzymes, resulting in plasmid pXY23. The plasmid was sequenced (ACGT Inc.) to confirm the sequence. The sequence obtained of the mtmW gene (and MtmW protein) has several differences from the prior

90 GenBank sequence for mtmW. This updated sequence was deposited with the GenBank accession number MK907881.60

Premithramycin B (PreB) production and isolation. Streptomyces argillaceus ΔmtmOIV mutant was plated onto M2 solid media and grown for 7 days at 28 ˚C. The solid media was used to make a 50 mL seed culture of TSB with 7 µL 50 mg/mL apramycin, which was grown for 3 days at 28 ˚C in a shaking incubator. 20 mL of the seed culture was used to seed 400 mL R5A liquid media with 10 µL of 50 mg/mL apramycin in a 2 L flask.

This culture was allowed to incubate for 7 days at 28 ˚C, 250 rpm after which it was centrifuged and the supernatant collected. 100 g/L of Amberlite® XAD16N was added to the supernatant was stirred overnight with 100 g/L of Amberlite XAD16N. The supernatant was filtered with a Whatman® 150 mm #4 filter until the Amberlite® was dry. The resin was stirred for 2 h with 400 mL HPLC-grade methanol and 4 filtered as before. The filtrate was concentrated and PreB was isolated via semi preparative HPLC.60

Enzyme production and isolation. pXY23 was transformed into BL21 (DE3) E. coli cells

(New England Biolabs). The cells were streaked on an LB agar plate with 50 µg/mL kanamycin, and an individual colony was used to start a 50 mL seed culture of LB

(lysogeny broth) and 50 µg/mL kanamycin which was incubated overnight at 18 ˚C, 250 rpms. 5 mL of the seed culture was added to 500 mL TB and 50 µg/mL kanamycin in a 2

L flask, grown at 18 ˚C to OD600: 0.5, and an induced with 0.1 mM IPTG. The temperature was reduced to 30 ˚C and grown overnight. The culture was centrifuged, the cells were lysed with sonication, and purified with TALON® Metal Affinity Resin

(Clontech Laboratories, Inc.). Dialysis was performed with a 20,000 MWCO Slide-A-

Lyzer Dialysis Cassettes (ThermoFisher Scientific) and a buffer, of 20 mM Tris-HCl, 100

91 mM NaCl, 10% glycerol, and 2 mM 2-mercaptoethanol, pH 7.5. The truncated protein,

MtmW-9, was expressed and isolated in the same way with the following exceptions: 5’-

AGCCATGGAGTTCCGAAGCCTTGGCCGA-3’ and 5’-

AAGGATCCCTACTCCTCGTCGTCGACGA -3’ primers were used to amplify mtmW, and BamHI and NcoI restriction enzymes were used. The TB culture was grown at 37 ˚C to the OD600 of 0.5 and the incubation temperature was reduced to 30 ˚C after IPTG induction. MtmOIV was produced as previously described.120 Briefly, mtmOIV was cloned into pRSETb and used to transform BL21 (DE3)pLysS. 1 L of NZCYM with 1 mM ampicillin and 1 mM chloramphenicol was cultured with the cells was grown to the

OD600 of 0.5-0.7, at which time IPTG was added to the final concentration of 1 mM.

The culture was centrifuged, the cell pellet was lysed by sonication, and the lysate was centrifuged and purified on TALON® Metal Affinity Resin (Clontech Laboratories,

Inc.).60

In vitro reaction. A 500 µL reaction was carried out in water in 2 mL microcentrifuge tubes with 50 mM Tris-HCl pH (7.5), 500 µM NADPH, 250 µM FAD, 5 µM MtmOIV,

15 µM MtmW, 300 µM PreB. The reaction was incubated at 30 ˚C for 1 hr. Iso-MTM was isolated with 10 mL by volume Amberlite® XAD16N in an Econo-Pac®

Chromatography Column (Bio Rad). The column was washed with 2 volumes of water and dried. iso-MTM was eluted in HPLC-grade methanol and purified by analytical

HPLC using a gradient of acetonitrile in water, both containing 0.1% formic acid. The

LC detection was carried out at 410 nm.60

LC/MS. All of the LC/MS data in this chapter was obtained from a Waters electrospray

LC/MS. Separation of compounds was performed using a mobile phase gradient of 30%

92 acetonitrile in water to 70% acetonitrile in water over 20 min. Unless otherwise indicated, the mobile phase also contained 0.1% formic acid.

EMSA assay. The EMSA assay samples were prepared in 40 mM Tris-HCl pH 8.5 with 1 mM NADPH, 0.5 mM PreB, 5 mM MgCl2, at concentrations of MtmOIV and MtmW specified in the text. The samples were incubated at 21 °C for 20 min and then run on a

1.7% agarose gel. The gel was prepared and run in 25 mM Tris pH 8.5 and 190 mM glycine. Each assay was performed in duplicate. Band intensities corresponding to the free MtmW were quantified with Volume Tools (Bio-Rad).60

Formaldehyde Crosslinking. All reactions were performed with 50 mM sodium phosphate pH (7.5), 500 µM NADPH, 250 µM FAD, 300 µM PreB, and 1% formaldehyde. Reactions are performed as indicated with 5 µM MtmOIV, 15 µM MtmW, and/or 15 µM TMK. Reactions were performed at 37 °C for the specified time period.

The reactions were quenched in 125 mM glycine and analyzed by SDS-PAGE using a 1 mm, 12% SDS-PAGE gel at 120 V for 75 min.

MgCl2 isomerization assay. 100 µL aliquots of 50 mM Tris-HCl pH 4, 6, and 8 were prepared. To these, iso-MTM was added to 10 µM with and without 50 mM MgCl2 (or at specified MgCl2 concentrations). These samples were incubated at 30 ˚C for 18 hours, extracted with butanol, dried, and brought to a volume of 50 µL with HPLC-grade methanol. The samples were then analyzed by LC/MS (at 410 nm).60

Antibiotic assays. A disk diffusion assay testing bacterial growth inhibition by iso-MTM was performed on LB agar media with Salmonella enterica ATCC 10708 and

Staphylococcus aureus ATCC 6538. iso-MTM, MTM and ampicillin were diluted to 1

93 mg/mL in methanol, each filter paper disk was sterilized before the addition of 10 µL of iso-MTM, MTM, ampicillin, or methanol. The disks were placed on LB agar plates, and

100 µL of cultures of S. enterica or S. aureus (at OD600 = 0.2) was evenly spread on the agar. The plates were incubated at 37 ˚C overnight and the zones of inhibition were measured by a metric ruler. An MIC assay was performed on iso-MTM by the double dilution method. Approximately 5 x 105 S. aureus cells were added to 1 mL LB cultures with iso-MTM at concentrations ranging 0.03-4 µg/mL. The cultures were incubated overnight at 37 ˚C and the OD600 was measured by spectrophotometer. The cultures originally containing 4 µg/mL iso-MTM were extracted with ethyl acetate and inspected by LC/MS. NMR. 1.8 mg of iso-MTM was dissolved in 180 µL acetone-d6 for NMR experiments with 600 MHz NMR. 1 H, HMBC, HSQC, COSY, and TOCSY NMR experiments were performed.60

Analytical ultracentrifugation. Sedimentation velocity (SV) experiments were performed in a Beckman Optima XL-I analytical ultracentrifuge (Beckman Coulter, Indianapolis,

IN) equipped with a 4-hole AnTi-60 rotor. Samples for analytical ultracentrifugation

(AUC) were in 40 mM Tris pH 8.0, 100 mM NaCl, 1 mM TCEP. AUC-SV studies were performed with MtmW at 6 µM and 17 µM. AUC-SV experiments were carried out at 20

°C with a rotor speed of 35,000 rpm and data collection performed in the intensity mode at 280 nm. About 500 scans were collected for each sample. Partial specific volumes, molecular masses, and van Holde-Weischet analysis were performed using the Ultrascan

III software.121 AUC-SV datasets were analyzed with SEDFIT 16.1122 using the continuous size distribution c(S) model. All c(S) distributions were 6 calculated with

94 fitted f/f0 and meniscus positions. Maximum entropy regularization for all fits was

0.95.60

Crystallization. X-ray diffraction data collection and crystal structure determination. For crystallization, MtmW (N-terminal three-residue truncation, ∆N3) was further purified on an S200 HR Sephacryl size exclusion column (GE Healthcare) by using a BioLogic

DuoFlow HPLC system (Bio-Rad), which had been equilibrated in 40 mM Tris-HCl, 100 mM NaCl, 2 mM β-mercaptoethanol, pH 8.0. Fractions were collected and concentrated to a final concentration of 12 mg/mL using Amicon Ultra-15 centrifugal filter device

(Millipore). The crystals of apo-MtmW were grown by vapor diffusion in hanging drops containing 1 μL of protein and 1 μL of the reservoir solution (0.1 M Na citrate, pH 6.0,

16-20% (w/v) PEG 1500) and incubated against 1 mL of the reservoir solution at 21 ˚C.

To obtain crystals of MtmW with bound NADP+ or NADP+ and PEG in the active site,

NADP+ (1 mM) or NADP+ (1 mM) and iso-MTM (0.5 mM) were added to the protein prior to setting the hanging drops. The crystals were gradually transferred into the reservoir solution additionally containing 20% (v/v) glycerol and then quickly immersed in liquid nitrogen. The X-ray diffraction data were collected on synchrotron beamline

22ID at the Advanced Photon Source at the Argonne National Laboratory (Argonne, IL).

The data were processed with HKL2000 suite.123 The crystal structure of apo-MtmW was determined by molecular replacement by using program Phaser with the crystal structure of putative oxidoreductase from Salmonella enterica (PDB ID: 3ERP)73 as a search model. The crystal structures of MtmW-NADP+ and MtmW-NADP+ -PEG complexes were determined analogously, by using the crystal structure of apo-MtmW as a search model. The ligands were built unambiguously into strong Fo-Fc omit electron density.

95 The model building and refinement was performed iteratively with Coot124 and

Refmac125, respectively. The crystal structures of apo-MtmW, MtmW-NADP+ and

MtmW-NADP+ -PEG complexes and respective structure factor data were deposited into the Protein Data Bank with accession codes 6OVQ, 6OVX and 6OW0.60

Total cell assay. Lysate cultures of S. argillaceus WT, S. argillaceus ΔmtmGIII, and S. argillaceus ΔmtmGI were grown to stationary phase 50 mL tryptic soy broth media in

250 mL baffled flasks. PreB was cultured and purified as previously described. The 1 mL reactions were carried out in protein lysis buffer (50 mM tris-HCl pH: 7.5, 50 mM NaCl,

10% glycerol) in 2 mL Eppendorf tubes with 500 µM NADPH, 250 µM FAD, and 300

µM PreB. The reactions were incubated at 30 ˚C for 1 hr. After the reactions were complete, they were extracted twice with an equal volume of butanol, evaporated and reconstituted with methanol. All reactions were analyzed by LC/MS in negative mode and analyzed at 405 nm and at 1083 m/z by extracted ion count.

MtmW Substrate Assay. A 500 µL reaction was carried out in water in 2 mL Eppendorf tubes with 50 mM Tris-HCl pH (7.5), 500 µM NADPH, 250 µM FAD, 5 µM MtmOIV,

300 µM PreB. The reaction was incubated at 30 ˚C for 1 hr. Iso-MTM was isolated with

10 mL by volume Amberlite® XAD16N in an Econo-Pac® Chromatography Column

(Bio Rad). The column was washed with 2 volumes of water and dried. iso-MTM was eluted in HPLC-grade methanol and purified by analytical HPLC. The mobile phase of

HPLC and LC/MS was gradient of acetonitrile in water, and the pH was measured at about 6.5, or a gradient of acetonitrile and 8 mM ammonium acetate in water adjusted to pH: 5.5, as indicated. The LC detection was carried out at 410 nm. In vitro reactions done

96 in MOPS were performed as above, replacing 50 mM Tris-HCl pH (7.5) with 50 mM

MOPS pH (7.5)

Extraction of the putative MTM DK-Tris adduct was done using pH adjusted butanol

(butanol was shaken overnight with an equal volume of 50 mM Tris-HCl pH (7.5), centrifuged and the butanol fraction reserved). Purification was done using preparative

HPLC, the mobile phase of which was a gradient of acetonitrile and 8 mM ammonium acetate in water adjusted to pH: 5.5 in vitro experiment with the putative MTM DK-Tris adduct and MtmW. 100 µL reactions were carried out in water in 2 mL Eppendorf tubes with 50 mM MOPS, 500 µM

NADPH, 250 µM FAD, 15 µM MtmW, 200 µM MTM DK-Tris adduct. The pH ranged from 8.25 to 6.5 as indicated.

Software. PyMol and QuteMol were used to display the structures of the proteins in this chapter. Chemical structures were made using ChemDraw 20.0. LC/MS chromatograms were generated using Waters' software, except in the case of Figure 2.23 where Prism software was used to compile several LC/MS experiments.

97 Interactions and Channeling of the Substrates in the Mithramycin Post-PKS Pathway

3.1 Background

Protein-protein interactions (PPI) are well characterized in primary metabolism but less studied in secondary metabolism, particularly in the case of post-PKS tailoring enzymes. The term metabolon describes a complex of enzymes that interact in such a way as to channel intermediates from one active site to another without releasing them into the surrounding environment. The benefits of metabolon formation include channeling and protecting labile intermediates, and protection of the organism from toxic intermediates. In cell respiration, enzymes of the glycolysis and citric acid cycle pathways each form complexes to sequester their respective intermediates, which can be used in other cell processes, for the vital role of energy production. Citrate synthase condenses acetyl-CoA and oxaloacetic acid into citric acid. The equilibrium of this reaction would make the forward direction unfeasible based on the low concentrations of oxaloacetic acid believed to be in the cell. This is overcome through channeling, which raises the local concentration of the substrates sufficiently for the reaction to proceed.127

The principle of metabolon formation is also established in plants. Sorghum bicolour produces dhurrin as a secondary metabolite. This cyanogenic glucoside is channeled to protect the plant from the hydrolysis of p-hydroxymandelonitrile into a shunt product and hydrogen cyanide.126-128

Secondary metabolon formation is of interest to the field of pharmaceutical sciences in the context of combinatorial biosynthesis, the process where enzymes are added, subtracted, or replaced to a secondary metabolite pathway with the intention of producing altered compounds. This process is hampered by low yields and low substrate affinity.

98 Increasing our knowledge of PPIs in medicinal secondary product pathways is important to facilitate combinatorial biochemistry, which in turn expands the chemical space of natural products and aids drug discovery strategies.129, 130

In the previous chapter, we discussed the characterization of the structure of

MtmW, the reaction it catalyzes, and its interaction with MtmOIV. The subject of this chapter will be an in-depth look at the interaction between MtmOIV and MtmW. From our previous studies, we know that MtmOIV binds to MtmW with greater affinity in the presence of its cofactors and substrate, PreB. We know that MtmW forms an octamer in vitro, and MtmOIV forms a dimer. And we also hypothesize that both MtmW and

MtmOIV interact with other enzymes in the MTM post-PKS pathway.120

The strategy for characterizing this part of the MTM post-PKS pathway includes identifying the MtmOIV, MtmW PPI interface through 19F NMR and protein footprinting, and identifying any additional PPI partners for each enzyme using a label transfer experiment.

3.2 Experimental Design and Results

3.2.1 19F NMR

In order to determine the location of the PPI interface between MtmOIV and

MtmW, 19F NMR was used. Fluorine is an excellent nucleus to study proteins using

NMR. It is approximately the same size as hydrogen, 1.35 Å and 1.2 Å, respectively, meaning it does not generally perturb the structure of the enzyme. Fluorine also is rarely present in biological systems, which reduces background and contaminating signals.

Fluorine also exists in 100% abundance in the 19F isotope, which has a 1/2 spin necessary for NMR. It also produces signals which are about 83% the intensity of 1H NMR. 99 Because 19F is the only natural isotope for fluorine, it is relatively inexpensive in comparison to 15N or 13C reagents. The chemical shift of fluorine is also influenced by a lone pair of electrons. This means that 19F NMR is exceptionally sensitive to its local environment. And finally, the chemical shift of 19F NMR stretches over nearly 1000 ppms, which somewhat ameliorates the peak broadening in larger proteins, as the peaks are generally farther apart than what is found in other NMR experiments using other nuclei.131-133

In our original experiment, we had planned to block the shikimate pathway of

MtmW-expressing E. coli BL21 with glyphosate, and feed 5-fluorotryptophan in place of non-fluorinated tryptophan, which would have resulted in 19F-labeled MtmW. This strategy suffered for two reasons. The first was low production, we were unable to induce growth of MtmW with fluorinated tryptophan. Preliminary NMR experiments showed that we had no incorporation of 5-fluorotryptophan in our purified MtmW. The second problem was coverage. MtmW has 5 tryptophans, and most of them are buried inside the protein structure according to its crystal structure. MtmOIV contains 9 tryptophan residues, most of which are clustered in the C-terminal domain, away from the active site.

Inspired by our lab’s success in covalent side chain labeling in our protein footprinting experiments, we formulated a labeling strategy to post-translationally add fluorine tags to the surface of MtmOIV and MtmW. The aim was to select two different tags, one for MtmOIV and one for MtmW, that would have different enough chemical shifts that we could obtain Nuclear Overhauser Enhancement (NOE) data if necessary to produce distance constraints to further enhance our data.

100 Acetic anhydride is often used to label the primary amine on the lysine side chain.

My hypothesis is that substituting trifluoroacetic anhydride (TFAA) for acetic anhydride would produce a similar label with a trifluorinated tag. MtmW has six lysine residues, two of which are buried in the interface of the octamer units, one in the cavity formed by the octamer, and three exposed to the solvent. Of the solvent exposed lysines, K218 and

K250 are adjacent to the active site where we expect MTM DK to be channeled.

MtmOIV also has four solvent exposed lysines, one of which, K81, is adjacent to the two aspartate residues whose labeling was blocked upon co-labeling MtmOIV and MtmW in our preliminary protein footprinting experiments (see below, section 3.2.2). Based on these data, we labeled both MtmOIV and MtmW with TFAA.

To test whether TFAA was able to label in the same manner as acetic anhydride, we followed a protocol for acetic anhydride labeling by Izumi et. al.134 To test the labeling efficiency we performed three and six rounds of TFAA labeling on MtmOIV and used an assay with 2,4,6-trinitrobenzene sulfonic acid (TNBSA), a reagent that detects free amines by the absorbance at 335 nm, and compared labeled and unlabeled MtmOIV.

We found that after 3 rounds of labeling, 85.0% of the free amines were labeled with

TNBSA, and after six rounds of labeling, only 78.0% of the primary amines were labeled with TNBSA as compared to unlabeled MtmOIV. This suggested that 15% and 22% of the primary amine sites were blocked from TNBSA labeling by the previous labeling reactions with TFAA, respectively.

MtmW was labeled using the same procedure as described above and showed a

TFAA incorporation of 49% of the free amines according to the TNBSA assay. However, when the labeled MtmW was analyzed by NMR, there were no peaks that corresponded

101 to labeled protein. As the size of a protein increases, its rotational correlation time (τc), the average time it takes for a molecule to rotate one radian, increases. As the τc increases, the peak width and signal-to-noise ratio both decrease due to more efficent transverse relaxation. In practical terms, this means that large proteins should have broad peaks when using NMR.135 The 15 kDa protein bovine protein disulphide isomerase has a peak width of 0.25 ppm, while the 94 kDa homodimer 3-hydroxybenzoate 6-hydroxylase has a peak width of about 3 ppm.136, 137 The MtmW octamer is over 280 kDa, so we expected our peak width to be considerably larger than 3 ppm. The results of our NMR experiments showed a multiplet of less than 0.1 ppm (Figure 3.1). For this reason, these peaks cannot be derived from our labeled MtmW, and we suspect they are from triflouroacetic acid that resulted from the labels disassociating from the lysines of MtmW

(Figure 3.2).

The absence of TFAA labeling in the literature also suggests that it is not a useful strategy. Acetic anhydride labeling is well documented,138, 139 and it would be reasonable to assume that other labs would have attempted TFAA labeling as it is the logical extension of acetic anhydride labeling. For these reasons, we suggest that TFAA labeling is not viable for 19F NMR studies.

102 Figure 3.1 19F NMR of TFAA labeled MtmW with and without MtmOIV

Figure 3.2 A. TFAA labeling reaction. B. Proposed mechanism of TFAA label disassociation.

The next attempt to covalently label MtmOIV and MtmW with a fluorinated tag was adapted from Abboud et. al.140 Briefly, cysteine mutants were generated on the surface of

103 MtmOIV on each face of the enzyme (Figure 3.3 A-C). These new cysteines were to be labeled with 3-bromo-1,1,1-trifluoroacetone (BTFA) and analyzed by 19F NMR (Figure

3.3d). While the generation of vectors with the MtmOIV cysteine mutants was successful, none of these mutants could be expressed in a soluble form. We hypothesize that this was a result of aggregation from the formation of disulfide bonds during protein expression or purification. However, this project was abandoned before the expression conditions could be optimized due to our discovery that MtmW formed an octamer, and the difficulties of performing NMR on large enzyme complexes made the experiments unfeasible.

Figure 3.3 A-C. Three sides of the MtmOIV dimer showing the proposed cysteine mutations (magenta). * indicates nearest amino acid if the mutation fell on a flexible loop that was not included in the crystal structure. PreB is shown as green sticks. D.

The 3-bromo-1,1,1-trifluoroacetone (BTFA) reaction with cysteine.

104 3.2.2 Protein Footprinting

Protein footprinting is the differential labeling of the exterior of a protein based on the loss of portions of surface area caused by PPIs. The methods of labeling include the exchange of surface exposed amide protons for deuterium, modification by hydroxyl radicals, and side chain covalent labeling. The labeled proteins are then digested into peptides and analyzed by mass spectrometry.

Hydrogen deuterium exchange (HDX) takes advantage of the natural exchange of the amide protons that occurs on the protein backbone at physiological conditions. By introducing deuterium to the solvent, typically in the form of D2O, the amide protons can exchange with deuterium, increasing the mass of each labeled position by 1.006 Da. This method has high resolution because the amide protons of all amino acids, apart from proline, can be exchanged, and as an isotope, deuteration has little effect on protein conformation. The method’s weaknesses chiefly involve the rapid back exchange of the deuterated amides to protons. This weakness can be mitigated by keeping the pH at 2.5, and by keeping the exchanged protein cold. Nevertheless, rapid, high-pressure liquid chromatography methods must be employed to capture HDX labeling, and a protease active in low pHs, such as pepsin, must be used for digestion into peptides.141, 142

Hydroxyl radicals are generated using radiolysis, photolysis, or chemical oxidation of water using Fenton’s reagent. The hydroxyl radicals attack the side chains of the amino acids forming the protein based on solvent accessibility and amino acid identity. For example, cysteine, methionine, and tryptophan are all sensitive to hydroxyl radical attack while asparagine, alanine and glycine are relatively insensitive. Unlike

HDX, the labeling is irreversible, which expands the options available for digestion and

105 analysis. Also unlike HDX, the changes in mass are more variable. Arginine, for example, can have a change in mass of -43.0534, +13.9793, or +15.9949 depending on the manner in which the hydroxyl radical attacks.143, 144 This variability can make the results of such experiments difficult to analyze.

Covalent labeling uses chemical reactions to label the amino acid side chains.

Different reactions are used to label different side chains, for example, vicinal dicarbonyl compounds such as 2,3-butanedione (BD) is used to label arginine, while tryptophans can be labeled using N-bromosuccinimide, and o‐nitrophenylsulfenyl. This specificity simplifies the analysis, as there are small subsets of amino acids available for labeling and the labels provide a single change in mass. Like hydroxyl radical labeling, there is very little back exchange, allowing for more freedom in choosing digesting and separating techniques.

With the resources and expertise at our disposal, we have elected to use a covalent labeling strategy to discover the PPI interface of MtmOIV and MtmW. In our experiments, we labeled aspartate and glutamate with N-(3-(dimethylamino)propyl)-N′- ethylcarbodiimide hydrochloride (EDC) and glycine ethyl ester (GEE), and histidine, lysine, tyrosine, serine, and threonine with diethylpyrocarbonate (DEPC).

Several pilot experiments were performed to optimize the conditions of DEPC labeling, yet we experienced persistent polyethylene glycol (PEG) contamination (Figure

3.4), which causes the suppression of the ion signal from our desired analytes and renders the LC/MS/MS experiment useless. There are three proposed methods by which PEG suppresses the ion signal, competition for charge, the decrease of evaporation of the mobile phase droplets, and interactions between PEG and the other analytes in the gas

106 phase.145 Regardless of the mechanism, PEG contamination prevented us from obtaining useful information from the DEPC labeling. We hypothesize that the PEG contamination came from the DEPC itself, because the materials and methods used to prepare DEPC samples for LC/MS/MS were common to the EDC/GEE and label transfer experiment samples prepared for mass spectrometry, yet only the DEPC samples contained PEG contamination.

Figure 3.4 PEG contamination in a DEPC labeled trypsin digest of MtmW. Red arrows highlight 44 Da intervals indicating PEG contamination.

The EDC/GEE labeling had promising results in our pilot experiments, there was a clear change in the labeling pattern depending on whether the enzymes were labeled together or separately. Unfortunately, because of time and budget restrictions, we were unable to confirm the results of the pilot experiments in a statistically significant way.

Therefore, the following merely describes our preliminary data, and the interpretation is contingent on further experiments confirming these results.

107 When labeled with EDC/GEE individually, 11 amino acids on MtmOIV, and 15 amino acids on MtmW were labeled. When the labeling took place during the

MtmOIV/MtmW reaction, 10 and 9 amino acids were labeled on MtmOIV and MtmW, respectively. When the enzymes were labeled together, three new labels appeared on

MtmOIV, and four were no longer present, while on MtmW no new labels appeared and six were not present (table 3.1).

Table 3.1 A summary of the EDC/GEE labeling results. The IonScore is the is the (10)-

Log10 of the probability that the observed match is a random event

Looking closer at the locations of the changes in EDC/GEE labeling of MtmOIV, we see that E114 is unavailable for labeling in the presence of MtmW. This is most likely, because it is in the active site of MtmOIV and not on the surface, due to long distance changes in the three dimensional structure of the enzyme as it binds to MtmW.

The two most interesting locations on MtmOIV are D94 and D99. They are away from the MtmOIV dimer interface, allowing us to rule out changes in the MtmOIV/MtmOIV

108 dimer interaction, and they are close to the active site of MtmOIV. This, we suspect, is the footprint that forms when MtmOIV and MtmW interact.

The final labeling site that was abolished was E123. Along with E75 and E119,

E123 forms part of the MtmOIV dimer interface, a location which we would expect to be obscured if our current dimer structures were correct. Additionally, E425 and E431 were not labeled until MtmOIV and MtmW were labeled together –meaning they became solvent-exposed upon MtmOIV binding to MtmW. These labels could arise from long range changes in tertiary structure, but they could also be an result of changes in the

MtmOIV dimer interface. We also hypothesize that the dimer interface is different in solution than has been shown in crystallization studies.120 If MtmOIV dimerized in a way that included E425 and E431 and then split into monomers when it bound to MtmW, this would explain the appearance of labels on E425 and E431.

109 Figure 3.4 Labeling of MtmOIV. Red spheres are amino acids that are labeled only when MtmOIV is alone. Yellow spheres are only labeled when MtmOIV is in contact with MtmW. Green spheres are labeled in both conditions. Supposed PPI interface involving D94 and D99 (black arrows)

110 Figure 3.5 Labeling results that question the established MtmOIV dimer. E75 and

E119 (green) appear to be blocked from labeling by the MtmOIV dimer interface. E123

(red) is only labeled when MtmOIV is alone in solution, and E425 and E431 (yellow) are labeled only when MtmOIV is in solution with MtmW. For colors see Figure 3.4

The labeling of MtmW is more intuitive but less reliable. For clarity, we can divide the labels on MtmW into four regions: the active site, the top, the interior, and the belt regions. The labels on the belt region are the least effected by the addition of

MtmOIV, it is unlikely that MtmOIV binds in this region. Changes in the interior are almost certainly from changes in tertiary structure or loss of access from MtmOIV binding close to the gaps that feed the interior, we do not expect MtmOIV to be able to access the interior of the octamer. The change in labeling around the active site is

111 expected and, we believe, evidence of a protein footprint (Figure 3.6a). The labels on the top region of MtmW are the most ambiguous because labeling is abolished next to labels that appear unaffected. A closer look at the IonScores (table 3.1) show that amino acids in the top region, E159, E190, E192, and E306, all have lower IonScores when MtmOIV is present. While the IonScores do not directly correlate with the proportion of labeled or unlabeled amino acid, lower scores indicate a lower probability of amino acid labeling, and can therefore provide qualitative data on labeling percentage. From these data, we suspect that these amino acids are loosely associated with the MtmOIV/MtmW PPI interface (Figure 3.6b). This work was done with the assistance of Dr. Redding Gober.

Figure 3.6 Labeling of MtmW. Red spheres are amino acids that are labeled only when

MtmW is alone. Green spheres are labeled in all conditions. A. Proposed PPI interface

B. Proposed PPI interface. Green amino acids in this area have lower IonScores when

MtmOIV is in solution.

112 3.2.3 Label Transfer

Label transfer is a modified crosslinking experiment where a tri-functional crosslinking reagent is used to transfer a label from one protein to its PPI binding partner.

In our experiments we used Sulfo-SBED (Thermofisher Scientific), which is used to transfer a biotin tag from one protein to another. The first functional group, N- hydroxysuccinimide (NHS), forms an amide bond with the primary amine group of a lysine on the “bait” protein. Aryl azide is the second functional group and under exposure to ultraviolet light, it forms covalent bonds with nucleophilic or active hydrogen groups on the “prey” protein. A reducible disulfide separates the NHS portion of the molecule from the aryl azide, biotin portion. Reduction of this bond leads to the labeling of the prey protein. The prey proteins can be purified from the reaction using streptavidin and characterized using western blotting or MS/MS. In our experiments, we planned to use this method to identify other proteins that interacted with the MtmOIV-MtmW complex.

In our labeling experiments MtmOIV and MtmW were used as bait proteins and S. argillaceus lysate was used as the source for prey proteins. In this experiment, we found that MtmW was able to pass the biotin label to MtmOIV, as expected (table 3.2). No other proteins were identified as PPI partners with MtmW, which appears to contradict our previous hypothesis that MtmW bound to the MtrAB channel. However, a closer look at the structure of MtmW reveals that only the N-terminal amine and three of the six lysines which react with the NHS group that is used to label the bait protein are exposed to the solvent on the exterior of the octamer. These primary amine groups are all clustered around the active site, where MtmOIV binds according to our preliminary

113 protein footprinting experiments. The remainder of the lysines are in the pocket formed by the octamer (Figure 3.7)

Figure 3.7 Two views of the MtmW octamer with one monomer removed for clarity. A.

The face of MtmW that we believe forms a PPI with MtrAB has no adjacent primary amines (magenta) B. The remainder of the primary amines are clustered around the active site defined here by the co-crystallized NADPH and polyethylene glycol

(orange).

114 Table 3.2 A summary of the label transfer results. Score is the measure of how well the

MS/MS data matches the sequence of the proteins that were queried. Coverage is the percentage of the amino acids in the protein that were identified. PSMs are the number of MS/MS spectra that were matched to the protein.

Therefore our results do not contradict our earlier hypothesis, they simply confirm

MtmOIV’s association to the MtmW active site. The label transfer experiment using

MtmOIV as the bait protein also showed that it had a PPI with MtmW, but unexpectedly also showed binding to MtrA. The data from the EDC/GEE labeling suggests that

MtmOIV interacts with the top region of MtmW (Figure 3.6b). Because of this, the N- terminus of MtmOIV should be in range to interact with MtrA. We should be cautious about assigning a PPI between MtmOIV and MtrA, however. Only a single, seven amino acid peptide was isolated from the MtmOIV bait experiments, and this was assigned to

MtrA at a low confidence level. Further experimentation will be necessary to confirm the

115 tentative results we found in these label transfer experiments. The label transfer experiments confirm the existence of the MtmOIV/MtmW PPI that was discovered in our previous work detailed in chapter two. Unfortunately, we were unable to establish definitive proof of a PPI between MtmOIV or MtmW with any adjacent enzymes in the pathway aside from MtrA. We believe that the primary reason for this is the number and location of the lysine residues that are available to react with the N-hydroxysuccinimide functional group of sulfo-SBED. MtmOIV and MtmW require NADPH for binding, it is possible that other enzymes in the complex form in a dynamic manner where the presence of their substrate is required to induce their PPI. For example, previous research shows that there is a likely PPI between MtmOIV and MtmGI and/or MtmGII (Figure

1.19b)52, however, this PPI may be dependent on the presence of premithramycin A3,

TDP-D-olivose, premithramycin B, or intermediates and substrates that are further downstream. Our experiments used crude lysate with added NADPH, FAD, and premithramycin B, and while we can assume all of the required components to form the

MTM complex were present, the quantity could have been insufficient to sustain the complex long enough to be captured by the label transfer experiment.

3.3 Conclusions and Future Directions

The results from the experiments in chapter three are all preliminary, so we must be cautious in assigning definitive results on the interactions and interfaces between

MtmOIV, MtmW, and other mithramycin post-PKS enzymes. We can say, however, that the existence of a PPI between MtmOIV and MtmW was confirmed by the label transfer and protein footprinting experiments that were described in this chapter. We can also

116 speculate on what our preliminary results would mean if they were supported by more robust data, as means to provide ideas for future research.

Examination of the proposed PPI interface between MtmOIV and MtmW suggests that MtmOIV binds as a monomer to the MtmW octamer (likely in a 1:1 or 1:2 ratio) with the MtmOIV C-terminal domain sitting on top of the MtmW top region (Figure 3.8).

This not only fulfills the constraints put upon the interface that were imposed by the protein footprinting experiment, but also puts the active sites in close proximity which is required for the protection of MTM DK (Figure 3.9). While there are gaps in the handmade docking structures, the structures do not account for the flexible nature of proteins in solution. The C-terminal domain and the middle domain, from T183 to A275, of MtmOIV appear to be especially mobile.

117 Figure 3.8 Three angles on the proposed PPI between MtmOIV (blue) and the MtmW octamer (grey). This Figure only shows a single MtmOIV for clarity. Red spheres are amino acids that are labeled only when MtmOIV or MtmW are alone. Yellow spheres are only labeled when MtmOIV is with MtmW. Green spheres are labeled in all conditions. Note: This Figure was docked by hand, the only data that were considered were the proximity of the active sites, and the results from the label transfer and protein footprinting experiments.

118 Figure 3.9 Detail on the active sites of MtmOIV and MtmW in the proposed PPI model.

MTM DK travels from the active site of MtmOIV (blue) to the active site of MtmW

(grey). The magenta sticks are co-crystallized PreB with MtmOIV and co-crystallized

NADPH and PEG with MtmW.

Expanding on the homology between MtmW and the β subunit of Kv1 potassium channels can help us explain the results of the appearance of MtrA in the label transfer experiments where MtmOIV is the bait protein. If we suppose that the MtmW interaction with MtrAB is similar to the structure of the Kv1 potassium ion channel, and the proposed PPI interface between MtmOIV and MtmW is correct, MtrA should interact with MtmOIV (see Figure 3.10 for a model). In summary, the proposed PPI interface agrees with both the experiments in this chapter, as well as the requirement to channel

MTM DK between the active site of MtmOIV and the active site of MtmW.

119 Figure 3.10 The Kv1 potassium ion channel (PMDB structure: 2A79). The Kvβ subunit

(green) is structurally homologous to MtmW. The remainder of the Kv1 ion channel

(purple) may also be homologous to the MTM pathway enzymes which are listed on the right. The proposed MtmOIV binding site on the left (red box) would fit into this system in a manner where is is accessible to MtrA, which is in agreement with our label transfer data

Future directions for this project include verifying the preliminary data, and expanding the experiments to other enzymes in the MTM pathway that are known to have

PPIs. To confirm the preliminary data, the protein footprinting experiments should be optimized and expanded to include BD labeling. A more rigorous crosslinking strategy could also be employed to find the PPI interface. A trypsin digestion and LC/MS/MS

120 analysis of the formaldehyde crosslinked MtmOIV/MtmW samples from chapter 2 may give interesting results, as would the mass spectrometry of the label transfer experiment before the crosslinker was reduced.

The other enzymes in the MTM that should be investigated for PPIs are MtmC and MtmGIV, and MtmGI and MtmGII. Recall from chapter 1 that MtmC produces an instable intermediate, TDP-D-4-keto-mycarose, that cyclizes if it is not protected. This vulnerable intermediate requires channeling from MtmC to MtmGIV to prevent the formation of shunt products. MtmC and MtmGIV are also bifunctional enzymes whose reactions are separated by MtmGIII and MtmMII. MtmTIII also acts directly after the second MtmGIV reaction. Because of the complex order in which these enzymes act, it is likely that one or more of these enzymes are also in a complex with MtmC and MtmGIV

(Figure 3.11).

Figure 3.11 A schematic of the proposed MTM trisaccharide synthesis complex.

We also know that MtmGI and MtmGII form a complex, MtmGI is nonfunctional without MtmGII. The MtmGI/MtmGII complex also interacts with MtmOIV because S.

121 argillaceus mtmOIV knockouts disrupt the interaction between MtmOIV and

MtmGI/MtmGII and 3A‐deolivosylpremithramycin B is released and is able to be detected in extracts, so it is likely that MtmGI and MtmGII are part of the MtmOIV and

MtmW complex as well (Figure 3.12)

Figure 3.12 A schematic of the proposed MTM complex involving MtmGI-MtrB. MtmW is shown as a tetramer for clarity, we currently hypothesize that MtmW remains an octamer in vivo.

3.4 Materials and Methods

Enzyme production and isolation. MtmOIV and MtmW were isolated as previously described. Briefly, mtmOIV and mtmW were ligated into pET28a, and transformed into E. coli BL21. Transformants were grown in LB medium at 37 °C to OD600: 0.5-0.7 when they were induced with 0.1 mM IPTG. The cultures were grown overnight at 30 °C and

122 when they were centrifuged and the cells lysed by sonication. The lysates were purified with TALON® Metal Affinity Resin (Clontech Laboratories, Inc.), and dialyzed to remove imidazole. All mutants were constructed using overlap PCR, and expressed and purified in the same manner

Enzyme Labeling. The enzymes that were labeled in the protein footprinting assays,

MtmOIV and MtmW, were labeled individually or labeled together in solution. To match the quality of the data from the LC/MS/MS of the co-labeled enzymes, individually labeled enzymes were pooled after quenching and before trypsin digestion. Unless otherwise specified, the ratio of MtmOIV to MtmW was 1:2 based on optimization of the in vitro activity assay of MtmW that is described on page 88 of this document. In the activity assay we found that iso-MTM was the sole product of the MtmOIV, MtmW complex at a ratio of 1:2 and formation of shunt products appeared as the proportion of

MtmW was reduced from this point. The use of the 1:3 ratio in the activity assay was to ensure that there were no shunt products present even if a portion of the MtmW degraded in storage.

Lysine residues of the enzymes were labeled with trifluoroacetic anhydride (TFAA).

MtmOIV or MtmW was suspended in 200 mM phosphate buffer pH: 8, with 3% glycerol,

100 mM NaCl, and diluted to 10 μM. 100 mM TFAA was freshly prepared by diluting into 200 mM phosphate buffer pH 8. Equal volumes of 100 mM TFAA and 10 μM enzyme solution were combined and allowed to react for 20 min at 4 °C, after which the enzyme was exchanged into fresh buffer by centrifugal filtration. This process was repeated 6 times. Labeling coverage was assayed using TNBS according to the manufacturers protocol (ThermoScientific Catalog # 28997).

123 Aspartate and glutamate were labeled using EDC/GEE according to the following protocol. Each 300 µL reaction was done in 25 mM MOPS, pH: 7.4 and had a final concentration of 6 µM of MtmOIV and/or 12 µM MtmW, 10 mM NADPH, 10 mM PreB, and had a GEE/EDC/total protein concentration of 1,500:60,000:1. Samples were incubated at room temperature for 30 min, and quenched by adding formic acid to a final concentration of 0.1% v/v.

Histidine, lysine, tyrosine, serine, and threonine residues were labeled using DEPC according to the following protocol. Each 500 µL reaction was done in 20 mM MOPS, pH: 7.4 and had a final concentration of 1 µM of MtmOIV and/or 2 µM MtmW, 10 mM

NADPH, 10 mM PreB, and 26.4 µL of DEPC. This was incubated at 37 °C for 1 minute, shaking. The reaction was quenched with 5 µL of 10 mM imidazole.

The labeled enzymes were digested with trypsin, desalted, and analyzed by LC/MS/MS in the manner described below.

Digestion. Enzymes were denatured and cysteine residues were blocked using 14 mM iodoacetamide at room temperature in the dark for 30 min, the reaction was quenched using 5.2 µL 0.5 M DTT for 15 min. 100 µL immobilized trypsin was washed and resuspended in 200 µL 25 mM tris pH 7.8. Labeled enzyme was concentrated to 100 µL and added to the immobilized trypsin and incubated at 37 °C for 2 hours. Immobilized trypsin was removed by centrifugation. 26 µL of 2.5% TFA was added to 650 µL of digested protein for a final concentration of 0.1%. The digested enzyme was desalted using a ZipTip (Sigma Catalog # ZTC04S096) using the manufacturer’s instructions.

124 19F NMR. In general, 2 mg of MtmW was suspended in 300 µL of NMR buffer (100 µM tris-HCl, 50 mM NaCl, 3% glycerol). MtmOIV was added to a ratio of 2:1 of MtmOIV to

MtmW. Experiments were attempted using the following combinations of labeled and unlabeled enzyme, unlabeled MtmW with TFAA labeled MtmOIV and TFAA labeled

MtmW with unlabeled MtmOIV.

Label Transfer. Label transfer experiments were performed using Sulfo-SBED (Sulfo-N- hydroxysuccinimidyl-2-(6-[biotinamido]-2-(p-azido benzamido)-hexanoamido) ethyl-

1,3'-dithioproprionate) Biotin Label Transfer Kit (ThermoScientific). Add 10 µL of dissolved Sulfo-SBED Reagent to 1 mg of bait protein in 200 mM phosphate buffer pH:

8, with 3% glycerol, 100 mM NaCl. The reaction was incubated for 60 minutes at 4 °C.

The buffer was exchanged 3 times using centrifugal filtration to remove excess non- reacted Sulfo-SBED from the labeled bait protein. The bait protein was added to 15 mL of S. argillaceus cell lysate. To encourage binding, The mixture was allowed to react for

5 min, after which it was exposed to 365 nm ultraviolet light for 15 minutes. An excess of DTT was added to the reaction to cleave the crosslinker, and the result was purified on streptavidin resin. The enzymes were then separated by SDS-PAGE (1 mm, 12% gel run at 120 V for 90 min.) or digested with immobilized trypsin (ThermoFisher Catalog

#20230) according to the manufacturer’s instructions. The digested enzyme was desalted using a ZipTip (Sigma Catalog # ZTC04S096) using the manufacturer’s instructions. The resulting peptides were sent to the University of Kentucky Proteomics Core Facility to be analyzed by LC/MS/MS.

LC/MS/MS. Samples were digested with trypsin and desalted as described above. The samples were submitted to the University of Kentucky’s Proteomics Core Facility for

125 LC/MS/MS analysis. The tryptic samples were used directly for LC-MS/MS analysis.

MS data sets were searched in MASCOT146 against a custom database containing only the sequences of MtmW and MtmOIV in the case of the protein footprinting experiments, and with MtmGI, MtmGII, MtmOIV, MtmW, MtrA, and MtrB in the case of the label transfer experiments.

126 References

[1] Baum, M. (1968) A clinical trial of mithramycin in the treatment of advanced malignant disease, Br J Cancer 22, 176-183. [2] Ellas, E. G., Reynoso, G., and Mittelman, A. (1972) Control of hypercalcemia with mithramycin, Ann Surg 175, 431-435. [3] Brown, J. H., and Kennedy, B. J. (1965) Mithramycin in the Treatment of Disseminated Testicular Neoplasms, N Engl J Med 272, 111-118. [4] Kennedy, B. J. (1970) Metabolic and toxic effects of mithramycin during tumor therapy, Am J Med 49, 494-503. [5] Grohar, P. J., Woldemichael, G. M., Griffin, L. B., Mendoza, A., Chen, Q. R., Yeung, C., Currier, D. G., Davis, S., Khanna, C., Khan, J., McMahon, J. B., and Helman, L. J. (2011) Identification of an inhibitor of the EWS-FLI1 oncogenic transcription factor by high-throughput screening, J Natl Cancer Inst 103, 962- 978. [6] Kofman, S., Perlia, C. P., and Economou, S. G. (1973) Mithramycin in the treatment of metastatic Ewing's sarcoma, Cancer 31, 889-893. [7] Bosserman, M. A., Downey, T., Noinaj, N., Buchanan, S. K., and Rohr, J. (2013) Molecular insight into substrate recognition and catalysis of Baeyer-Villiger monooxygenase MtmOIV, the key frame-modifying enzyme in the biosynthesis of anticancer agent mithramycin, ACS Chem Biol 8, 2466-2477. [8] Weidenbach, S., Hou, C., Chen, J. M., Tsodikov, O. V., and Rohr, J. (2016) Dimerization and DNA recognition rules of mithramycin and its analogues, J Inorg Biochem 156, 40-47. [9] Vellingiri, B., Iyer, M., Devi Subramaniam, M., Jayaramayya, K., Siama, Z., Giridharan, B., Narayanasamy, A., Abdal Dayem, A., and Cho, S. G. (2020) Understanding the Role of the Transcription Factor Sp1 in Ovarian Cancer: from Theory to Practice, Int J Mol Sci 21. [10] Tornin, J., Martinez-Cruzado, L., Santos, L., Rodriguez, A., Nunez, L. E., Oro, P., Hermosilla, M. A., Allonca, E., Fernandez-Garcia, M. T., Astudillo, A., Suarez, C., Moris, F., and Rodriguez, R. (2016) Inhibition of SP1 by the mithramycin analog EC-8042 efficiently targets tumor initiating cells in sarcoma, Oncotarget 7, 30935-30950. [11] Gao, Y., Jia, Z., Kong, X., Li, Q., Chang, D. Z., Wei, D., Le, X., Suyun, H., Huang, S., Wang, L., and Xie, K. (2011) Combining betulinic acid and mithramycin a effectively suppresses pancreatic cancer by inhibiting proliferation, invasion, and angiogenesis, Cancer Res 71, 5182-5193. [12] Quarni, W., Dutta, R., Green, R., Katiri, S., Patel, B., Mohapatra, S. S., and Mohapatra, S. (2019) Mithramycin A Inhibits Colorectal Cancer Growth by Targeting Cancer Stem Cells, Sci Rep 9, 15202. [13] Sankar, S., Bell, R., Stephens, B., Zhuo, R., Sharma, S., Bearss, D. J., and Lessnick, S. L. (2016) Mechanism and relevance of EWS/FLI-mediated transcriptional repression in Ewing sarcoma, Oncogene 35, 6155-6156. [14] Ross, K. A., Smyth, N. A., Murawski, C. D., and Kennedy, J. G. (2013) The biology of ewing sarcoma, ISRN Oncol 2013, 759725.

127 [15] Hou, C., Weidenbach, S., Cano, K. E., Wang, Z., Mitra, P., Ivanov, D. N., Rohr, J., and Tsodikov, O. V. (2016) Structures of mithramycin analogues bound to DNA and implications for targeting transcription factor FLI1, Nucleic Acids Res 44, 8990-9004. [16] Osgood, C. L., Maloney, N., Kidd, C. G., Kitchen-Goosen, S., Segars, L., Gebregiorgis, M., Woldemichael, G. M., He, M., Sankar, S., Lessnick, S. L., Kang, M., Smith, M., Turner, L., Madaj, Z. B., Winn, M. E., Nunez, L. E., Gonzalez-Sabin, J., Helman, L. J., Moris, F., and Grohar, P. J. (2016) Identification of Mithramycin Analogues with Improved Targeting of the EWS- FLI1 Transcription Factor, Clin Cancer Res 22, 4105-4118. [17] Nunez, L. E., Nybo, S. E., Gonzalez-Sabin, J., Perez, M., Menendez, N., Brana, A. F., Shaaban, K. A., He, M., Moris, F., Salas, J. A., Rohr, J., and Mendez, C. (2012) A novel mithramycin analogue with high antitumor activity and less toxicity generated by combinatorial biosynthesis, J Med Chem 55, 5813-5825. [18] Lumachi, F., Brunello, A., Roma, A., and Basso, U. (2009) Cancer-induced hypercalcemia, Anticancer Res 29, 1551-1555. [19] Mundy, G. R., and Guise, T. A. (1997) Hypercalcemia of malignancy, Am J Med 103, 134-145. [20] Perlia, C. P., Gubisch, N. J., Wolter, J., Edelberg, D., Dederick, M. M., and Taylor, S. G., 3rd. (1970) Mithramycin treatment of hypercalcemia, Cancer 25, 389-394. [21] Rosol, T. J., Chew, D. J., Couto, C. G., Ayl, R. D., Nagode, L. A., and Capen, C. C. (1992) Effects of mithramycin on calcium metabolism and bone in dogs, Vet Pathol 29, 223-229. [22] Hertweck, C., Luzhetskyy, A., Rebets, Y., and Bechthold, A. (2007) Type II polyketide synthases: gaining a deeper insight into enzymatic teamwork, Nat Prod Rep 24, 162-190. [23] Dewick, P. M. (2002) Medicinal Natural Products, John Wiley & Sons Ltd. [24] Chan, Y. A., Podevels, A. M., Kevany, B. M., and Thomas, M. G. (2009) Biosynthesis of polyketide synthase extender units, Nat Prod Rep 26, 90-114. [25] Remsing, L. L., Garcia-Bernardo, J., Gonzalez, A., Kunzel, E., Rix, U., Brana, A. F., Bearden, D. W., Mendez, C., Salas, J. A., and Rohr, J. (2002) Ketopremithramycins and ketomithramycins, four new aureolic acid-type compounds obtained upon inactivation of two genes involved in the biosynthesis of the deoxysugar moieties of the antitumor drug mithramycin by Streptomyces argillaceus, reveal novel insights into post-PKS tailoring steps of the mithramycin biosynthetic pathway, J Am Chem Soc 124, 1606-1614. [26] Lombo, F., Blanco, G., Fernandez, E., Mendez, C., and Salas, J. A. (1996) Characterization of Streptomyces argillaceus genes encoding a polyketide synthase involved in the biosynthesis of the antitumor mithramycin, Gene 172, 87-91. [27] Keatinge-Clay, A. T., Maltby, D. A., Medzihradszky, K. F., Khosla, C., and Stroud, R. M. (2004) An antibiotic factory caught in action, Nat Struct Mol Biol 11, 888- 893. [28] Katz, L. (2009) The DEBS paradigm for type I modular polyketide synthases and beyond, Methods Enzymol 459, 113-142.

128 [29] Kapur, S., Lowry, B., Yuzawa, S., Kenthirapalan, S., Chen, A. Y., Cane, D. E., and Khosla, C. (2012) Reprogramming a module of the 6-deoxyerythronolide B synthase for iterative chain elongation, Proc Natl Acad Sci U S A 109, 4110-4115. [30] Saxena, P., Yadav, G., Mohanty, D., and Gokhale, R. S. (2003) A new family of type III polyketide synthases in Mycobacterium tuberculosis, J Biol Chem 278, 44780-44790. [31] Hashimoto, M., Nonaka, T., and Fujii, I. (2014) Fungal type III polyketide synthases, Nat Prod Rep 31, 1306-1317. [32] Weymouth-Wilson, A. C. (1997) The role of carbohydrates in biologically active natural products, Nat Prod Rep 14, 99-110. [33] Liu, H. W., and Thorson, J. S. (1994) Pathways and mechanisms in the biogenesis of novel deoxysugars by bacteria, Annu Rev Microbiol 48, 223-256. [34] Olano, C., Mendez, C., and Salas, J. A. (2010) Post-PKS tailoring steps in natural product-producing actinomycetes from the perspective of combinatorial biosynthesis, Nat Prod Rep 27, 571-616. [35] Perez, M., Lombo, F., Baig, I., Brana, A. F., Rohr, J., Salas, J. A., and Mendez, C. (2006) Combinatorial biosynthesis of antitumor deoxysugar pathways in Streptomyces griseus: Reconstitution of "unnatural natural gene clusters" for the biosynthesis of four 2,6-D-dideoxyhexoses, Appl Environ Microbiol 72, 6644- 6652. [36] Lombo, F., Siems, K., Brana, A. F., Mendez, C., Bindseil, K., and Salas, J. A. (1997) Cloning and insertional inactivation of Streptomyces argillaceus genes involved in the earliest steps of biosynthesis of the sugar moieties of the antitumor polyketide mithramycin, J Bacteriol 179, 3354-3357. [37] Gonzalez, A., Remsing, L. L., Lombo, F., Fernandez, M. J., Prado, L., Brana, A. F., Kunzel, E., Rohr, J., Mendez, C., and Salas, J. A. (2001) The mtmVUC genes of the mithramycin gene cluster in Streptomyces argillaceus are involved in the biosynthesis of the sugar moieties, Mol Gen Genet 264, 827-835. [38] Wang, G., Kharel, M. K., Pahari, P., and Rohr, J. (2011) Investigating Mithramycin deoxysugar biosynthesis: enzymatic total synthesis of TDP-D-olivose, Chembiochem 12, 2568-2571. [39] Wang, G., Pahari, P., Kharel, M. K., Chen, J., Zhu, H., Van Lanen, S. G., and Rohr, J. (2012) Cooperation of two bifunctional enzymes in the biosynthesis and attachment of deoxysugars of the antitumor antibiotic mithramycin, Angew Chem Int Ed Engl 51, 10638-10642. [40] Lozano, M. J., Remsing, L. L., Quiros, L. M., Brana, A. F., Fernandez, E., Sanchez, C., Mendez, C., Rohr, J., and Salas, J. A. (2000) Characterization of two polyketide methyltransferases involved in the biosynthesis of the antitumor drug mithramycin by Streptomyces argillaceus, J Biol Chem 275, 3065-3074. [41] Tevyashova, A. N. (2016) Olivomycin A – an Antitumor Antibiotic of the Aureolic Acid Group, Pharmaceutical Chemistry Journal 50, 425-430. [42] Rodriguez, D., Quiros, L. M., and Salas, J. A. (2004) MtmMII-mediated C- methylation during biosynthesis of the antitumor drug mithramycin is essential for biological activity and DNA-drug interaction, J Biol Chem 279, 8149-8158. [43] Mendez, C., Gonzalez-Sabin, J., Moris, F., and Salas, J. A. (2015) Expanding the Chemical Diversity of the Antitumoral Compound Mithramycin by Combinatorial

129 Biosynthesis and Biocatalysis: The Quest for Mithralogs with Improved Therapeutic Window, Planta Med 81, 1326-1338. [44] La Ferla, B., Airoldi, C., Zona, C., Orsato, A., Cardona, F., Merlo, S., Sironi, E., D'Orazio, G., and Nicotra, F. (2011) Natural glycoconjugates with antitumor activity, Nat Prod Rep 28, 630-648. [45] Trefzer, A., Blanco, G., Remsing, L., Kunzel, E., Rix, U., Lipata, F., Brana, A. F., Mendez, C., Rohr, J., Bechthold, A., and Salas, J. A. (2002) Rationally designed glycosylated premithramycins: hybrid aromatic polyketides using genes from three different biosynthetic pathways, J Am Chem Soc 124, 6056-6062. [46] Rohr, J., and Zeeck, A. (1987) Metabolic products of microorganisms. 240. Urdamycins, new angucycline antibiotics from Streptomyces fradiae. II. Structural studies of urdamycins B to F, J Antibiot (Tokyo) 40, 459-467. [47] Williams, G. J., and Thorson, J. S. (2009) Natural product glycosyltransferases: properties and applications, Adv Enzymol Relat Areas Mol Biol 76, 55-119. [48] Schuman, B., Evans, S. V., and Fyles, T. M. (2013) Geometric attributes of retaining glycosyltransferase enzymes favor an orthogonal mechanism, PLoS One 8, e71077. [49] Blanco, G., Fernandez, E., Fernandez, M. J., Brana, A. F., Weissbach, U., Kunzel, E., Rohr, J., Mendez, C., and Salas, J. A. (2000) Characterization of two glycosyltransferases involved in early glycosylation steps during biosynthesis of the antitumor polyketide mithramycin by Streptomyces argillaceus, Mol Gen Genet 262, 991-1000. [50] Fernandez, E., Weissbach, U., Sanchez Reillo, C., Brana, A. F., Mendez, C., Rohr, J., and Salas, J. A. (1998) Identification of two genes from Streptomyces argillaceus encoding glycosyltransferases involved in transfer of a disaccharide during biosynthesis of the antitumor drug mithramycin, J Bacteriol 180, 4929- 4937. [51] Blanco, G., Patallo, E. P., Brana, A. F., Trefzer, A., Bechthold, A., Rohr, J., Mendez, C., and Salas, J. A. (2001) Identification of a sugar flexible glycosyltransferase from Streptomyces olivaceus, the producer of the antitumor polyketide elloramycin, Chem Biol 8, 253-263. [52] Nur-e-Alam, M., Mendez, C., Salas, J. A., and Rohr, J. (2005) Elucidation of the glycosylation sequence of mithramycin biosynthesis: isolation of 3A- deolivosylpremithramycin B and its conversion to premithramycin B by glycosyltransferase MtmGII, Chembiochem 6, 632-636. [53] Lombo, F., Menendez, N., Salas, J. A., and Mendez, C. (2006) The aureolic acid family of antitumor compounds: structure, mode of action, biosynthesis, and novel derivatives, Appl Microbiol Biotechnol 73, 1-14. [54] Menendez, N., Nur-e-Alam, M., Fischer, C., Brana, A. F., Salas, J. A., Rohr, J., and Mendez, C. (2006) Deoxysugar transfer during chromomycin A3 biosynthesis in Streptomyces griseus subsp. griseus: new derivatives with antitumor activity, Appl Environ Microbiol 72, 167-177. [55] Baig, I., Perez, M., Brana, A. F., Gomathinayagam, R., Damodaran, C., Salas, J. A., Mendez, C., and Rohr, J. (2008) Mithramycin analogues generated by combinatorial biosynthesis show improved bioactivity, J Nat Prod 71, 199-207.

130

[56] Menendez, N., Nur, E. A. M., Brana, A. F., Rohr, J., Salas, J. A., and Mendez, C. (2004) Tailoring modification of deoxysugars during biosynthesis of the antitumour drug chromomycin A by Streptomyces griseus ssp. griseus, Mol Microbiol 53, 903-915. [57] Garcia, B., Gonzalez-Sabin, J., Menendez, N., Brana, A. F., Nunez, L. E., Moris, F., Salas, J. A., and Mendez, C. (2011) The chromomycin CmmA acetyltransferase: a membrane-bound enzyme as a tool for increasing structural diversity of the antitumour mithramycin, Microb Biotechnol 4, 226-238. [58] Gibson, M., Nur-e-alam, M., Lipata, F., Oliveira, M. A., and Rohr, J. (2005) Characterization of kinetics and products of the Baeyer-Villiger oxygenase MtmOIV, the key enzyme of the biosynthetic pathway toward the natural product anticancer drug mithramycin from Streptomyces argillaceus, J Am Chem Soc 127, 17594-17595. [59] Penning, T. M. (2015) The aldo-keto reductases (AKRs): Overview, Chem Biol Interact 234, 236-246. [60] Wheeler, R., Yu, X., Hou, C., Mitra, P., Chen, J. M., Herkules, F., Ivanov, D. N., Tsodikov, O. V., and Rohr, J. (2020) Discovery of a Cryptic Intermediate in Late Steps of Mithramycin Biosynthesis, Angew Chem Int Ed Engl 59, 826-832. [61] Albertini, V., Jain, A., Vignati, S., Napoli, S., Rinaldi, A., Kwee, I., Nur-e-Alam, M., Bergant, J., Bertoni, F., Carbone, G. M., Rohr, J., and Catapano, C. V. (2006) Novel GC-rich DNA-binding compound produced by a genetically engineered mutant of the mithramycin producer Streptomyces argillaceus exhibits improved transcriptional repressor activity: implications for cancer therapy, Nucleic Acids Res 34, 1721-1734. [62] Remsing, L. L., Gonzalez, A. M., Nur-e-Alam, M., Fernandez-Lozano, M. J., Brana, A. F., Rix, U., Oliveira, M. A., Mendez, C., Salas, J. A., and Rohr, J. (2003) Mithramycin SK, a novel antitumor drug with improved therapeutic index, mithramycin SA, and demycarosyl-mithramycin SK: three new products generated in the mithramycin producer Streptomyces argillaceus through combinatorial biosynthesis, J Am Chem Soc 125, 5745-5753. [63] Scott, D., Chen, J. M., Bae, Y., and Rohr, J. (2013) Semi-synthetic mithramycin SA derivatives with improved anticancer activity, Chem Biol Drug Des 81, 615-624. [64] Chen, J. M. (2015) Chemoenzymatic studies to enhance the chemical space of natural products, University of Kentucky. [65] Fernandez, E., Lombo, F., Mendez, C., and Salas, J. A. (1996) An ABC transporter is essential for resistance to the antitumor agent mithramycin in the producer Streptomyces argillaceus, Mol Gen Genet 251, 692-698. [66] Locher, K. P. (2016) Mechanistic diversity in ATP-binding cassette (ABC) transporters, Nat Struct Mol Biol 23, 487-493. [67] Florez, A. B., Alvarez, S., Zabala, D., Brana, A. F., Salas, J. A., and Mendez, C. (2015) Transcriptional regulation of mithramycin biosynthesis in Streptomyces argillaceus: dual role as activator and repressor of the PadR-like regulator MtrY, Microbiology 161, 272-284. [68] Kormanec, J., Novakova, R., Mingyar, E., and Feckova, L. (2014) Intriguing properties of the angucycline antibiotic auricin and complex regulation of its biosynthesis, Appl Microbiol Biotechnol 98, 45-60.

131

[69] Vasanthakumar, A., Kattusamy, K., and Prasad, R. (2013) Regulation of daunorubicin biosynthesis in Streptomyces peucetius - feed forward and feedback transcriptional control, J Basic Microbiol 53, 636-644. [70] Xu, G., Wang, J., Wang, L., Tian, X., Yang, H., Fan, K., Yang, K., and Tan, H. (2010) "Pseudo" gamma-butyrolactone receptors respond to antibiotic signals to coordinate antibiotic biosynthesis, J Biol Chem 285, 27440-27448. [71] Rebets, Y., Ostash, B., Luzhetskyy, A., Kushnir, S., Fukuhara, M., Bechthold, A., Nashimoto, M., Nakamura, T., and Fedorenko, V. (2005) DNA-binding activity of LndI protein and temporal expression of the gene that upregulates landomycin E production in Streptomyces globisporus 1912, Microbiology 151, 281-290. [72] Miller, M. B., and Bassler, B. L. (2001) Quorum sensing in bacteria, Annu Rev Microbiol 55, 165-199. [73] Kim, T., Flick, R., Brunzelle, J., Singer, A., Evdokimova, E., Brown, G., Joo, J. C., Minasov, G. A., Anderson, W. F., Mahadevan, R., Savchenko, A., and Yakunin, A. F. (2017) Novel Aldo-Keto Reductases for the Biocatalytic Conversion of 3- Hydroxybutanal to 1,3-Butanediol: Structural and Biochemical Studies, Appl Environ Microbiol 83. [74] Schlegel, B. P., Jez, J. M., and Penning, T. M. (1998) Mutagenesis of 3 alpha- hydroxysteroid dehydrogenase reveals a "push-pull" mechanism for proton transfer in aldo-keto reductases, Biochemistry 37, 3538-3548. [75] Gulbis, J. M., Zhou, M., Mann, S., and MacKinnon, R. (2000) Structure of the cytoplasmic beta subunit-T1 assembly of voltage-dependent K+ channels, Science 289, 123-127. [76] Long, S. B., Campbell, E. B., and Mackinnon, R. (2005) Crystal structure of a mammalian voltage-dependent Shaker family K+ channel, Science 309, 897-903. [77] Pan, Y., Weng, J., Cao, Y., Bhosle, R. C., and Zhou, M. (2008) Functional coupling between the Kv1.1 channel and aldoketoreductase Kvbeta1, J Biol Chem 283, 8634-8642. [78] Nakatani, H., Goda, S., Unno, H., Nagai, T., Yoshimura, T., and Hemmi, H. (2012) Substrate-induced change in the quaternary structure of type 2 isopentenyl diphosphate isomerase from Sulfolobus shibatae, J Bacteriol 194, 3216-3224. [79] Xu, J., Oshima, T., and Yoshida, M. (1990) Tetramer-dimer conversion of phosphofructokinase from Thermus thermophilus induced by its allosteric effectors, J Mol Biol 215, 597-606. [80] Hanau, S., d'Empaire, L. P., Capone, I., Alberighi, S., Montioli, R., and Dallocchio, F. (2013) Evidence for dimer/tetramer equilibrium in Trypanosoma brucei 6- phosphogluconate dehydrogenase, Biochim Biophys Acta 1834, 2647-2652. [81] Wohlert, S. E., Kunzel, E., Machinek, R., Mendez, C., Salas, J. A., and Rohr, J. (1999) The structure of mithramycin reinvestigated, J Nat Prod 62, 119-121. [82] Zimmerman, S. B., and Minton, A. P. (1993) Macromolecular crowding: biochemical, biophysical, and physiological consequences, Annu Rev Biophys Biomol Struct 22, 27-65. [83] Christiansen, A., Wang, Q., Samiotakis, A., Cheung, M. S., and Wittung-Stafshede, P. (2010) Factors defining effects of macromolecular crowding on protein stability: an in vitro/in silico case study using cytochrome c, Biochemistry 49, 6519-6530.

132

[84] Minton, A. P., and Wilf, J. (1981) Effect of macromolecular crowding upon the structure and function of an enzyme: glyceraldehyde-3-phosphate dehydrogenase, Biochemistry 20, 4821-4826. [85] Homouz, D., Perham, M., Samiotakis, A., Cheung, M. S., and Wittung-Stafshede, P. (2008) Crowded, cell-like environment induces shape changes in aspherical protein, Proc Natl Acad Sci U S A 105, 11754-11759. [86] Stagg, L., Zhang, S. Q., Cheung, M. S., and Wittung-Stafshede, P. (2007) Molecular crowding enhances native structure and stability of alpha/beta protein flavodoxin, Proc Natl Acad Sci U S A 104, 18976-18981. [87] Rivas, G., Ferrone, F., and Herzfeld, J. (2004) Life in a crowded world, EMBO Rep 5, 23-27. [88] Ramos, A., Lombo, F., Brana, A. F., Rohr, J., Mendez, C., and Salas, J. A. (2008) Biosynthesis of elloramycin in Streptomyces olivaceus requires glycosylation by enzymes encoded outside the aglycon cluster, Microbiology 154, 781-788. [89] Zakharian, E., and Reusch, R. N. (2004) Streptomyces lividans potassium channel KcsA is regulated by the potassium electrochemical gradient, Biochem Biophys Res Commun 316, 429-436. [90] Papanikou, E., Karamanou, S., and Economou, A. (2007) Bacterial protein secretion through the translocase nanomachine, Nat Rev Microbiol 5, 839-851. [91] Srinivasan MC, V. H., Powar VK, Khire JM. (1983) High activity extracellular glucose (xylose) isomerase from a Chainia species, Biotechnology letters 5, 611- 614. [92] Srinivasan MC, V. H., Powar VK, Rele MV, Bastawde KB. (1984) Studies on xylan degrading enzyme from Chainia, Biotechnology letters 6, 715--718. [93] Amedei, A., Munari, F., Bella, C. D., Niccolai, E., Benagiano, M., Bencini, L., Cianchi, F., Farsi, M., Emmi, G., Zanotti, G., de Bernard, M., Kundu, M., and D'Elios, M. M. (2014) Helicobacter pylori secreted peptidyl prolyl cis, trans- isomerase drives Th17 inflammation in gastric adenocarcinoma, Intern Emerg Med 9, 303-309. [94] Halder, P., Datta, C., Kumar, R., Sharma, A. K., Basu, J., and Kundu, M. (2015) The secreted antigen, HP0175, of Helicobacter pylori links the unfolded protein response (UPR) to autophagy in gastric epithelial cells, Cell Microbiol 17, 714- 729. [95] Aich, P., and Dasgupta, D. (1995) Role of magnesium ion in mithramycin-DNA interaction: binding of mithramycin-Mg2+ complexes with DNA, Biochemistry 34, 1376-1385. [96] Huang, H. W., Li, D., and Cowan, J. A. (1995) Biostructural chemistry of magnesium. regulation of mithramycin-DNA interactions by Mg2+ coordination, Biochimie 77, 729-738. [97] Lin, C. I., McCarty, R. M., and Liu, H. W. (2017) The Enzymology of Organic Transformations: A Survey of Name Reactions in Biological Systems, Angew Chem Int Ed Engl 56, 3446-3489. [98] Joerger, A. C., Gosse, C., Fessner, W. D., and Schulz, G. E. (2000) Catalytic action of fuculose 1-phosphate aldolase (class II) as derived from structure-directed mutagenesis, Biochemistry 39, 6033-6041.

133

[99] Thankachan AP, A. S., Sindhu KS, Anilkumar G. (2015) An overview of Zn- catalyzed enantioselective aldol type C–C bond formation, RSC Advances 5, 62179-62193. [100] Kim, H. J., Hyun, E. K., Kim, Y. S., Lee, Y. J., and Oh, D. K. (2006) Characterization of an Agrobacterium tumefaciens D-psicose 3-epimerase that converts D-fructose to D-psicose, Appl Environ Microbiol 72, 981-985. [101] Zhang, W., Fang, D., Xing, Q., Zhou, L., Jiang, B., and Mu, W. (2013) Characterization of a novel metal-dependent D-psicose 3-epimerase from Clostridium scindens 35704, PLoS One 8, e62987. [102] Izumori K, R. A., Okaya H, Tsumura T. (1993) A new enzyme, D-ketohexose 3- epimerase, from Pseudomonas sp. ST-24, Bioscience, biotechnology, and biochemistry 57, 1037-1039. [103] Yoshida, H., Yoshihara, A., Ishii, T., Izumori, K., and Kamitori, S. (2016) X-ray structures of the Pseudomonas cichorii D-tagatose 3-epimerase mutant form C66S recognizing deoxy sugars as substrates, Appl Microbiol Biotechnol 100, 10403- 10415. [104] Tang J, G. X., Zhu L, Hu C. (2015) Mechanistic Study of Glucose-to-Fructose Isomerization in Water Catalyzed by [Al(OH)2(aq)]+, ACS Catalysis 5, 5097- 5103. [105] Rellán-Piñeiro M, G.-R. M., López N. (2017) A mechanism for the selective epimerization of the glucose mannose pair by Mo-based compounds: towards catalyst optimization, Green Chemistry 19, 5932-5939. [106] Tanase T, S. F., Yano S, Yoshikawa S. (1986) Novel C-2 epimerization of aldoses and stereoselective uptake of one of the epimeric aldoses by nickel (II) complexes, Journal of the Chemical Society, Chemical Communications, 1001- 1003. [107] Tanase T, S. F., Kuse M, Yano S, Hidai M, Yoshikawa S. (1988) Novel C-2 epimerization of aldoses promoted by nickel (II) diamine complexes, involving a stereospecific pinacol-type 1, 2-carbon shiftc, Inorganic Chemistry 27, 4085- 4094. [108] Grubbs RD, M. M. (1987) Magnesium as a regulatory cation: criteria and evaluation, Magnesium 6, 113-127. [109] Bystrykh LV, F.-M. M., Herrema JK, Malpartida F, Hopwood DA, Dijkhuizen L. (1996) Production of actinorhodin-related "blue pigments" by Streptomyces coelicolor A3(2), Journal of bacteriology 178, 2238-2244. [110] Decho, A. W., Norman, R. S., and Visscher, P. T. (2010) Quorum sensing in natural environments: emerging views from microbial mats, Trends Microbiol 18, 73-80. [111] Hou, M. H., and Wang, A. H. (2005) Mithramycin forms a stable dimeric complex by chelating with Fe(II): DNA-interacting characteristics, cellular permeation and cytotoxicity, Nucleic Acids Res 33, 1352-1361. [112] Quiros, L. M., Aguirrezabalaga, I., Olano, C., Mendez, C., and Salas, J. A. (1998) Two glycosyltransferases and a glycosidase are involved in oleandomycin modification during its biosynthesis by Streptomyces antibioticus, Mol Microbiol 28, 1177-1185.

134

[113] Walker, M. S., and Walker, J. B. (1970) Streptomycin biosynthesis and metabolism. Enzymatic phosphorylation of dihydrostreptobiosamine moieties of dihydro-streptomycin-(streptidino) phosphate and dihydrostreptomycin by Streptomyces extracts, J Biol Chem 245, 6683-6689. [114] Cundliffe, E. (1989) How antibiotic-producing organisms avoid suicide, Annu Rev Microbiol 43, 207-233. [115] Xu, Y., Willems, A., Au-Yeung, C., Tahlan, K., and Nodwell, J. R. (2012) A two- step mechanism for the activation of actinorhodin export and resistance in Streptomyces coelicolor, mBio 3, e00191-00112. [116] Coisne, S., Bechet, M., and Blondeau, R. (1999) Actinorhodin production by Streptomyces coelicolor A3(2) in iron-restricted media, Lett Appl Microbiol 28, 199-202. [117] Kharel, M. K., Pahari, P., Shaaban, K. A., Wang, G., Morris, C., and Rohr, J. (2012) Elucidation of post-PKS tailoring steps involved in landomycin biosynthesis, Org Biomol Chem 10, 4256-4265. [118] Kallio, P., Liu, Z., Mantsala, P., Niemi, J., and Metsa-Ketela, M. (2008) Sequential action of two flavoenzymes, PgaE and PgaM, in angucycline biosynthesis: chemoenzymatic synthesis of gaudimycin C, Chem Biol 15, 157-166. [119] Tibrewal, N., Downey, T. E., Van Lanen, S. G., Ul Sharif, E., O'Doherty, G. A., and Rohr, J. (2012) Roles of the synergistic reductive O-methyltransferase GilM and of O-methyltransferase GilMT in the gilvocarcin biosynthetic pathway, J Am Chem Soc 134, 12402-12405. [120] Wang, C., Gibson, M., Rohr, J., and Oliveira, M. A. (2005) Crystallization and X- ray diffraction properties of Baeyer-Villiger monooxygenase MtmOIV from the mithramycin biosynthetic pathway in Streptomyces argillaceus, Acta Crystallogr Sect F Struct Biol Cryst Commun 61, 1023-1026. [121] Brookes, E., Cao, W., and Demeler, B. (2010) A two-dimensional spectrum analysis for sedimentation velocity experiments of mixtures with heterogeneity in molecular weight and shape, Eur Biophys J 39, 405-414. [122] Schuck, P. (2000) Size-distribution analysis of macromolecules by sedimentation velocity ultracentrifugation and lamm equation modeling, Biophys J 78, 1606- 1619. [123] Otwinowski, Z., and Minor, W. (1997) Processing of X-ray diffraction data collected in oscillation mode, Methods Enzymol 276, 307-326. [124] Emsley, P., Lohkamp, B., Scott, W. G., and Cowtan, K. (2010) Features and development of Coot, Acta Crystallogr D Biol Crystallogr 66, 486-501. [125] Murshudov, G. N., Vagin, A. A., and Dodson, E. J. (1997) Refinement of macromolecular structures by the maximum-likelihood method, Acta Crystallogr D Biol Crystallogr 53, 240-255. [126] Graham, J. W., Williams, T. C., Morgan, M., Fernie, A. R., Ratcliffe, R. G., and Sweetlove, L. J. (2007) Glycolytic enzymes associate dynamically with mitochondria in response to respiratory demand and support substrate channeling, Plant Cell 19, 3723-3738. [127] Lindbladh, C., Rault, M., Hagglund, C., Small, W. C., Mosbach, K., Bulow, L., Evans, C., and Srere, P. A. (1994) Preparation and kinetic characterization of a

135

fusion protein of yeast mitochondrial citrate synthase and malate dehydrogenase, Biochemistry 33, 11692-11698. [128] Jorgensen, K., Rasmussen, A. V., Morant, M., Nielsen, A. H., Bjarnholt, N., Zagrobelny, M., Bak, S., and Moller, B. L. (2005) Metabolon formation and metabolic channeling in the biosynthesis of plant natural products, Curr Opin Plant Biol 8, 280-291. [129] Lee, T. S., Khosla, C., and Tang, Y. (2005) Orthogonal protein interactions in spore pigment producing and antibiotic producing polyketide synthases, J Antibiot (Tokyo) 58, 663-666. [130] Cai, W., and Zhang, W. (2018) Engineering modular polyketide synthases for production of biofuels and industrial chemicals, Curr Opin Biotechnol 50, 32-38. [131] Danielson, M. A., and Falke, J. J. (1996) Use of 19F NMR to probe protein structure and conformational changes, Annu Rev Biophys Biomol Struct 25, 163- 195. [132] Gettins, P. G. (1994) 1H- and 19F-NMR approaches to the study of the structure of proteins larger than 25 kDa, Int J Biol Macromol 16, 227-235. [133] Gerig, J. T. (1989) Fluorine nuclear magnetic resonance of fluorinated ligands, Methods Enzymol 177, 3-23. [134] Izumi, S., Kaneko, H., Yamazaki, T., Hirata, T., and Kominami, S. (2003) Membrane topology of guinea pig cytochrome P450 17 alpha revealed by a combination of chemical modifications and mass spectrometry, Biochemistry 42, 14663-14669. [135] Frueh, D. P. (2014) Practical aspects of NMR signal assignment in larger and challenging proteins, Prog Nucl Magn Reson Spectrosc 78, 47-75. [136] Sucharitakul, J., Wongnate, T., Montersino, S., van Berkel, W. J., and Chaiyen, P. (2012) Reduction kinetics of 3-hydroxybenzoate 6-hydroxylase from Rhodococcus jostii RHA1, Biochemistry 51, 4309-4321. [137] Pitsawong, W., Chenprakhon, P., Dhammaraj, T., Medhanavyn, D., Sucharitakul, J., Tongsook, C., van Berkel, W. J. H., Chaiyen, P., and Miller, A. F. (2020) Tuning of pKa values activates substrates in flavin-dependent aromatic hydroxylases, J Biol Chem 295, 3965-3981. [138] Soldi, M., Cuomo, A., Bremang, M., and Bonaldi, T. (2013) Mass spectrometry- based proteomics for the analysis of chromatin structure and dynamics, Int J Mol Sci 14, 5402-5431. [139] Rainwater, D. L., and Kolattukudy, P. E. (1982) Specific acetylation of essential lysine residues in malonyl-CoA decarboxylase, Int J Biochem 14, 609-614. [140] Abboud, M. I., Hinchliffe, P., Brem, J., Macsics, R., Pfeffer, I., Makena, A., Umland, K. D., Rydzik, A. M., Li, G. B., Spencer, J., Claridge, T. D., and Schofield, C. J. (2017) (19) F-NMR Reveals the Role of Mobile Loops in Product and Inhibitor Binding by the Sao Paulo Metallo-beta-Lactamase, Angew Chem Int Ed Engl 56, 3862-3866. [141] Percy, A. J., Rey, M., Burns, K. M., and Schriemer, D. C. (2012) Probing protein interactions with hydrogen/deuterium exchange and mass spectrometry-a review, Anal Chim Acta 721, 7-21.

136

[142] Chalmers, M. J., Busby, S. A., Pascal, B. D., West, G. M., and Griffin, P. R. (2011) Differential hydrogen/deuterium exchange mass spectrometry analysis of protein- ligand interactions, Expert Rev Proteomics 8, 43-59. [143] Wang, L., and Chance, M. R. (2011) Structural mass spectrometry of proteins using hydroxyl radical based protein footprinting, Anal Chem 83, 7234-7241. [144] Takamoto, K., and Chance, M. R. (2006) Radiolytic protein footprinting with mass spectrometry to probe the structure of macromolecular complexes, Annu Rev Biophys Biomol Struct 35, 251-276. [145] Weaver, R., and Riley, R. J. (2006) Identification and reduction of ion suppression effects on pharmacokinetic parameters by polyethylene glycol 400, Rapid Commun Mass Spectrom 20, 2559-2564. [146] Perkins, D. N., Pappin, D. J., Creasy, D. M., and Cottrell, J. S. (1999) Probability- based protein identification by searching sequence databases using mass spectrometry data, Electrophoresis 20, 3551-3567.

137 Vita

Degrees Awarded:

Northwestern Connecticut Community College Associates of Science

University of Connecticut Bachelor of Science

Professional Publications

Wheeler, R., Yu, X., Hou, C., Mitra, P., Chen, J.M., Herkules, F., Ivanov, D.N.,

Tsodikov, O.V. and Rohr, J., 2020. Angewandte Chemie International Edition, 59(2), pp.826-832.

Gober, R., Wheeler, R. and Rohr, J., 2019. MedChemComm, 10(11), pp.1855-1866.

Fullmer, M.S., Soucy, S.M., Swithers, K.S., Makkay, A.M., Wheeler, R., Ventosa, A.,

Gogarten, J.P. and Papke, R.T., 2014. Frontiers in microbiology, 5, p.140.

Ram Mohan, N., Fullmer, M.S., Makkay, A.M., Wheeler, R., Ventosa, A., Naor, A.,

Gogarten, J.P. and Papke, R.T., 2014 Frontiers in microbiology, 5, p.143.

Ryan Wheeler

138