The Catholic University of America

Home , ARDB

THE CATHOLIC UNIVERSITY OF AMERICA

Serine Residues in Linker-2 of the Yeast Multidrug Transporter Pdr5 Modulate Drug Resistance

A DISSERTATION

Submitted to the Faculty of the

Department of Biology

School of Arts and Sciences

Of The Catholic University of America

In Partial Fulfillment of the Requirements

For the Degree

Doctor of Philosophy

Hadiar Rahman

Washington, D.C.

2020

Serine Residues in Linker-2 of the Yeast Multidrug Transporter Pdr5 Modulate Drug Resistance

Hadiar Rahman, Ph.D.

Director: John E. Golin, Ph.D.

Multidrug resistance to antifungals and chemotherapeutic drugs is a huge clinical problem for treatment of fungal infections and cancers. Overexpression or alteration of multidrug efflux ATP Binding Cassette (ABC) transporters is often associated with this broad- spectrum of resistance. Pleiotropic Drug Response 5 gene (PDR5) in Saccharomyces cerevisiae defines a subclass of important fungal efflux pumps. The characterization of mutations in this transporter uncovered two novel mechanism of increased resistance.

The first mutation, A666G, resulted in enhanced resistance without increasing either the amount of protein in the plasma membrane or the ATPase activity. In fluorescence-quenching transport assays with rhodamine 6G in purified plasma membrane vesicles, the initial rates of rhodamine 6G fluorescence quenching of both the wild-type and mutant showed a strong dependence on ATP concentration, but were about twice as high in the latter. Plots of initial rate of fluorescence quenching versus ATP concentration exhibited strong cooperativity that was significantly increased in the A666G mutant and accounted for the observed enhancement in resistance.

Alanine substitution mutations in six serine residues of linker-2 all exhibited increased resistance to xenobiotic agents. Biochemical studies of these nonsynonymous mutants demonstrated that they have increased steady-state levels of Pdr5 protein expression and thus enhanced resistance. Quantitative-RT PCR and metabolic labelling experiments demonstrated

that the mutants had levels of PDR5 mRNA that were two to three times as high as in the isogenic wild-type strain because the transcript half-life was increased. These data demonstrate that the nucleotides encoding unconserved amino acids may be used to regulate expression and suggest that Pdr5 has a newly discovered RNA stability element within its coding region.

This dissertation by Hadiar Rahman fulfills the dissertation requirement for the doctoral degree in Cellular and Microbial Biology by Dr. John E. Golin Ph.D., as Director and by Dr. Pamela L. Tuma Ph.D., and Dr. John S. Choy Ph.D., as Readers.

John E. Golin, Ph.D., Director

Pamela L. Tuma, Ph.D., Reader

John S. Choy, Ph.D., Reader

DEDICATION

This work is dedicated to my parents, my wife, my family members and all my well-wishers who

always encourage me to move forward

iii

TABLE OF CONTENTS

List of Figures viii List of Tables xi List of Abbreviations xii Acknowledgements xiv

INTRODUCTION 1 Historical perspective and impact of drug resistance 1 Mechanism of multidrug resistance (MDR) 3 ABC transporters 5 Saccharomyces cerevisiae as a model organism for studying drug resistance 9 ABC transporter in S. cerevisiae 12 Pleotropic drug resistance 5 (Pdr5) 12 The signaling interface of Pdr5 17 Suppressor analysis of an S558Y mutant that uncouples transport and 18

MATERIALS AND METHODS 27 Yeast strains 27 Chemicals and drugs 29 Media 29 Site directed mutagenesis 29 Bacterial transformation of XL10-Gold ultra-competent cells 32 Plasmid isolation from E. coli 33 Yeast transformation 34

Determining the inhibitory concentration 50 (IC50) of a 37 Isolation of genomic DNA from yeast strain 37 PCR amplification of PDR5 38

Long Range PCR 40 DNA sequencing 41 Gel Electrophoresis of PCR products 42 Whole cell protein extraction 42 Preparation of plasma membrane vesicles 43 Determination of protein concentration using Coomassie Plus 47 Determination of protein concentration using BCA method 48 Protein gel electrophoresis 49 Western blotting 49 Assay of ATPase activity 50 Quenching assay 50 R6G transport assay 51 Cycloheximide chase 51 RNA isolation using Zymo-Research kit 52 RNA isolation using Ribopure yeast RNA purification kit 53 cDNA synthesis 55 Q-RT PCR 55 Metabolic labeling of mRNA with 4-thiouracil 56 Mass spectrometry 59

RESULTS 61 Part I: Alanine 666 glycine: a mutation that enhances cooperativity 61 Between transport sites and increases multidrug resistance Isolation of the A666G mutation 62 The steady-state level of Pdr5 was not increased in A666G mutant PM vesicles 63 The ATPase activities of the wild-type and A666G PM vesicle preparations 65 are similar The A666G mutant increased resistance to many Pdr5 transport substrates 66

The A666G mutant had increased R6G transport in whole cells 70 Imazalil sulfate inhibits R6G transport in whole cells 71 The A666G mutant enhanced R6G fluorescence quenching in purified 73 PM vesicles Fluorescence quenching was directionally proportional to ATPase activity 76 No large changes occurred in the IRs of fluorescence quenching over an 8x 77 range of R6G concentration The enhanced fluorescence quenching observed in the A666G PM vesicles is 78 attributable to increased cooperativity between transport sites The inhibition of R6G fluorescence quenching by IMZ in A666G PM vesicles 85 also exhibited enhanced cooperativity

Part II: Alanine substitutions in phosphoserine residues lead 88 to increased PDR5 expression Alanine substitution mutants in six phosphoserine residues were 89 hyper-resistant to Pdr5 substrates An S837A, S854A double mutant does not exhibit greater resistance than 93 the S837A and S854A single mutants The S837A and S854A mutants exhibited enhanced R6G transport 97 An S837D mutant strain was also hyper-resistance to Pdr5 substrates 99 Mass spectrometry of Pdr5 phosphoserine residues suggested that Ser-837 is infrequently 102 phosphorylated Increased level of Pdr5 in purified PM vesicles in the S837A mutant strain 104 The S837A mutant does not increased the stability of Pdr5 107 The S837A mutant has a higher level of Pdr5 mRNA 109 The S837A mutation increases the half-life of the PDR5 transcript 110 The hyper-resistance exhibited by the mutants is not a non-Pdr5 mutation in the R-1 112

strain or pSS607 plasmid

DISCUSSION 115

APPENDIX 126 Appendix-I: 126 Appendix-II: 130 REFERENCES 131 Letter of permission to use figure 7 Letter of permission to use figure 9

vii

LIST OF FIGURES

Figure 1: Possible mechanism of drug resistance 4 Figure 2: Schematic representation of ABC transporter 7 Figure 3: Nucleotide binding motifs in ABC transporters 8 Figure 4: Life cycle of yeast 10 Figure 5: Schematic secondary structure of Pdr5 based on the topology 15 Figure 6: Canonical and non-canonical ATP binding sites 16 Figure 7: Homology model of Pdr5 shows spatial arrangement of TMDs and NBDs 20 Figure 8: Location of suppressor mutation in Pdr5 21 Figure 9: Conformational change upon phosphorylation of R domain leads CFTR 25 channel to open Figure 10: Selection of transformants with PDR5 gene integration 36 Figure 11: The steady-state level of Pdr5 in PM vesicles is unaltered in the A666G 64 mutant Figure 12: The wild-type and mutant PM vesicle preparations have indistinguishable 66 ATPase activities Figure 13: The A666G mutant exhibits strong hyperresistance to multiple Pdr5 substrates 68 Figure 14: The A666G mutant increases resistance to transport substrates 69 Figure 15: The A666G mutant enhances R6G transport in whole cells. 70 Figure 16: Imazalil sulfate inhibition of R6G transport is concentration dependent 72 Figure 17: Pdr5-mediated R6G transport is enhanced in PM vesicles prepared from 75 A666G cells Figure 18: The initial rates of fluorescence quenching are directly proportional to ATPase 76 activity Figure 19: IRs of fluorescence quenching measured over a range of R6G concentration 78 Figure 20: Representative plots of fluorescence quenching performed with 1.5 80 3.0, and 5.0mM ATP viii

Figure 21: The A666G mutant enhances cooperativity between transport sites 81

Figure 22: The A666G mutant exhibits increased cooperativity between transport sites 84 Figure 23: Inhibition of R6G fluorescence quenching by IMZ mirrors the results with 86 whole-cell transport studies Figure 24: Relative location of the linker-2 residues 89 Figure 25: Preliminary testing of phosphoserine mutants indicated that they all increase 91 drug resistance Figure 26: The S837A and S854A mutants exhibits strong hyper-resistance to multiple 92 Pdr5 substrates Figure 27: The S837A mutation results in enhanced resistance to transport substrates 94 Figure 28: An S837A, S854A double mutant is no more hyper-resistant than single 95 mutants Figure 29: An S837A, S850A double mutant has a resistance profile similar to single 96 mutants Figure 30: The A666G mutant exhibits stronger hyper-resistance to cycloheximide than 97 the S837A Figure 31: Rhodamine 6G transport is enhanced in the S837A and S854A mutant strains 98 Figure 32: An S837D substitution mutant is also hyper-resistance to Pdr5 transport 100 substrates Figure 33: The S837D mutant strain exhibits enhanced R6G transport 101 Figure 34: Plasma membrane vesicles made from the S837A strain have more Pdr5 in the 104 membrane. Figure 35: The ATPase activity is doubled in PM vesicles prepared from the S837A 106 mutant. Figure 36: The alanine mutants exhibited elevated levels of Pdr5 in whole-cell extracts. 107 Figure 37: Pdr5 is a relatively stable protein 108 Figure 38: The S837A mutant has enhanced levels of PDR5 transcript. 110

Figure 39: The S837A mutation increases the PDR5 transcript half-life. 111 Figure 40: Recreation of the WT and S837A mutant strains recreated their phenotypic 113 differences Figure 41: mRNA expression level of Pdr5 from the WT and S837A transformants 114

LIST OF TABLES

Table1: list of clinically important ABC transporter present in humans 5 Table 2: Interaction and function of NBD motifs 9 Table 3: Members of ABCC and ABCG subfamily implicated in multidrug resistance 13 Table 4: Yeast Strains 28 Table 5: Quick Change Primers 30 Table 6: Quick change PCR protocol 31 Table 7: PDR5 amplification PCR protocol 39 Table 8: Primer Sequences for PCR amplification of PDR5 39 Table 9: Recipe for Long Range PCR reaction 40 Table 10: Thermal cycle for Long Range PCR reaction 41 Table 11: Preparation of diluted albumin (BSA) Standards 47

Table 12: cDNA synthesis recipe for 1µg of RNA sample 55 Table 13: List of primers used for qRT-PCR 56 Table 14: IR of fluorescence quenching versus ATP concentration: kinetic parameters 83 Table 15: Summary of h values from various experiments 87 Table 16: Nucleotide changes resulting in alanine and aspartate substitutions in linker-2 90 Table 17: Twenty serine residues are phosphorylated in Pdr5 107

LIST OF ABBREVIATIONS

5-FOA 5-Fluoroorotic acid β-ME Β-mercaptoethanol mix ABC ATP binding cassette ATP adenosine triphosphate BCA bicinchronic acid CDC Center for Disease Control DMSO dimethyl sulfoxide ECL extracellular loops G418R resistant to G418 Glu glutamic acid Gln glutamine Gly glycine HRP horse radish peroxidase

IC50 drug concentration in which 50% of growth or activity is inhibited ICL intracellular loops LB lysogeny broth (Miller) NBD nucleotide binding domain NPBST PBST with 5% nonfat dry milk PBS phosphate buffered saline PBST PBS with 1% tween PCR polymerase chain reaction Pdr plieotropic drug response Pgp P-glycoprotien PM plasma membrane R6G rhodamine-6-G

xii

Ser serine SD synthetic dextrose SDS sodium dodecyl sulfate TMD transmembrane domain TMH transmembrane helix Ura uracil WHO World Health Organization WT wild-type YPD yeast peptone dextrose

xiii

ACKNOWLEDGEMENTS

First and foremost, I would like to thank the Almighty God for giving me the strength to pursue this journey to my PhD. I am so much indebted to my thesis supervisor, Dr. John Golin, who has provided me with excellent guidance and support throughout this whole journey. I couldn’t imagine having finished this work without his help and constant motivation. He is such a great person and mentor whose presence in my life is a great blessing. I would also like to thank Dr. Pamela Tuma and Dr. John S Choy, who have provided me with guidance and have helped me to mature scientifically. I am grateful to them for their time and the contributions they made throughout this journey.

I would also like to express my appreciation for a number of members of the Golin lab for their support and encouragement. I am particularly thankful to have worked with Dr. Stephen Patrick Joly, whose personality and knowledge gave me much strength. It was also a great pleasure to have worked with many undergraduate students who have contributed significantly to my work, most notably Michael Robertello, Dante Nicotera, and especially Joshua Carneglia and Andrew Rudrow, who have become very excellent colleagues, friends, and researchers.

I am very grateful to my colleagues and friends here in the Biology Department. Special thanks to Sameer Shah who is always a very good friend to me. I am also grateful to current and previous CUA Biology Department graduate students who have supported, encouraged, and befriended me. I would like to acknowledge Wavell Pereira and Trinh Thuan for their great help and service to the Golin lab and to the whole department. Many thanks to my friend Kabir, Iqbal, Rafi, Zaman, Amdad, Masum and many others who are always a source of inspiration.

Finally, I am very much thankful to my parents for their constant prayers, encouragement and mental support. I am thankful to all my siblings for their support as well. Finally, I am truly grateful for my wife Nabila for her great support and patience throughout this endeavor.

xiv

INTRODUCTION

Historical perspective and impact of drug resistance

Infectious agents or pathogens might be the costliest in terms of number of lives that have been lost. Since the discovery of infectious agents back in the late 19th century, scientists have worked hard to find the appropriate therapeutic regimen. Paul Ehrlich, the father of modern chemotherapy, when he used agents to treat trypanosome infections, he observed that the agents were not working and were resistant to the microbes (George, 2009). His idea of resistance was that one strain is susceptible to one agent but is resistant to another. He also pointed out that resistance is inherited.

Alexander Fleming discovered penicillin in 1928, and its first medical application in the

1930 to treat gonococcal infection provided hope for successful treatment of bacterial infections

(Wainwright and Swan, 1986). Mass production of penicillin by Merck & Co. began in 1942 and was used to treat streptococcal sepsis (Grossman, 2008). The War Protection Board at that time asked for mass distribution of penicillin stocks to Allied troops fighting in Europe. By June 1945 the mass production of penicillin increased to over 646 billion units per year (Parascandola,

1980). After World War II, penicillin was available for civilian use. Mass production of penicillin, along with the discovery of sulfonamides in 1937 and subsequently tetracycline, bacitracin, and many other antibiotics, undoubtedly, revolutionized the field of medicine (Davies and Davies, 2010). However, excessive and unnecessary use of the antibiotics was accompanied by the emergence of resistant strains. Sulfonamide resistance was reported in the late 1930s 1

(Davies and Davies, 2010). When other classes of antibiotics including β-lactams, tetracyclines and phenicol’s were used, the resistant bacterial pathogens also exhibited resistance to these other antibiotics even though they had not been exposed to these drugs. Thus, these organisms developed multidrug resistance (MDR).

Juliano and Ling (1976) first reported a case of multidrug resistance (MDR) to antitumor agents in Chinese Hamster ovary cells. They observed that cells that were resistant to colchicine also showed resistance to wide variety of amphiphilic drugs. Surface labelling of resistant cell membranes revealed a glycoprotein present in the plasma membrane that was absent in wild type cells. As this glycoprotein altered drug permeability, they named it P glycoprotein (P-gp). They observed that the amount of P-gp expression correlated with the degree of drug resistance.

Genomic sequencing of P-gp revealed that this protein belong to ATP binding cassette (ABC) transporter protein (Dean and Allikmets, 1995). Multidrug resistance (MDR) is not only confined to bacterial species or mammalian cells; parasites, and fungi also develop resistance to the drug substrates used to treat infection.

MDR became a critical issue of public health concern. Fungal infection alone cost millions of lives globally each year. Based on the recent data, WHO estimated around 1.7 million deaths occur globally due to fungal infections (WHO, 2018). The Centers for Disease Control and Prevention (CDC) reported that in the United States of America alone, more than two million people are diagnosed with serious fungal infections and approximately 23,000 people die as a result (Li et al. 2016; CDC, 2016). Fungal multidrug resistance therefore remains a major

3 clinical problem (Borges and Walmsley, 2003). PDR5, an ABC transporter gene in

Saccharomyces cerevisiae, was discovered through gene amplification. It showed multidrug resistant, when tested against cycloheximide and sulfometuron methyl (Leppert et al., 1990).

Later, the Cdr1 ABC transporter implicated in multidrug resistance was discovered in Candida albicans (Prasad et al., 1995). Overexpression of Cdr1 in C. albicans conferred resistance to fluconazole (Hiller et al., 2006). Because multidrug resistance is associated with treatment failure and increased mortality, it is important to detect resistance, and develop new therapeutics to circumvent it (Ghannoum and Rice, 1999).

Mechanism of multidrug resistance (MDR)

Multiple mechanisms are responsible for multidrug resistance (MDR). The mechanisms by which MDR can be acquired include impaired uptake of the drugs, sequestration of drugs, and modulation of drug targets, drug inactivation, or overexpression or alteration of multidrug transporters (Figure 1). Often, MDR results from the overexpression of ABC transporters in fungal pathogens and cancers (Kolotoylannis and Lewis, 2002; Goodman, Fojo, and Bates, 2002;

Lage, 2003; Prasad et al., 2014; Kathawala et al., 2015). For example, azole antifungals are commonly used to treat fungal infection. However, overexpression of the Cdr1 ABC transporter in C. albicans and C. neoformans resulted in increased resistance to fluconazole and other antifungals with an azole group. Cdr1 is a member of ABCG and Pdr subfamilies, and the members of the latter are found mostly in fungi and plants where they are responsible for efflux of toxic substances from the cell. After the sequence of Cdr1 became available, it was shown to be similar in structure to human P-gp and to share 53% sequence identity with another fungal

ABC transporters Pdr5, which is found in the common baker’s yeast S. cerevisiae (Ghannoum and Rice, 1999). As all these ABC transporters are predominantly involved in MDR, understanding the phylogeny, structure and function of these ABC transporters will greatly enhance our knowledge of how these membrane proteins result in MDR.

Figure 1: Possible mechanism of drug resistance

Here are multiple ways that MDR is developed in bacteria, pathogenic fungi and even cancer cells. This illustration indicates the possible mechanisms including impaired uptake of the drug,sequestration of the drug, modulation of drug targets, drug inactivation, and enhanced expression ABC transporters in the cell membrane, or alteration of transporter function.

ABC transporters

ATP Binding Cassette (ABC) transporters are one of the oldest and the largest superfamilies of membrane proteins (Henikoff et al., 1997). ABC transporter proteins participate in the transport of a broad range of substrates across cells and are present in all living organisms ranging from prokaryotic bacteria to multicellular eukaryotic organisms (Dean et al., 2001).

Phylogenetic analysis divided the ABC transporter superfamily into seven subfamilies. There are

49 ABC transporters present in humans (Vasiliou et al., 2009). Many complex and Mendelian disorders including cystic fibrosis, retinal degeneration, cholestasis and hypercholesterolemia are the result of mutations in these protein (Dean, 2005). A list of clinically relevant transporters is found in Table 1. The first four pathologies are the result of loss-of function mutations.

Table1: list of clinically important ABC transporter present in humans1

Transporter Function Medical condition

CFTR Removal of chloride ions cystic fibrosis ABCA1 Transport of cholesterol Tangier’s disease

ABCB11 Removal of bile salts from the liver Severe liver damage

ABCG5/G8 Cholesterol transport Sitosterolemia

P-gp Multidrug efflux from cells Overexpression leads to multidrug resistance to chemotherapeutic treatment ABCG2 Multidrug efflux from cells Overexpression leads to multidrug resistance in breast cancer patients

1 Information in this table compiled from various sources including Zhang et al., 2017; Dean and Allikmets, 1995.

ABC transporters bind and hydrolyze ATP in the cytosol to move substrates either inside or outside of the cell or to flip molecules from the inner to the outer leaflet of the membrane

(Cremers et al., 1998; Cserepes et al., 2004; Dean and Annilo, 2005). Based on the direction substrate translocation by, the transporters are functionally categorized into two types: importers and exporters. Importers transport substrates, organic ions and nutrients into the cell and are predominantly present in prokaryotes. The vitamin B12 importer, BtuCD, and the methionine importer, MetNI, from E. coli are good examples (Cui and Davidson, 2011). Exporter efflux various transport substrates including drugs, ions, toxins outside of the cells and are present in both prokaryotes and eukaryotes. Pdr5, Cdr1, Mdr1 and P-gp are examples of multidrug exporters.

Recently, structures of both bacterial and mammalian ABC transporters have been solved

(Chang and Roth, 2001). There are also numerous sequence alignments of ABC transporters.

Both the structures and alignments give us important information regarding their domain structures. ABC transporters all contain two nucleotide binding domains (NBD) and two transmembrane domains (TMD) (Figure 2). All of the ABC transporters share these common four domains and these structural domains are conserved across species. Thus, a functional full length ABC transporter consists of two NBD and two TMDs or a dimer of two half transporters

(Dean and Annilo, 2005). Each of the TMDs consists of six alpha helices which cross the lipid bilayer multiple times and have low sequence conservation. In contrast, NBDs are cytosolic and have highly conserved sequence motifs.

Figure 2: Schematic representation of ABC transporter This figure illustrates the four domain presents in a functional ABC transporter and their spatial arrangement. This organization of domain is based on a 2-D topological model structure. Overall, the transporter has two transmembrane domains (TMD) and two nucleotide binding domains (NBD) with the amino terminal (N) and carboxy terminal (C) ends of the protein noted. In a forward topology, the NBD is followed by a TMD. Two TMDs with 12 α-helices span the lipid bilayer. NBDs are in the cytosol and are connected with TMDs through a long stretch of amino acids. This schematic of an ABC transporter is based on the Sav1866 and P-gp crystal structures (Cannon et al., 2009. and Aller et al., 2009).

Typically, an NBD in an ABC transporter has at least 6 sequence motifs which include two Walker motifs (A and B), a much conserved signature motif S (LSGGQ) (also known as the

C-loop) an H-loop, a Q-and a D-loop (Lewis et al., 2004). Four consensus motifs (A, B, H, Q) from one NBD and two motifs (S, D) from the second NBD create each of the ATP binding sites

(Figure 3).

Figure 3: Nucleotide binding motifs in ABC transporters In this figure, each letter designates an individual motif (described in detail in Table 2). Two NBDs, each with six sequence motifs, are shown in NBD1 in light green and NBD2 in blue. NBDs bind together head to tail and form two sites where two ATP molecule can bind.

Multiple residues from all these motifs play crucial roles in positioning of ATP and liberating the γ-phosphate through enzymatic catalysis. Walker A, Walker B, H loop and Q-loop of one NBD and the D-loop and Signature motif of the opposite NBD co-ordinate the positioning of the ATP molecule, water molecule and magnesium. The glutamine and histidine residues in theQ loop and H-loop participate in ATP recognition and hydrolysis (Table 2; Lewis et al., 2004; ter Beek et al., 2014).

Table 2: Interaction and function of NBD motifs

Motif Function Walker A (P-loop) Coordinates Mg2+ and the β- and γ-phosphate (GXXGXGK(S/T))

Walker B The glutamatic acid residue is critical for ATP hydrolysis (ϕϕϕϕDE)

D-loop Helps in the geometry of the catalytic site and stabilizes NBD: NBD (SALD) interactions

H-loop Assists in geometric positioning of the attacking water, ATP, and Mg2+ molecules

Saccharomyces cerevisiae as a model organism for studying drug resistance

Saccharomyces cerevisiae is single cell eukaryotic organism commonly known as baker’s yeast or budding yeast. It is one of the most intensively studied model organisms for several reasons. It has a short generation time and doubles every 90-120 minutes at the optimal growth temperature 300C (Boekhout and Robert 2003). It is easy to culture in the laboratory and to manipulate genetically. It has ~6,275 genes compactly organized on 16 linear chromosomes

(Goffeau, et al., 1996). It was the first eukaryotic genome to be completely sequenced (Botstein, et al., 1997).

Yeast can live in culture as a haploid or diploid. It can exist in three specialized cell mating types; a, α and a/α. When a and α haploid cells are placed adjacent to each other, they mate with ~100% efficiency and generate a/α diploid cells. Both a and α cell types undergo

10 mitotic cell division under normal growth condition. In addition to mitotic growth, a/α diploids can undergo meiotic cell divisions when starved for nitrogen. This process, known as sporulation, gives rise to four haploid spores. Two products are MAT a and two are MAT α. This facilitates standard genetic analysis. The life cycle is illustrated in Figure 4.

Figure 4: Life cycle of yeast2

Saccharomyces cerevisiae has two life cycle forms, a and α. A haploid cell type can easily grow in the laboratory and propagates through asexual (mitotic cell division) budding. These haploid cells mate with each other and form diploid cells. Under nutrient stress, the diploid yeast will undergo sporulation and through meiosis, give rise to four haploid cells which can subsequently propagate and again mate each other to continue its life cycle. (1) Budding (2) Conjugation (3) Spore.

2 This figure adapted from Wikipedia open source with modification

Research using S. cerevisiae has contributed to the discovery of biologically important human proteins by studying their homologs in yeast. After yeast genome sequencing was completed in 1996, the data was placed in the public domain through the Saccharomyces cerevisiae genome database (SGD). The availability of the yeast sequence database and a collection of isogenic deletion mutants made it even more valuable as a model organism for understanding the regulation and structure of eukaryotic cells. Nearly 31% of the protein encoding genes of yeast are human homologues. No other model organism has contributed more than yeast to the discovery of biologically significant mammalian genes related to the aging process ((Longo et al., 2012). Yeast has been extensively used to understand the fundamental processes of meiotic recombination and DNA repair (Ruderfer et al., 2006). The robust homologous recombination machinery in yeast cells make it possible to easily knock-in or knockout genes. The effect on the fitness of a gene knockout is tested against an isogenic wild- type stain. S. cerevisiae is a noninvasive human pathogen with very low virulence and therefore rarely causes infection (Murphy and Kavanagh, 1999). However, related pathogenic yeast species, including Candida albicans and Cryptococcus neoformans, cause invasive fungal infection. Treatment failure of these fungal infections affects more than 1.5 million lives globally. One of the great contributing factors is drug resistance which is often caused by the overexpression of ABC transporter protein in these pathogenic yeast species. Due to the genetic similarity of these proteins, studying ABC transporters in S. cerevisiae will not only elucidate the mechanism of drug resistance observed in pathogenic fungal infections, but also will provide great insight in understanding the mechanism of multidrug resistance in the chemotherapeutic treatment of human disease.

ABC transporters in S. cerevisiae

Due to genetic, cell biological and biochemical tractability, yeast is a well-suited model organism for studying protein functions and protein interactions. Based on BLAST searches, the yeast genome contains 30 ABC proteins; 22 of them are true ABC transporters and the remaining eight are considered non-transporters (Paumi et al., 2009). Based on phylogenetic analysis, five subfamilies represent all the of the ABC proteins: Subfamily ABCB, ABCC, ABCD, ABCE,

ABCF and ABCG ((Lamping et al., 2010; Paumi et al., 2009). Of these subfamilies, MDR

(multidrug resistance) proteins are in ABCB transporters, MRP (multidrug resistance-associated protein) is a member of the ABCC subfamily, and PDR (pleiotropic drug resistance) proteins are classified with other ABCG transporters and are involved in antifungal resistance (Taglicht and

Michaelis 1998). Members of ABCC subfamily play important roles in the efflux of xenobiotic compounds (Paumi et al., 2009). Most members of ABCG (Pdr) subfamily expressed at the plasma membrane in S. cerevisiae and play important roles in multidrug resistance (Golin and

Ambudkar, 2015; Paumi et al., 2009). Table 3 displays two of the five S. cerevisiae ABC transporter subfamily members implicated in multidrug resistance.

Pleotropic Drug Resistance 5 (Pdr5)

Pdr5 is the founding member of the pdr subfamily of efflux pumps found only in fungi and slime molds (Lamping et al. 2010). Pdr5 was discovered as a DNA sequence which when overexpressed via a high copy plasmid, conferred increased resistance to cycloheximide and sulfometuron methyl (Leppert et al., 1990).

Table 3: Members of ABCC and ABCG subfamily implicated in multidrug resistance3.

Subfamily Transporter Localization Function

ABCC Yor1 Plasma membrane Efflux oligomycin, organic anions and other (MRP) compounds ABCG Pdr5 Plasma membrane Multidrug resistance to xenobiotic compounds (Pdr)

Pdr10 Plasma membrane Pleiotropic drug resistance network

Pdr11 Plasma membrane Multidrug resistance and sterol uptake

Pdr12 Plasma membrane Weak organic acid efflux

Pdr15 Plasma membrane Drug efflux

Pdr18 Plasma membrane Pleiotropic drug resistance

Snq2 Plasma membrane Multidrug resistance

Aus1 Plasma membrane Sterol uptake

Yol075c Plasma membrane Unknown

Adp1 Plasma membrane Unknown

Genetic mapping identified Pdr5 on the chromosome XV between ADE2 and HIS3. Tn5 insertion mutations in Pdr5 resulted in a hypersensitive phenotype when compared to the isogenic wild type strain. Insertion mutations also ruled out the possibility that these genes were essential (Leppert et al., 1990). The sequence of Pdr5 later appeared through three other genetic screens before it was completely sequenced (Balzi et al., 1994). Pdr5 mediates resistance to

3 Information presented in this table compiled from Lamping et al., 2010; Paumi et al., 2009.

14 hundreds of structurally and functionally diverse compounds. Importantly, like its human homolog ABC transporter P-gp, it also known to transport anticancer substrates (Kolaczkowski, et al., 1996).

At the time when Pdr5 was discovered, it was not known how the Pdr5 gene was regulated. Meyers et al., (1992) reported that the gain of-function allele PDR1-3 conferred enhanced cycloheximide (cyh) resistance through overexpression of Pdr5 (Meyers, et al., 1992).

Later it was reported that functionally overlapping zinc-fingered transcription factors, Pdr1 and

Pdr3, regulate the transcription of Pdr5 (Balzi et al., 1994; Katzmann et al., 1994).

Pdr5 has 1511 amino acids in a single polypeptide chain with a molecular weight of 168 kDa.

Like all other members of the Pdr subfamily, Pdr5 has 12 transmembrane α-helices (TMHs); the first six of the helices make up transmembrane domain 1 (TMD1) and the remaining six helices make the TMD2. The TMDs show low sequence similarity. Each of the TMHs are connected with each other through a series of four intracellular (ICL) loops and six extracellular loops

(ECLs). It is important to note that ABCG and Pdr subfamily members, including Pdr5, have a reverse orientation of NBDs and TMDs compared to the standard arrangement in other ABC transporters. They also have very short ICLs like the bacterial ABC importers. A schematic secondary structure based on the topology studies of Pdr5 is found in Figure 5.

Figure 5: Schematic secondary structure of Pdr5 based on the topology. Like other Pdr subfamily members, Pdr5 has twelve transmembrane alpha helices (TMHs) forming two transmembrane domains (TMDs) in the plasma membrane which connect with two nucleotide binding domains (NBDs) located in the cytoplasm. The intracellular loops (ICLs) are very short. The ECL3 and ECL6 extracellular loops are relatively long (Golin and Ambudkar, 2015).

Pdr5 is an asymmetric ABC transporter. Like other ABC transporters it has two nucleotide binding domains. These undoubtedly come together in a head-to-toe orientation to create the two ATP binding sites. A striking feature of the ABCG (Pdr) subfamily is its motif deviance. Each NBD has a canonical portion and a non-canonical portion (Fig 6). As a result, one ATP site is catalytic, the other is not.

Canonical Non-canonical Motif ATP site ATP site Walker A GKTT GCTT

Walker B DE DN Q-loop Q E Signature VSGGE LNVEQ

D-loop GLD GLD

Figure 6: Canonical and non-canonical ATP binding sites. Pdr5 is an asymmetric ABC transporter as both the ATP binding sites are not identical. The canonical ATP binding site is composed of the signature (S) and D motifs from NBD1 and the Walker A, Walker B and Q-loop motifs from NBD2. The non-canonical ATP binding site is composed of the Walker A, Walker B and Q-loop motifs from NBD1 and the signature (S), D- motifs from NBD2. The non-canonical ATP binding site cannot hydrolyze ATP. Changes in amino acids in the motifs in canonical vs non-canonical are also presented (Golin and Ambudkar, 2015).

The canonical ATP binding site has a catalytic glutamic acid (E) residue in the Walker B motif whereas in the non-canonical ATP binding site, the glutamic acid residue is replaced with asparagine (N). Pdr5 also has a non-canonical sequence in the Walker A, Signature (C-loop) and

Q-loop motifs. Overall, the non-canonical site has 11 nucleotide substitutions. Importantly, the

Walker A lysine residue is replaced by cysteine in the non-canonical site. The whole signature

(c-loop) motif is changed in the non-canonical site (Figure 6). These observed changes in sequence motifs are unique compared to other ABC transporters (Golin and Ambudkar, 2015).

The ATP hydrolysis in the canonical site is the key energy source required to transport drug and thereby develops drug resistance (Ernst et al., 2008). The communication between ATP hydrolysis in the ATP binding sites and drug binding and release in the TMDs is regulated through a signaling pathway that involves the non-canonical site (Furman et al., 2013).

The signaling interface of Pdr5

Because Pdr5 is a polytopic protein, its transport activity is mediated by a signaling interface between the drug binding and ATP hydrolysis sites. The drug binding cavity in the

TMDs is composed of a very diverse sets of amino acids allowing structurally and functionally diverse drugs to bind. From the drug binding sites in the TMDs, there is a channel used to transport drugs across the membrane. (Seeger and van Veen, 2009).

Amino acid residues in the transport channel play a critical role in substrate recognition, binding, and transport. (Rutledge, et al., 2011). The high-resolution atomic structure of Sav1866 provided foundation for studying the signaling pathway in Pdr5. The ADP bound structure

18 revealed a large cavity, which is exposed to the extracellular space, and is considered to be part of the signal transmission pathway (Seeger and van Veen, 2009). The ADP bound homodimer of the Sav1866 structure provides evidence that ICL2, which connects TMH2 and TMH3, contacts the X-loop and Q-loop residues of NBD2 in a trans-orientation (Zolnerciks et al., 2007). In

MsbA, a lipid transporter in E. coli, a similar observation revealed that ICL2 interacts with the

Q-loop NBD of the opposite monomer suggesting a signal interface through these regions. The transporter associated with antigen processing (TAP) protein belongs to the ABC transporter family and transports a diverse spectrum of peptides across the membranes (Oancea et al., 2009).

A similar picture emerges with this eukaryotic transporter. ICL2 contacts the trans X-loop and

Q-loop residues in NBD2 thus providing further structural support for a conserved signaling pathway (Golin and Ambudkar, 2015). These studies, however, do not provide functional evidence for a signaling interface. This was obtained using the power of the yeast genetic system which allowed for suppressor mutation analysis. This approach, which is not easily performed in higher eukaryotes, established an in vivo map of the Pdr5 transmission interface (Golin and

Ambudkar, 2015).

Suppressor analysis of an S558Y mutant that uncouples transport and ATPase activity

Study of the S558Y phenotypically null mutation, which was hypersensitive to all tested drugs, provided strong support for a signal interface in Pdr5 (Sauna et al., 2008). The S558Y mutation in TMH2 retained significant drug binding capacity and was expressed at normal levels in purified plasma membrane (PM) vesicles. Furthermore, the ATPase activity of S558Y was reduced compared to WT but should have supported at least some transport. This mutant,

19 however, was completely transport deficient and exhibited drug hypersensitivity that was similar as Δpdr5 (Sauna, et al., 2008). In order to investigate if the S558Y ATPase activity had altered sensitivity to inhibition, clotrimazole, which acts as s a non-competitive inhibitor, was used

(Golin et al., 2007). The inhibition of The S558Y ATPase was more resistant to inhibition than the WT enzyme. This observation suggested that the signal for allosteric inhibition of ATPase activity was not communicated properly to the NBDs from the TMDs. Taken together, these data indicated that the S558Y mutant behaved as a transmission-interface defective mutant. Because the S558Y mutant was in TMH2, these data suggested that the signaling between this helix and the attached ICL1 was disrupted (Sauna et al., 2008). Sauna et al. (2008) proposed that the large size of the tyrosine residue side chain in the S558Y mutant kinked TMH2 and as a result, the

ICL1 no longer reached the Q-loop of NBD1. To identify the signal interface, they screened for suppressor mutants of S558Y that restored significant resistance.

Sequence analysis of the suppressor mutants revealed that many of these were in the Q- loop (N242k, E244G, and D246Δ), ICL1 (S597X) and TMH5 (M679L) which is connected to

ICL2. The N242K mutant was analyzed in detail. When N242K was coupled with S558Y, resistance to clotrimazole and cycloheximide was restored without further increasing ATPase activity. Also, the double mutant exhibited a wild-type like inhibition of ATPase activity by clotrimazole. From these data, it was hypothesized that the mutation of the asparagine residue to the large, bulky lysine residue re-established the connection between ICL1 and NBD1. An analysis of the E244G Q-loop mutant yielded similar results (Ananthaswamy et al., 2010). These observations, plus the location of the other suppressor mutants gave strong support to a cis

(rather than trans) interface involving ICL-1, TMH2, the Q-loop, TMH5 and ICL2. A homology model of Pdr5 based on the solved ABCG5/8 structure is found in Figure 7. The cis interface predicted earlier by genetics is readily apparent.

Figure 7: Homology model of Pdr5 shows spatial arrangement of TMDs and NBDs. This homology model of Pdr5 based on ABCG5/G8 shows spatial arrangement of TMDs and NBDs in 3-dimensitional structure. Extracellular loops (ECLs) are on top and with a different color. ECL3 (cyan) and ECL6 (light pink) are clearly larger than other ECLs. All 12-α helices are identical. TMD1 consists of six α-helices that are in yellow color and the six helices of TMD2 are in blue color. NBD1 (brown) that is connected with TMD1 through a long stretch of amino acids. NBD2 (cyan) is also connected with TMD2 through another long stretch of amino acids (Tanabe et al., 2019).

Further suppressor analysis demonstrated that when an N242K mutant was constructed in an otherwise WT background, the resulting strain was hypersensitive to multiple drugs (Sauna et al., 2008). Thus, drug-resistant suppressors of N242K cycloheximide hypersensitivity were later obtained and sequenced. Remarkably, bioinformatic analysis indicated that two of these (V656L and A666G) were in ICL2 (Figure 8). These two mutants also appeared again when the WT overexpressing strain was exposed to a lethal concentration of cycloheximide. These mutants were termed ultra-resistant because their IC50s exceeded the WT (Downes et al., 2013).

Figure 8: Location of suppressor mutation in Pdr5. Gray circled residues (N242, E244, D246) in the Q-loop suppress the hypersensitivity of S558Y (light green circlse in TMH2). These residues in gray appeared in the genetic screen which either suppressed clotrimazole sensitivity or showed enhance resistance when yeast cells were plated on lethal doses of cycloheximide. Suppressors of N242K cycloheximide hypersensitivity are indicated in light green (Golin and Ambudkar, 2015).

The V656L mutant was studied extensively (Downes et al., 2013). The V656L was 1.4 to

2.1 times more resistant to seven compounds than the WT. When ATPase activity of this mutant was tested, it become readily apparent that V656L showed hyperresistance without a further increase in the ATPase activity. These data provided evidence that V656L in ICL2 plays an important role in transmitting the signal from the NBDs where ATP is hydrolyzed to the TMDs where the drug is bound. Furthermore, the hypersensitive E244G Q-loop mutation was suppressed by the V656L mutation. The Q-loop was known to play a critical role in the intradomain crosstalk of ABC transporters, including Pdr5 (Urbatsch et al., 2000; Dalmas et al,

2005; Ananthaswamy et al. 2010). Later, Kolaczkowski et al. (2013) characterized a series of hyperresistant mutants in the CaCdr1 transporter of the pathogenic fungus C. albicans in an important study. This ABC efflux pump has 56% amino acid identity to Pdr5. These mutations also were in the signal transmission interface, and these investigators analyzed both single- mutant and multiple-mutant combinations with respect to drug resistance and ATPase activity in the presence of transport substrates. Several of these mutants were similar phenotypically to the

Pdr5 collection. However, the mechanism responsible for the increased resistance in the

Saccharomyces and Candida mutants was not elucidated.

The V656L mutant was intriguing because it increased resistance without altering Pdr5 expression. However, resistance can also be elevated by simply increasing the level of Pdr5 in the plasma membrane. Balzi et al., (1987) first reported that PDR1 gene isolated from S. cerevisiae was responsible for the pleiotropic drug resistance. They proposed that Pdr1 controlled expression of several target genes mediating drug resistance. This was because the

23 sequence of Pdr1 contained a zinc finger suggesting Pdr1 peptide might bind to DNA. Later,

Leppert et al., (1990) identified PDR5 as a target sequence, which when amplified on a high copy plasmid, increased resistance. Meyers et al., (1992) demonstrated that gain-of-function mutations in PDR1 lead to overexpression of Pdr5 in yeast resulting in increased multidrug resistance. The gain of function mutation in Pdr1-3 which resulted in overexpression of Pdr5 also increased the level of the Yor1 and Snq2 transporters in the plasma membrane and thus increased resistance to 4-nitroquinoline.

My dissertation explores two classes of Pdr5 mutations that increase resistance. The first class is represented by the A666G mutant (which resembles and was isolated with the V656L hyper resistant mutant found in ICl-2) is due to a novel mechanism of resistance that does not require enhanced expression. The second class of mutants are in the linker-2 region of Pdr5 and further increase expression.

Bioinformatics analysis of Pdr5 indicates that it has very long and relatively unconserved linker regions that connect portions of the transmembrane domains (TMDs) with the nucleotide- binding domains (Rutledge et al., 2011). In this study, we focused on linker-2 which connects

TMH6 to the canonical portion of NBD2. Linker-2 had never been studied extensively. The lab was attracted to it because a stretch of six serine residues that are phosphorylated (Chi et al.,

2007; Li et al., 2007; Albuquerque et al., 2008; Holt et al., 2009). This was especially interesting because the role of phosphorylation depends on the particular ABC transporter.

A relatively early study of Pdr5 indicated that phosphorylation of the transporter is mediated by the overlapping casein kinase-1 isoforms, Yck1 and Yck2. The double mutant is temperature-sensitive lethal that exhibited reduced localization of Pdr5 to the plasma membrane

(Decottignes et al., 1999). Several of the serine residues in linker-2 are targets of these kinases.

In the case of the cystic fibrosis transmembrane conductance regulator (CFTR), phosphorylation of its regulatory region is central to channel function (Gadsby and Nairn, 1996; Mense et al.,

2006). Both CFTR and Pdr5 are asymmetric ABC proteins, however, CFTR is unique in that it has cytosolic regulatory (R) domain which controls channel opening. Recent structures of CFTR from both human and zebrafish have been solved by Zhang et al., (2017). The CFTR structure was in a dephosphorylated state, the R domain created a wedge between the two nucleotide binding domains and prevented dimerization of the NBDs (Figure 9).

Figure 9: Conformational change upon phosphorylation of R domain leads CFTR channel to open. CFTR has an R domain which creates a wedge between the two transmembrane domains. Without phosphorylation of the R domain, the nucleotide binding domains do not make dimers. Upon phosphorylation of the R domain, it disengages, and the nucleotide domains dimerize and causing the conformation of the outer domain to change allowing the channel to open outside of the cell (Zhang et al., 2017)

In cryo-EM structure, a single gate near the extracellular surface is closed, and thus block ions to pass across the membrane. They proposed that upon phosphorylation of the regulatory

(R) domain and presence of ATP, both the NBDs bind with ATP and dimerize, and the transporter opens to the outside to allow ions to pass across the membrane. This also seems to be a feature of other members of the ABCB1 family (Aryal et al., 2015).

In human P-glycoprotein (P-gp), however, the linker region is a long stretch of 60 amino acids that connect the two homologous halves of P-gp. A cluster of serine residues in the linker region was subjected to analysis by German et al., (1996) to understand the role of phosphorylation in the modulation of drug resistance. In order to characterize the role of phosphorylation, they substituted all serine residues in the linker region with monophosphorylated alanine which prevented phosphorylation and aspartic acid which mimicked permanent phosphorylation. They found that both the alanine and aspartic mutants exhibited similar levels of cell surface expression of P-gp and conferred drug resistance at WT levels. Their study provided evidence that phosphorylation in the linker region of P-gp does not play an essential role in the multidrug resistance phenotype mediated by P-gp. The role of phosphorylation in the linker region of P-gp therefore contrasts with ABCB1 and CFTR ABC transporter proteins where phosphorylation plays a critical role in either efflux of drugs or in channel opening.

In order to characterize the role of phosphorylation in the liner-2 region, we constructed single-alanine substitutions in each of its six phosphoserine residues. Surprisingly, the resulting mutants exhibited strong multidrug hyper-resistance and enhanced whole-cell rhodamine 6G

(R6G) drug transport. Detailed analyses described in this dissertation demonstrated that this phenotype was due to enhanced transcription even though the alterations were in the coding region.

MATERIALS AND METHODS

Yeast strains

Table 4 shows the strains used in this study. Except for JG2001, all of the Saccharomyces cerevisiae strains were derived from R-1, which lacks all plasma membrane ABC transporters and contains a PDR1-3 mutation that causes overexpression of PDR5 when this gene is inserted at its chromosomal location. Thus, virtually all drug resistance to the particular compounds that we tested was mediated by the Pdr5 efflux pump. The R-1 strain which served as a negative control for most of the experiments in this dissertation. We also used a phenotypically null mutation, G312A, as a negative control. We cultured the strains at 30°C. We used the pSS607- integrating plasmid for site-directed mutagenesis as previously described (Golin et al., 2007).

This plasmid has a WT PDR5 gene under the transcriptional control of its own upstream region, as well as a URA3-selectable marker. In general, we cultured cells in yeast extract, peptone, and dextrose (YPD) medium at 30°C. Cultures used to perform whole-cell-transport assays were grown in synthetic dextrose, yeast nitrogen base medium (SD) supplemented with uracil and histidine.

Table 4: Yeast Strains

Strain Pertinent genotype Reference source Designation

MATα PDR1-3 pdr5::KANMX4 ura3 his1 yor1 pdr20 R-1 Sauna et al., 2008 pdr11 ycf1 pdr3

JG2001 Sauna et al., 2008 JG2015 R-1 + pSS607 (Wild type) Sauna et al., 2008 Isogenic to JG2015, but the pdr5::KanMX4 cassette JG2004 was replaced with a second copy of PDR5 Isogenic to JG2015 but containing a G312A mutation in Furman et al., JG2063 the insertion plasmid pSS607 instead of the WT allele 2013. Isogenic to JG2015, but containing an A666G mutation JG2133 in the insertion plasmid pSS607 instead of the WT This study allele Isogenic to JG2133, but the strain contains two copies JG2153 This study of the A666G mutation A666G R-1 + A666G in pSS607 Arya et al., 2008 S837A R-1 + S837A in pSS607 This study S840A R-1 + S840A in pSS607 This study S841A R-1 + S841A in pSS607 This study S849A R-1 + S849A in pSS607 This study S850A R-1 + S850A in pSS607 This study S854A R-1 + S854A in pSS607 This study S837A+S850A R-1 + S837A and S850A in pSS607 This study S837A+S854A R-1 + S837A+S854A in pSS607 This study Winston et al., BY4741 MATa his3Δ1 leu2Δ0 met15Δ0 ura3Δ0 1995

Chemicals and drugs

All chemicals were purchased from Sigma Aldrich (St. Louis, MO) except for 5- fluoroorotic acid (5-FOA) and G-418, which we obtained from Research Products International

(Mt. Prospect, IL), aprotinin, which we purchased from Fisher Scientific (Waltham, MA), and climbazole, cerulenin, cyproconazole, tebuconazole, and imazalil, which we obtained from LKT laboratories (St. Paul, MN). All chemicals were dissolved in DMSO except for cycloheximide, which was dissolved in water, and 5-FOA and G-418, which were dissolved in sterilized media.

Thiolutin was purchased from Cayman Chemicals.

Media

All nutrient media were purchased from Midsci (St. Louis, Mo) and prepared according to package directions with milli-Q water. Strains were maintained on yeast peptone dextrose

(YPD) agar prepared from YPD broth base with 2% bacto-agar. Yeast transformants were selected on synthetic dextrose with histidine (SD+his) media prepared with 20 g dextrose, 20 g bacto-agar, 7 g yeast nitrogen base without amino acids (purchased from Research Products

International, Mt. Prospect Il) and 0.1g histidine per liter milli-Q water, sterilized by autoclaving.

G-418 (0.2 g/L) and 5-FOA (1.0 g/L) were added in powdered form to sterilized and cooled YPD or SD+his with 0.1 g/L uracil (SD+his+ura) media, respectively.

Site- directed mutagenesis Site-specific mutations were introduced into the plasmid pSS607 using a Stratagene

QuikChange Lightning Site-Directed Mutagenesis Kit (Agilent Technologies La Jolla, CA).

Primers were designed using the QuikChange primer design program at www.agilent.com/genomics/qcpd. These primers were supplied by Integrated DNA Technology.

Primers used in this study are listed in Table 5.

Table 5: Quick Change Primers

Name Primer sequence Strains created using these primers S837A-F 5’-ctatcgctggataagtcagcacgttccccacgttttcgg-3” S837A

S837A-R 5’-ccgaaaacgttggggaacgtgctgacttatccagcgatag-3” S837A+S854A,

S837D-F 5’-ctatcgctggataagtcatcacgttccccaacgttttcgg-3” S837D

S837D-R 5’-ccgaaaacgttggggaacgtgatgacttatccagcgatag-3” S837D

S840AF 5’-cctatcgctggctaagtcactacgttccccaa-3” S840A

S840AR 5’-gttggggaacgtagtgacttagccagcgatagg-3” S840A+S850A

S841AF 5’-cttgtagcattttcctatcggcggataagtcactacgttccc-3” S841A

S841AF 5’-gggaacgtagtgacttatccgccgataggaaaatgctacaag-3”

S849AF 5’-attcctcttcagaggcttcttgtagcattttcctatcgctgg-3” S849A

S849AR 5’-ccagcgataggaaaatgctacaagaagcctctgaagaggaat-3”

S850AF 5’-gtatcggattcctcttcagcgctttcttgtagcattttc-3” S850A

S850AR 5’-gaaaatgctacaagaaagcgctgaagaggaatccgatac-3”

S854AF 5’-ctccgtaagtatcggcttcctcttcagagctttctt-3” S854A

S854AR 5’-aagaaagctctgaagaggaagccgatacttacggag-3” S837A+S854A

The reaction mixture was made up of 1.0 µL of dNTP mix, 5.0 µL 10x reaction buffer,

1.5 µL Quik Solution reagent and 1.0 µL of QuikChange lightning enzyme (supplied in kit) which were added to 1.0 µL forward and reverse primers at 125 ng/µL each, and 1.0 µL pSS607 template DNA at ~100 ng/µL concentration. Reactions were carried out in TECHNE TC-3000 thermocycler (Bibby Scientific Ltd, Staffordshire,UK) with the protocol as in the Table 6.

Table 6: Quick change PCR protocol

Segment Cycles Temperature Time

1. Initial Denaturation 1 95ºC 2 minutes

2. Amplification 18 95ºC 20 seconds

60ºC 10 seconds

68ºC 8 minutes

3. Final Extension 1 68ºC 5 minutes

After PCR, the reaction mixture was treated with 2.0 µL of DpnI to remove the methlyated parental plasmid and incubated for 5 minutes in a 37ºC heat block. The remaining mixture was used to transform XL10-Gold ultra-competent Escherichia coli cells.

Bacterial transformation of XL10-Gold ultracompetent cells

XL10-Gold ultracompetent cells supplied with the Stratagene QuikChange Lightning

Site-Directed Mutagenesis Kit (Agilent Technologies, La Jolla, CA) were transformed according to the manufacturer’s protocol with modification only in the volume spread onto LB agar plates with ampicillin (50 μg/mL).

XL10-Gold ultra-competent cells were thawed on ice and 45 µL of the cell suspension was aliquoted into pre-chilled 14 mL polypropylene round-bottom tubes. 2.0 µL of β- mercaptoethanol mix (β-ME) (supplied with kit) was added, swirled to mix, and incubated on ice for 2-10 minutes. 2.0 µL of DpnI-treated DNA from the QuikChange reaction mixture, or 1.0

µL of pUC18 positive control plasmid (supplied in kit) were added to separate aliquots of β-ME treated XL10-Gold ultra-competent cells. Nothing was added to an aliquot of ultra-competent cells for a negative control. The tubes were swirled to mix and incubated on ice for 30 minutes.

The reactions were heat-pulsed in a 42ºC water bath for 30 seconds and returned to ice for 2 minutes. 500 µL of LB broth preheated to 42ºC was added to the reactions, swirled to mix, and incubated in a 37ºC incubator with shaking for 1 hour. 250 µL of the sample reactions and negative control and 10 µL of the positive control were spread onto LB agar plates containing 50

μg/mL ampicillin. The plates were inverted and incubated in a 37ºC incubator for up to 24 hours. Colonies obtained after 18-24 hours incubation were sub-cultured onto LB agar plates containing 50 μg/mL ampicillin and incubated for 24 hours at 37ºC. The resulting cultures were used to inoculate 10 mL LB broth cultures with 50 μg/mL ampicillin for plasmid isolation described below.

Plasmid isolation from E. coli

Plasmid pSS607 was isolated from E. coli using an IBI High-Speed Plasmid Mini Kit

(IBI scientific, Peosta, IA) spin column protocol with variation in the volume of starting sample and one additional centrifugation step. All reagents used were supplied with the kit. 10 mL LB with ampicillin broth cultures were inoculated with the transformants of interest. The cultures were grown overnight at 37ºC in 15 mL sterile conical tubes with shaking. The cells were pelleted by centrifugation at top speed in a standard clinical centrifuge for 3 minutes. The supernatant was discarded, and the pellet was resuspended in 200 µL PD1 buffer and transferred to a sterile 1.7 mL microcentrifuge tube. Samples were spun at 14,000 revolutions per minute

(rpm) for 1 minute, the supernatant was decanted, and the pellet resuspended in 200 µL PD1 buffer by pipetting up and down. Then 200 µL of PD2 buffer were added and the solution was mixed by inversion 10 times. The cells were allowed to lyse for 2 minutes at room temperature and 300 µL of PD3 buffer were added and mixed by inversion 10 times then centrifuged for 3 minutes at 14,000 rpm. The supernatant was decanted into a PD Column (supplied in kit) and the pellet discarded. The PD Column was centrifuged at 14,000 rpm for 30 seconds, the flow through discarded and 400 µL of W1 buffer added to the top of the column. The PD Column was centrifuged for 14,000 rpm for 30 seconds, the flow through discarded and 600 µL Wash buffer was added to the top of the column. The PD Column was centrifuged at 14,000 rpm for 30 seconds and the flow through discarded. The PD Column was then dried in an empty collection tube by centrifugation at 14,000 rpm for 3 minutes. The dry PD column was then transferred to a sterile microcentrifuge tube and 50 µL of elution buffer were added to the column matrix. After two minutes at room temperature, the column was centrifuged for 2 minutes at 14,000 rpm to

34 collect the DNA in the microcentrifuge tube. The plasmid concentration and DNA purity were assessed using a Thermo scientific NanoDrop 2000 spectrophotometer and analysis software

(Thermo Fisher Scientific Inc. Wilmington, DE).

Yeast transformation

The R-1 S. cerevisiae strain was transformed by the lithium acetate method using a

Sigma-Aldrich kit and protocol (St. Louis, MO). We transformed yeast with a pSS607 integrating plasmid containing mutations in PDR5 introduced by site-directed mutagenesis as described above. A sterile 250 mL vented baffled Erlenmeyer flask containing 20 mL YPD broth was inoculated with R-1 cells and incubated overnight in a shaking incubator at. Six to ten mL of the overnight culture were added to 80 ml sterile YPD in a 500 ml sterile Erlenmeyer flask to an A600 of ~0.3. The culture was incubated at 30ºC with shaking for 5-6 hours until reaching an

A600 of >0.8 and <1.5. The cells were collected by centrifugation of 50 mL aliquots at 5,000 rpm for 5 minutes. The supernatant was discarded, and pellets combined and washed in 50 mL sterile water before centrifugation at 5,000 rpm for 5 minutes. The supernatant was discarded, and the cells were resuspended in lithium acetate transformation buffer (supplied with kit) to make cells competent for transformation.

The competent cells were immediately transformed. We added 100-200 ng of plasmid

DNA to microcentrifuge tubes containing 10 µL of 10 mg/mL single stranded salmon testes

DNA (supplied with kit) followed by the addition of 100 µL competent R-1 cells. The mixture was vortexed (no plasmid DNA was added for the negative control). Six hundred µL of

35 polyethylene glycol/lithium acetate/tris/ EDTA (Plate) buffer (supplied with kit) was added to the cell suspensions and vortexed to mix. The tubes were incubated at 30ºC with shaking for 30 minutes. Tubes were heat-shocked for 15 minutes at 42ºC in a heat block. The tubes were centrifuged at 14,000 rpm for 3 seconds. The supernatant was discarded before resuspending the cells in 500 µL of sterile water. 100-150 µL of the cell suspension were spread onto synthetic dextrose agar with histidine (SD+His) plates and incubated for up to 5 days at 30ºC. Resulting colonies were subcultured onto YPD and incubated for 24-48 hours at 30ºC. To determine if

PDR5 integrated properly within the R-1 strain, a recombination scheme using 5-FOA

(previously described by Sauna et al, 2008 and illustrated in Figure 10) was utilized. Cells from each transformant were inoculated into 2 mL YPD broth cultures and grown overnight at 30ºC with shaking. Cells were collected by centrifugation in a standard clinical centrifuge at 6000 rpm for 3 minutes. The supernatant was poured off, and cells were washed one time in sterile water then resuspended in 1 mL sterile water. Cell suspensions were diluted 1:100 and 20-50 µL were spread onto SD+Ura+His plates containing 1.0 g/L 5-FOA. Due to the toxicity of 5- fluorouracil, the product of URA3 enzymatic activity on 5-FOA, only cells lacking the URA3 gene through recombination will produce colonies. Segregants were then plated onto YPD and

YPD with 0.2 mg/mL G418. KANMX4 confers resistance to G418 (G418R) and those colonies that lose the KANMX4 gene will have retained PDR5 and be sensitive to G418. These G418 sensitive (G418S) colonies were retained for sequence verification and further testing.

Figure 10: Selection of transformants with PDR5 gene integration (Sauna et al., 2008). In this illustration, a selection procedure was used to identify colonies in which the cells successfully integrated PDR5 through selection on media containing 5-FOA. Colonies appeared on media containing 5-FOA must have undergone one of two potential recombination events resulting in the loss of URA3 and either PDR5 or KANMX4. We further tested colonies, which appeared on 5-FOA, to G-418 plate and identified colonies that were sensitive to G-418. Colonies sensitive to G-418 indicated for PDR5 integration were used for further study.

Determining the inhibitory concentration 50 (IC50) of a drug in liquid culture

In order to test the effect of drug on different strains, 5.0 x 104 cells from a 2-3 mL YPD broth overnight culture were transferred to tubes containing 2 mL of sterile YPD and drug at the concentrations indicated. Tubes were then incubated for 48 hours at 30ºC with shaking. The percent inhibition was calculated by determining the ratio of A600 of culture grown at a particular drug concentration against A600 of the same strain grown in YPD without drug and repeated from separate inoculations for n=3 unless otherwise indicated. Points were plotted and graphed using

GraphPad Prism 8.0 software and the IC50’s were determined from the graph. The IC50 is defined as the drug concentration where growth (A600) was one-half that of cells cultured without drug.

Isolation of genomic DNA from yeast strain

Fifty ml of yeast cell cultures were grown in SD+ His broth for overnight (O/N). The cells were pelleted by centrifugation at 5000 rpm for 3 minutes before resuspending the cells in

1.0 ml of spheroplast buffer (final concentration is 1 M sorbitol, 50 mM sodium phosphate buffer pH7.5 0.1% β-mercaptoethanol, and 100 μg/ml zymolaselyticase). After adding 1.0 ml spheroplast buffer, cells were centrifuged for 5000 rpm for 5 minutes. The supernatant was discarded, and an additional 1.0 ml of spheroplast buffer was added. The pellets were resuspended by pipetting and then incubated at 30°C for 30 minutes. Into each tube, 400 μl of 0.2

N NaOH +1% SDS were added and the tubes mixed by inversion before placing on ice for 5 minutes. The solution was transferred to two microfuge tubes. Then, 750ul of samples were added into each microcentrifuge tube and 300 μl of 5 M potassium acetate, pH 4.8 were added.

The tubes were mixed by inversion and incubated for 2 minutes at room temperature. The tubes were centrifuged at 15000 rpm for 30 minutes. Following this, 750 μl of supernatant were transferred to a fresh microfuge tube and 750 μl of 1:1 phenolchloroform solution was added.

The samples were mixed with a vortex and centrifuged at 15000 rpm for 5 minutes. The supernatant above the meniscus level was removed and 450 μl of isoproponal was added. The tubes were mixed by inverting 10 times and placed at -20°C for 30 minutes. The samples were centrifuged for 10 minutes at 15000 rpm. The supernatants were discarded, and the pellets dried for 1-2 hour before resuspension in TE buffer plus RNase. The pellets were solubilized at room temperature for overnight before determining the absorption at 260 nM.

PCR amplification of PDR5

PDR5 was amplified by the polymerase chain reaction (PCR) using a Qiagen Long

Range PCR kit (Qiagen, Germantown, MD) The reaction mixture was made up of 2.5 µL of dNTP mix, 5.0 µL 10x Long Range PCR buffer, and 0.5 µL of Long Range PCR Enzyme mix

(supplied in kit) to 4.0 µL forward and reverse primers at 5 µM each, and 1.0 µL template DNA at 80-120 ng/µL and 33 µL RNase free water (supplied in kit). The reactions were carried out in a TECHNE TC-3000 thermocycler (Bibby Scientific Ltd, Staffordshire,UK) with the recipe found in Table 7.

Table 7: PDR5 amplification PCR protocol

Segment Cycles Temperature Time

1. Initial Denaturation 1 93ºC 3 minutes

2. Amplification 40 93ºC 15 seconds

55ºC 30 seconds

68ºC 6 minutes

The PDR5 primers were supplied by Eurofins MWG Operon (Huntsville, AL) and are shown in

Table 8 below. These primers were used to amplify genomic PDR5, which was integrated in the yeast chromosome.

Table 8: Primer Sequences for PCR amplification of PDR5

Primer Sequence

Pdr5 Forward 5’-caaaagaaaaagtcacgcaagttg-3’

Pdr5 Reverse 5’-ttgaaatgtagaaagctcgctgaa-3’

Successful PCR was confirmed by the presence of a 4.7 kb band by agarose gel electrophoresis of the PCR product. A 1% agarose gel with ethidium bromide was loaded with 3-

9 µL PCR product mixed with 1µL 6x gel loading dye (IBI, Peosta IA). Electrophoresis was carried out at 90-120 V for 45 minutes to 1 hour.

Long Range PCR

In order to amplify the Pdr5 gene, which is around 4.5kb, we used the Long-Range PCR reaction kit (Qiagen). The PCR reaction was set as shown in the Table 9. The following thermal cycle (Table 10) was used to amplify the Pdr5 fragment.

Table 9: Recipe for Long Range PCR reaction

SL # Components Amount (µl)

1 Long Range PCR buffer 10x, 25mM Mg+2 5

2 dNTP 2.5

3 Primer (Pdr5 Forward primer 5uM) 4

4 Primer (Pdr5 Reverse primer 5uM) 4

5 RNase Free H2O 33.1

6 Long Range PCR Enzyme Mix 0.4

7 Template DNA 1

Total 50

Table 10: Thermal cycle for Long Range PCR reaction

Comments

Initial activation step 3 min 93 °C Initial denaturation of template DNA

3-step cycling:

Denaturaiton 15 s 93 °C

Annealing 30 s 62 °C 5 °C below Tm of primers.

Extension 1 min/kb 68 °C Extension time of 1 min per kilobase

of genomic DNA targets was used

Number of Cycles 35 35 cycles gives the best amplification

results

End of PCR cycling: Indefinite 4 °C

DNA Sequencing

The PCR product from the PDR5 amplification reaction or pSS607 plasmid (isolation described above) was sent to Retrogen (SanDiego, CA) or SeqWright (Houston, TX) for PCR product purification and sequencing by primer extension. In order to get the amino acid sequence, the nucleotide sequences were translated to amino acid sequences by ExPASy translation tool (http://web.expasy.org/translate/). The resulting sequence was used in a S.

42 cerevisiae WU-BLAST2 search (http://www.yeastgenome.org/cgi-bin/blast-sgd.pl) to compare the results of sequencing to the canonical PDR5 sequence.

Gel Electrophoresis of PCR Products

PCR products were analyzed using gel electrophoresis to determine whether the PCR reaction produced the desired products. We prepared 1% agarose gel solution with 1x TAE buffer in a 500 ml Erlenmeyer flask. The flask was heated until the solution mixture boiled.

Following this, the flask was cooled to 550C and 5 µl of 50 mg/ml ethidium bromide was added per 100 ml of agarose gel solution. The gel solution was poured onto the assembled gel apparatus and kept it at room temperature until the gel hardened. Ten microliters (10 µl) of master mixture for each sample was prepared by diluting 5 µl of sample into 3.5 µl of milliQ water and 1.5 µl of

6x loading dye. The mixture for each sample was loaded onto gel along with the DNA ladder.

Electrophoresis was performed at 100 v for an hour. Once the gel electrophoresis was completed, the gel was scanned using a ChemiDoc Touch imager (BioRad).

Whole cell protein extraction

Two days prior to total lysate preparation, 10, 20 ml cultures of each strain were started in YPD broth. Afive hundred ml erlenmyer flask containing 50 ml of YPD was inoculated with starter cultures. The number of cells inoculated into the cultures was calculated with the following mathematical formula 0.25X 105 cells/ml X 50= required amount in 1 ml. Five hundred microliters of PMSF was added to each 50 ml culture. The cultures were shaking for 20 minutes. While the cultures were shaking, a standard set of protease inhibitors was added to the

43 homogenization buffer (see Golin et al., 2007). After 20 minutes of shaking, bulk cultures were transferred to 50 ml Falcon tubes and spun at 5000 rpm for 5 minutes.

15 ml culture tubes were weighed and 2 mL of miliQ H2O added to each pellet followed by centrifugation at 5500 rpm for 5 minutes in a clinical centrifuge machine. The supernatant was decanted, and each wet pellet was weighed (usually 0.4g). Twice the volume of the pellet of homogenization buffer with protease inhibitors was added. The micro mini bead beating (MBB) vials, buffer and eppendorf tubes were placed on ice 20 minutes before beating to chill. Glass beads (1.1ml) were put in the MBB vials, followed by the cells which were in the homogenization buffer. Cells were lysed during 5 cycles with 1 min bead beating followed by 1 minute of chilling in a water bath. Each vial was then disassembled and turned upright to let the beads settle. Once the beads had settled, roughly 800 ul of total lysate was recovered and placed into a pre-chilled 1.5ml Eppendorf tube. The solution was then spun at 15000 rpm for 6-7 min at

(4°C). The supernatant was collected and aliquoted into Eppendorf tubes. Samples were stored in the -85°C ultra-freezer.

Preparation of plasma membrane vesicles a) Preparation of media and starter culture

Four liters (L) of YP media (10 g of yeast extract, 20 g of peptone/L) were prepared and divided into 5 x 2 L flasks with 720 ml of YP media placed into each flask. 500 mL of 20% glucose, 1.5 L of miliQ water and 500 mL Boost media (50 g of yeast extract, 100 g of peptone/L) and were prepared. Starter cultures were grown in 50ml of YPD and shook at 110 or

130 rpm at 30°C overnight. Eighty mL of sterile 20% glucose were added to each 2 L flask containing 720 mL YP media.

b) Cell harvesting

5 The following morning, the starter culture was measured at OD600 and 5x10 cells were placed in each of the 800 ml flasks containing YP and glucose. These flasks were incubated over night at 25°C with shaking at 140 rpm. After 15-16 hours of incubation, the OD600 was measured to determine if an OD of 1.5 had been reached. Once the OD600 reached 1.5, 80 mL of boost media were added to each flask and incubated at 140 rpm at 25°C. After about 4~ 4.5 hours of incubation, the OD600 was measured. If OD600 was 3.5, we harvested the cells at 5000xg for 15 min at 4°C. The cells were washed with 15 mL of MilliQ per 800 ml culture flask and put into a

50 mL Falcon tube. The cells were centrifuged. Each pellet was put on ice. The cell pellets were flash frozen and kept at -800C.

c) Cell disruption

All purification steps were done at 4 °C. The cell pellets from flash frozen samples were thawed on ice for at least 3-4 hours before the purification steps. The pellets were resuspended in milliQ water to 84 ml in a 100 ml cylinderthen transfered into a pre-chilled 250 ml Glass Beaker.

The beaker was stirred by hand, and 5 ml of 1 M Tris-acetate pH 7.5 and 1 ml 0.5 EDTA pH 8.0 were added. Approximately 270 ml of ice-cold glass beads and 2 ultra O-complete protease inhibitor tablets (SigmaAldrich, St. Louis, MO) were added to a bead beater container. The

45 protease inhibitors were given time to dissovle, and a resuspension solution containing the cells was transfered into the container. For better lysis of the cells, the container was filled with miliQ water. To lyse the cells, a bead beater was used. The bead beater was cooled with an ice bath for the entire time to avoid heat production during stirring. The bead beating was set on a mode of 5 x 1 min, with a 1 min break between every beating cycle.

d) Enrichment of the plasma membrane

After disruption, the homogenate was filtered using a glass filter to remove the glass beads.

The total lysates were then transferred into a pre-chilled 250 ml beaker and filterred through 60-

75 ml filter funnel. The beads were washed two times with 70 ml TAEG buffer. At each step, 70 ml of TAEG buffer were added. To remove non-disrupted cells, nuclei and the rest of the cell wall, three centrifugation steps were done using an SLA-1500 rotor in a Sorvall centrifuge. For the first step, we centrifuged for 5 minutes at 2600 rpm to get rid of cells that did not lyse. The resultant supernatant was transferred into fresh 250 ml tubes and centrifuged for the second step for 5 minutes at 2600 rpm. The supernatant was transferred into fresh 250ml tubes and centrifuged for 5 minutes at 4500 rpm. The supernatant from this step was transfered into 50 ml transparent tubes with a closed lid.

e) Extraction of plasma membrane vesicle and mitochondria via centrifugation.

The final supernatant was centrifuged for 40 minutes at 13000 rpm using an SS34 rotor. At the same time, one protease inhibitor tablet was added to 50 ml TAE buffer and kept on ice.

After centrifugation, the supernatant was discarded and the pellet (mostly mitochondria and

46 plasma membrane vesicles) was resuspended in 12 ml TAE and protease inhibitor buffer solution. The centrifuge tubes were washed with 4 ml TAE and protease inhibitor buffer solution. Both parts of the solution were collected and combined into a 50 ml falcon tube. The protein concentration was determined with the CoomassiePlus assay (Thermofisher Scientific) according to the manual (n=3). The absorption was read at OD595.

f) Precipitationof mitochondria

The plasma membrane vesicles were stirred in a prechilled 125/150 ml glass beaker in a cold room (at 4 °C) and adjusted to a protein concentration of 5 mg/ml (50 μl sample was removed and then stored at -20 °C for SDS-PAGE analysis). A few ml (5-7 ml) of 1 M sodium acetate

(pH 5.2) were used to adjust the solution to pH 5.2. The precipitated mitochondria were centrifuged in an SS34 rotor at 7700 rpm for 5 min. The supernatant was transfered to a prechilled 125/150 ml glass beaker with magnetic stirrer and the pH quickly adjusted to 7.5 with

2.5 M Tris-acetate pH 7.5.

g) Extraction of the purified plasma membrane

The purified plasma membrane vesicles were centrifuged with an SS34 rotor at 14900 rpm for 30 minutes. At the same time, one protease inhibitor tablet was added to 50 ml of 50 mM

Hepes pH 7.0 and stored on ice. The supernatant was discarded, and the pellet was resuspended in 2 ml of 50 mM Hepes (pH 7.0) with protease inhibitors. The protein concentration was determined with a 1:10 and a 1:4 dilution of the plasma membrane vesicles in 50 mM Hepes buffer. The protein concentration was adjusted to 1 mg/ml with 50 mM Hepes pH 7.0 containing

47 protease inhibitor, 140 µl plasma membrane samples were aliquoted and flash frozen in liquid nitrogen and stored at -80°C.

Determination of protein concentration using Coomassie plus

a) Preparation of Diluted Albumin (BSA) Standards

Bovine serum albumin (BSA) standards were diluted as shown in the Table 11. The

standards were used to generate a curve to measure the protein concentration of the unknown

samples.

Table 11: Preparation of diluted albumin (BSA) Standards

Vial Volume of Diluent Volume and Source of BSA Final BSA Concentration

(TAE buffer PM prep)

A 0 300 μL of Stock 2000 μg/mL

B 125 μL 375 μL of Stock 1500 μg/mL

C 325 μL 325 μL of Stock 1000 μg/mL

D 175 μL 175 μL of vial B dilution 750 μg/mL

E 325 μL 325 μL of vial C dilution 500 μg/mL

F 325 μL 325 μL of vial E dilution 250 μg/mL

G 325 μL 325 μL of vial F dilution 125 μg/mL

H 400 μL 100 μL of vial G dilution 25 μg/mL

I 400 μL 0 0 μg/mL = Blank

48 b) Standard test tube procedure to measure protein concentration

We mixed the Coomassie Plus Reagent solution (Thermofisher Scientific) immediately before use by gently inverting the bottle several times. We then removed the amount of reagent needed to quantify the plasma membrane and let the reagent equilibrate to room temperature

(RT) before use. Fifty microliters of each standard (from A to I vial) were pipetted into test tubes. For the plasma membrane vesicle, 50 µl were also pipetted into appropriately labeled test tubes (n=3 for each sample). We added 1.5 ml of the Coomassie Plus reagent to each tube and mixed the reaction well. Samples were incubated at room temperature for 10 minutes (RT). With the spectrophotometer set to 595 nm, water was used as blank and for calibration. We measured the absorbance of all the samples and blank. The average 595 nm measurement for the blank replicates subtracted from the 595 nm measurements of all other individual standards and the plasma membrane sample replicates. A standard curve was prepared by plotting the average blank-corrected 595 nm measurement for each BSA standard against its concentration in μg/ml.

The standard curve was used to determine the protein concentration of each plasma membrane sample.

Determination of protein concentration using BCA method

The protein concentration from purified plasma membranes was determined using a bicinchoninic acid (BCA) protein determination kit (Thermo Scientific, Rockford, IL). Standard curves were created using 0, 5, 10, 15, 20, 25, and 30 µg/µL samples plotted using GraphPad

Prism 8.0 software. Purified plasma membrane samples (3 at 10 µL each) were measured against

49 the standard curve and protein concentration was determined using GraphPad Prism 8.0 software.

Protein gel electrophoresis

Membrane proteins (15-30µg) were denatured in 5% sodium dodecyl sulfate (SDS) at

37ºC for 30 minutes then loaded into precast NuPage 7% tris acetate gels (Life Technologies,

Carlsbad, Ca). Proteins were separated by electrophoresis at 150 V in Novex XCell SureLock

Mini-Cell electrophoresis chambers. Tris acetate gels were stained in colloidal blue stain

(Invitrogen, Carlsbad, CA) for 1-2 hours, and then rinsed overnight before scanning for relative concentration of 160 kDa bands (the molecular weight of Pdr5) against control strains.

Western blotting

After protein separation by electrophoresis membrane proteins were transferred to nitrocellulose membranes (Life Technologies Carlsbad, Ca) in Novex XCEll II Blot Module

(San Diego, Ca) under constant 400 mA current for one hour. Membranes were blocked for 40 minutes in PBS buffer with 1% tween (PBST) and 5% nonfat milk (NPBST) and then incubated at 4ºC overnight with gentle rocking with Pdr5 rabbit polyclonal antibodies (Thermofisher

Scientific) diluted 1:5000 and Pma1 mouse monoclonal antibody (1:10000), respectively.

Membranes were then washed 3 times in PBST for 15 minutes before incubating at room temperature with gentle rocking for 2 hours in a 1:10,000 dilution of horse radish peroxidase

(HRP) conjugated anti-rabbit IgG for Pdr5 (Sigma) and 1:3000 dilution of anti-mouse for

Pma1in NPBST. The chemiluminescent signal was developed using Novex ECL HRP

50 chemiluminescent Substrate Reagent Kit (Life Technologies Carlsbad, CA). The membrane was incubated in developer for 1 minute then scan using a ChemiDoc Touch (Bio Rad). Bands were analyzed using Image J software (NIH).

Assay of ATPase activity

To measure the ATPase activity of the PM vesicles, we used 300 mM Tris-glycine buffer

(pH 9.5) in a final volume of 100 µl for 8 min at 35°C. To reduce background, we added 0.2 mM ammonium molybdate, 50mM KNO3, and 10mM NaN3 was added, respectively (Goffeau et al.,

1988). The non-Pdr5 activity observed in ΔPdr5 negative control PM vesicles was subtracted as background before calculating activity.

Quenching Assay

Each assay had a total volume of 2 ml containing 30 µg PM vesicles and 100 nM R6G, 5 mM ATP. All reactions were done at 35°C. The final concentration of the buffer was 50 mM

Hepes and 5 mM MgCl2. We removed and acclimatized the appropriate amount of Hepes buffer into a 15 ml culture tube and equilibrated it to room temperature. Room temperature (RT) sterile miliQ water was used. A master mixture was made and added into corresponding reaction tubes.

At the very last step 2ul of 100M of R6G was added to the reactions and mixed gently. The reaction was incubated for 4-5 minutes at RT and 1.96 ml placed into each cuvette. The cuvette was placed in the fluorometer chamber for 5 minutes for temperature equilibrium. Following this, 100 µl of 100 mM of ATP was added to all of the tubes and fluorescence intensity was

51 monitored for 20 minutes. The excitation wavelength was set to 529 nm and emission wavelength to 553 nm.

R6G transport assay

We grew overnight cultures for the specific strains to be tested in 5 ml YPD. On the following afternoon, three cultures per strain were started in 5 ml of SD+His+Ura at 0.2 x 105 cells/ml and grown overnight. There were three independent cultures for each strain. We determined the cell concentration at OD600 and then calculated the volume of culture needed to have 0.3 x 107 cells in eppendorf tube. The calculated volume was added to the empty Eppendorf tube. These were centrifuged for 60 seconds at top speed in the microcentrifuge at 40C. The supernatant was carefully removed, and the pellets were resuspended in 500 µl of Hepes glucose buffer containing 10 µM R6G. The cell pellets were vortexed and placed in a circulating water bath set at 300Cfor 90 minutes. The cells were pellted by centrifugation for 1 minute at 15000 rpm. The pelleted cells were resuspendedby vertexing in 1 ml of cold Hepes (minus glucose) buffer without R6G and transferred into pre-chilled FACS tubes to determine the retained fluorescence. FACSort was used to determine the fluorescence retention with an excitation wavelength set at 529 nm and emission wavelength at 553 nm (Becton-Dickinson, Franklin

Lakes, NJ). The data were analyzed using CellQuest program.

Cycloheximide chase

250 mL cultures in YPD for each strain were grown overnight. The OD value of the culture suspension was measured at ABS600. Once the cells reached expected OD600 (0.6-0.8),

52 equal number of cells were placed in 250 mL opaque tubes and spun at 5000 rpm for 5 minutes using a GSA rotor. The supernatant was discarded and the pellet for each strain was resuspended in 50 ml YPD. The suspended cells were transferred into a 2 L flask containing 250 mL of fresh

YPD. To each of these 300 mL cultures, cycloheximide was added to 250 µM. Immediately after adding the cycloheximide, the cultures were mixed, and 50 ml of culture was transferred into a

50mL Falcon tube labelled as 0 minute. Samples (50 ml) were removed at 60 minutes interval for five hours. Following collection, the cells were immediately centrifuged and washed with sterile miliQ water and flash frozen in liquid nitrogen for subsequent analysis.

RNA isolation using Zymo-Research Kit

For small scale RNA isolation, a Zymo-research RNA kit (Zymo Research) were used.

Yeast strains were grown overnight in 10 ml YPD. The next morning, cell growth was assessed

7 at OD600, and the volume of cells needed to obtain 2.5 x 10 cells was calculated. 1.4 ml of RNA lysis buffer was added to the cells and the suspension mixed with gentle pipetting.

Approximately 1 ml of 0.5 mm glass beads were added to a 2 ml mini bead beater tube. The cells were then transferred into these tubes and beat for 5 minutes with a 1 min break at room temperature (RT). Seven hundred microliter (700 µl) of supernatant were transferred to the

Zymo-spin IIICG column in a collection tube and centrifuged for 30 seconds. The flow-through was saved and 700 ml of 95% ethanol was added to it. This mixture was transferred to a second

Zymo-spin ICG column in a collection tube and centrifuged for 30 seconds. The flow-through was discarded. 400 ul RNA prep buffer was applied to the column which was then centrifuged for 30 seconds. The flow through was discarded and 700 µl of RNA wash buffer was added to

53 the column and centrifuged for 30 seconds. The flow through was again discarded and 400 µl

RNA wash buffer was added to the column and centrifuged for 2 minutes. The column was transferred to an RNase free tube and 30 µl of DNase/RNase free DEPC water were added to the column and centrifuged for 30 seconds. The RNA was transferred to an RNase free tube. The concentration of RNA was measured with a nanodrop 2000 spectrophotometer (Thermofisher

Scientific) and purified RNA was aliquoted and stored at -800C for subsequent analysis.

RNA isolation using RiboPure Yeast RNA purification Kit

RiboPure Yeast Kit (Thermofisher Scientific) was used for large scale RNA isolation of up to 3 x 108 yeast cells and provided a maximum yield of 300 µg of RNA. First, we removed the cell pellets from -80C were thawed on ice for approximately 20–30 minutes. RNA extraction was performed using a Ribopure yeast-RNA extraction kit with a few adaptations. For each sample, 750 μL of ice-cold zirconia beads were placed into a 1.5-mL screw cap tube supplied with the kit. To the cell pellets, 480 μL of the lysis buffer, 48 μL of 10% SDS, and 480 μL of phenol: chloroform: isoamyl alcohol (25:24:1, v/v/v) were added. The cells were vortexed and transferred to the tubes containing the 750 µl ice-cold Zirconia beads. Lysis was performed on the cells using mini bead beater (MBB) plus at 40C refrigerator. The tubes were centrifuged at

16,000 x g for 5 min at room temperature. The upper phase (RNA-containing phase) was collected and transferred to a fresh 15 mL culture tube. Typically, the volume recovered for each tube was about 530 μl. To the 15 mL tubes containing the partially purified RNA, the binding buffer was added thoroughly. For each 100 μl of RNA solution, 350 μl of binding buffer was added. To the previous mixture, 100% ethanol was added and mixed thoroughly. For each 100

μL of RNA solution, 235 μL of 100% ethanol was added. From the RNA and ethanol mixture,

700 μl was applied to a filter cartridge assembled in a collection tube, both provided with the kit.

Centrifugation was performed for 1 min at 16,000 x g. If the centrifugation duration was not sufficient for the total volume to pass through the filter, centrifugation was repeated for another

30 s. The flow-through was discarded and the same collection tube was reused. Another 700 μL of RNA-binding buffer-ethanol solution was added to the filter and centrifuged again at 16,000 x g for 1 min until the RNA solution eluted. The filter was washed 1x with 700 μl of washing solution1. Centrifugation was performed at 16,000 x g for 1 min and the flow through discarded.

The filter was washed 2x with 500 μL of washing solution 2/3, via centrifugation at 16,000 x g for 1 min, and the flow through discarded. The tubes were centrifuged at 16,000 x g for 1 min to completely dry the filter. The filter cartridge was transferred to a fresh collection tube (RNA- appropriate tube) and RNA eluted with 50 μL of DEPC-treated, RNase-free H2O (preheated to

100°C). The column was then centrifuged for 1 min at 16,000 x g. The RNA elution step was repeated with 50 μl of preheated DEPC-treated, RNase-free H2O. To confirm that all the total volume had passed through the filter, it was centrifuged and additional 3 min at 16,000 x g. The contents of the tubes were pooled into and RNA concentration determined, and purity assessed using a Nanodrop 2000 spectrophotometer (Thermofisher Scientific). RNA samples were stored at -80c.

55 cDNA synthesis

To make cDNA from the RNA, iScript cDNA synthesis kit (Biorad) was used and according to the manufacturer protocol. For 1µg of purified RNA sample, the recipe in Table 12 was used. Athermal cycler was used to incubate the reaction mixture; priming 5 minutes at 250C, reverse transcription 20 minutes at 460C, reverse transcriptase inactivation 1 minute at 950C and holding at 40C.

Table 12: cDNA synthesis recipe for 1µg of RNA sample

Sl # Compoonent Volume (in μl)

1 5x iScript Reaction Mix 4

2 iScript Reverse Transcriptase 1

3 Nuclease Free Water 5

4 DNase treated RNA JG2015 (ID: RN102) 10

Total 20

Q-RT PCR

We performed q-RT PCR with PowerUp SYBR green master mix (Thermofisher

Scientific) according to the manufacturer protocol. The DNA polymerase was activated at 50°C during a 2-min holding stage, followed by another 2-min holding stage at 95°C. Amplification

56 was achieved with 40 cycles at 95°C for 15 seconds and 60°C for 1 minute. The melting curve was performed for 15 seconds at 95°C and 1 min at 60°C. The primers for PDR5 and the three reference genes are found in the Table 13.

Table 13: List of primers used for qRT-PCR

Primer Sequence

PDR5-F ‘5-caaaactccactcaatcggcacccaac-3’ PDR5-R ‘5-agccatattcttaacccaggcggcac-3’ ALG9-F ‘5-cacggatagtggctttggtgaacaattac-3’ ALG9-R ‘5-tatgattatctggcagcaggaaagaacttggg-3’ TAF10-F ‘5-atattccaggatcaggtcttccgtagc-3’ TAF10-R ‘5-gtagtcttctcattctgttgatgttgttgttg-3’ GAPDH-F ‘5-cggtagatacgctggtgaagtttc-3’ GAPDH-R ‘5-tggaagatggagcagtgataacaac-3’

Metabolic labeling of mRNA with 4-thiouracil a) Cell culture and mRNA labelling with 4-TU

For 4-thiouracil (4-TU) pulse experiments to evaluate transcription rate, logarithmic phase cells were grown to an O.D. 600 nm of 0.33 in SD medium containing 1 mM histidine and 0.1 mM uracil before adding 4-TU at a final concentration of 5 mM. After 3 h growth at 30 °C in a

57 shaking incubator, the cells were collected by centrifugation at 3000 x g at 4 °C. The cells were washed once and resuspended in a volume of SD containing 1 mM histidine and 20 mM uracil.

The cultures were incubated at 30 °C. Aliquots of cells were removed at intervals of 0, 15, 30, and 40 min. The cells were washed once with 10 ml of cold phosphate-buffered saline, and after discarding the supernatant, the pellets were flash frozen in liquid nitrogen and stored at -80 °C.

To extract RNA, the cell pellets were thawed on ice for 20-30 min. RNA was extracted from up to 109 cells/sample with a Ribopure yeast RNA extraction kit (Thermo Fisher).

Following DNase treatment, RNA was quantified with a Nanodrop 2000 spectrophotometer and the concentration adjusted to 2 mg/ml. For each time point, 200-µg samples were stored for subsequent biotinylation and purification of 4-TU-labeled RNA at -80 °C.

b) Thiol-specific biotinylation and subsequent recovery of 4-TU RNA

Biotinylation of 4-TU RNA was performed with an EZ-link HPDP-Biotin kit (Thermofisher

Scientific) with minor modifications. A 200-μg aliquot of extracted RNA for each time point was heated for 1 min at 60 °C and chilled on ice for 2 min before transferring the material to a 2-ml spin tube. Following this, we added 600 μl of DEPC-treated, DNase- and RNase-free water and

100 μl of biotinylation buffer (final concentration was 100 mM Tris-HCl, 10 mM EDTA, pH 7.5, made up in DEPC-treated, DNase- and RNase-free water) and 200 μl of biotin-HPDP (1 mg/ml

DMSO). The sample was incubated for 3 h at room temperature and protected from light.

Following this, we added an equal volume of chloroform to the tube and mixed the contents vigorously. The sample was centrifuged at 13000 x g for 5 min at 4 °C, and the upper phase was

58 transferred to 2-ml microcentrifuge tubes. We added 1/10 the volume of 5M sodium chloride and mixed the sample before adding an equal volume of isopropanol and centrifuging at 13000 x g for 35 min at 4 °C. The supernatant was removed and 1 ml of ice-cold 75% ethanol was added.

The tubes were centrifuged at 13,000 x g for 10 min at 4 °C. The supernatant was removed and discarded. The tubes were recentrifuged for 30 seconds to remove any residual alcohol. The pellets were not allowed to dry before they were resuspended in 100 μl of DEPC-treated, DNase- and RNase-free water.

The biotinylated RNA was heated for 10 min at 65 °C and the samples were chilled on ice for 5 min. One hundred µl of streptavidin-coated magnetic beads (Miltenyi Biotec) were added to the biotinylated RNA (volume 200 μl). The sample was incubated at room temperature with slight shaking for 90 min. The magnetic μ columns provided with the kit (Miltenyi Biotec) were placed in the magnetic MACS MultiStand, and 100 μl nucleic acid equilibration buffer was added. Following this, 900 μl of room-temperature biotynlation buffer (100 mM Tris-HCl [pH

7.5], 10 mM EDTA, 1 M NaCl in DEPC-treated, RNase-free water) was added to the columns.

The bead/RNA mixture (200 μl) was then added to the columns. The flow-through was collected in 1.5-mL tubes and applied again to the same magnetic column. The flow-through was retained as it represented the unlabeled RNA fraction.

The columns were washed 5x with increasing volumes of washing buffer (600, 700, 800,

900, and 1,000 μl). The newly synthesized RNA was eluted with 200 μl of 0.1 M DTT. Three minutes later, a second elution was performed with an equal volume of 0.1 M DTT. After eluting

59 the RNA, 0.1 volumes of 3 M NaOAc (pH 5.2), 3 volumes of ice-cold 100% ethanol, and 2 µl of

20 mg/mL glycogen, RNA grade (Thermo Scientific) were added, and the RNA was allowed to precipitate overnight, at -20 °C. The RNA was recovered by centrifugation (13,000 x g for 10 min, at 4 °C) and resuspended in 15 μl of DEPC-treated, RNase-free water.

Mass spectrometry

Pdr5 PM vesicle proteins were solubilized in 5X SDS PAGE buffer. Forty micrograms of protein were loaded into each well and the proteins were separated by electrophoresis. The in- gel samples were reduced with DTT and then alkylated with iodoacetamide, washed and then digested overnight using trypsin. After extraction, the tryptic digest was dried down and re- dissolved in 15 μL of 2.5% acetonitrile/0.1% formic acid. 5 μL of the digest were run by nano

LC-MS/MS using a 2h gradient on a 0.075mmx250mm Waters CSH C18 column feeding into a

Q-Exactive HF mass spectrometer.

All MS/MS samples were analyzed using Mascot (Matrix Science, London, UK; version

2.6.1). Mascot was set up to search the cRAP_20150130.fasta (123 entries); Swiss-Prot database selected for Saccharomyces cerevisae, (7,905 entries) assuming the digestion enzyme trypsin.

Mascot was searched with a fragment ion mass tolerance of 0.060 Da and a parent ion tolerance of 10.0 PPM. Deamidated of asparagine and glutamine, oxidation of methionine and carbamidomethyl of cysteine, phospho of serine, threonine and tyrosine were specified in Mascot as variable modifications.

Scaffold (version Scaffold_4.8.9, Proteome Software Inc., Portland, OR) was used to validate MS/MS based peptide and protein identifications. Peptide identifications were accepted if they could be established at greater than 80.0% probability by the Peptide Prophet algorithm

(Keller, A et al Anal. Chem. 2002; 74 (20):5383-92=34) with Scaffold delta-mass correction.

Protein identifications were accepted if they could be established at greater than 99.0% probability and contained at least 2 identified peptides. Protein probabilities were assigned by the

Protein Prophet algorithm (Nesvizhskii, Al et al Anal. Chem. 2003; 75 (17):4646-58=35).

Proteins that contained similar peptides and could not be differentiated based on MS/MS analysis alone were grouped to satisfy the principles of parsimony. Proteins sharing significant peptide evidence were grouped into clusters.

RESULTS

Part I

Alanine 666 glycine: a mutation that enhances cooperativity between transport sites and increases multidrug resistance

Resistance to fungal and chemotherapeutic treatment is a huge clinical problem.

Overexpression of ABC efflux pumps often associated with resistance in treatment.

Understanding the mechanism of the resistance is critical in developing better therapeutics. In this study we focused, as mentioned in the introduction, on the yeast ABC transporter Pdr5 to understand the mechanism of drug resistance. Here I describe a mutation, A666G, in Pdr5 that shows greater resistance to most of the tested compounds than does an isogenic wild-type strain.

This mutant exhibited enhanced resistance without increasing either the amount of protein in the plasma membrane or the ATPase activity. In this part of my thesis, I investigated the mechanism responsible for enhanced resistance in A666G mutant. The result was unanticipated and novel and is a new paradigm for enhanced resistance.

Isolation of the A666G mutation

The A666G mutation appeared in two separate genetic screens. In the first screen, it was isolated as a suppressor of a Q-loop region mutant N242K, which resulted in increased hypersensitivity to Pdr5 transport substrates (Sauna et al., 2008). In the second screen, yeast cells already overexpressing Pdr5 were plated on a lethal concentration of cycloheximide (Downes et al., 2013). The A666G substitution arose here as well. We recreated this mutation on the integrating plasmid pSS607, which was placed in the ΔPdr5 strain R-1. We confirmed that the resulting strains were hyperresistant to cycloheximide. We performed all of the work in this study with the recreated mutation. We made two additional substitutions: A666V and A666L.

The former was phenotypically identical to the A666G mutant. Western bloting demonstrated that the latter failed to localize to the PM.

The evolutionary relationships between fungal ABC transporters was the subject of a detailed study (Gebelska, Krijger, and Breunig, 2006). The Pdr subfamily was initially described as one of eight clusters found in fungi. It was further subdivided into four groups of ABC transporters, each resembling Pdr5, Snq2, Pdr12, or Pdr11. A more recent study divided the Pdr subfamily into nine distinct clusters (Lamping et al., 2010). Among transporters of this subfamily, Ala-666 shows 97% conservation (Lamping et al., 2010).

The steady-state level of Pdr5 was not increased in A666G mutant PM vesicles

To determine whether the steady-state level of Pdr5 in PM vesicles could account for the large increase in drug resistance, we performed a Western blot analysis with three sets of purified

PM vesicles from the A666G strain with wild-type and ΔPdr5 preparations as controls. We obtained a Pdr5/Pma1 ratio for each of the samples in each blot where Pdr5 could be detected.

We then compared these values from the wild-type and the A666G mutant preparations. The

A666G had 1.03 + 0.26 the amount of Pdr5 as the wild-type. A t-test indicated no significant difference between the two sets of preparations. The results therefore demonstrate that steady- state levels of Pdr5 in vesicles prepared from the wild-type and A666G strains do not differ

(Figure 11A). During the course of this study, we switched to a vastly improved method for purifying vesicles. This ensured preparations of much higher enzyme and transport activity.

Coomassie blue-stained samples of wild-type, A666G, and ΔPdr5 preparations are shown in

Figure 11B. The Pdr5 band, which was absent in the ΔPdr5 sample, was estimated to make up roughly 10% of the total PM vesicle protein. We also determined that the ΔPdr5 vesicles consistently had more Pma1 in PM vesicles than either the wild-type or A666G mutant preparations.

Figure 11. The steady-state level of Pdr5 in PM vesicles is unaltered in the A666G mutant. (A) We performed Western blotting as described in the Materials and Methods with 10 µg of purified PM vesicle protein solubilized for 30 min at 37 °C in SDS-PAGE. All lanes are from the same gel, but a lane with a blot of another strain was cropped and is not shown. The ratio of Pdr5/ Pma1 is shown below the wild-type and A666G lanes. (B) We performed gel electrophoresis with 10 μg samples of solubilized PM vesicle protein prepared according to Kolaczkowski et al. (1996) as modified by Ernst et al. (2008). Following electrophoresis as described in the Materials and Methods, the gel was stained in SimplyBlue Safe Stain solution for one hour and destained in reverse osmosis water overnight. A different set of molecular weight markers was used in each panel.

The ATPase activities of the wild-type and A666G PM vesicle preparations are similar

We tested the ATPase activity as a function of ATP concentration at 35 ˚C in four independent sets of PM vesicle preparations from both strains, with the same Hepes transport buffer (pH 7.0) employed in fluorescence-quenching studies described below. We made each pair, consisting of PM vesicles from the wild-type and mutant, on the same day or on successive days. Under these conditions we noted little difference in the ATPase activities of the wild-type and mutant preparations (Figure 12A). Therefore, the increased resistance of the A666G mutant cannot be attributed to increased ATPase activity. Similarly, the Km values were not significantly different (Figure 12B).

These vesicles have a large amount of Pma1, which retains significant activity at a pH of

7.0. This results in significant background, especially at ATP concentrations that are 3 mM or greater. We therefore probably underestimated the Pdr5-specific ATPase, although we saw no difference between the strains. Pdr5 ATPase activity has a broad pH range (Ernst et al., 2008).

When Tris-glycine (pH 9.5) served as the assay buffer to measure the activities in two sets of PM vesicles (Figure 12C), the background observed in the negative control was considerably reduced. In this particular experiment, the Vmax of the wild-type ATPase activity was ~1.5

μmoles/min/mg and the Km was ~1.1 mM. The corresponding values for the A666G mutant were

1.9 μmoles/min/mg and 0.7 mM.

Figure 12. The wild-type and mutant PM vesicle preparations have indistinguishable ATPase activities. Vmax (A) and Km (B) values from assays of four PM vesicles preparations from wild-type (green box) and A666G (blue box) mutant strains. The assays were performed in Hepes buffer (pH 7.0) as described in the Materials and Methods. The horizontal bars indicate the median values. (C) ATPase activity was also assayed in Tris-glycine buffer (pH 9.5). No other parameters of the assay were altered. In this panel: green line (■) = wild-type; blue line = (□) A666G mutant.

The A666G mutant increased resistance to many Pdr5 transport substrates

We compared the resistance of the wild-type and A666G strains to the six Pdr5 transport substrates shown in Figure 13. Five additional plots are found in the Figure 14. The entire collection encompassed compounds that are distinct in structure and mechanism of action

(tamoxifen, two trialkyltin chlorides, cycloheximide, and cerulenin) and a set of structurally similar compounds (bifonazole, clotrimazole, imazalil sulfate, and cyproconazole) that inhibit ergosterol (and therefore membrane) biosynthesis. Thus, we tested 11 Pdr5 transport substrates.

These varied considerably in hydrophobicity and size. Cycloheximide was the most polar (logP =

0.56); tamoxifen was the most hydrophobic (log p = 7.88) and largest. Cerulenin was the smallest of the Pdr5 substrates tested.

The A666G mutation created robust resistance to 10 of the 11 tested compounds.

Depending on the compound, the resistance to these substrates in the A666G mutant strain was roughly 2-4x higher than in the wild-type. The mutant did not exhibit enhanced clotrimazole resistance. In two-way ANOVA tests only the plots for clotrimazole showed no significant difference between the wild-type and A666G strains.

Identification of substrates whose transport is not enhanced by the A666G mutation may have bearing on the mechanism of resistance. The clotrimazole data are complicated by the fact that this transport substrate is a potent inhibitor of Pdr5 ATPase in vitro, which might limit any additional resistance above the wild-type level in vivo. We identified coumarin 6 as a strong Pdr5 substrate whose transport was not enhanced further by the A666G mutant and did not inhibit

Pdr5-specific ATPase activity.

Figure 13. The A666G mutant exhibits strong hyperresistance to multiple Pdr5 substrates. Cells were cultured in YPD broth at 30 °C for 48 h in the presence of drugs as described in the Materials and Methods. YPD cultures of each strain that contained no drug served as an untreated control for growth comparisons. Cell concentration was determined at 600 nm. In this figure: ■, green, WT; ▲, red, ΔPdr5; and □, blue, A666G (n = 3).

Figure 14. The A666G mutant increases resistance to transport substrates. Cells were cultured in YPD broth at 30 °C for 48 h in the presence of drugs as described in the Materials and Methods. YPD cultures of each strain that contained no drug served as an untreated control for growth comparisons. Cell concentration was determined at 600 nm. In this figure: ■, green WT; ▲, red, ΔPdr5; and □, blue, A666G (n = 3).

The A666G mutant had increased R6G transport in whole cells

We also compared the whole-cell R6G transport capability of the wild-type, A666G, and

G312A mutant strains with 5 µM (Figure 15A) and 10 µM R6G (Figure 15B). The G312A null mutant and the isogenic ∆Pdr5 strain served as negative controls.

Figure 15. The A666G mutant enhances R6G transport in whole cells. Transport of 5 µM (A) or 10 µM (B) R6G against a concentration gradient was performed as described in the Materials and Methods at 30 °C for 90 min. The median fluorescence (a.u.) obtained from sorting 10,000 cells/sample is shown (n = 3). We also compared transport at 30 °C and 35 °C in the same strains (B). Each independent culture (n = 3) was split into two portions and transport was monitored at the two temperatures with the whole-cell transport protocol described in the Materials and Methods.

In an assay with 5µM R6G, the wild-type strain accumulated a median value of 94.4 arbitrary fluorescence units (a.u.); the ∆Pdr5 strain retained 1920 a.u. Cells treated with 50 mM

2-deoxyglucose to deplete ATP levels retained levels of fluorescence that were comparable to the negative control. Fluorescence was 6x as great in the wild-type as in the A666G strain (15.7 a.u.). A similar result appeared in a transport assay conducted with 10 µM R6G. In this case, the differential between the A666G and wild-type strains was about 5x (Figure 15B). Whole-cell transport experiments and measurements of drug resistance were carried out at 30 ˚C, and various in vitro assays were performed at 35 ˚C. We therefore performed one set of whole-cell transport experiments at the higher temperature to determine whether the relative difference between the strains was maintained (Figure 15B). Although transport was reduced overall, the wild-type still accumulated about 5x as much R6G as the A666G mutant strain. We concluded that the A666G mutant was not temperature sensitive and that the results obtained at the two temperatures were comparable.

Imazalil sulfate inhibits R6G transport in whole cells

In a previous study, imazalil sulfate (IMZ) exhibited concentration-dependent inhibition of R6G whole-cell transport (Mehla et al., 2014). IMZ was also useful because it did not inhibit

Pdr5 ATPase activity at the concentrations used in our transport assays (Downes et al., 2013).

When 5 µM R6G was the substrate, IMZ caused concentration-dependent inhibition with levels of retained fluorescence reaching that of the ∆Pdr5 control (Figure 16A). The IC50 of IMZ in the wild-type strain was ~60-70 µM. The A666G mutant, however, behaved differently. It took a larger concentration of IMZ to begin to see inhibition.

Figure 16. Imazalil sulfate inhibition of R6G transport is concentration dependent. (A) Inhibition of 5 μM R6G transport was performed in the presence of IMZ. The assay was the same as the one used throughout the whole-cell transport studies except that during inhibition assays, IMZ was added at the same time as R6G and remained throughout the entire incubation period. The red line indicates the level of fluorescence in the ΔPdr5 control strain. In these experiments, n = 6 for WT (green line) and n = 4 for the A666G mutant (blue line). (B) A logarithmic plot constructed from the same data.

At higher concentrations, however, the mutant curve rose sharply so that the IC50 (about

100 µM) was similar to that of the wild-type. When we used a nonlinear transformation to make plots of the log of the IMZ concentration versus the fluorescence (Figure 16B), the wild-type

73 curve had a Hill (h) coefficient of 0.9, but the mutant value was 2.4, suggesting cooperativity between transport sites.

The concentrations of IMZ used to inhibit R6G transport also inhibited wild-type growth in culture. Although the gating profiles of IMZ-treated cells obtained during fluorescence cell sorting were similar to the untreated controls and gave no evidence of increased cellular damage or death during the relatively short incubation period, caution is required in interpreting these results. Studies carried out by Siegel and Ragsdale (1978) demonstrated effects of imazalil on ergosterol precursor pools in as early as 30 minutes. Therefore, it was important to perform the in vitro studies described below.

The A666G mutant enhanced R6G fluorescence quenching in purified PM vesicles

We developed a fluorescence-quenching assay suitable for measuring R6G transport for

Pdr5 (Kolaczkowski et al., 1996) and previously used it to analyze the transport capability of a

D1042N mutant in the Pdr5 D-loop (Furman et al., 2013). R6G is a known inhibitor of Pdr5

ATPase, however. It was therefore important to establish whether the wild-type and A666G mutant enzyme activities were inhibited to the same degree. If, for example, the A666G mutant enzyme was more sensitive to inhibition than the wild-type, the initial rate (IR) of fluorescence quenching would be underestimated in the mutant. In Figure 8A, Arya et al., (2019) demonstrated that at the concentration of R6G used in the quenching experiments, the inhibition is negligible.

We initially assayed a single preparation of PM vesicles from the wild-type and A666G strains (Figure 17A). We performed reactions with 3 mM ATP, a concentration thought to be physiological for Saccharomyces (Ozalp et al., 2010). As expected, the kinetics of R6G quenching in the wild-type and A666G strains were first order (R-squared values = 0.9926 and

0.9965, respectively) when we performed a linear regression on a plot of the natural logarithm

(ln) of the fluorescence a.u. versus time (Figure 17B). PM vesicles from the isogenic ΔPdr5 served as a negative control and showed no quenching. The plot from this preparation yielded a slope that was not significantly different from zero. We obtained a similar result when we omitted ATP from a reaction containing either wild-type or A666G (data not shown). Because the ATPase activities of the wild-type and A666G PM vesicles used in this experiment were similar, we concluded that the IR of fluorescence quenching was about twice as fast in the mutant. ATPase activities varied somewhat from one PM vesicle preparation to another; furthermore, we switched to a much-improved method of PM vesicle preparation that yielded significantly higher ATPase activities. We used this method to prepare the PM vesicles used exclusively in the experiments described in Figures 21, 22 and 23.

Figure 17. Pdr5-mediated R6G transport is enhanced in PM vesicles prepared from A666G cells. In these experiments: ■, green, WT; ▲, red, ΔPdr5; and □, blue, A666G. Fluorescence quenching was carried out as described in the Materials and Methods at 35 °C with 30 µg PM vesicle protein suspended in transport buffer containing 100 nM R6G and 3 mM ATP in a final volume of 2 ml, as described by Furman et al. (2013). (A) Quenching was carried out with PM vesicles prepared from WT, ΔPdr5, and the A666G mutant strains as described in the Materials and Methods. The plot shows the fluorescence in a.u. at 1-min intervals. (B) A linear regression was performed on the same data shown in Fig. 19 by plotting the ln fluorescence (a.u.) against the time with GraphPad software. Fluorescence-quenching experiments used independent PM preparations from the WT and A666G mutant strains. Initial quenching rates were determined by linear regression performed on each plot shown in panel B. We plotted the rates were plotted against ATPase activity determined in Hepes transport buffer with 3mM ATP, with the assay described in the Materials and Methods.

Fluorescence quenching was directionally proportional to ATPase activity We evaluated the relationship between fluorescence quenching and ATPase activities, which we measured in the Hepes buffer (pH 7.0) using the same (3 mM) concentration of ATP (Figure 18). We did this for six preparations of A666G mutant and eight preparations of WT vesicles. The data demonstrate that although ATPase activities can vary considerably, the IRs of

Figure 18. The initial rates of fluorescence quenching are directly proportional to ATPase activity. Fluorescence quenching experiments were performed with independent PM preparations of the WT and A666G mutant. Initial quenching rates were determined by linear regression performed on each plot shown in panel B of Figure 17. The rates were plotted versus ATPase activity determined in Hepes transport buffer with 3mM ATP using the assay described in the Materials and Methods.

fluorescence quenching in the A666G mutant (blue line) were roughly 2x faster than the WT

(green line) regardless of enzyme activity. Thus, a linear regression was performed and yielded

77 slopes of -0.0092 and -0.0209 for the WT and A666G mutant respectively. This indicated that the IRs of fluorescence quenching of the A666G mutant are about 2.3x faster than the WT. The

R-squared values were 0.7888 and 0.8558 respectively indicating acceptable fits. To determine whether the slopes of the lines were different, a t-test was performed. We obtained a t-value of

0.0106 indicating a very high probability that the lines were different (0.9918).

No large changes occurred in the IRs of fluorescence quenching over an 8x range of R6G concentrations

We looked at the IRs of fluorescence quenching over an 8x range in R6G concentration in preparations of wild-type and the A666G mutant (12.5 nM–100 nM). We observed no consistent change in the quenching rates over this range of concentrations (Figure 19). If anything, the mutant rates decreased very slightly with increased concentrations of R6G.

Figure 19. IRs of fluorescence quenching measured over a range of R6G concentrations. Fluorescence quenching experiments with WT (green bar) and A666G (blue bar) PM vesicles were performed,as described in the Materials and Methods, with different concentrations of R6G (12.5nM – 100nM).

The enhanced fluorescence quenching observed in the A666G PM vesicles is attributable to increased cooperativity between transport sites

To evaluate whether the A666G phenotype is attributable to altered kinetics of the drug transport cycle, we initially compared the IRs of fluorescence quenching of the wild-type and mutant PM vesicles with similar preparation dates and ATPase activities over a range of ATP concentrations (1.5-10 mM).

With the improved method of purifying PM vesicles, we obtained preparations of much higher ATPase activity and fluorescence-quenching capability. We observed that at higher IRs of

R6G fluorescence quenching, the curves became nonlinear with increasing time in both wild- type and A666G vesicles. Simply doing a linear regression on the entire data set (20 minutes) would have resulted in an underestimation of the IRs of fluorescence quenching once we performed a linear transformation (and the R-squared values in some cases would have been poor). Therefore, we determined the IRs of R6G fluorescence quenching for these from the linear portion of the curves, with linear regression. Representative plots of the ln fluorescence versus time are shown in here (Figure 20) from assays with 1.5 mM, 3.0 mM, and 5.0 mM ATP. The linear portions used to determine rates are indicated. The R-squared values for the linear portions were all >0.99.

Figure 20. Representative plots of fluorescence quenching performed with 1.5, 3.0, and 5.0 mM ATP. Fluorescence quenching experiments were performed as described in the Materials and Methods. We plotted the ln of the fluorescence values taken at one-minute intervals. WT: green; A666G: blue. The linear portions of the curves used to determine the IRs have solid symbols.

When we plotted the rates versus ATP concentration (Figure 21A), several features were readily apparent. The curves for both the mutant and the wild-type fit a nonlinear transformation that used an allosteric sigmoid equation. The R-squared values for the wild-type and A666G plots were 0.9776 and 0.9921, respectively, indicating a reasonably strong fit.

Figure 21. The A666G mutant enhances cooperativity between transport sites. Fluorescence quenching was performed with 100 mM R6G at 35 °C for 20 min as described in the Materials and Methods except that 60 μg of purified PM vesicles were used in each reaction with the first set of PM vesicles tested (A) and 30 μg were used with a second set of WT and mutant PM vesicles (B). In each panel: ■, green line = WT; □, blue line = the A666G mutant. The ratio of the A666G mutant and WT IRs are compared above and below the Km (C).

The Km (ATP)s necessary to reach half the Vmax of fluorescence quenching were similar: 3.5 mM for the wild-type and 3.4 mM for the A666G mutant (Table 14). There was a remarkable increase in the IRs of fluorescence quenching between 3- and 4-mM ATP in both cases. This largely accounts for the high h coefficients of 5.9 for the wild-type and 8.6 for the mutant. This cooperativity represents interactions between the drug transport sites rather than the ATP hydrolysis sites. Pdr5-mediated ATPase activity is well studied, follows strict Michaelis-Menten kinetics, and therefore exhibits no cooperativity (Golin et al., 2007; see also Figure 2C).

The higher h coefficient observed in the plot from the A666G mutant was potentially of importance because it suggested a mechanism by which resistance could be enhanced. For this reason, we prepared a new set of wild-type and A666G PM vesicles and again tested the effect of varying ATP concentrations on the IRs of fluorescence quenching. In this experiment, however, we also measured the IRs of R6G fluorescence quenching at the lower ATP concentrations of 0.5 and 1.0 mM. The results in the two experiments were qualitatively similar (Figure 21B).

Pertinent kinetic data for the experiments shown in Figure 21 are averaged and presented in

Table 14.

We compared the enhancement of quenching in the A666G PM vesicles at each of the

ATP concentrations that we tested (Figure 21C). Overall, we saw no significant difference in enhancement at concentrations that were above or below the Km values. However, in both experiments, the ATP concentration directly above the Km resulted in the largest enhancement.

Table 14: IR of fluorescence quenching versus ATP concentration: kinetic parameters†

Kinetic parameters WT A666G

-1 -4 Vmax Quenching (S x 10 ) -19 (3.5) -33 (1.4)

Km Quenching (mM) 2.9 (0.9) 2.2 (0.4)

Hill values (h) 4.7 (1.8) 6.8 (2.6)

‡ Vmax ATPase (μmol / min / mg) 1.4 (0.2) 1.4 (0.7)

Km ATPase (mM) 2.1 (1.3) 2.1 (0.9)

† The kinetic parameters are the average values for the experiments shown in Fig. 8A and 8B. The kinetic parameters were determined with GraphPad Prism 8.0 software. The standard deviations are included in parentheses. Each point on a curve is the average IR of R6G fluorescence quenching for two quenching reactions. One of these (2 mM ATP, A666G) was recognized as an outlier by Graphpad software and removed.

‡The ATPase activity was measured in Tris-glycine buffer (pH 9.5).

Figure 22A-B presents results from a set of wild-type and A666G mutant preparations that were made and tested independently of each other. In addition, we performed an earlier experiment that used a pair of wild-type and A666G PM vesicles prepared with the original purification method and therefore of lower ATPase and quenching activity (Figure 22C). In all of these, the h coefficient was noticeably higher for the mutant than the wild-type.

Figure 22. The A666G mutant exhibits increased cooperativity between transport sites. The experiments shown in this figure are analogous to those illustrated in Figure 20. The quenching reaction was performed as described in the Materials and Methods. In addition to the data found in Figure 20, we did the analogous experiment with (A) an additional WT and (B) A666G mutant preparations that were made and assayed independently of each other. (C) We also performed the same assay with PM vesicles that were made using the older protocol and were therefore were not as active. The number at the bottom of each curve is the Hill coefficient. In this figure: ■, green WT; and □, blue, A666G.

The inhibition of R6G fluorescence quenching by IMZ in A666G PM vesicles also exhibited enhanced cooperativity

We also observed enhanced cooperativity in the A666G mutant PM vesicles when we used IMZ to inhibit R6G fluorescence quenching. We compared the inhibition of R6G fluorescence quenching with 50, 75, and 100 nM R6G in new pairs of PM vesicles prepared from the wild-type and A666G strains (Figure 23A-C). The lowest concentration of IMZ used (0.5

µM) resulted in 0%-15% inhibition relative to the untreated control depending on the strain and

R6G concentration. The mutant curves exhibited greater cooperativity than the wild-type ones at all three concentrations. The data are summarized in Table 15, where they are compared to the values obtained from experiments in which we monitored the IR of R6G fluorescence quenching as a function of ATP concentration. The h coefficients for both wild-type and mutant were higher in the experiments where ATP was varied than in those in which we used imazalil sulfate as an inhibitor.

Figure 23. Inhibition of R6G fluorescence quenching by IMZ mirrors the results with whole-cell transport studies. Fluorescence quenching experiments were performed with 25 µg of purified PM vesicle protein for 20 min at 35 °C as described in the Materials and Methods. A different pair of PM vesicles was used to monitor fluorescence quenching at 75 nM R6G. Inhibition studies were performed with IMZ with (A) 50 nM, (B) 75 nM, and (C) 100 nM R6G with various concentrations of IMZ (0.5-30 µM). The IRs were determined with Graph Pad software through the same linear transformation that was applied to the tamoxifen data. The resulting curves were fitted by a nonlinear transformation (IR versus log [inhibitor], variable slope with four parameters). In all panels: WT = ■, green line; A666G = □, blue line.

Table 15: Summary of h values from various experiments.

The h values were determined for two different types of experiments. One set of experiments determined the IRs as a function of ATP concentration. A second set of experiments compared the inhibition of R6G fluorescence quenching in WT and A666G PM vesicles by imazalil sulfate.

Strain Conditions Mean h n SD

value

WT (JG2015) Varying the ATP concentration 4.1 3† 1.6

WT (JG2015) Imazalil sulfate inhibition 2.3 3 0.5

A666G Varying the ATP concentration 6.6 3† 1.8

A666G Imazalil sulfate inhibition 3.7 3 1.1

†These calculations include an experiment performed with an additional PM preparation of WT and A666G PM vesicles made and tested separately.

In summary, analysis of the A666G mutant established a new and novel mechanism of drug resistance. Instead of increasing expression, the A666G mutation improves the efficacy with which ATP is used by increasing the cooperativity between drug transport sites.

Part II

Alanine substitutions in phosphoserine residues lead to increased PDR5 expression

After we determine the novel mechanism of cooperativity for enhanced resistance in A666G, we shifted our focus to the serine residues of linker-2 of Pdr5. Our initial hypothesis was that phosphorylation of these serine residues could modulate the activity of Pdr5. Studies conducted in the linker region of the P-gp, human homolog of Pdr5, however, revealed that phosphorylation does not play any role in P-gp mediated drug efflux. However, in the regulatory domain of

CFTR, an asymmetric ABC protein like Pdr5, phosphorylation plays important role in channel opening. In order to examine if the phosphorylation of serine residues modulate Pdr5 drug resistance, we made alanine substitutions for all serine residues in linker-2. Our data demonstrate that dephosphorylation of these serine residues enhanced resistance relative to the WT. The behavior of an S837D phosphomimic mutant and mass spectrometry data indicated phosphorylation of these serine residues was unlikely to have any role in Pdr5 mediated drug resistance. In order to investigate the mechanism of drug resistance due to alanine substitutions, we studied protein expression in whole cell and in the plasma membrane level. Our data indicate increased expression of Pdr5 in the plasma membrane. We quantified the whole cell-level of

Pdr5 transcripts and determined the PDR5 half-life in mutant and WT cells to investigate the mechanism of increased expression and as a result enhanced resistance.

Alanine substitution mutants in six phosphoserine residues were hyper-resistant to Pdr5 substrates

We constructed six single-alanine substitutions in the linker serines: S837A, S840A,

S841A, S849A, S850A, and S854A. A schematic representation is found in Figure 24. The nucleotide changes that were made are found in Table 16. In this set of residues, four of six possible triplet codons for serine were used. At three of these positions (Ser-840, -850, and -

854), a single nucleotide change resulted in the alanine substitution. It is therefore highly likely that such mutants arise in vivo.

Figure 24. Relative location of the linker-2 residues. The schematic illustrates the location of the six serine residues analyzed in this study. Linker (L)- 2 connects TMH6 to the canonical portion of NBD2.

Table 16: Nucleotide changes resulting in alanine and aspartate substitutions in linker-2

Amino acid Serine DNA codon Substitution New DNA codon

Ser-837 AGT alanine GCT

Ser-837 AGT aspartate GAT

Ser-840 TCC alanine GCC

Ser-841 AGC alanine GCC

Ser-849 AGC alanine GCC Ser-850 TCT alanine GCT

Ser-854 TCC alanine GCC

Our initial evaluation of drug resistance used climbazole, cycloheximide, cyproconazole, and cerulenin is shown (Figure 25). Remarkably, all of the mutants were significantly more resistant than the WT control strain to all four drugs. We selected S837A and S854A for more in- depth study. We compared their relative resistance to eight Pdr5 transport substrates over a range of concentrations, and we used the isogenic WT strain as a control. The data for eight substrates are found in Figure 26. The S837A mutant exhibited modestly greater resistance to clotrimazole, cyproconazole and imazalil sulfate great as in the WT than did S854A. In general, the S837A mutant had estimated IC50 values that were about 2-4x greater than WT

Figure 25. Preliminary testing of phosphoserine mutants indicated that they all increase drug resistance.Cultures of six alanine substitution mutants were tested for resistance to clotrimazole, cycloheximide, cyproconazole, and cerulenin in liquid YPD cultures (48 h at 30 °C) at several concentrations as described in the Materials and Methods. In these experiments, n > 4.

Figure 26. The S837A and S854A mutants exhibits strong hyper-resistance to multiple Pdr5 substrates. Note. ■ = WT (green line); ● = S837A (orange line); ▼ = S854A (blue line). Cells were cultured in YPD broth at 30 °C for 48 h in the presence of drugs, as described in the Materials and Methods. YPD cultures of each strain that contained no drug served as an untreated control for growth comparisons. Cell concentration was determined at 600 nm. (n > 3).

We noticed toward the end of this analysis that the WT strain (JG2001) that we used in this study was more sensitive to some of the transport substrates than the other WT strain that was also employed in our laboratory. When we sequenced the Pdr5 gene in the former, we observed no changes in the Pdr5 sequence. Nevertheless, we also compared the JG2015 and

S837A strains for their relative resistance to six of the transport substrates (Figure 27). In each case, The S837A strain was significantly more resistant than the WT control. The JG2015 strain and its Ura-derivative were used exclusively in the whole-cell transport and biochemical experiments described below.

An S837A, S854A double mutant does not exhibit greater resistance than the S837A and

S854A single mutants

We initially constructed an S837A, S854A double mutant. We compared the resistance of single and double mutants to six of the substrates (Figure 28). These plots clearly demonstrate that in every case, the double-mutant strains had relative resistance that was no greater than either single-mutant counterpart. An analogous series of experiments with an S837A, S850A double mutant gave similar results for the four substrates that were tested (Figure 29).

Figure 27. The S837A mutation results in enhanced resistance to transport substrates. The experiments are analogous to those described in Figure 25 except that the JG2015 WT strain (instead of the JG2001 strain) was used. The WT and S837A curves have green and orange lines respectively. In these experiments, n = 4.

Figure 28. An S837A, S854A double mutant is no more hyper-resistant than single mutants. Resistance to xenobiotic agents was also monitored in single and double mutants as described in the Materials and Methods. Cells were cultured in YPD broth for 48 h at 30 °C. In these experiments, n > 3. In all panels, the S837A (orange line), S854A (blue line), and double mutant (brick red) were compared to each other and to the WT control (green line).

Figure 29. An S837A, S850A double mutant has a resistance profile similar to single mutants. Here we used S837A+S850A double mutant to test for resistance to xenobiotic agents with the IC50 protocol as described in the Materials and Methods. Cells were cultured in YPD broth for 48 h at 30 °C. In these experiments, n > 3. In all panels, the S837A (orange line), S850A (light brown), and double mutant (pink line) were compared to each other and to the WT control (green line).

One possibility we had to consider was that the single mutants reached the maximum resistance that Pdr5 could mediate for a substrate. If that were true, even mutations in two distinct biochemical steps would yield a non-additive phenotype. This is clearly not the case.

We tested the cycloheximide resistance of an A666G gain-of-function mutation that was previously characterized (Arya et al., 2019) along with the S837A mutant. The resistance of the former was roughly twice that of the latter (Figure 30). This demonstrated that Pdr5 is clearly capable of mediating a level of resistance that is greater than is exhibited by the S837A mutant.

Figure 30. The A666G mutant exhibits stronger hyperresistance to cycloheximide than the S837A mutant. Growth of each strain was measured in liquid culture as described in Figure 25 (n=3).

The S837A and S854A mutants exhibited enhanced R6G transport

We evaluated the R6G transport capability in whole cells of the S837A and S854A mutants relative to a WT and a G312A null mutant strain (Figure 31). The WT (JG2015) had

a median fluorescence of 336.8 a.u., which was about 13.5x less than the catalytically dead

G312A mutant (median value: 4539 a.u.). The median retained fluorescence values observed

98 with the S837A and S854A mutants were 65.38 and 108.1, respectively. Thus, relative to the

WT, these two mutants increased transport capability 5.2x and 3.1x, respectively. This differential was greater than most of those observed with other transport substrates when drug resistance was evaluated. An S837A, S854A double mutant showed the same level of retained fluorescence as the S854A mutant. A t test found no significant difference in the fluorescence levels in the mutants.

Figure 31. Rhodamine 6G transport is enhanced in the S837A and S854A mutant strains. R6G transport in whole cells was measured as described in the Materials and Methods. Incubations were performed for 90 min at 30 °C with 3 x 106 cells in Hepes (0.0M) buffer containing 1 mM glucose and 10 μM R6G.The retained fluorescence in 10,000 cells was determined with a fluorescence cell sorter. In these experiments, n = 3.

An S837D mutant strain was also hyper-resistant to Pdr5 substrates

The use of serine-to-aspartate substitutions to create phosphomimics is a standard strategy. We made and tested an S837D substitution and predicted it would probably be phenotypically WT because Ser-837 is known to be phosphorylated from several mass spectrometry studies and because the S837A mutant, which can’t be phosphorylated, was hyper- resistant. However, our evaluation of the S837D mutant strain for resistance to clotrimazole, cycloheximide, and cerulenin revealed a phenotype similar to that of the S837A mutant (Figure

32).

100

Figure 32. An S837D substitution mutant is also hyper-resistant to Pdr5 transport substrates. The S837D (□, blue line) mutant strain was compared to the S837A mutant (●, orange line) and an isogenic WT control strain (■, green line) for relative resistance to (A) cycloheximide and (B) cerulenin.

101

We also compared the R6G transport capability of the S837D mutant to the S837A and

WT strains (Figure 33). The S837D mutant retained significantly less fluorescence (about half the amount) than the WT strain (p = 0.027).

These results were unexpected. We considered the possibility that the pSS607 plasmid where all of the mutants were constructed had acquired a mutation that increased resistance and was shared by all of the mutants. We therefore recovered the integrated S837D mutant sequence through PCR and sequenced the entire PDR5 gene. We found no additional mutations.

Figure 33. The S837D mutant strain exhibits enhanced R6G transport. R6G transport was performed in whole cells as described in the Materials and Methods. In these experiments, n =3.

102

Mass spectrometry of Pdr5 phosphoserine residues suggested that Ser-837 is infrequently phosphorylated

We carried out mass spectrometry on Pdr5 peptide fragments to better evaluate the role of linker-2 phosphorylation on Pdr5 activity. Our observation that an alanine mutation at any of the six serines increased resistance indicated that the putative modulation by phosphorylation would require that most Pdr5 molecules be completely modified at these residues. We analyzed Pdr5 protein from purified PM vesicles prepared in the presence of phosphatase inhibitors.

First, we determined which serine residues were phosphorylated. The mass spectrometry covered nearly 70% of the amino acids in Pdr5 (1040/1511). The majority of the phosphoserines were in the large amino-terminal tail and linker-2 (Table 17). The NetPhosYeast program

(Technical University of Denmark) predicted that 14/20 had a greater than 50% probability of being phosphorylated. Four global phosphorylation studies of yeast phosphopeptides identified only 10/20.

Table 17: Twenty serine residues are phosphorylated in Pdr5

Number of Location Predicted Present in global phosphorylated serine probability analyses (√)

17 Amino terminus 0.638 18 Amino terminus 0.600 20 Amino terminus 0.519 21 Amino terminus 0.674 22 Amino terminus 0.595 √ 54 Amino terminus 0.791 √ 58 Amino terminus 0.639 √

103

61 Amino terminus 0.673 √ 70 Amino terminus 0.501 74 Amino terminus 0.290 104 Amino terminus 0.203 126 Amino terminus 0.073 318 NBD-1 0.573 837 Linker 2 0.542 √ 840 Linker 2 0.514 √ 841 Linker 2 0.444 √ 849 Linker 2 0.667 √ 850 Linker 2 0.755 √ 854 Linker 2 0.427 √ 945 NBD2 0.224

Further analysis of the results from mass spectrometry revealed 10 distinct phosphopeptides that included Ser-849, 850, and 854 and had a high probability (> 99%) of matching Pdr5 in the data base (there is a trypsin site between Ser-841 and Ser-849). The relative abundance of a particular pattern was estimated from the area of each peptide’s chromatographic peak. Ser-850 was phosphorylated in all cases; Ser-854 in a majority of the cases. For the sequence covering Ser-837, Ser-840, and Ser-841, three distinct cases were recovered. None of the three phosphopeptides was phosphorylated at all three sites. The two most abundant phosphopeptides lacked a phosphoserine at residue 837. Although other explanations are plausible, these data suggested that a significant portion of the Pdr5 molecules present in vivo contain at least one linker-2 serine that remains unmodified.

104

Increased levels of Pdr5 in purified PM vesicles in the S837A mutant strain

Over the course of these studies, we made two independent PM vesicle preparations from the S837A mutant and four from the WT strain. Gel electrophoresis was performed, and the separated proteins were stained with Coommassie blue. The uniform density of the Pma1 band

(except in the ΔPdr5 lane) across the gel indicates equivalent loading of solubilized protein from each sample. The S837A PM vesicles clearly contained more Pdr5 than the controls (Figure

34A). This result was also confirmed by Western blotting of PM vesicles (Figure 34B). The

S837A mutant membranes had about twice as much Pdr5 as the WT control.

Figure 34. Plasma membrane vesicles made from the S837A strain have more Pdr5 in the membrane. Purified PM vesicles were prepared as described by Kolaczkowski et al. (1996) and modified by Ernst et al. (2008). (A) PM (6 μg) protein samples from three preparations of S837A and three preparations of WT PM vesicles were solubilized in SDS PAGE and subjected to electrophoresis in 7% tris-acetate gels for 80 min at 150 V before staining with Coommassie

105 blue.(B) Samples (3 μg) from the same PM vesicle preparations were also immunoblotted as described in the Materials and Methods.

These results also suggested that the PM vesicles prepared from the S837A mutant strain would have about twice the ATPase activity of the WT control. This was the case when we assayed independent preparations for ATPase activity with 3mM ATP (Figure 35A). The average activity of mutant PM vesicles was ~1.8x as high as the WT (1.95 and 3.47

μmol/min/mg, respectively). We also measured ATPase activity with varying ATP concentrations with a WT and S837A mutant preparation. As expected, the kinetics fit the

Michaelis-Menten equation. The Vmax for the WT and S837A mutant preparations were 1.9 and

4.8 μmol/min/mg. respectively (Figure 35B). The increase in ATPase activity was therefore proportional to the increased level of Pdr5.

106

Figure 35. The ATPase activity is doubled in PM vesicles prepared from the S837A mutant. PM vesicles were prepared as described by Kolaczkowski et al. (1996). (A) The assay of ATPase activity was performed as previously described (Arya et al., 2019) with 3 mM ATP. Activity was measured in Tris-glycine buffer (pH 9.5). Activity was assayed in two different PM vesicle preparations prepared from the S837A mutant and three from the WT strain JG2015. (B) ATPase activity was measured as a function of ATP concentration in PM vesicles prepared from WT (■, green line) and S837A (●, orange line) strains in the conditions described in the Materials and Methods.

We also monitored the level of Pdr5 in whole-cell extracts prepared from the WT and all of the alanine substitution mutants (Figure 35A). If the increased resistance was caused by improved trafficking to the PM, the enhancement in Pdr5 level would not have been the same in a whole-cell lysate, where PM vesicles are not the only source of Pdr5 protein. In all cases, the enhancement for all of the mutants was about twice the WT level. In a separate experiment

(Figure 35B), we also blotted samples from two independent preparations of the S837D mutant, along with a lysate from S854A and several from S837A and the WT.

107

In this set of experiments, the S837D and S837A mutants showed a comparable level of enhancement. In contrast, the level of Pdr5 in the S854A extract was lower than those observed with the other mutants, but clearly higher than the WT.

Figure 36. The alanine mutants exhibited elevated levels of Pdr5 in whole-cell extracts. (A) Whole-cell extracts were prepared, proteins (10 μg/sample) were solubilized in SDS PAGE, and gel electrophoresis and Western blotting were performed as described by Rahman et al. (2018). The relative amounts of Pdr5 were determined and normalized with GAPDH as the loading standard. The bar graph at the bottom of each lane represent the average of the ratio of GAPDH to Pdr5 protein. (B) Extracts were also prepared from the S837D mutant and, for comparison, from the WT, S837A, and S854A strains. They were blotted as described in the Materials and Methods and in panel A. The bar graph contains the average enhancement found in multiple WT, S837A, and S837D blots.

The S837A mutant does not increase the stability of Pdr5

An obvious potential explanation for the increased expression of alanine and aspartate substitution mutants was that they had increased the stability of Pdr5. Two relatively early

108 studies of Pdr5 turnover that used pulse-chase experiments yielded different estimates of its half- life. It is possible that the different genetic backgrounds used in the two studies were responsible.

The first (Egner and Kuchler, 1995) indicated the Pdr5 half-life was 60-90 min; the second

(Decottignies, Owsianik, and Ghislain, 1999) showed that it was in excess of 2 h. We performed a cycloheximide-chase experiment to determine whether increased stability of the S837A mutant was a reasonable explanation for the increased level of mutant Pdr5 in the PM (Figure 36A, B).

Figure 37. Pdr5 is a relatively stable protein. A cycloheximide chase experiment was performed with the WT (A) and S837A (B) mutant cells as described in the Materials and Methods. Following the addition of 125 μM cycloheximide to the cultures, aliquots of cells were lysed at 1-h intervals and subjected to electrophoresis and Western blotting. (C) The amount of Pdr5 protein was normalized with respect to GAPDH protein that was run on the same gel. The filter was cut and the Pdr5 and GAPDH signals developed separately. A second time course experiment reproduced the results in this figure.

109

The cell-division time for the WT strain was ~90 min. There was no reduction in the level of Pdr5 for at least 2 h. A significant decrease in the amount of Pdr5 was not observed until the 4-h time point. At 5 h, roughly 63% and 72% of the WT and S837A protein respectively, remained. A repeat of the experiment gave similar results, although there was no difference between the WT and S837A mutant preparations. It didn’t seem likely that increased stability of mutant proteins was responsible for the hyper-resistant phenotype of the alanine substitutions.

The S837A mutant has a higher level of Pdr5 mRNA

The cycloheximide chase experiment strongly indicated that the increased levels of Pdr5 could not be attributed to enhanced stability of the transporter. We therefore considered the possibility that these mutants had enhanced levels of Pdr5 transcript. We used q-RT PCR to compare the levels in the WT and mutant with three reference genes: ALG9, TAF10, and TDH3 (Teste et al.,

2009). Following this, we calculated the difference between the strains from the ΔΔCT for each

-ΔΔc reference gene and averaged the three values before calculating the enhancement with the 2 T method (Livak and Schmittgen, 2001). The high CT for the negative controls, which lacked reverse transcriptase, indicated that we were amplifying mRNA rather than contaminating DNA

(which was eliminated by DNAse treatment). The results from four independent experiments

(Figure 37) indicated that the S837A mutant showed a roughly 2.5-3.0x enhancement in PDR5 transcript level regardless of which reference gene served as a comparison. This increase was similar in magnitude to the enhancements in Pdr5 level, drug resistance, and R6G transport that we observed in the alanine substitution mutants. Statistical analysis indicated that the

110 enhancements observed with the ALG9 and TAF10 reference genes were greater than 1.0x the

WT value with greater than 95% confidence

Figure 38. The S837A mutant has enhanced levels of PDR5 transcript. Q-RT-PCR was performed as described in the Materials and Methods. The ΔCT values for the three references genes were determined and these values were used to obtain the ΔΔCT between the WT and mutant strains for each reference gene. The times difference was calculated from the formula: ΔΔC fold change = 2 T. For the ALG9 and TAF10 experiments, n =4. For the TDH3 determination (used in experiment 3# and #4), n =2. The ΔCT values the first two experiments were the average of two determinations; the values in experiments 3 and 4 are calculated from the median (n=3).

The S837A mutation increases the half-life of the PDR5 transcript

A possible explanation for the increased PDR5 transcript levels was that their half-lives were extended by these mutations. To evaluate whether this was the case with the S837A mutant strain, we labeled RNA continuously with 4-TU for 3 hours and chased with cold uracil for up to

111

45 min with samples taken at 15-min intervals. RNA molecules containing 4-TU were recovered by crosslinking and column purification as described in the Materials and Methods. We then performed q-RT PCR and determined the CT for PDR5 and the three reference genes.

Figure 39. The S837A mutation increases the PDR5 transcript half-life. RNA was labeled continuously with 4-TU for three hrs. Following this, cells were chased with 5 mM uracil for various times before extracting the RNA and recovering the 4-TU-containing transcripts as described in the Materials and Methods. Following q-RT PCR, the CT values were used to calculate the % 4-TU remaining at 15, 30, and 45 min with the value at 0 min as 100%. Data were plotted using the one-phase exponential decay equation: y= span. e-k.x + plateau where k is the rate constant, x is the time, y is the log% 4-tu remaining after the uracil chase. In all of these plots, ■ = WT (green line); ● = S837A (orange line). In these experiments, n = 3.

112

The amount of transcript remaining relative to the time zero point (100%) was determined from three independent experiments and plotted with a one-phase exponential decay equation (Figure 38). The R2 values exceeded 90% in all cases (and 99% in a majority).

Comparison of the plots of the mutant and WT versions of the three reference genes revealed that the small differences in their half-lives were not significant. In contrast, however, the half-life of the PDR5 transcript was 3.4 times as long as in the mutant (12.7 min) as in the WT (3.70 min).

The 95% confidence intervals did not overlap. A two-way ANOVA test indicated that the row factor (comparison between strains of the remaining 4-tu at each time point) was a source of significant variation (p = 0.0045). This suggests that the increase in steady-state level can be completely attributed to the increased half-life of the S837A transcript.

The hyper-resistance exhibited by the mutants is not attributable to a non-Pdr5 mutation in the R-1 strain or pSS607 plasmid

The JG2015 WT strain was constructed from R-1 (see materials and methods) and has been used in our laboratory since 2007. The S837A strain was constructed in R-1 in 2017. One explanation for the similar phenotypes of the seven mutants analyzed in this study is that their hyper-resistance was caused by another alteration that took place during the 10-year interval separating these strains. This seemed unlikely because non-linker 2 mutants made in pSS607 and placed in R-1 after S837A was constructed had WT levels of Pdr5 in PM vesicles. Nevertheless, to be sure, we retransformed the strain with the pSS607 and pS837A plasmids. We compared the two original strains and two WT (JG2015*) and S837A mutant transformants for resistance to

113 cycloheximide and climbazole (Figure 39), these had phenotypes that were indistinguishable from their original counterparts.

Figure 40. Recreation of the WT and S837A mutant strains recreated their phenotypic differences. Resistance to cycloheximide and climbazole was measured in liquid culture at several concentrations of each drug (n=4) as described in the Materials and Methods. The original strains and two transformants of R-1 with the WT (JG2015*) and mutant allele were tested.

114

The expression level of the PDR5 transcript was evaluated in a WT and S837A transformants (Figure 40). As expected, S837A mutant strains exhibited higher levels than the

WT. The lower enhancement in expression observed with the S854A mutant is consistent with its less robust hyper-resistance and R6G transport relative to the S837A mutant.

Figure 41: mRNA expression level of Pdr5 from the WT and S837A transformants. Here using q-RT PCR technique we quantified the whole cells expression of Pdr5 transcripts from the transformants. The fold expression for both strains is normalized to two reference genes. S837A (●, orange bar graph) showed significantly higher expression level than WT (■, green bar). In this experiment, n=3.

In summary, the S837A mutation like the other phosphoserine alterations increased resistance relative to the WT by increasing the quantity of cellular Pdr5. Further analysis of the

S837A mutant strain demonstrated that this was due to an increase in the stability of the PDR5 mRNA.

DISCUSSION

My dissertation work uncovered two distinct strategies for enhancing drug resistance through gene mutation. The first mechanism exemplified by the A666G mutant is without precedent in the literature and was not even conceived as a possibility by us when we began our analyses. This mutant increased the cooperativity of Pdr5 transport sites. The second strategy was exhibited by the S837A mutant in linker 2. This mutation led to increased expression by increasing the half-life of Pdr5 mRNA.

Traditionally, cancers and pathogenic organisms are screened for the overexpression of

ABC efflux pumps. In this report, we describe the phenotypic features of a novel mutation:

A666G, which resulting a robust enhancement in drug resistance. The A666G mutant was acquired under selection and generally had IC50 values that were 2.5-4x as high as the WT depending on the substrate. When we measured R6G efflux in whole cells, the WT retained

~3.4-6x as much fluorescence as the mutant. Significantly, these enhancements occurred despite no further increase in the ATPase activity or the steady-state level of the transporter in the PM.

These observations have major implications for therapeutic treatment of cancers and fungal pathogens. They further suggest that robust resistance mediated by ABC transporters may be obtained without overexpression. Because Ala-666 is a highly conserved residue in the Pdr subfamily, the mechanism uncovered for the A666G mutant may operate broadly under selection. The transport of clotrimazole and coumarin 6, however, was not enhanced by this alteration. We observed no difference in clotrimazole resistance or coumarin 6-trans 115

116 capability between the WT and A666G mutant strains. Understanding the mechanism behind the increased resistance that defines the A666G mutant phenotype might offer improved treatment of fungal resistance and drug-resistant cancer. It may help identify therapeutic compounds that are more effective because mutants similar to A666G are unable to enhance their transport.

Both whole-cell transport and fluorescence-quenching assays in the presence and absence of competing transport substrates indicate that this mutant makes the efflux process more efficient by increasing cooperativity between transport sites. It is important to note that cooperativity between drug-binding sites in P-gp has been known for some time (Shapiro and

Ling, 1997). That a genetic modification can result in greater resistance because of increased cooperativity is a novel and important observation. Although there are numerous examples of multidrug transporter overexpression leading to hyperresistance, the mechanism behind mutants such as A666G remained unknown. When the IRs of R6G fluorescence quenching were plotted as a function of ATP concentration, both mutant and WT preparations showed cooperativity.

However, it was greater in the mutant. We observed a similar phenomenon with IMZ inhibition of R6G fluorescence quenching. It should be noted, however, that small changes in the IRs of

R6G fluorescence quenching can result in relatively large changes in h coefficients. Therefore, a single experiment is probably not definitive. For instance, in the plot shown in Fig. 20A, the 95% confidence interval for the WT gives a range for the h coefficient of 4.2-8.6. For the A666G mutant the range is 6.7-11. When we pooled all of the data including that found in the

Supporting Information, the difference in h coefficients between WT and mutant was significant according to a t-test (p = 0.036). Furthermore, in all of the experiments in which we determined

117 the h coefficient, the mutant value was always higher than the corresponding WT coefficient obtained at the same time. Thus, the evidence that the A666G mutant enhances resistance by increasing cooperativity between transport sites is striking.

Interestingly, when we compared the kinetics of IMZ inhibition of R6G transport to ATP- dependent R6G fluorescence quenching, the former exhibited less cooperativity in both the mutant and WT preparations. The WT differential was statistically significant (p = 0.022). This observation suggested that although Pdr5 ATPase is unstimulated by its substrates, they affect the interaction between transport sites. Consistent with this idea is the observation that IMZ is a weaker substrate than R6G. For instance, the ΔPdr5 strain was only 10 times as sensitive to IMZ as the WT (Downes et al., 2013). In contrast, the strain lacking Pdr5 retained 20 times as much

R6G fluorescence as the WT strain. Furthermore, it took micromolar amounts of IMZ to completely inhibit nanomolar amounts of R6G quenching.

We considered the possibility that the cooperativity we observed could be accounted for by the kinetic drug selection model first proposed by Ernst et al. (2008) to explain the behavior of a substrate-specific H-loop mutation of Pdr5. This model proposes that the time spent in each conformational state during the transport cycle is subject to genetic control. Therefore, a mutation could lead to a longer period spent in the inward-facing, drug-binding structure. This might also result in a longer period for a substrate to interact with Pdr5 and perhaps allow increased cooperativity. This model may very well explain some of the interesting FK506 hyperreistant mutants in Cdr1 and Pdr5 (Tanabe et al., 2018). Although we have no concrete

118 kinetic evidence that explains the behavior of the A666G mutant, some observations from this study lend support to this model. The kinetic drug selection model predicts that when the IRs of the quenching reaction are below the Km (which is ~2.0-3.5 mM ATP), the differential between the mutant and WT should be reduced or perhaps even eliminated because the ATP is rate limiting. Under such conditions, the entire transport cycle would be moving relatively slowly and the substrate (R6G) would therefore have more time to interact with even the WT transporter. It is not clear how far below the Km the ATP concentration would need to be to observe such an effect. Although we saw no significant difference when we compared the quenching enhancement above and below the Km, the difference at the 0.5 mM ATP concentration was lower (1.5x faster in the mutant preparations) than that observed directly above the Km (2.3x faster in the mutant preparations). Furthermore, the lack of an enhancement in coumarin 6 transport in the A666G strain makes us cautious about discarding kinetic substrate selection as the explanation for this mutant’s behavior. One prediction of this model is that strong Pdr5 substrates (for instance, those that equilibrate rapidly) might interact so quickly with Pdr5 that the time spent in the drug-binding conformation is not rate limiting. The transport of such compounds would therefore not be enhanced further by increasing the proportion of time spent in the drug-binding (inward-facing) conformation.

The recently reported cryo-EM structure of ABCG2 (Manolaridis et al., 2018) suggests that substrates bind to an inner pocket containing drug-binding sites. During the ATP-driven conformational switch, drug molecules pass through a gate to an outer binding pocket before release. It is possible that Pdr5 has an analogous structure. Pdr5 has a gating function and Ser-

119

1368 is critical (Mehla et al., 2014). An alignment of ABCG2 and Pdr5 places Ser-1368 quite close to Leu-554 and Leu-555, which make up part of the gate in the former efflux pump. It is possible that the A666G mutant fosters even greater cooperativity between two pockets during this two-step exit process.

Following the analysis of the A666G mutant, we began to characterize the linker-2 phosphoserine mutants. The initial goal for this portion of the dissertation research was to investigate the role of phosphorylation in Pdr5 function. Once the alanine substitutions were analyzed, we assumed that their hyperresistance phenotype was the result of a lack of phosphorylation at the missing serine residue. Remarkably, all six of these mutants had increased levels of Pdr5. A possible explanation was that linker-2 modulated Pdr5 activity by its reduction through phosphorylation of these serines. Several observations ruled out this explanation.

Modulation would have required that many copies of the Pdr5 transporter be phosphorylated at all six residues, but the mass spectrometry data indicated that some residues-notably Ser-837 was often missing a phosphate group. Furthermore, the S837A and S837D mutants were both hyper- resistant to several Pdr5 substrates.

Instead, we observed that the increase in steady-state levels of Pdr5 in purified PM vesicles could be explained by the surprising observation that the S837A mutant had a higher steady-state level of mRNA. Experiments with the uracil analog 4-TU demonstrated that the enhancement observed in the mutant was attributable to an increased half-life of the PDR5 transcript. Our results are in reasonably good agreement with a global analysis of WT yeast half-

120 lives that also used the 4-TU technique (Chan et al., 2018). Those investigators calculated a

PDR5 half-life of 4.35 min, which is similar to our observed value (3.70 min). Their values for

TAF10 (6.11 min) and TDH3 (23.4 min) were also reasonably close to the values we obtained for these transcripts (5.58 and 18.5 min, respectively). Our estimated half-life for ALG9, however was significantly longer than theirs (12.8 versus 3.67 min). The reason for this disparity is not obvious.

Yeast mRNA half-lives are the subject of numerous studies in many genes. Global studies of yeast transcription identified crucial sequences in both the 3’ untranslated portions and the 5’ end (Shalgi et al., 2005; Geisberg et al., 2014). Translation initiation is also extremely important (Chan et al., 2018).

It is well known that synonymous mutations can alter phenotypes through a variety of mechanisms affecting transcription or translation (Duan et al., 2003; Chamary et al., 2005;

Pagani, Raponi, and Baralle, 2005; Kimchi-Sarfaty et al., 2007; Hunt et al., 2014; Gotea et al.,

2015; Rauscher and Ignatova, 2018). Synonymous mutations can result in either loss- or gain-of- function mutations. Synonymous mutations also alter the mammalian drug transporter P-gp

(Kimchi-Sarfaty et al., 2007; Fung et al., 2014). Sarfaty et al., (2007) in their study of human P- gp demonstrate that synonymous single-nucleotide polymorphisms (SNPs) do not change either the mRNA levels or the cell surface expression of P-gp. However, the mutant P-gp showed a structural alteration at the substrate and inhibitor interaction site. They argued that this alteration reduced the speed of translation and as a result reduced protein folding.

121

Among different ethnic groups, numerous SNPs have been identified in human P-gp resulting in a number of haplotypes with varying abundance (Gerloff, 2004; Hoffmeyer, 2000).

A synonymous SNP 3435C>T in human P-gp is associated with multiple diseases including

Parkinson, leukemia and inflammatory bowel disease (Schwab et al., 2003). Thus, important determinants of these conditions are probably genetically linked to the P-gp gene. Both clinical and pre-clinical data support the idea that polymorphisms in human P-gp affect the response to drug treatment (Marzolini et al., 2004). In order to investigate the mechanism of how 3435C>T

SNP in human P-gp altered the pharmacotherapy, Wang et al., (2005) studied the mechanism of the allelic variation and disease susceptibility. They looked at the level of mRNA expression level and secondary structure. They found that 3435C allele had a significantly higher mRNA expression level compared to the 3435T allele. Also, the transcription data indicated that the mRNA stability of 3435C was higher than 3435T. Analysis of mRNA secondary structure supported the idea that 3435C>T affects the secondary structure of mRNA. Synonymous studies conducted with the dopamine D2 receptor indicate that a synonymous SNP (957C>T) can alter mRNA secondary structure which ultimately can change the mRNA stability and protein expression (Duan et al., 2003).

Synonymous mutations in bacteria can also influence gene expression. Studies conducted by Kudla et al., (2009) in Escherichia coli indicated that synonymous mutation can affect stability of mRNA folding near ribosomal binding site resulting in variation of protein level.

Their study provides a mechanistic explanation of how mRNA folding affect translation initiation and thereby protein expression. Tightly folded mRNA impeded translation initiation

122 and reduced protein synthesis (Kozak, 2005). Synonymous mutations can also affect organism fitness through enhance gene expression and as a result, drive adaptive evolution. Other studies in E coli K-12 strain, revealed that synonymous mutations can play important roles in adaptive evolution under positive selection (Conrad et al., 2009).

Reports of nonsynonymous mutations affecting transcription are not found in abundance.

Two studies with other ABC transporters are of interest, however, Manoharial et al. (2007) demonstrated that Cdr1 mutants selected for increased resistance to azoles exhibited both increased transcription rate and mRNA stability. These investigators ruled out alterations in the promoter regions but did not provide sequencing data. Thus, it was not clear whether these phenotypes were attributable to synonymous or nonsynonymous alterations or whether more than one mutation was involved. However, a study using Pseudomonas fluorescens was particularly revealing (Aryal, et al., 2014). When a strain of this bacterium was grown for 1000 generations with growth-limiting concentrations of glucose both spontaneous synonymous and nonsynonymous mutations were recovered in the ABC glucose transporter permease subunit that increased transcription. Although the mechanism behind the increase was not determined, these data clearly show that positive selection acts at the nucleotide level. There is also a report of a missense mutation in the mouse inositol-5-phosphatase gene that reduces transcription and creates a profound loss-of-function phenotype (Nguyen et al.2011).

It may be the case that most bifunctional codons are found in relatively unconserved amino acids where mutations are less likely to cause significant perturbations in protein

123 structure. In that regard, these changes would be similar to synonymous mutations. Because unconserved residues are not studied as extensively as conserved ones, missense mutations affecting transcription in the former might be overlooked.

Linker-2, a region of approximately 60 amino acids connecting TMD-1 (via TMH6) to the canonical portion of NBD2, is not conserved. For instance, Pdr5 shares 75% and 67% amino acid identity respectively with its two paralogs Pdr10 and Pdr15. Nevertheless, the amino acid identity in the linker-2 region is only 47% and 45%, respectively, in these transporters. In both cases, amino acid substitutions are found in the positions corresponding to the Pdr5 serines that we analyzed. In Pdr10, the residues equivalent to Ser-840 and Ser-841 are replaced with an alanine and glutamic acid. The changes in Pdr15 are even more dramatic. Ser-837 is replaced by proline, and the end of linker-2 containing Ser-449 to Ser-854 is missing entirely. Among the 12 strains the Saccharomyces Genome Data Base surveyed for variants, four alterations were found in the linker-2 region among three strains (although none were in the six serines). This information, and the surprising observation that seven substitutions in six serine residues create similar resistance phenotypes give rise to some interesting speculation. First, these observations suggest that unconserved nucleotides may be fertile ground for regulating mRNA abundance through altered secondary structure. This is because the resulting amino acid changes often have little effect on protein structure.

Furthermore, it would appear that the shorter WT PDR5 transcript half-life is the default, favorable option. This assumes that the mechanism for the increased resistance of the Ala-837

124 mutant applies to the other substitutions as well. Their PDR5 half-lives were not determined in this study, but it is a reasonable assumption that they are longer than that of the WT. The observation that double mutants are no more resistant than single mutants is also consistent with the idea that the short half-life of PDR5 transcript is the result of a naturally selected secondary structure. Many departures from that structure in the linker-2 region of the RNA molecule default to longer half-life regardless of whether there are one (S841A or S854A, for example), two (S837A), or three (S37A, S854A) altered nucleotides. Our data suggest that some of the linker-2 codons define an RNA stability element that reduces the lifetime of the PDR5 transcript.

There are at least two, non-mutually exclusive plausible explanations for the short half- life of the PDR5 transcript. The first is prompted by the kinetics of Pdr5 ATPase activity. A striking feature of the activity is that although the Km (ATP) is quite high (about 2 mM), the basal ATPase activity is enormous (roughly 2 µmole Pi released / mg / min) in strains that overexpress this protein. This means that the transporter is using a substantial quantity of ATP.

This might be detrimental especially under nutrient-limiting conditions where Pdr5 is known to function in robust fashion (Rahman et al. 2018). Perhaps continued overexpression of the transporter is selectively disadvantageous over time even though it results in increased resistance. Alternatively, overexpressed Pdr5 might deplete the cell of important metabolites.

These ideas are consistent with the observation that PDR5 transcription is temporarily elevated in the presence of a drug via the Pdr1 transcriptional regulator (Fardeau, et al., 2007). Thus, in most common yeast strains, Pdr5 production appears to be highly regulated.

125

In this dissertation work, I investigated two phenotypically similar mutants in the Pdr5 transporter: an A666G mutant in the transmembrane helix 5 and six alanine mutants in six serine residues in the linker-2 region of Pdr5. Although the enhancement of resistance in these mutants was similar, the mechanisms responsible were very different. Our studies on A666G revealed a novel mechanism of drug resistance in Pdr5. Our data demonstrate that cooperativity in the transport sites of A666G mutant enhanced the efflux and thereby increased the drug resistance relative to the WT. In our study of linker-2 region of Pdr5, we found that phosphorylation of the serine residues in this region of Pdr5 doesn’t play any role in drug transport. Further investigation revealed that the S837A mutant had increased expression of Pdr5 transcript and mRNA half-life which increased the Pdr5 protein and thereby enhanced resistance. There does not seem to be an evolutionary drawback to the A666G mutant since there is no further increase in Pdr5 protein in the membrane and no need to use additional ATP to achieve the increased transport. In contrast, the results from the linker-2 analysis suggests that the short, half life of

Pdr5 observed in the WT strain might be preferable in the long term. One reason for this is that although drug resistance is increased, a large increase in the constitutive ATPase of Pdr5 results from increased expression and this may be a very inefficient use of energy.

APPENDIX

Appendix I: Preparation of Buffer

Spheroplast Buffer:

Stock Volume added for 50 ml Final Concentration

Solution

1 M Sodium Phosphate 2.5 ml 50 mM

Buffer PH 7.5

2 M Sorbitol 25 ml 1M

β-mercaptoethanol 50 µl 0.1%

Lyticase (Sigma) 0.005 gm 100 µg/ml

1M NaPO4 (Sodium Phosphate) PH 7.5:

PH 1M Na2HPO4 (ml) 1M NaH2PO4 (ml)

7.6 84.5 15.5

Adjust the PH to 7.5 with Concentrated HCL

126

0.2N NaOH+ 1% SDS:

Stock Amount added for 20 ml Final Concentration

10 N NaOH 0.4 ml 0.2 N

SDS 0.2 g 1%

TE buffer PH 8.0:

Stock Amount added for 100 ml Final Concentration

1 M Tris PH 8.0 1 ml 10 mM

0.5 M EDTA 200 µl 1 mM

Adjust PH to 8.0

1M Tris Acetate Buffer PH 7.5:

Components Prep for 500ml Final Concentration

Tris Base (MW 121.14g) 60.57 g 1M

Glacial Acetic Acid (17.48M) 28.60 ml 1M

Note: Dissolved the components in 350 ml sterile MQ and adjusted PH 7.5 then brought the volume up to 500ml with miliQ water.

127

2.5M Tris Acetate Buffer PH 7.5

Components Prep for 500 ml Final Concentration

Tris Base (MW 121.14 g) 151.43 g 2.5 M

Glacial Acetic Acid (7.48M) 71.5 ml 2.5 M

TAEG Buffer:

Components Prep for 500 ml Final Concentration

Stock 1M Tris-acetate pH 7.5 5 mL 10 mM

0.5M EDTA 200 µL 0.2 mM

100% glycerol 100 mL 20%

TAE Buffer:

Components Prep for 500ml Final Concentration

Stock 1M Tris-acetate pH 7.5 5 mL 10mM

0.5M EDTA 200 µL 0.2mM

128

1 M Sodium acetate pH 5.2:

Components Prep for 500ml Final Concentration

Sodium-acetate Trihydrate 68.05 g 1 M

Note: Added glacial acetic acid to bring the pH to 5.2

0.5 M HEPES pH 7.0

Dissolved in water

Components Prep for 500ml Final Concentration

HEPES sodium Salt 65.075 g 0.5M

129

Appendix II: Preparation of Media

Boost Media

Components For 500 mL

Yeast Extract 25g

Peptone 50g

YP Media (500 mL)

Components 500 mL

Yeast Extract 5g

Peptone 10g

20% Glucose (500 mL)

Component 500 mL

D-Glucose 100g

130

REFERENCES

Agashe, D., Martinez-Gomez, N.C., Drummond, D.A., and Marx, C.J. (2013). Good codons, bad transcript: large reductions in gene expression and fitness arising from synonymous mutations in a key enzyme. Mol Biol Evol, 30: 549-560.

Albuquerque, C. P., Smolka, M. B., Payne, S. H., Bafna, V., Eng, J., and Zhou, H. (2008) A multidimensional chromatography methodology for in depth phosphoproteome analysis. Mol. Cell. Proteomics, 7: 1389-1396.

Aller, S. G., Yu, J., Ward, A., Weng, Y., Chittabiona, S., Zhuo, R., Harrell, P. M., Trinh, Y. T., Zhang, Q., Urbatsch, I. L., and Chang, G. 2009. Structure of P-glycoprotein reveals a molecular basis for poly-specific drug binding. Science, 323: 1718-1722.

Ananthaswamy, N., Rutledge, R., Sauna, Z. E., Ambudkar, S. V., Dine, E., Nelson, E., Xia, D., and Golin, J. (2010) The signaling interface of the yeast multidrug transporter Pdr5 adopts a cis configuration and there are functional overlap and equivalence of the deviant and canonical Q- loop residues. Biochemistry 49: 4440-4449.

Anthony, V., and Skach, W.R. (2002). Molecular mechanism of P-glycoprotein assembly into cellular membranes. Curr Protein Pept Sci 3: 485-501.

Arya, N., Rahman, H., Rudrow, A., Wagner, M., Schmitt, L., Ambudkar, S. V., and Golin, J. (2019) An A666G mutation in transmembrane-helix 5 of the yeast multidrug transporter Pdr5 increases drug efflux by enhancing cooperativity between transport sites. Molecular Microbiology, 112 (4):1131-1144.

131

Aryal, B., Laurent, C., and Geisler, M. (2015) Learning from each other: ABC transporter regulation by protein phosphorylation in plant and mammalian systems. Biochem. Soc. Trans. 43: 966-974.

Bailey, S. F., Hinz, A., and Kassen, R. (2014) Adaptive synonymous mutations in an experimentally evolved Pseudomonous fluorescens population. Nature Comm. DOI: 10.1038/ncomms5076.

Balzi, E., Wang, M., Leterme, S., Van Dyck, I., and Goffeau, A. (1994) Pdr5, a novel yeast multidrug resistance conferring transporter controlled by the transcription regulator Pdr1. Journal of Biological Chemistry, 269: 2206-2214.

Bissinger, P. H., and Kuchler, K. (1994) Molecular cloning and expression of the Saccharomyces cerevisiae STS1 gene product. A yeast ABC transporter conferring mycotoxin resistance. Journal of Biological Chemistry, 269: 4180-4186.

Borges-Walmsley, M. I., McKeegan, K. S., and Walmsley, A. R. (2003) Structure and function of efflux pumps that confer resistance to drugs. Biochemical Journal, 376: 313-338.

Bruggemann, E. P., Germann, U. A., Gottesman, M. M., and Pastan, I. (1989) Two different regions of P-glycoprotein are photoaffinity labeled by azidopine. Journal of Biological Chemistry, 264, 15483-15488.

Brunak, S., and Engelbrecht, J. (1996) Protein structure and the sequential structure of mRNA: alpha-helix and beta-sheet signals at the nucleotide level. Proteins, 25: 237-252.

Cannon, R. D., Lamping, E., Holmes, A. R., Niimi, K., Baret, P. V., Keniya, M. V., Tanabe, K., Niimi, M., Goffeau, A., and Monk, B. C. (2009) Efflux mediated antifungal drug resistance. Clin. Microbiol. Rev., 22 (2): 291-321.

132

Chamary, J. V., and Hurst, L. D. (2005) Evidence for selection of synonymous mutations affecting stability of mRNA secondary structure in mammals. Genome Biology 6: R75.

Chan, L. Y., Mugler, C. F., Heinrich, S., Vallotton, P., and Weis, K. (2018) Non-invasive measurement of mRNA decay reveals translational initiation as the major determinant of mRNA stability. Elife,7 e32536

Chang, G., and Roth, C.B. (2001) Structure of MsbA from E. coli: a homolog of the multidrug resistance ATP binding cassette (ABC) transporters. Science, 293:1793-1800.

Chi, A., Huttenhower, C., Geer, V. L., Coon, J.J., Syka, J. E. P., Bar, D., Shabanowitz, J., Burke, D. J.,Troyanskaya, O. G., and Hunt, D. F. (2007) Analysis of phosphorylation sites on proteins from Saccharomyces cerevisiae by electron transfer dissociation (ETD) mass spectrometry. Proc. Natl. Acad. Sci (USA), 104: 2193-2198.

Cremers, F.P., van de Pol, D.J., van Driel, M., den Hollander, A.I., van Haren, F.J., Knoers, N.V., Tijmes, N., Bergen, A.A., Rohrschneider, K., Blankenagel, A., et al. (1998) Autosomal recessive retinitis pigmentosa and cone-rod dystrophy caused by splice site mutations in the Stargardt's disease gene ABCR. Hum Mol Genet, 7: 355-362.

Conrad, T.M., Joyce, A.R., Applebee, M.K., Barrett, C.L., Xie, B., Gao, Y., and Palsson, B.O. (2009) Whole-genome resequencing of Escherichia coli K-12 MG1655 undergoing short- term laboratory evolution in lactate minimal media reveals flexible selection of adaptive mutations. Genome Biol, 10: 118.

Cui, J., and Davidson, A.L. (2011) ABC solute importers in bacteria. Essays Biochem, 50: 85-99.

Dalmas, O., Orelle, C., Foucher, A. E., Geourjon, C., Crouzy, S., Pietro, A. D., and Jault,

133

J. M. (2005) The Q-loop Disengages from the First Intracellular Loop during the Catalytic Cycle of the Multidrug ABC Transporter BmrA, Journal of Biological Chemistry, 280: 36857– 36864.

Davidson, A. L., Dassa, E., Orelle, C., and Chen, J. (2008) Structure, function, and evolution of bacterial ATP-binding cassette systems. Microbiol Mol Biol Rev., 72 (2): 317-364.

Davies, J., and Davies, D. (2010) Origins and evolution of antibiotic resistance. Microbiol Mol Biol Rev, 74: 417-433.

Dawson, R. J. P., and Locher, K. P. (2006) Structure of a bacterial multidrug ABC transporter. Nature, 443: 180-185.

Dawson R. J. P., and Locher, K. P. (2007) Structure of the multidrug ABC transporter Sav1866 from Staphylococcus aureus in complex with AMP-PNP. FEBS Letters, 581: 935-938.

Dean M, Allikmets R. (1995) Evolution of ATP-binding cassette transporter genes. Curr. Opin. Genet. Dev, 5:779–85.

Dean, M., and Annilo, T. (2005) Evolution of the ATP-binding cassette (ABC) transporter superfamily in vertebrates. Annu Rev Genomics Hum Genet, 6: 123-142.

DeCottignes, A., Owsianik, G., and Ghislain, M. (1999) Caesin kinase 1-dependent phosphorylation and stability of the yeast multidrug transporter Pdr5p. J. Biol. Chem., 274: 37139-37146.

Downes, M.T., Mehla, J., Ananthaswamy, N., Wakschlag, A., LaMonde, M., Dine, E., Ambudkar, S.V., and Golin, J. (2013) The transmembrane interface of the Saccharomyces cerevisiae multidrug transporter Pdr5: Val-656 located in intracellular-

134

loop 2 plays a major role in drug resistance. Antimicrob. Agents Chemother. 57: 1025- 1034.

Duan, J., Wainwright, M.S., Comeron, J. M., Saitou, N., Sanders, A. R., Gelernter, J., and Gejman, P. V. (2003) Synonymous mutations in the human dopamine receptor D2 (DRD2) affect mRNA stability and synthesis of the receptor. Human Mol. Genet. 12: 205-216.

Egner, R. and Kuchler, K. (1995) Endocytosis and vacuolar degradation of the plasma membrane-localized Pdr5 ATP-binding cassette transporter in Saccharomyces cerevisiae. Mol. Cell. Biol,.15: 5879-5887.

Ernst, R., Kueppers, P., Klein, C. H., Schwarzmueller, T., Kuchler, K., and Schmitt, L. (2008) A mutation of the H-loop selectively affects rhodamine transport by the yeast ABC transporter Pdr5. Proc. Natl. Acad. Sci. (USA), 105: 5069-5074.

Fardeau, V., Lelandais, G., Oldfield, A., Salin, H., Lemoine, S., Garcia, M., Tanty, V., Le Crom, S., Jacq, C., and Devaux, F. (2007) The central role of Pdr1 in the foundation of yeast drug resistance. J. Biol. Chem. 282: 5063-5074.

Fojo, A. T., Whang-Peng, J., Gottesman, M. M., and Pastan, I. (1985) Amplification of DNA sequences in human multidrug-resistance KB carcinoma cells. Proc. Natl. Acad. Sci. U.S.A., 82: 7661-7665.

Fung, K. L., Pan, J., Ohnuma, S., Lund, P., Pixley, J. N., Kimchy-Safarty, C., Ambudkar, S. V., and Gottesman, M. M. (2014) Mdr1 synonymous polymorphisms alter transporter specificity and protein stability in a stable epithelial monomer. Cancer Res. 74: 598-608.

Furman, C., Mehla, J., Ananthaswamy, N., Arya, N., Kulesh, B., Kovach, I., and Golin,

135

J. (2013). The deviant ATP-binding site of the multidrug efflux pump Pdr5 plays an active role in the transport cycle. Journal of Biological Chemistry, 288: 30420-30431.

Gadsby, D. C. and Nairn, A. C.(1996) Control of CFTR channel gating by phosphorylation and nucleotide hydrolysis. Physiol. Rev., 79: S77-S107.

Gbelska, Y., Krjger, J-J, and Breunig, K.D. (2006) Evolution of gene families: the multidrug transporter genes in five related yeast species. FEMS Yeast Research, 6: 345-355.

Geisberg, J.V., Moqtaderi, Z., Fan, X., Ozsolak, F., and Struhl, K. (2014) Global analysis of mRNA isoform half-lives reveal stabilizing and destabilizing elements in yeast. Cell, 156: 812-824.

George A. Jacoby. History of Drug resistant microbes. Antimicrobial Drug resistance. Pp 3-7.

Gerloff, T. (2004). Impact of genetic polymorphisms in transmembrane carriersystems on drug and xenobiotic distribution. Naunyn Schmiedebergs Arch Pharmacol, 369:69–77.

Germann, U. A., Chambers, T., Ambudkar, S. V., Licht, T., Cardarelli, C. O., Pastan, I., and Gottesman, M. M. (1996) Characterization of phosphorylation-defective mutants of human P-glycoprotein expressed in mammalian cells. J. Biol.Chem. 271: 1708-1716.

Glavinas, H., Krajcsi, P., Cserepes, J., and Sarkadi, B. (2004) The role of ABC transporters in drug resistance, metabolism and toxicity. Current Drug Delivery, 1 (1): 27-42.

Goffeau, A. and Dufour, J.P. (1988) Plasma membrane ATPase from the yeast Saccharomyces cerevisiae. Methods in Enzymology, 157:528–533.

136

Golin, J., Barkatt, A., Cronin, S., Eng, G., and May, L. (2000) Chemical specificity of the Pdr5 multidrug resistance gene product of Saccharomyces cerevisiae based on studies with tri- n-alkyltin chlorides. Antimicrob Agents Chemother, 44 (1): 134-138.

Golin, J., Ambudkar, S. V., Gottesman, M. M., Habib, A. D., Sczepanski, J., Ziccardi, W., and May, L. (2003) Studies with novel Pdr5p substrates demonstrate a strong size dependence for xenobiotic efflux. Journal of Biological chemistry, 278 (8): 5963-5969.

Golin, J., Kon, Z. N., Chung-Pu, W., Martello, J, Hanson, L., Supernavage, S., Ambudkar, S. V., and Sauna, Z. E. (2007) Complete inhibition of the Pdr5p multidrug efflux pump ATPase activity by its transport substrate clotrimazole suggests that GTP as well as ATP may be used as an energy source. Biochemistry, 46: 13109-13119.

Gotea, V., Gartner, J. J., Quotob, N., Einitski, L., and Samuels, Y. (2015) The functional relevance of somatic synonymous mutations in melanoma and other cancers. Pigment Cell and Melanoma, 28: 673-684.

Gottesman, M. M., Fojo, T, and Bates, S. E. (2002) Multidrug resistance in cancer: role of ATP- dependent transporters. Nature Review of Cancer, 2:46-58.

Greiberg, J. V., Moqtaderi, Z., Fan, X., Ozsolak, F., Struhl, K. (2014) Global analysis of mRNA isoform half-lives reveals stabilizing and destabilizing elements in yeast cells. Cell, 156: 812-824.

Gros, P., Croop, J., Roninson, I., Varshavsky, A., and Housman, D. E. (1986) Isolation and characterization of DNA sequences amplified in multidrug resistant human cells. Proc. Natl. Acad. Sci. U.S.A, 83: 337-341

137

Grossman, C.M. (2008). The First Use of Penicillin in the United States. Annals of Internal Medicine, 149: 135-136.

Guo, X., Li, J., Wang, T., Liu, Z., Chen, X., Li, Y., Gu, Z., Mau, X., Guan, W., and Li, Y. (2012) A mutation in intracellular loop 4 affects the drug efflux activity of the yeast multidrug transporter Pdr5p. PLoS One, 7:e29520. doi:10.1371/journal.pone.0029520.

Gupta, R. P., Keuppers, P., Schmitt, L., and Ernst, R. (2011) The multidrug transporter Pdr5: a molecular diode? Journal of Biological Chemistry, 392: 53-60.

Heimer, S., Langmann, T., Moehle, C., Mauerer, R., Dean, M., Bell, F. U., von Bergmann, K., and Schmitz, G. (2002) Mutation in the human ATP-binding cassette transporte ABCG5 and ABCG8 in sitosterolema. Human Mutation, 20: 151-151.

Henikoff, S., Greene, E.A., Pietrokovski, S., Bork, P., Attwood, T.K., and Hood, L. (1997) Gene families: the taxonomy of protein paralogs and chimeras. Science, 278: 609-614.

Hirata, D., Yano, K. Miyahara, K., Miyakawa, R. (1994) Saccharomyces cerevisiae Ydr1, which encodes a member of the ATP-binding cassette superfamily is required for multidrug resistance. Curr. Genet, 26: 285-294.

Hoffmeyer S, Burk O, von Richter O, Arnold HP, Brockmoller J, Johne A. (2000) Functional polymorphisms of the human multidrug-resistance gene:multiple sequence variations and correlation of one allele with P-glycoprotein expression and activity in vivo. Proc Natl Acad Sci USA., 97:3473–3478.

Holt, L. J., Touch, B. B., Villen, J., Johnson, A. D., Gygi, S. P., and Morgan, D. O. (2009) Global analysis of Cdk1 substrate phosphorylation sites provides insights into evolution. Science, 325: 1682-1686.

138

Hunt, R. C., Simhardi, V. L., Iandole, M., and Sauna, Z.E. (2014) Exposing synonymous mutations. Trends in Genetics, 30: 308-321.

Jungwirth, H., and Kuchler, K. (2006) Yeast ABC transporters- a tale of sex, stress, drugs and aging. FEBS Letters, 580: 1131-1138

Kathawala, R.J., Gupta, P., Ashley, C. R., and Chen, Z-S. (2015) The modulation of ABC transporter- mediated multidrug resistance in cancer: a review of the past decade. Drug Resist. Updates, 18: 1-17.

Katzmann, D. J., Hallstrom, T. C., Voet, M., Wysock, W., Golin, J., Volckaret, G., and Moye- Rowley, W. S. (1995) Expression of an ATP-binding cassette transporter gene YOR1 is required for oligomycin resistance in Saccharomyces cerevisiae. Mol. Cell. Biol., 15: 6875-6883.

Keller, A., Nesvizhskii, A., Kolker, E., and Abersold, R. (2002) Empirical statistical model for estimating the accuracy of peptide identifications made by MS/MS and data base search. Analyt. Chem, 74: 5583-5392.

Kimchi-Sarfaty, C., Oh, J. M., Kim, I-W, Sauna, Z., Calcagano, A. M., Ambudkar, S. V., and Gottesman, M. M. (2007) A silent polymorphism in the MDR1 gene changes substrate specificity. Science, 315: 525-528.

Kolaczkowski, M., van der Rest, M., Cybularz-Kolaczkowski, A., Soumillion, J.P., Konigs, W.N., and Goffeau, A. (1996) Anticancer drugs, ionophoric peptides, and steroids as substrates of the yeast multidrug transporter Pdr5. J. Biol. Chem., 271:31543- 31548.

139

Kolaczkowski,, M., Sroda-Pomianek, K., Kolaczkowski, A., and Michalak, K. (2013) A conserved interdomain communication pathway of pseudosymmetrically distributed residues affects substrate specificity of the fungal multidrug transporter Cdr1p. Biochimica and Biophysica Acta-Biomembrane, 1828: 479-490.

Konotoylannis, D and Lewis, R. E. (2002) Antifungal drug resistance of pathogenic fungi. Lancet, 355:1135-1144.

Kozak, M. (2005). Regulation of translation via mRNA structure in prokaryotes and eukaryotes. Gene, 361:13-37.

Kudla, G., Murray, A.W., Tollervey, D., and Plotkin, J.B. (2009) Coding-sequence determinants of gene expression in Escherichia coli. Science, 324:255-258.

Lage, H. (2003) ABC transporters: implications on drug resistance from microorganisms to human cancers. Intl. Jour. Antimicrob. Agents, 22:188-199.

Lamping, E., Baret, P. V., Holmes, A. R., Monk, B. C., Goffeau, A., and Cannon, R. (2010) Fungal PDR transporters: phylogeny, topology, motifs, and function. Fungal Genet. Biol., 47 (2): 127-142.

Leppert, G., McDevitt, R., Folco, S. C., Van Dyk, T., Ficke, M. B., and Golin, J. (1990) Cloning by gene amplification of two loci conferring multiple drug resistance in Saccharomyces. Genetics, 125: 13-20.

Lewis, H.A., Buchanan, S.G., Burley, S.K., Conners, K., Dickey, M., Dorwart, M., Fowler, R., Gao, X., Guggino, W.B., Hendrickson, W.A., et al. (2004). Structure of nucleotide- binding domain 1 of the cystic fibrosis transmembrane conductance regulator. EMBO J., 23:282-293.

140

Li, D., J. Zhang, M. Anpo, M. Xue, and Liu, Y. (2005) Photophysical and photochemical properties of coumarin-6 molecules incorporated within MCM-48. Materials Letters, 59:2120-2123.

Li, X., Gerber, S. A., Rudner, A. D., Beausoleil, S. A., Hass, W., Villen, J., Elias, J. E., and Gygi, S. P. (2007) Large-scale phosphorylation of analysis of alpha factor-arrested Saccharomyces cerevisiae. J. Proteome Res. 6: 1190-1197.

Liu, B., and M. Pop. (2009) ARDB—Antibiotic Resistance Genes Database. Nucleic Acids Res. 37: D443–D447.

Livak, K. J. and Schmittgen, T. D. (2001) Analysis of relative gene expression data using real- -ΔΔc time Quantitative PCR and the 2 T method. Methods, 25: 402-408.

Locher, K. P. Structure and mechanism of ATP-binding cassette transporters. (2009) Phil. Trans. R. Soc. B., 364: 239-245

Longo, V.D., Shadel, G.S., Kaeberlein, M., and Kennedy, B. (2012) Replicative and chronological aging in Saccharomyces cerevisiae. Cell Metab 16: 18-31. Loo, T. W., Bartlett, M. C., and Clarke, D. M. (2003) Simultaneous binding of two drugs in the binding pocket of human multidrug resistance protein P-glycoprotein. Journal of Biological Chemistry, 278, 39706-39710.

Lugo, M. and Sharom, F. J. (2005) Interactions of LDS-751 and rhodamine 123 with P- glycoprotein. Evidence for simultaneous binding of both drugs. Biochemistry, 44:14020- 14029.

141

Manoharal, R., Gaur, N. A., Panwar, S. L., Morschhauser, J., and Prasad, R. P. (2008) Transcription activation and increased mRNA stability of CDR1 in azole-resistant Candida albicans. Antimicrobial Agents and Chemotherapy, 52: 1481-1492.

Manolaridis, I., Jackson, S.M., Taylor, N. M.I., Korval, J., Stahlberg, H., and Locker, K. P. (2018) Cryo-EM structures of a human ABCG2 mutant trapped in ATP-bound and substrate bound states. Nature, 563: 426-430.

Martin, C., Berridge, C., Higgins, C. F. Mistry, P., Charlton, P., and Callaghan, R. (2000) Communication between multiple drug-binding sites pn P-glycoprotein. Molecular Pharmaceutics, 56: 624-632.

Marzolini C, Paus E, Buclin T, Kim RB. (2004). Polymorphisms in human MDR1 (P-glycoprotein): recent advances and clinical relevance. Clin Pharmacol Ther, 75:13–33.

Mehla, J., Ernst, R., Moore, R., Wakschlag, A., Marquis, M. K. , Ambudkar, S. V., and Golin, J. (2014) Evidence for a diode-baseed mechanism in a multispecific ATP-binding cassette (ABC) transporter Pdr5. Journal of Biological Chemistry, 289:26597-26606.

Mense, M., Vergani, P., White, D. M., Altberg, G., Nairn, A. C., and Gadsby, D. C. (2006) In vivo phosphorylation of CFTR promotes formation of a nucleotide-binding domain heterodimer. EMBO J., 25: 4728-4739.

Meyers, S., Schaur, W., Balzi, E., Wagner, M., Goffeau, A., and Golin, J. (1992) Interaction of the yeast pleiotropic drug resistance genes PDR1 and PDR5. Curr. Genet., 21: 431-436.

142

Nesvizhskii, A., Keller, A., Kolker, E., and Aebersold, R. (2003) A statistical model for identifying proteins by tandem mass spectrometry. Analyt. Chem. 75: 4646-4658.

Nikolaou, K. C., and T. Montagnon. (2008) Molecules that changed the world. Wiley VCH, Weinheim, Germany.

Oancea, G., O’Mara, M. L., Drew Bennett, W. F., Tieleman, D. P., Abele, R., and Tampé, R. (2009) Structural arrangement of the transmission interface in the antigen ABC transport complex TAP. Proc Natl Acad Sci (USA), 106 (14): 5551-5556.

Oldham M. L., and Chen, J. (2011) Crystal Structure of the maltose transporter in a pre- translocation intermediate state. Science, 332: 1202-1205.

Oldham, M. L., Khare, D., Quiocho, F. A., Davidson, A. L., and Chen, J. (2007) Crystal structure of a catalytic intermediate of the maltose transporter. Nature, 450: 515-521.

Ozalp, V. C., Pedersen, T. R., Nielsen, L. J., and Olsen, L. F. (2010) Time-resolved measurements of intracellular ATP in the yeast, Saccharomyces cerevisiae using a new type of nanosensor. Journal of Biological Chemistry, 285: 37579-37588.

Pagani, F., Raponi, M., and Baralle, F. E. (2005) Synonymous mutations in CFTR exon 12 affect splicing and are not neutral in evolution. Proc Natl Acad Sci (USA), 102: 6368- 6372.

Parascandola, J. (1980). The History of antibiotics: a symposium. American Institute of the History of Pharmacy No. 5. ISBN 978-0-931292-08-8.

143

Paumi, C.M., Chuk, M., Snider, J., Stagljar, I., and Michaelis, S. (2009). ABC transporters in Saccharomyces cerevisiae and their interactors: new technology advances the biology of the ABCC (MRP) subfamily. Microbiol Mol Biol Rev 73, 577-593.

Pfaller, M. A. (2012) Antifungal drug resistance: mechanisms, epidemiology, and consequences for treatment. Amer. J. Med. 125: S3-S13.

Plotkin, J.B., and Kudla, G. (2011). Synonymous but not the same: the causes and consequences of codon bias. Nature reviews Genetics, 12: 32-42.

Prasad, R., De Wergifosse, P., Goffeau, A., and Balzi, E. (1995). Molecular cloning and characterization of a novel gene of Candida albicans, CDR1, conferring multiple resistance to drugs and antifungals. Curr Genet 27, 320-329.

Prasad, R. P., Banerjee, A, Khandewal, N. K., and Dhamagaye, S. (2014) The ABCs of Candida albicans multidrug transporter Cdr1p. Eukaryotic Cell, 14:1154-1164.

Rahman, H., Carneglia, J., Lausten,M., Robertello, M., Choy, J., and Golin, J. (2018) Robust, pleiotropic drug resistance-5 (Pdr5) is vigorously maintained during glucose and nitrogen starvation. FEMS Yeast Res. 18: fy029

Rank G. H., and Bech-Hansen, N. J. (1973) Single nuclear gene inherited cross-resistance and collateral sensitivity to 17 inhibitors of mitochondrial function in Saccharomyces cerevisiae. Mol. Gen. Genet. 126: 93-102.

Rauscher, R. and Ignatova, Z. (2018) Timing during translation matters: synonymous mutations in human pathologies influence protein folding and function. Biochem. Soc.Trans. 46:937-944.

144

Rogers, B., Decottignies, A., Kolaczkowski, M., Caravajal, E., Balzi, E., and Goffeau, A. (2001) The pleiotropic drug ABC transporters from Saccharomyces cerevisiae. J. Mol. Microbiol. Biotechnol. 3: 207-214.

Roninson, I. B., Abelson, H.T., Houseman, D. E., Howell, N., and Varshavsky, A. (1984) Amplification of specific DNA sequences correlates with multidrug resistance in Chinese hamster cells. Nature. 309: 626-628.

Rutledge, R., Esser, L., Ma, J., and Xia, L. (2011) Toward an understanding of the mechanism of action of the yeast multidrug resistance transporter Pdr5p: a molecular modeling study. J. Struct. Biol. 173: 333-344.

Sauna, Z. E., and Ambudkar, S. V. (2001) Characterization of the catalytic cycle of ATP hydrolysis by human P-glycoprotein. J. Biol. Chem. 276 (15): 11653-11661.

Sauna, Z. E., Bohn, S. S., Rutledge, R., Dougherty, M. P., Cronin, S., May, L., Xia, D., Ambudkar, S.V., and Golin, J. (2008) Mutations define cross-talk between the N-terminal nucleotide-binding domain and transmembrane-helix 2 of the yeast multidrug resistance transporter Pdr5: Possible conservation of a signaling interface for coupling ATP hydrolysis to drug transport. Journal of Biological Chemistry, 283:35010-35023.

Schwab M, Schaeffeler E, Marx C, Fromm MF, Kaskas B, Metzler J. (2003) Association between the C3435T MDR1 gene polymorphism and susceptibility for ulcerative colitis. Gastroenterology, 124:26–33.

Seeger, M.A., Bordignon, E., and Hohl, M. (2015) ABC exporters from a structural perspective in George, A. M. (ed) ABC Transporters: Forty Years Springer, p. 65-84.

145

Servos, J., Hasse, E., and Brendel, M. (1993) Gene SNQ2 of Saccharomyces cerevisiae, which confers resistance to 4-nitroquinoline-N-oxide and other chemicals, encodes a 169 kDa protein homologous to ATP-dependent permeases. Mol. Gen. Genet., 236: 214-218.

Shalgi, R., Lapidot, M., Shamir, R., and Pipel, Y. (2005) A catalog of stability associated sequence elements in 3’ UTRs of yeast mRNAs. Genome Biology, 6: R86.

Shapiro, A. And Ling, V. (1997) Positively cooperative sites for drug transport by P- glycoprotein with distinct drug specificities. European Journal of Biochemistry, 250:130- 137.

Shukla, S., Saini, P., Jha, S., Ambudkar, S. V., and Prasad, R. (2003) Functional Characterization of C. albicans ABC transporter Cdr1. Eukaryotic Cell, 2:1361-1375.

Siegel, M. and Ragsdale, N. Antifungal mode of action of imazalil. (1978) Plant Biochemistry and Physiology 9: 48-56.

Siemion, I.Z., and Siemion, P.J. (1994). The informational context of the third base in amino acid codons. Biosystems 33:139-148.

Supek, F., and Vlahovicek, K. (2005) Comparison of codon usage measures and their applicability in prediction of microbial gene expressivity. BMC Bioinformatics, 6:182.

Tanabe, K., Bonus, M., Tomryama, S., Miyoshi, K., Nagi, M., Nimi, K., Chindamporn, A, et al. (2018) FK506 resistance of S. cerevisiae Pdr5 and C. Albicans Cdr1 involves mutants in transmembrane domains and extracellular loops. Antimicrobial Agents and Chemotherapy, Doi 10.1128/AAC.01146-18.

146

ter Beek, J., Guskov, A., and Slotboom, D.J. (2014) Structural diversity of ABC transporters. J Gen Physiol, 143: 419-435.

Teste, M. A., Duquenne, M., Franscois, J. M., and Parrou, J. L. (2009) Validation of reference genes for quantitative expression analysis by real-time-RT-PCR in Saccharomyces cerevisiae. BMC Mol Biol. 10: 99-113.

Urbatsch, I. L., Gimi, K., Wilke-Mounts, S., and Senior, A. E. (2000) Investigation of the role of glutamine-471 and glutamine-1114 in the catalytic sites of P-glycoprotein. Biochemistry 39:11921-11927.

Van der Bliek, A. M., Van der Velde-Koerts, Ling, T. V., Borst, P. 1986. Overexpression and amplification of five genes in a multidrug resistant Chinese hamster ovary cell line. Mol. Cell. Biol. 6: 1671-1678.

Vasiliou, V., Vasiliou, K. and Nebert, D.W. (2009). Human ATP-binding cassette (ABC) transporter family. Human Genomics, 3(3): 281–290.

Zolnerciks, J. K., Wooding, C., and Linton, K. J., 2007. Evidence for a Sav1866-like architecture for the human multidrug transporter P-glycoprotein. FASEB J., 21: 3937-3948.

Wainwright, M., and Swan, H. (1986). C.G. Paine and the earliest surviving clinical records of penicillin therapy. Medical History, 30(1): 42-56. doi:10.1017/S0025727300045026.

Zhang, Z., Liu, F., and Chen, J. (2017). Conformational Changes of CFTR upon Phosphorylation and ATP Binding. Cell, 170: 483-491 e488.

147

Permission to use figure 7

Permission to use figure-9