Doctorate of Philosophy (Ph.D.)

Development of Analytical Methods to Assist with the Purification & Characterization of Novel Endogenous Cardiotonic Steroids Extracted from Sus domesticas Skeletal Muscle

A dissertation submitted to the Division of Research and Advanced Studies of the University of Cincinnati

In partial fulfillment of the requirements for the degree of Doctorate of Philosophy (Ph.D.)

In the Department of Chemistry of the College of Arts and Sciences March 2018 by Cory Stiner

B.S., Chemical Technology, University of Cincinnati, Cincinnati, OH 2012

Committee Chair: Dr. Julio Landero Committee Co-Chair: Dr. Edward Merino Committee Member: Dr. Judith Heiny Committee Member: Dr. Pearl Tsang

Abstract

A new analytical method was designed to quantify the activity of the Na+, K+ ATPase in skeletal muscle and red blood cells based on ICP-MS-MS. This new approach will aid in determining the physiological effects of novel endogenous cardiotonic steroids extracted from pig skeletal muscle. The Na+, K+ ATPase is a vital enzyme in all eukaryotic cells. The Na+, K+

ATPase is also the receptor for cardiotonic steroids, a class of compounds that are used in the treatment of heart failure as well as many other medical conditions. Endogenous cardiotonic steroids are present at low levels in tissues and blood, and these compounds play a significant role in blood pressure regulation and cardiac function. The presence of uncharacterized cardiotonic steroids in tissues presents a relevant area for research.

The first part of my dissertation focuses on metal ion transport by the Na+, K+ ATPase in cells which were quantified by ICP-MS-MS. The activity of the Na+, K+ ATPase in mouse skeletal muscles and human red blood cells was measured based on the amount of Rb, as a K analog and K present upon quantification by ICP-MS-MS. The previous method used to quantify Rb and measure Na+, K+ ATPase activity in tissues and cells required the use of radioactive materials.

The instrument used for the previous method is called a scintillation counter which can detect and measure the ionizing radiation from 86Rb. The new approach using ICP-MS-MS does not require the use of radioactive 86Rb. Instead, natural abundance 85Rb and 87Rb can be used to quantify Rb within the cells to determine the activity of the Na+, K+ ATPase. This new method helps to study physiological processes that are metal-dependent and can be used for various cell types and conditions. This work was done in collaboration with Dr. Judith Heiny and Dr.

Tatiana Radzyukevich from the Department of Molecular and Cellular Physiology, University of

Cincinnati.

The second part of my dissertation focuses on ICP-MS-MS analysis of micro-sized biological samples with heteroatoms as an internal tag for mass-free quantification of selected elements. ICP-MS-MS has been used for elemental analysis at trace and ultra-trace levels in various research areas. In biomedical research, the application of ICP-MS-MS has been limited by the sample volumes obtained. A new approach was developed which allowed for mass-free quantification of selected elements in biological samples that have sub-microgram mass, processed into low microliter volumes. The biological samples used were H. capsulatum, B. dermatitidis yeast, RAW macrophages, and skeletal myofiber cells. This work was done in collaboration with Dr. Judith Heiny and Hesamedin Hakimjavadi (Department of Molecular and

Cellular Physiology, University of Cincinnati), and 1Dr. Kavitha Subramanian Vignesh and 1,2Dr.

George Deepe Jr. (1Department of Internal Medicine – Infectious Diseases Division, University of Cincinnati) (2Veterans Affairs Hospital, Cincinnati, OH).

The third part of my dissertation focuses on the isolation and purification of novel endogenous cardiotonic steroids from pig skeletal muscle. Cardiotonic steroids are a class of compounds that can be extracted from various living organisms. It is known that these compounds can bind to the Na+, K+ ATPase and inhibit ion transportation across the plasma membrane. The goal of this project is to isolate and purify novel endogenous cardiotonic steroids from mammalian tissue. A variety of techniques were used to refine these endogenous cardiotonic steroids from pig skeletal muscle, but a few that were used are the batch affinity extraction (BAE), size exclusion-high performance liquid chromatography (SEC-HPLC) and

iii

reverse phase-high performance liquid chromatography (RP-HPLC). Throughout the purification process, multiple assays are conducted to ensure that the compound is present in each phase of the purification process. The results have shown that these endogenous compounds are very active and have a high affinity for the Na+, K+ ATPase, which is one of the traits for these compounds. This work was done in collaboration with Dr. Judith Heiny (Department of

Molecular and Cellular Physiology, University of Cincinnati), Jiawei Gong (Department of

Chemistry, University of Cincinnati), and David Cowart (Ohio University).

The fourth portion of my dissertation focuses on the physiological effects of novel endogenous cardiotonic steroids extracted from pig skeletal muscle tissue. Various assays were used such as the ATPase assay, red blood cell assay, and the reversibility red blood cell assay.

These assays help to determine if the unknown compounds have inhibitory effects on the Na+,

K+ ATPase, if they work in higher order mammals, and if they can be easily unbound from the enzyme. This work was done in collaboration with Dr. Judith Heiny and Hesamedin Hakimjavadi

(Department of Molecular and Cellular Physiology, Univerisity of Cincinnati) and David Cowart

(Ohio University).

The final portion of my dissertation focuses on the characterization of novel endogenous cardiotonic steroids extracted from pig skeletal muscle. There are two classes of cardiotonic steroids which are cardenolides and bufadienolides. These compounds are differentiated by their lactone ring, steroid core, and sugar moiety. The endogenous cardiotonic steroids will be characterized by determining their molar weight and structure using mass spectrometry and NMR. The mass spectrometry results have shown mass-to-charge ratios that do not match any other known cardiotonic steroids. This work was done in collaboration with

Dr. Judith Heiny (Department of Molecular and Cellular Physiology, University of Cincinnati).

Others have helped in the process of characterizing this compound, but are not a part of the collaboration such as Dr. Balu Addepalli & Dr. Robert Ross (Department of Chemistry – Mass

Spec Facility, University of Cincinnati) and Dr. Miki Watanabe (NMR Metabolomics Core –

Division of Pathology, Cincinnati Children’s Hospital Medical Center).

Dedication

I dedicate this dissertation to my parents: my mother, Ms. Sarita L. Stiner and my late father, Mr. Clyde Lewis; my grandparents: Ms. JoAnn Stiner, Mr. Colvin L. Stiner, and Ms. Hazel

Woodson; my aunt: Ms. Kimberely A. Andrews; my uncle: Mr. Eric Stiner; my cousins: Ms.

Danielle Stiner and Mr. Frank M. Andrews.

vii

Acknowledgments

First, I would like to thank my advisor Dr. Julio Landero. He has been a great mentor throughout my Ph.D. studies. He is very knowledgeable, kind, patient, and a good leader which are great qualities of a Ph.D. advisor. Under his guidance throughout the years have taught me how to be a better analytical chemist, problem solver, critical thinker, and troubleshooter. I am genuinely grateful for all of the support provided by Dr. Julio Landero; I could not have made it this far without his help. Thank you for believing in me.

Second, I would like to thank my collaborator, Dr. Judith Heiny. I always saw her as a second advisor and mentor. She took the time to train me on the physiological aspects of my research projects. Dr. Judith Heiny has been a tremendous support and has provided many opportunities for me throughout my Ph.D. studies. I am very thankful for her time and efforts in helping me become a better scientist.

Third, I would like to thank my committee members, Dr. Edward Merino and Dr. Pearl

Tsang. Thanks for all of the advice and challenging questions that made me think of my research projects from different perspectives. Also, I thank both professors for teaching me biochemistry. The extra knowledge gained from both professors has helped me to understand some aspects of my research projects better, and this could not have happened without both of them. Thanks again for all the help.

I would also like to thank my initial co-advisor, Dr. Joseph Caruso. Sadly Dr. Joseph

Caruso passed away during my studies here at the University of Cincinnati, but he was an excellent research advisor, leader, and mentor. He was an active advocate for me during my time of need, and he truly believed in me when others did not. I thank him so much for

viii

everything that he has done to help me grow and develop while being a part of his lab. I would like to repeat thanks, and he will truly be missed by many.

I would also like to thank members of the Landero and Heiny group: Jiawei Gong, for teaching me HPLC and working with me on a research project; Keaton Nahan, for always helping me around the lab and giving me advice; Amberlie Clutterbuck, for giving advice and sharing materials for HPLC; Skyler Smith, for sharing information on ICP-MS; Megan Stanton, for providing feedback on my presentations; David Cowart, for helping me scale up and purify my

OLC samples; Hesamedin Hakimjavadi and Dr. Tatiana Radzyukevich, for working with me on projects and providing much advice.

I also acknowledge other members of the Department of Chemistry: Dr. Keyang Ding for running samples using NMR for my research project. I would like to thank Dr. Balu Addepalli and Dr. Robert Ross for analyzing my samples using mass spectrometry. I would also like to acknowledge Kellee Adams, Jill Hulsman, Kimberly Carey, Kristina Ament, John Baker, Dr.

Stephen Macha, Dr. Peter Padolik and Dr. Anne Vonderheide for their help throughout the years within the chemistry department.

I would also like to thank all of my family and friends who have supported me throughout my Ph.D. studies. I thank Shad Williams, Lisa Watkins, Khahlia Sanders, Cassandra

McPherson, Shelia Williams, Shelina Williams, Chase Lewis, The Simmions Family, Gregory Lee,

Kolade Ojo, Desirae Bowen, Annitra Bailey, Arielle Smith, Tempestt Young, Bianca Calloway,

Rayon Fulton, Don Jason and many others for all of their help and support in many ways. I am also thankful for the financial support from the Department of Chemistry, National Institutes of

Health NIAMS grants R01-AR063710 & R01-AI106269, National Institutes of Health grant award

to promote diversity in health-related research, the Advanced Multidisciplinary Training

Program in Systems Biology grant award from the Department of Molecular & Cellular

Physiology, and Agilent Technologies.

Ultimately, I thank the Lord for being with me every step of the way throughout this

Ph.D. program. I faced many challenges while being in the program which only made me a stronger individual. I learned many valuable lessons during my Ph.D., but one lesson that intrigued me the most was that the Lord equips people to do his work. I believe the challenges that I faced were only preparing me for something more significant in God’s plan. I am grateful for his guidance, and I could not have done this without the Lord; I cannot thank him enough.

Table of Contents

Abstract ………………………………………………………………………………………………...... ii

Dedication …………………………………………………………………………………………………………………………. vii

Acknowledgments…………………………………………………………………………………………………………….. viii

List of Figures …………………………………………………………………………………………………………………….. xv

List of Tables …………………………………………………………………………………………………………………… xviii

Chapter 1: Introduction ………………………………………………………………………………………………………. 1

1.1 Overview of this Chapter ……………………………………………………………………………………. 1

1.2 Na+, K+ ATPase ……………………………………………………………………………………………………. 1

1.3 Cardiotonic Steroids …………………………………………………………………………………………… 4

1.4 Activity of the Na+, K+ ATPase & Overview of ICP-MS …………………………………………. 9

1.5 Objective of this Research ………………………………………………………………………………… 13

1.6 Overview of this Dissertation …………………………………………………………………………… 14

Chapter 2: Metal Ion Transport Quantified by ICP-MS in Intact Cells …………………………………. 16

2.1 Introduction …………………………………………………………………………………………………….. 16

2.2 Experimental ……………………………………………………………………………………………………. 18 2.2.1 Chemicals and Materials ……………………………………………………………………….. 18 2.2.2 Sample Preparation ………………………………………………………………………………. 19 2.2.2.1 Mouse Skeletal Muscles …………………………………………………………. 19 2.2.2.2 Human Red Blood Cells …………………………………………………………… 21 2.2.3 Instrumentation ……………………………………………………………………………………. 22 2.2.4 Statistical Analysis …………………………………………………………………………………. 25

2.3 Results and Discussion ……………………………………………………………………………………… 25 2.3.1 Reference values for endogenous Rb content in full skeletal muscles …… 26 2.3.2 Equimolar replacement of RbCl for KCl to measure Na+, K+ ATPase activity 2.3.3 Muscle sulfur content as an index of muscle mass ………………………………… 28 2.3.4 Tracer RbCl used to measure Na+, K+ ATPase activity …………………………….. 30

2.3.5 Na+, K+ ATPase transport in human RBCs analyzed by ICP-MS ………………. 31

2.4 Conclusion ………………………………………………………………………………………………………. 34 Acknowledgments ………………………………………………………………………………………………… 36

Chapter 3: ICP-MS-MS Analysis of Biological Micro Samples with Heteroatoms as an Internal Tag for Mass-Free Quantification of Selected Elements ……………………………………………………. 37

3.1 Introduction ……………………………………………………………………………………………………. 37

3.2 Experimental …………………………………………………………………………………………………… 42 3.2.1 Chemicals and Materials ………………………………………………………………………. 42 3.2.2 Sample Preparation ……………………………………………………………………………… 42 3.2.2.1 Culture & quantification of RAW macrophages ……………………… 42 3.2.2.2 Culture & quantification of yeasts …………………………………………. 43 3.2.2.3 Dissociation & quantification of single muscle cells (fibers) …… 44 3.2.2.4 Sample preparation of total metal analysis ……………………………. 45 3.2.3 Instrumentation …………………………………………………………………………………… 46 3.2.4 Statistical Analysis ……………………………………………………………………………….. 51

3.3 Results and Discussion ……………………………………………………………………………………. 51 3.3.1 S or P content provides better index of cell mass than cell number …….. 54 3.3.2 Prediction of cell mass from the S or P content of H. capsulatum ………… 58 3.3.3 Normalization by S or P provides accurate quantification of metal elements in small cell numbers 3.3.4 Quantification of metal ion transport in single skeletal muscle cells ……. 62

3.4 Conclusion ………………………………………………………………………………………………………. 68 Acknowledgments ………………………………………………………………………………………………… 69

Chapter 4: Isolation & Purification of Novel Endogenous Cardiotonic Like Steroids from Sus domesticas Skeletal Muscle Tissue …………………………………………………………………………………… 70

4.1 Introduction ……………………………………………………………………………………………………. 70

4.2 Experimental …………………………………………………………………………………………………… 71 4.2.1 Chemicals and Materials ………………………………………………………………………. 71 4.2.2 Sample Preparation ……………………………………………………………………………… 72 4.2.2.1 Homogenization of pig skeletal muscle ………………………………….. 72 4.2.2.2 Batch affinity extraction (BAE) ……………………………………………….. 73 4.2.2.3 3H-Ouabain Competition binding assay ………………………………….. 74 4.2.3 Instrumentation ……………………………………………………………………………………. 75 4.2.3.1 Size exclusion chromatography (analytical) ……………………………. 75 4.2.3.2 Reverse phase chromatography (analytical) …………………………... 76

xii

4.2.3.3 Hydrophilic interaction liquid chromatography (analytical) ……. 77 4.2.3.4 Size exclusion chromatography (preparatory) ………………………… 77 4.2.3.5 Reverse phase chromatography (preparatory) ……………………….. 78 4.2.3.6 Hydrophilic interaction liquid chromatography (semi-prep) …… 79 4.2.4 Data Analysis ………………………………………………………………………………………… 79

4.3 Results and Discussion …………………………………………………………………………………….. 80 4.3.1 HPLC Fraction Naming System ………………………………………………………………. 80 4.3.2 OLC SEC-HPLC separations analyzed by CB activity assay ………………………. 81 4.3.3 OLC RP-HPLC separations analyzed by CB activity assay ………………………… 86 4.3.4 Preparative SEC-HPLC fractions …………………………………………………………….. 91 4.3.5 Preparative RP-HPLC fractions ………………………………………………………………. 92 4.3.6 HILIC Separations ………………………………………………………………………………….. 95

4.4 Conclusion ……………………………………………………………………………………………………….. 99 Acknowledgments ……………………………………………………………………………………………….. 100

Chapter 5: The Physiological Effects of Novel Endogenous Cardiotonic Like Steroids Extracted from Sus domesticas Skeletal Muscle Tissue ……………………………………………………………………. 101

5.1 Introduction …………………………………………………………………………………………………… 101

5.2 Experimental ………………………………………………………………………………………………….. 102 5.2.1 Chemicals and Materials ……………………………………………………………………… 102 5.2.2 Sample Preparation …………………………………………………………………………….. 104 5.2.3 Instrumentation …………………………………………………………………………………… 107 5.2.4 Statistical Analysis ……………………………………………………………………………….. 107

5.3 Results and Discussion …………………………………………………………………………………… 108 5.3.1 ATPase assay ………………………………………………………………………………………. 108 5.3.2 Red Blood Cell assay …………………………………………………………………………… 111 5.3.3 Reversibility red blood cell assay ………………………………………………………… 114

5.4 Conclusion …………………………………………………………………………………………………….. 117 Acknowledgments ………………………………………………………………………………………………. 119

Chapter 6: Characterization of Novel Endogenous Cardiotonic Like Steroids Extracted & Purified from Sus domesticas Skeletal Muscle Tissue ……………………………………………………… 120

6.1 Introduction ………………………………………………………………………………………………….. 120

6.2 Experimental …………………………………………………………………………………………………. 121

6.3 Results and Discussion …………………………………………………………………………………… 122

xiii

6.3.1 Mass spectrometry …………………………………………………………………………….. 122 6.3.2 LC-MS …………………………………………………………………………………………………. 126 6.3.3 Structure elucidation by NMR …………………………………………………………….. 133 6.4 Conclusion …………………………………………………………………………………………………….. 139 Acknowledgments ………………………………………………………………………………………………. 140

Chapter 7: Summary, Conclusions, and Future Work ………………………………………………………. 141

Glossary: Abbreviations ………………………………………………………………………………………………….. 147

References ……………………………………………………………………………………………………………………… 149

xiv

List of Figures

Figure 1.1: Diagram of the Na+, K+ ATPase cycle. This diagram shows a cycle of how ions are transported across the cell membrane. The orange hexagons: represent sodium ions. The yellow ovals: represent potassium ions. The blue pentagons: represent phosphate groups. Adapted from reference [1] …………………………………………………………………………………………………….. 4 Figure 1.2: Cardiotonic Steroids. This diagram shows different cardiotonic steroid chemical structures that have been extracted from a few living organisms. Adapted from reference [5] . 5 Figure 1.3 Part 1: Cardenolides. This diagram shows various structural motifs of cardenolides as well as the family and species it was extracted from, the molecular weight, and the molecular formula of the compound. Adapted from reference [7] ………………………………………………………….. 6 Figure 1.3 Part 2: Bufadienolides. This diagram shows various structural motifs of one cardenolide and three bufadienolides as well as the family and species it was extracted from, the molecular weight, and the molecular formula of the compound. Adapted from reference [7]…………………………………………………………………………………………………………………………………………….. 7 Figure 1.4: Schematic of Scintillation Counter. This schematic shows the parts of the instrument that contribute to its operation. Adapted from reference [14] ……………………………. 10 Figure 1.5: Schematic diagram of the ICP-QQQ. Adapted from reference [17] ……………………… 12 Figure 2.1: Rubidium Calibration Curve Obtained from ICP-MS QQQ …………………………………. 26 Figure 2.2: The S content of mouse skeletal muscles is an accurate index of the tissue mass. 30 Figure 2.3: Rb taken up by RBCs in the absence and presence of 1mM ouabain, measured by ICP-MS and referred to Fe based mass. The RSD for each group was 1.6% …………………………… 34 Fig. 3.1: A) P and B) S content (total ng per sample) plotted versus the number of H. capsulatum cells in each sample ………………………………………………………………………………………….. 56 Fig. 3.2: P versus S content (total ng per sample) plotted for different cell numbers of H. capsulatum cells …………………………………………………………………………………………………………………… 57 Figure 3.3: Cell mass computed from P content obtained by ICP-MS-MS versus experimental cell mass obtained in an analytical balance ……………………………………………………………………….. 59 Figure 3.4: Ratio of Zinc/Sulfur calculated from the total ng of each element in different samples of H. capsulatum and B. dermatitidis …………………………………………………………………… 61 Figure 3.5: Correlations between the K, S, P and Rb content (ng) and number of muscle fibers in each sample ……………………………………………………………………………………………………………………… 64 Figure 3.6: Correlations among the Rb, P, and S content (total ng) measured by ICP-MS-MS in samples of 1, 2, 3, and 5, 7 and 10 EDL fibers ………………………………………………………………………. 65 Figure 3.7: Transport activity of Na+ , K+ ATPase in dissociated mouse EDL fibers ………………. 67

Figure 4.0: Roadmap to better understand how each fraction was collected and processed…81

Figure 4.1: Size Exclusion Chromatogram & Isoabsorbance Plot of Ouabain Standard ……………………. 83 Figure 4.2: Size Exclusion Chromatogram of OLC compound ………………………………………………. 84 Figure 4.3: CB assay of SEC fractions ………………………………………………………………………………….. 85 Figure 4.4: Reverse Phase Chromatogram & Isoabsorbance Plot of Fraction B …………………… 87 Figure 4.5: Reverse Phase Chromatogram & Isoabsorbance Plot of Fraction C …………………… 88 Figure 4.6: Reverse Phase Chromatogram & Isoabsorbacne Plot of Fraction D …………………… 89 Figure 4.7: Competition Binding Assay of Reverse Phase Fractions …………………………………….. 90 Figure 4.8: Preparatory Size Exclusion Chromatogram of OLC compound …………………………… 91 Figure 4.9: Preparatory Reverse Phase Chromatograms of OLC compound ………………………… 94 Figure 4.10: HILIC Separation of Fraction B1 ………………………………………………………………………. 96 Figure 4.11: Semi-preparative HILIC Chromatogram & Isoabsorbance Plot of Fraction B1 ….. 97 Figure 4.12: Competition Binding Assay of HILIC Fractions …………………………………………………. 98 Figure 5.1: ATPase Assay of Ouabain ……………………………………………………………………………….. 109 Figure 5.2: ATPase Assay of OLC B1.1 & B1.2 ……………………………………………………………………. 111 Figure 5.3: Red Blood Cell Assay of Negative Controls ……………………………………………………… 113 Figure 5.4: Red Blood Cell Assay of OLC B1.1 & B1.2 ………………………………………………………… 114 Figure 5.5: Reversibility Red Blood Cell Assay of OLC B1.1 & B1.2 ……………………………………. 116 Figure 6.1: Direct Infusion of Ouabain using a ThermoFinnigan LTQ XL Mass Spectrometer.123 Figure 6.2: MSMS of Ouabain at 585.17 m/z ……………………………………………………………………. 124 Figure 6.3: Mass spectra of OLC B1 ………………………………………………………………………………….. 125 Figure 6.4: Total Ion and Extracted Ion Chromatogram of Ouabain ………………………………….. 127

Figure 6.5: Mass Spectra of Ouabain Obtained from the EIC ……………………………………………. 128

Figure 6.6: Total Ion & Extracted Ion Chromatogram of OLC B1 ……………………………………….. 130

Figure 6.7: MSMS of mass 497.1688 from OLC B1 …………………………………………………………….. 131

Figure 6.8: MSMS of mass 237.0885 from OLC B1 …………………………………………………………….. 132

Figure 6.9: NMR spectra of unknown OLC compounds …………………………………………………….. 133

xvi

Figure 6.10: Comparison between Ouabain & B1.2 ………………………………………………………….. 134

Figure 6.11: Comparison between Bufalin & B1.2 ……………………………………………………………. 135

Figure 6.12: Comparison between Tris & B1.2 ………………………………………………………………….. 136

Figure 6.13: Comparison of B1.2 with possible contaminants …………………………………………… 137

xvii

List of Tables

Table 2.1: Instrument tune parameters for ICP-MS QQQ …………………………………………………… 23 Table 2.2: Rb uptake by the Na+, K+ ATPase in quiescent mouse EDL muscle measured by ICP- MS using equimolar replacement of RbCl for KCl in the uptake buffer ………………………………. 28 Table 2.3: Rb uptake by the Na+, K+ ATPase in quiescent mouse EDL muscle measured by ICP- MS using 200μM RbCl as a tracer ………………………………………………………………………………………. 31 Table 2.4: Comparison of Rb transport by human RBCs measured by ICP-MS and 86Rb tracer…………………………………………………………………………………………………………………………………. 33 Table 3.1: Instrument tune parameters for ICP-MS-QQQ using micro concentric nebulizer… 49

Table 3.2: Instrument tune parameters for ICP-MS-QQQ using a high-efficiency low-flow nebulizer ………………………………………………………………………………………………………………………….. 50

Table 3.3 Calibration Summary: Summary of common ICP-MS-MS calibration parameters …. 53

xviii

Chapter 1: Introduction

1.1 Overview of this Chapter

Chapter one will start by covering the background of the Na+, K+ ATPase and the physiology of how it functions. The next section will include the background of cardiotonic steroids and how they interact with the Na+, K+ ATPase. Afterward, the following section will discuss the activity of the Na+, K+ ATPase, and a brief overview of ICP-MS-MS. This chapter will also include the objectives of this research project. Finally, a short summary of this dissertation will be provided.

1.2 Na+, K+ ATPase

Eukaryotic cells have different components that contribute to the structure and functionally of the cell. Each piece of the cell has a different role in maintaining homeostasis, which is the ability to preserve chemical and biological stability within a cell [1]. The ionic composition in the cellular environment is necessary for maintaining homeostasis because it influences osmotic pressure, which can affect the flux of water, keeping a specific ratio of water inside and outside of the cell based on the natural gradient of solutes. This natural occurrence of water exchange must be directed to the cells to perform its functions and maintain homeostasis, and controlling the solute concentrations is the primary mechanism used for this purpose. Besides the control of osmotic pressure, the cells require transmembrane gradients of specific solutes, like sodium, to perform essential biological functions. Active transport is necessary to create such up-hill gradients. Along the plasma membrane are various types of transmembrane proteins that allow the transport of ions or molecules into and out of the cell.

One of the most vital transmembrane proteins, expressed in all eukaryotic cells, is the sodium- potassium pump (Na+, K+ ATPase).

The Na+, K+ ATPase is composed of an α, β, and γ (also termed FXYD) subunit. The α subunit is the primary catalytic moiety of the transporter and is present in four isoforms in humans. The β subunit is essential for the recruitment of the α subunit to the plasma membrane and the occlusion of K+ ions [2]. The γ subunit functions as a regulator of the Na+, K+

ATPase to ensure appropriate tissue functions such as Na+ reabsorption, muscle contractility, and neuronal excitability [3]. The Na+, K+ ATPase transports three sodium ions (Na+) outside of the cell and two potassium ions (K+) inside of the cell [1]. It can create and maintain a steep transmembrane gradient of sodium and potassium, keeping most of the Na ions outside the cells and K ions inside. The transmembrane K+ gradient underlies the resting membrane electrical potential, and the transmembrane Na+ gradient provides the driving force for many

Na-dependent secondary transporters that mediate the uptake and extrusion of glucose, calcium, amino acids, vitamins, and other metabolites and ions [2]. Through these critical transport roles, the Na+, K+ ATPase is an essential regulator of cell volume, body fluid, and electrolyte balance. In addition to these crucial functions in ion transport, the Na+, K+ ATPase plays a receptor signaling role through its highly conserved cardiotonic steroid (also termed cardiac glycosides) binding site on the extracellular domain of the enzyme. Binding of exogenous or endogenous cardiotonic steroids to this site facilitates interactions between Na+,

K+ ATPase and multiple kinases, phosphatases and other protein partners, to regulate blood pressure, gene expression, metabolism, cell growth and other vital cell functions [4].

Figure 1.1 shows a simplified diagram of the Na+, K+ ATPase cycle which displays how the ions are transported. The Na+, K+ ATPase has two main conformations. The first conformation of the Na+, K+ ATPase is open to the intracellular space of the cell and the second conformation is open to the extracellular space. The binding sites within the first conformation have a high affinity for sodium ions. Once the sodium ions bind to their sites, the Na+, K+ ATPase must go through phosphorylation so it can close off to the intracellular space and open to the extracellular space. The Na+, K+ ATPase goes through phosphorylation by the hydrolysis of adenosine triphosphate (ATP). During this process, ATP is converted to adenosine diphosphate

(ADP). Now that the Na+, K+ ATPase is bound to a phosphate group it changes its shape to the second conformation [1].

The second conformation has a strong affinity for potassium ions while it is open to the extracellular space. Since it has a high affinity for potassium ions, the sodium ions that are bound to the inside of the Na+, K+ ATPase is now released into the extracellular space, and then potassium ions attach to the inside of the Na+, K+ ATPase. Once the potassium ions are bound to the Na+, K+ ATPase while it is in the second conformation, it causes the Na+, K+ ATPase to go through dephosphorylation. When dephosphorylation occurs, the Na+, K+ ATPase changes shape again, going from the second conformation back to the first. Once the Na+, K+ ATPase is back in the primary conformation, the potassium ions that are bound to the Na+, K+ ATPase are released into the intracellular space. When this process is complete, the cycle starts over again.

This cycle maintains the concentration gradients for Na and K across the plasma membrane [1].

1.3 Cardiotonic Steroids Cardiotonic steroids (CTS) are a class of compounds that can be extracted from various living organisms. Figure 1.2 shows an example of different cardiotonic steroid chemical structures that have been derived from a few living organisms. Ouabain is found in the common foxglove (Digitalis purpurea) as well as other plants. Marinobufagenin and proscillaridin A are present in the cane toad (Bufo marinus) and sea squill (Urginea maritima), respectively [5]. A large variety of cardiotonic steroids are known to exist, and many cardiotonic steroids have yet to be characterized. Cardiotonic steroids have many names such as ouabain-like compounds

(OLC), cardiac glycosides, and endogenous digitalis factor (EDLF). It is known that these compounds can bind to the Na+, K+ ATPase and inhibit ion transport across the plasma membrane [6].

Figure 1.2: Cardiotonic Steroids. This diagram shows different cardiotonic steroid chemical structures that have been extracted from a few living organisms. Adapted from reference [5].

There are two main classes of cardiotonic steroids, cardenolides and bufadienolides.

These compounds have three distinct structural motifs which are differentiated by their lactone ring, steroid core, and sugar moiety. Cardenolides contain a five-membered lactone ring and a sugar moiety, whereas the bufadienolides contain a six-membered lactone ring and no sugar moiety [7]. Figure 1.3 Part 1 shows a few examples of cardenolides and Figure 1.3 Part 2 shows examples of bufadienolides. When cardiotonic steroids bind to the Na+, K+ ATPase, the lactone ring is placed near the K+ binding sites, and the sugar moiety is exposed to the extracellular space outside of the cell [7].

Figure 1.3 Part 1: Cardenolides. This diagram shows various structural motifs of cardenolides as well as the family and species they are extracted from, their molecular weights, and molecular formulas. Adapted from reference [7].

Figure 1.3 Part 2: Bufadienolides. This diagram shows various structural motifs of one cardenolide and three bufadienolides as well as the family and species they are extracted from, their molecular weights, and molecular formulas. Adapted from reference [7].

Cardiotonic steroids are used in medicines to treat numerous medical conditions such as congestive heart failure, hypertension, and arrhythmia. As a treatment of congestive heart failure, these compounds can increase the force of contraction of the heart muscle by reversing the Ca exchanger mechanisms. Inhibition occurs because cardiotonic steroids block the Na+, K+

ATPase from going through dephosphorylation. By preventing dephosphorylation, it causes the intracellular sodium concentration to increase, and the intracellular potassium concentration to decrease. As a result, the sodium-calcium exchanger loses its sodium gradient slowing removal of calcium ions. As the concentration of calcium ions inside the cell increase, it enhances the ability of the heart muscle to contract [6].

Blood pressure regulation is affected when the muscular tone of the blood vessels are changed by the same effects described before, where cardiotonic steroids bound to the Na+, K+

ATPase. As CTS bind to the Na+, K+ ATPase, blood pressure increases [8]. The rise in blood pressure occurs because the sodium ion concentration increases inside of the cell due to inhibition of the Na+, K+ ATPase [9]. A research study compared, how blood pressure was affected in pregnant and non-pregnant mammals as they were exposed to increased concentrations of ouabain [10]. The results showed that the pregnant mammals were not affected by the increased levels of CTS, but the non-pregnant mammals were affected. Blood pressure increased significantly in non-pregnant mammals, in proportion to their exposure to

CTS [10]. This research study shows that exposure to CTS increases blood pressure which can lead to hypertension.

1.4 Activity of the Na+, K+ ATPase & Overview of ICP-MS

The activity of the Na+, K+ ATPase is routinely measured in laboratories by quantifying the amount of radioactive rubidium, 86Rb, which is imported from the extracellular environment as the result of Na+, K+ ATPase stimulation [11,12,13]. Rubidium is chemically similar to potassium as they are both alkali metals found within the first group on the periodic table and they both share one electron within their valence shell. Potassium is found in the intracellular and extracellular spaces around the cell; this high base level concentration make it difficult to detect small differences in potassium’s concentration resulting from Na+, K+ ATPase activity. On the other hand, rubidium is not naturally present in cells in large amounts. Its low base level concentration make it a better element for measuring and detecting small variations in its concentration. For this purpose, the use of rubidium for testing the activity of the Na+, K+

ATPase is required. High concentrations of rubidium transported into the cell represent high activity of the Na+, K+ ATPase and low concentrations of it indicate minimal activity of the Na+,

K+ ATPase.

The previous instrument used to quantify the amount of radioactive 86Rb transported into the cell by the Na+, K+ ATPase is called a scintillation counter [14]. A scintillation counter is an instrument used to detect and measure ionizing radiation, such as the ionizing radiation from 86Rb. Figure 1.4 shows a schematic of the scintillation counter which labels each component that contributes to its operation. The scintillator or phosphor, which is a fluorescent material, is bombarded with radiation, in this case  particles emitted by 86Rb. The radiation or incident particles that collide with the scintillator are converted to photons. The photons then travel through the photocathode, when this occurs, the photons are converted into electrons

which then go through a photomultiplier tube. As the electrons go through the photomultiplier tube, the electrons are multiplied each time they collide with a new dynode. Once the electrons reach the end of the photomultiplier tube, the electrons are amplified by the high voltage source to become an electrical pulse. The electrical pulses are then amplified and recorded into a data storage system [14].

Figure 1.4: Schematic of Scintillation Counter. This schematic shows the parts of the instrument that contribute to its operation. Adapted from reference [14]

The use of radioactive materials such as 86Rb requires the implementation of many safety protocols and procedures because exposure to radiation can cause damage to biomolecules reflected as mutations, teratogenesis, and cancer. If these damages occur to the genetic material, and the altered sequences or mutations are not repaired, uncontrolled cellular multiplication can lead to cancer [15]. Also, the cost of 86Rb is very high. 86Rb at $600 for one mCi has a half-life of 18 days. When working with radioactive materials, there is also a lack of flexibility for sample preparation. Radioactive materials must remain in a designated area within a lab to reduce contamination; this places limitations on various types of testing that can

be done on radioactive samples. Using triple quadrupole ICP-MS (inductively coupled plasma mass spectrometry) has many advantages compared to using a scintillation counter.

Triple quadrupole ICP-MS is an analytical instrument that is used to conduct trace or ultra-trace elemental composition analysis for various sample types [16]. Figure 1.5 shows a schematic of the triple quadrupole ICP-MS. The aqueous sample flows through the nebulizer to convert the liquid sample to a fine mist; then the fine mist is sprayed into the spray chamber where the mist is filtered to obtain a specific droplet size before going to the plasma. Once the aerosol reaches the plasma, it is ionized, and the cones and lens are used as an ion guide for ions to enter the quadrupole mass filter. The triple quadrupole ICP-MS contains two quadrupoles and an octopole reaction cell that rests between the quadrupoles. The first quadrupole (Q1) is used as a mass filter for the m/z ratios selected. The octopole reaction cell is used to remove any isobaric and polyatomic interferences, by use of He, O2, or H2 gas to react with ions present in the octopole reaction cell. When analyzing phosphorus or sulfur, it is essential to use the octopole reaction cell with O2 gas. It is vital because both phosphorus and sulfur have many polyatomic and isobaric interferences such as 14N16O1H+, 15N15N1H+, 15N16O+,

14N17O+, 13C18O+, 12C18O1H+ for phosphorus (31P) and 16O2+, 14N18O+, 15N17O+, 14N17O1H+,

15 16 1 + 32 N O H for sulfur ( S) [18]. The O2 gas used in the reaction will add on to the phosphorus and sulfur ions giving them a new mass such as 31P being converted to 31P16O which offers a unique m/z ratio of 47 and 32S being converted to 32S16O which provides a new m/z ratio of 48. The new masses 47 and 48 would then be selected by the second quadrupole (Q2) mass filter to obtain the concentration (ppb) for phosphorus and sulfur, respectively. After the chosen ions travel through Q2, the ions will reach the detector and be quantified.

Figure 1.5: Schematic diagram of the ICP-QQQ. Adapted from reference [17]

ICP-MS can be used to quantify the amount of 85Rb and 87Rb (naturally abundant isotopes) inside of cells and does not require the use of radioactive isotopes. It can also simultaneously analyze multiple elements of interest to gather more information about the samples. ICP-MS is a fully quantitative technique which allows a precise calculation of the mass of Rb flux, as opposed to the scintillation counter where a normalization against the control group is typically reported. ICP-MS, offers very low detection limits to the parts per trillion (ppt) range, high spectral resolution for multi-elemental isotope detection, and has a minimal number of interferences [16]. These advantages of ICP-MS compared to measurements with a scintillation counter significantly expand the options for experiment design and can change the way laboratories measure the activity of the Na+, K+ ATPase, and any other ATPases that transport metal ions.

1.5 Objective of this Research

Endogenous cardiotonic steroids need to be further investigated considering they can be extracted from various living organisms, and are known to have important physiological functions that have not yet been characterized. The physiology behind many of these compounds is still being investigated as they can be used to treat various medical conditions such as congestive heart failure, hypertension, and arrhythmia. Each cardiotonic steroid is unique and may be better suited for treating a medical condition more efficiently and effectively than another cardiotonic steroid. Dr. Heiny’s lab in the Department of Molecular and

Cellular Physiology at the University of Cincinnati focuses on the functionality of skeletal muscle at various levels. We proposed to find novel endogenous cardiotonic steroids in pig skeletal muscle because cardiotonic steroids are known to be stored in tissues and organs throughout the body. Pigs are also a close mammalian model to humans, making it a system that can be useful in physiological studies.

The first goal of this project was to develop and validate a new method to measure the activity of the Na+, K+ ATPase using triple quad ICP-MS. Afterward, this technique is used to determine how the novel endogenous compound would affect the Na+, K+ ATPase compared to known cardiotonic steroids. When it comes to the physiology, this method will show how effective the novel endogenous cardiotonic steroids are in inhibiting the Na+, K+ ATPase. This approach can also be used to study various cell types to determine their elemental concentrations and physiology.

The main goal of this research is to characterize a novel endogenous cardiotonic steroid from pig skeletal muscle. The cardiotonic steroid must first be purified from pig skeletal muscle and tested for activity before the compound is characterized. Characterization of this novel molecule, will aid in understanding its physiological role and molecular interactions with the

Na+, K+ ATPase. Once we understand its physiology, we can compare it to the functionality of known cardiotonic steroids. Along with characterizing this novel compound, the development of the method was also a significant part of this research project. Having a traditional method for purifying and testing the activity of compounds is essential, especially for studies in living organisms. After characterizing the compound present in pig skeletal muscles, it will be used to identify cardiotonic steroids in humans. The endogenous cardiotonic steroids in humans will be used to study their role in the regulation of gene expression in the kidney and tissue during chronic hypertension, its mechanisms of action, and the structure-activity relationships.

1.6 Overview of this Dissertation

This introductory chapter gives a background on the Na+, K+ ATPase, cardiotonic steroids, the activity of the Na+, K+ ATPase, and an overview of triple quad ICP-MS. The second chapter covers metal ion transport quantified by ICP-MS in intact cells. The method developed in this section is used to measure the activity of the Na+, K+ ATPase in mouse skeletal muscle and human red blood cells to determine the uptake rates for Rb under various experimental conditions. The third chapter covers ICP-MS-MS analysis of microbiological samples with heteroatoms as an internal tag for mass-free quantification of selected elements. The method used in this section was modified to analyze samples with submicrogram weights. The third

chapter also shows how this approach using ICP-MS-MS can be used to analyze physiological functions in various cell types.

The fourth chapter covers the purification of novel endogenous cardiotonic steroids from pig skeletal muscle. There are multiple techniques utilized in this section to purify the OLC samples such as the batch affinity extraction, SEC-HPLC, RP-HPLC, and HILIC-HPLC. Methods were developed for these techniques to ensure optimal purification of the novel CTS product.

The fifth chapter covers the physiological effects of the novel endogenous cardiotonic steroids in pig skeletal muscle and human red blood cells. The sixth chapter addresses the characterization of novel endogenous cardiotonic steroids extracted from pig skeletal muscle.

To characterize the compound, I must determine the molecular weight and structure of the novel CTS compound. I used multiple techniques such as flow injection mass spectrometry and

LC-MS to assess molar weight, and NMR to elucidate the structure. The details of these various methods are discussed in more depth in chapter six.

The work presented in chapters two through six were done in collaboration with Dr.

Heiny, Dr. Tatiana Radzyukevich, Hesamedin Hakimjavadi, and David Cowart from the

Department of Molecular and Cellular Physiology at the College of Medicine, University of

Cincinnati. Other people that I worked with to obtain results for various chapters are Dr.

Kavitha Subramanian Vignesh, Dr. Geogre Deepe Jr., Dr. Balu Addepali, Dr. Robert Ross, and Dr.

Miki Watanabe.

Chapter 2: Metal Ion Transport Quantified by ICP-MS in Intact Cells 2.1 Introduction In all eukaryotic cell membranes, metal ion transport proteins are present and perform essential cell functions. Measuring the tracer flux of the ion transport proteins determines the transport rate of metal ion transporters which represents the enzyme activity. During a standard experiment, the cells being analyzed are incubated in a medium containing the transported ion and a trace amount of 86Rb (radioactive isotope) that is an analog for the transported ion. The transport rate is quantified by the amount of tracer taken up per unit time and scaling it by the molar ratio of tracer to the transported ion. An accurate transport rate per molecule of enzyme is gained by normalizing to the number of transporters in the sample.

Alternatively, the flux in multicellular preparations is normalized to cell number, protein content, or tissue mass. Efflux of ions is a common trait of membrane transporters which is assayed similarly after the cells are loaded with tracer. 86Rb tracer flux measurements are highly sensitive, and a useful quantification can be obtained for minuscule changes in flux with a good signal-to-noise ratio. However, safety and other issues place limits on the experimental design.

In this chapter, I describe a non-radioactive method to measure transmembrane metal ion flux using Inductively Coupled Plasma Mass Spectrometry (ICP-MS). ICP-MS is a reliable and established technique used to quantify metals, metalloids, and some non-metals in various sample types with ease. Some benefits the ICP-MS has is that it has a sensitivity of parts-per- trillion, the wide working range of concentrations and low interferences. ICP-MS was introduced in the 1980s [19,20], and it has evolved to become the most sensitive and versatile tool for elemental detection and quantification. The application of ICP-MS to complex biological

samples has increased due to recent technical improvements that remove spectral interferences and improve reproducibility [21-26].

There are significant advantages for biological applications when ICP-MS is used to measure metal ion transport. The ICP-MS can be utilized in a broader range of physiological studies, and it can detect metal ions for which no viable radioactive isotopes exist. Importantly,

ICP-MS allows simultaneous measurement of multiple elements and isotopes in the same sample in a short period. With this capability, it is possible to monitor simultaneously various metal-dependent physiological processes such as coupled and secondary ion transport. Being able to monitor elements simultaneously can also facilitate the study of transport proteins that do not have identified transport ions. Also, this capability opens the possibility of using internal and external elemental tags to normalize results in small samples, to improve precision.

As proof-of-principle, I measured the transport activity of the Na+, K+ ATPase in intact mouse skeletal muscle and human red blood cells (RBCs). The Na+, K+ ATPase is an essential transmembrane protein and enzyme in all animal cells. It maintains the intracellular and extracellular gradients for Na and K ions that underlie membrane excitation, electrolyte balance, and osmotic pressure; and the Na+, K+ ATPase provides the driving force for the majority of secondary transporters that transfer other ions, nutrients, and essential metal co- factors across the membrane. I measured its transport activity or uptake rate using naturally abundant rubidium such as 85Rb and 87Rb, which is an established analog for K transport by the

Na+, K+ ATPase [27]. To increase precision among sample sets, I used Sulfur content of mouse skeletal muscle and Iron content of human RBCs, measured simultaneously in each sample, as an elemental index of sample mass in place of wet tissue weight. There is an ample amount of

Sulfur in muscle tissue which is present in the cysteine-rich contractile proteins, glutathione, and other proteins. Hemoglobin is the primary iron-containing metalloprotein of RBCs which comprises approximately 35% of RBC wet mass.

2.2 Experimental 2.2.1 Chemicals and Materials All reagents, standards, and chemicals used for this project were trace metal grade

(Thermo Fisher Scientific, USA; or Sigma-Aldrich, USA). Water of 18 MΩ-cm purity (Milli-Q

Academic, EMD Millipore, USA) was used for all solutions. Drinking water certified reference materials were obtained from High-Purity Standards (USA). Elemental standards were obtained from various companies (High-Purity Standards, USA; Claritas PPT, SpexCertiPrep, USA;

PlasmaCal, SCP Science, USA; and Thermo Fisher Scientific, USA). The standard reference material RM 8414 (bovine muscle powder) was obtained from the National Institute of Science and Technology (NIST, USA). Ouabain octahydrate was from Sigma-Aldrich (USA).

C57/B6 mice (The Jackson Laboratory, USA) were used as a source of muscle tissue. The mice were anesthetized (2.5% Avertin, 17 mL/kg) during tissue extraction and euthanized after tissue removal. All procedures were performed in accordance with the Guide For the Care and

Use of Laboratory Animals (National Research Council of the National Academies, USA). The experimental protocols were approved by the University of Cincinnati Institutional Animal Care and Use Committee.

The Equilibration Buffer contained (mM): 118 NaCl, 4.7 KCl, 2.5 CaCl2, 1.2 MgCl2, 1.2

NaH2PO4, 11 d-glucose, 25 NaHCO3; gassed with 95% O2, 5% CO2; pH 7.4, 32°C. The Uptake

Buffer was identical to Equilibration Buffer except that either 4.7 mM RbCl was used in place of

KCl; or, 200 µM RbCl was added as a tracer for KCl. The Wash Buffer contained (mM): 15 Tris-Cl,

2.5 CaCl2, 1.2 MgCl2, 263 sucrose; pH 7.4, 0 – 2 °C. Buffers were prepared fresh and used within one week. Solutions were perfused through the chamber at 2 ml/min. The temperature was maintained to within ± 0.5°C by an in-line solution heater and controller (Warner Instruments,

USA) and monitored by a bath thermistor positioned near the muscle. The content of Rb and S in the equilibrium and was solutions were below the limit of detection by ICP-MS for Rb and 3 ppb for S, while the signal for the samples was in the 2500 and 3500 range. This validates the effectiveness of the wash procedure and allowed us to neglect the trace Rb and S content of the solutions (before addition of exogenous RbCl).

2.2.2 Sample Preparation

2.2.2.1 Mouse Skeletal Muscles

To measure basal Rb uptake in skeletal muscle, the extensor digitorum longus (EDL) muscle was surgically removed and positioned between parallel platinum plate electrodes in a recording chamber perfused with a physiological saline. One tendon was fixed, and the other was attached to an isometric force transducer (BG-50, Kulite Semiconductor Products, Inc.,

USA) connected to a bridge amplifier (TB-4, WPI, Inc., USA). Two EDL muscles were obtained from each animal. Basal Na+, K+ ATPase turnover under steady-state conditions was measured as follows. The muscle was: i) perfused in Equilibration Buffer at 32 °C and stimulated with brief pulses to set muscles length to Lo, the length at which the muscle produces maximal twitch force; ii) rested for 15 min in Equilibration Buffer; iii) incubated for 10 min in RbCl -containing

Uptake Buffer; iv) immediately washed with cold K-, Rb-, and Na-free Wash Buffer (4 x 15

minute immersions with shaking) to stop enzyme cycling and remove excess cations from the extracellular space; and v) gently blotted and frozen in acid-washed Eppendorf tubes for subsequent processing by ICP-MS. Non-specific uptake was determined in separate muscles using the same protocol, but with 1 mM ouabain included in all solutions. Untreated muscles were removed, weighed, and processed directly for determination of S content by ICP-MS. Dry mass was determined after freeze-drying the muscle so the results of wet and dry mass are viable to report.

Muscle samples were subjected to acidic mineralization to oxidize the organic matter, solubilize all metals, and simplify the matrix. The samples were weighed on an analytical balance and placed in acid-washed 10 mL glass digestion vials with Teflon-lined caps, to which 1 mL of 1:1 HNO3:H2O and 0.1 mL of internal standard mix were added. A pre-digestion step was carried out for 1 h at 120 °C using a heating block. Samples volumes were brought to 2 mL with doubly deionized water, and a microwave assisted digestion was performed (CEM Corp., USA).

The microwave program consisted of a ramp to 120 °C for 5 min and a holding time of 30 seconds; followed by a ramp to 190 °C for 5 min and holding time of 7 minutes. Once the samples were mineralized, sample volumes were brought to 10 mL with doubly deionized water. A standard reference material, bovine muscle powder (NIST RM 8414, 0.0070 g – 0.0100 g) was treated identically and assayed to establish the accuracy of the measurement. The concentration of all metals in the reference material was within the NIST range.

2.2.2.2 Human Red Blood Cells

To measure the Rb uptake of human RBCs, eight units of unfrozen, leucocyte-depleted, packed human RBCs preserved with ADSOL (adenine, dextrose, sorbitol, sodium chloride and mannitol) were obtained from the Hoxworth Blood Center, University of Cincinnati. The units were collected on two dates from healthy volunteers of mixed gender and stored at 4 °C for up to 42 days. The units contained an average 7 x 109 RBC/mL and 180 g Hb/L. All procedures were approved by the Institutional Biosafety Committee of the University of Cincinnati. Before experiments, an aliquot was taken, washed to remove plasma, and re-suspended in an Rb-free

RBC Buffer containing (mM): 130 NaCl, 2 CaCl2, 1 MgCl2, 0.5 NaH2PO4, 10 d-glucose, 12 NaHCO3,

10 HEPES; pH 7.4 at 37 °C, 290 mOs/kg H2O. The wash consisted of 3 cycles of dilution in 5 volumes of buffer followed by centrifugation at 500 x g for 10 min at 4 °C and removal of supernatant. After the final wash, the sample was resuspended in the buffer at a hematocrit of

0.5.

Uptake Buffer was prepared by adding 240 µM RbCl to the RBC Buffer. Rb uptake was measured under 4 conditions in the tubes containing: 150 µl of washed and suspended RBCs,

50 µL of either 10 mM ouabain (1 mM final concentration) or H2O, 800 µL of Uptake Buffer, and

0 or 4 nCi/mL 86Rb. Uptake was carried out for 2 h at 37 ± 1 °C. Each condition was measured in triplicate, and all samples were processed in parallel.

After the incubation, samples were washed 3 times as described above to remove

86 extracellular Rb and Rb. After the final wash, the sample was resuspended in H2O and taken either for counting, or determination of Rb by ICP-MS. 86Rb activity was determined by liquid

scintillation counting of beta decay (Filter-Count, Perkin-Elmer, USA; LS-6500 Liquid Scintillation

Counter, Beckman, USA). Aliquots of Uptake buffers without and with 86Rb were taken in each experiment and used to obtain a calibration factor in CPM/nMole of Rb. ICP-MS determination of Rb and Fe concentrations was carried out as described for skeletal muscle samples.

The content of Rb in the initial, nominally Rb-free, RBC buffer and the final supernatant was below the limit of detection by ICP-MS for Rb, while the signals for the samples were in the

12,000 ppb range. This validates the effectiveness of the wash procedure and allowed us to neglect the trace Rb content of the RBC buffer.

For RBC analysis, the final volume after the Rb uptake experiment was digested. For this, the samples were analytically transferred to a glass digestion vial by using two portions of 1 mL of doubly deionized water to rinse the original vial. All transferred volume was digested after adding 1 mL of concentrated trace metal grade nitric acid and 0.4 mL of internal standard mix.

The mixture was heated to 140 °C for 5 h in a dry heating bath. 250 µl of H2O2 were added to complete the digestion and heated for 1 h at 140 °C. After complete mineralization, the samples were brought to 40 mL by adding doubly deionized water before the analysis by ICP-

MS.

2.2.3 Instrumentation

The selected elements - 23Na, 24Mg, 31→47P, 32→48S, 39K, 43, 44Ca, 55Mn, 56Fe, 60Ni, 63Cu, 66Zn,

85Rb, and 208Pb were determined by ICP-MS by the external calibration method, using an Agilent

8800 inductively coupled plasma mass spectrometer triple quadrupole (ICP-MS QQQ, Agilent

Technologies, USA equipped with a micro concentric nebulizer, (Meinhard MircroMist,

Meinhard, USA), a Peltier cooled double pass spray chamber, standard torch, and autosampler.

Data analysis was performed using Mass Hunter workstation version 4.1 of ICP-MS QQQ software (Agilent Technologies, USA) to determine the total concentration of elements in the tissues and solutions. Two tune modes were used sequentially to ensure proper ionization and interference removal. Table 2.1 shows the different instrument parameters utilized for each gas mode. Internal standards for Sc, Y, In, & Ce were used to represent the full mass range. Results are given as the mean and RSD of measurements from independent samples.

Table 2.1: Instrument tune parameters for ICP-MS QQQ.

Instrument calibration curves for detection of Rb in a NIST standard sample of bovine skeletal muscle were linear with a regression coefficient of 1.0000 in the range of 0.1-400 ppb.

In no gas mode, the instrument limit of detection is 1 femtogram ml-1 or 6 ppt. This implies a detection limit of 6 femtograms or 0.07 femtomoles of Rb per milligram wet tissue. The calibration was linear for all detection modes.

The ICP-MS signal for Rb is not compromised by isobaric interferences, but this is not the case for S. Nitrogen, oxygen and hydrogen-based polyatomic interferences from atmospheric plasma represent a major challenge for direct analysis of S at its most abundant isotope. High-resolution instruments can be used to solve this problem, but they are costly and not as robust as the quadrupole based low-resolution instruments. Helium is commonly used to remove the majority of the 34S interferences, but a significant loss of sensitivity is observed. A reaction with oxygen before the quadrupole has also been employed, as the polyatomic interferences do not react with oxygen, but the reaction is matrix-dependent and the analytes at the oxidized S product become interferents. A new approach was developed in early 2013 by

Agilent Technologies which consists of the addition of an extra quadrupole that allows filtering out all ions except the m/z = 32 (S, interferences), and then reacting them with oxygen to filter out the original m/z, which is now conformed by the un-reacted polyatomic interferents. This new technology allows the quantification of S with less interference without signal loss, resulting in lower limits of detection and background signal. Although we used this technology for the Rb transport assay described here, we note the very intense signals of S in skeletal muscles of the studied size would also allow the use of helium in collision mode to remove S interferents.

2.2.4 Statistical Analysis

Origin 8.5 was used to perform all statistical analysis. The comparisons between muscle samples, un-treated EDL vs. TA, were performed with a t-test. The comparison of Rb values between Rb treated, and Rb + Ouabain treated samples were made with a t-test, and showed a p < 0.01 with a power ≥0.97 for the reported sample sizes.

2.3 Results and Discussion

In this study, 85Rb was used, from natural abundance Rb, to measure the activity of the

Na+, K+ ATPase in mouse skeletal muscles and human red blood cells using ICP-MS QQQ. The instrument must be calibrated for Rb as well as other elements of interest before any sample analysis can be conducted. Figure 2.1 shows the calibration for Rb obtained by ICP-MS QQQ.

The instrument's response gives a correlation coefficient or R-value of 1, and the detection limit is approximately 2 ppt. These values prove that the ICP-MS QQQ is capable of quantifying trace amounts of Rb present within treated samples. Phosphorus, sulfur, and iron are three elements that were also important to monitor because they are needed to normalize the Rb concentration for each sample or sample set. Without phosphorus, sulfur, and iron the rubidium concentration or uptake rate for each sample set would be based on a wet mass, which gives a larger variance within the same sample groups. The discrepancies occur due to rinsing, freezing, and dehydration of the samples which makes it hard to keep the water content constant in sample groups. To correct the inconsistencies obtained from the wet mass, we found a commonality from sample to sample based on the P, S, and Fe content within each sample. With the correction, different sample sets can be compared to each other.

Figure 2.1: Rubidium Calibration Curve Obtained from ICP-MS QQQ.

2.3.1 Reference values for endogenous Rb content in full skeletal muscle

Determination of Rb concentration in muscle tissue was possible with a high signal-to- noise ratio and detection levels at low as sub-parts per billion or sub-nanograms per gram of tissue. Reference values for the endogenous Rb concentration in untreated skeletal muscles was 7,112.3 ± 193.3 ng/g (n=20) in the mouse extensor digitorum longus muscle (EDL) and

7,371.0 ± 796.3 ng/g based on the wet mass of the tissues (n=10) in the tibialis anterior (TA) muscle. These values were not significantly different, and therefore reference levels could be determined from various muscles in the same animal when needed. These values for mouse muscles are in the range of reference values for the Rb content of bovine skeletal muscle (NIST

RM 8414) and were subtracted from measurements of the muscle Rb content taken up during active Na+, K+ ATPase transport.

2.3.2 Equimolar replacement of RbCl for KCl to measure Na+, K+ ATPase activity

Next, we measured the amount of Rb taken up by quiescent (non-contracting) EDL muscles during a 10 min incubation in a modified physiological buffer in which KCl was replaced on an equimolar basis with 4.7 mM RbCl (Table 2.2). Rb transport rate by the Na+, K+ ATPase, computed as the ouabain-sensitive component and normalized to wet muscle weight, was

322.0nMol Rb/g-min at 32 °C. Ouabain is an established, specific inhibitor of Na+, K+ ATPase transport. The relative standard deviation (RSD) was 35.9% in control and 19.0% in the presence of the inhibitor, ouabain. This transport rate is close to published values for resting transport by the Na+, K+ ATPase in mouse EDL obtained from measurements of 86Rb flux [28].

The RSD is attributed to animal-to-animal differences and the difficulty of getting an accurate wet mass of samples weighing only 5–12 mg.

Table 2.2: Rb uptake by the Na+, K+ ATPase in quiescent mouse EDL muscle measured by ICP- MS using equimolar replacement of RbCl for KCl in the uptake buffer. Rb transport rate was measured using a 10min incubation time in physiological saline containing 4.7mM RbCl in place of KCl, and nominally 0mM KCl, at 32 °C. Non-specific Rb uptake contributed by all other K and Rb transport pathways was measured in independent muscles using 1mM ouabain. Using S- based muscle mass, non-specific Rb uptake was 26.2% of total uptake and was subtracted to obtain net ouabain-sensitive uptake by the Na+, K+ ATPase. Relative standard deviation, RSD. Rb content was normalized either to weighed wet tissue mass or the S-based mass.

2.3.3 Muscle sulfur content as an index of muscle mass

We tested whether we could reduce the RSD of the measurements by using the muscle

Sulfur content as an index of muscle mass. To calculate muscle mass based on Sulfur content, a pooled sample of 20 untreated EDL muscles having a combined wet weight of ~200mg was used to obtain the average S concentration, average wet weight obtained by weighing the untreated samples, and average dry mass obtained after freeze drying. The average S-to-wet mass ratio of the mouse EDL was 5.49 ± 0.25 mg g−1. Using the average S content and an accurate value for the average EDL wet mass, a ratio-to-mass factor, Rb/S, was calculated and used to express the Rb concentration based on S content (S-based mass). The total S content of each sample correlated exactly with the wet mass of the untreated tissues, the same

correlation was maintained for EDL and TA and was independent of sample treatment (Figure

2.2a). The S-based mass also correlated linearly with the weighed wet mass of untreated samples (Figure 2.2b). However, the slope of 0.80 indicates a 20% decrease when S is used to calculate the sample mass. Muscle is a multicellular tissue, and the difference represents the mass contributed by fluid retained in the extracellular spaces. The S concentrations were identical whether measured using a 10 min or a 1 min incubation (data not shown), indicating that S is not actively transported into or out of the muscle. These results validate the use of S- based mass, obtained from a simultaneous measurement of the S content of each sample, for normalization. When this normalization is applied to the same samples, the Rb concentrations show a much smaller variation between samples (Table 1, S-based mass). The RSD decreased to

8.0% and 5.2% for control and ouabain-treated samples, respectively. These values are within the expected animal-to-animal variability, indicating that the measurement using S-based mass is not limited by weighing errors. Therefore, the mass determined from S content gives a more precise determination of Rb content with less variability and greater statistical power than obtained using weighed wet mass. Notably, the net transport rate obtained using S-based mass increased to 399.1 nMolRb/g-min at 32 °C, indicating that transport activity obtained using wet mass underestimates the actual transport rate of Na+, K+ ATPase for Rb by ~24%.

Figure 2.2: The S content of mouse skeletal muscles is an accurate index of the tissue mass. (a) Total S content of each sample vs. weighed mass. Different symbols represent different muscles and conditions. (Blue Squares), EDL muscles incubated for 10min in physiological saline containing 4.7mM RbCl in place of KCl; (Red Circles), EDL muscles incubated for 10min in the same solution+1mM ouabain; (Green Triangles), EDL muscles incubated in 200μM RbCl and 4.7mM KCl; (Red Diamonds), EDL muscles incubated in the same solution+1mM ouabain; ο, untreated EDL muscle removed from the animal without incubation; (White Triangles), untreated TA muscles. Slope= 3.82× 103, correlation coefficient= 0.9868 (b) Muscle mass computed from the S content of each sample (S-based mass) vs. weighed mass. Slope= 0.786, correlation coefficient= 0.987.

2.3.4 Tracer RbCl used to measure Na+, K+ ATPase activity

Given the low detection limits of ICP-MS for Rb and our low blank contents for it, we investigated whether Na+, K+ ATPase transport could be measured at physiological concentrations of extracellular K, using Rb as a tracer ion (Table 2.3). Total K uptake determined using a mole fraction of 0.2 mM RbCl to 4.7 mM KCl was 685.0± 76.6 (n= 12) ng/g, with an RSD of 11.2%. The slightly increased RSD is due to the much lower Rb concentration in the uptake buffer compared to measurements with equimolar RbCl. Non-specific uptake in the presence of ouabain was 88.4± 11.3 (n= 10), with an RSD of 12.8%. However, non-specific

uptake of tracer Rb is less than the reference amount in untreated muscles and therefore unreliable. A more reliable estimate of nonspecific tracer Rb is the average fractional non- specific uptake obtained using equimolar Rb (26.2% of total uptake; from Table 2.2). Net ouabain-sensitive K transport rate by the Na+, K+ ATPase obtained using tracer Rb, and physiological K concentration is 505.5 nMolK/g-min. Notably, this rate is 27% greater (P=0.01) than that obtained using equimolar Rb as a congener for K, and suggests that the Na+, K+

ATPase in muscle may transport K more efficiently than Rb.

Table 2.3: Rb uptake by the Na+, K+ ATPase in quiescent mouse EDL muscle measured by ICP- MS using 200μM RbCl as a tracer. Rb uptake was measured using a 10 min incubation time in Uptake Buffer containing 4.7mM KCl and 200μM RbCl, at 32C. Ouabain-sensitive Na+, K+ ATPase transport was obtained as described in Table 1. The amount of Rb taken up by the muscle was scaled by the molar ratio of K to Rb in the Uptake Buffer (23.5) to obtain the transport rate of the Na+, K+ ATPase for K. NET ouabain-specific transport by Na+, K+ ATPase was obtained after subtracting the fractional non-specific uptake computed as 26.2% of total uptake (from Fig 1) as described in text.

2.3.5 Na+, K+ ATPase transport in human RBCs analyzed by ICP-MS

To quantitatively compare Na+, K+ ATPase transport rates obtained from ICP-MS and

86Rb flux in the same preparation, and to test the broader applicability of the method, we

measured Rb uptake in human red blood cells by both techniques. RBCs are an established model for studies of Na+, K+ ATPase kinetics, and transport cycle [29]. RBCs have a 30–40-fold lower membrane density of Na+, K+ ATPase than mammalian skeletal muscles and thereby provide a useful model to test the limits of detection of the method. Additionally, use of a homogeneous suspension of single cells allows a direct comparison of wet weight and Fe-based weight without the dilution effect of retained interstitial water. The endogenous Rb concentration of untreated RBCs was 2,204.7± 60.8 ng/g (n = 8 independent units of blood), with an RSD of 2.76%, and it was subtracted for calculations of Rb uptake by the Na+, K+ ATPase.

The ouabain-sensitive K transport rate of RBCs obtained by ICP-MS under conditions of physiological K concentration and tracer RbCl was 16.0 nMolK/g-min, referred to weighed wet mass (Table 2.4). A more precise determination of transport rate was obtained by ICP-MS using the Fe content of RBCs as an index of cell mass. The average Fe content of 8 units of untreated

RBCs having an average Hb concentration of 180 g/L was 5,684.14± 276.51 ng/g. The total Fe content of each sample correlated precisely with the mass of the untreated RBCs. The average concentration of Rb was not statistically different when calculated by Fe content.

Na+, K+ ATPase transport rate measured by ICP-MS-MS using the Fe-based mass of each sample was 18.6 nMolK/g at 37 °C, and the RSD decreased to only 1.6% (Figure 2.3 and

Table 2.4). The significantly higher precision obtained using Fe-based mass is attributed to having both the Fe-based mass of the pooled sample before any treatment and the Fe content of the samples after incubation in uptake buffer. Any loss of cell mass by hemolysis or pipetting during the assay is corrected for since Fe is lost in proportion to cell number; while the Rb

content retained in the cells after washing and subtraction of the Fe-based endogenous Rb content reflects Rb taken up by enzyme transport of the remaining intact cells.

The transport rate measured in parallel using tracer 86Rb was 11.5 nMolK/g-min based on wet weight. The lower transport rate obtained using tracer 86Rb is mainly attributed to inaccuracies in collecting an accurate weight for normalization since, in this case, we have only the initial wet weight of the RBC suspension without any correction for cell loss during the assay. The accuracy of the 86Rb tracer measurement could be improved using a larger sample size, an 86Rb lot with higher activity, and counting gamma emission directly to avoid sample transfer into scintillant.

Table 2.4: Comparison of Rb transport by human RBCs measured by ICP-MS and 86Rb tracer. All measurements were run in parallel under identical conditions using the same pooled batch of RBCs. The batch was pooled from 4 units of leucocyte-depleted RBCs, washed, and resuspended in Rb-free buffer at a hematocrit of 50%. The incubation mix contained 150μL RBCs, 5mMK, 200μM RbCl, and Uptake Buffer without or with 1mM ouabain, in a volume of 1ml. Radioisotope tracer assays contained, in addition, 4nCi/ml 86Rb. Uptake was carried out for 2h at 37C. K transport rate was computed by multiplying the Rb transport rate by the molar ratio of K to Rb in the Uptake buffer. The endogenous Rb concentration of the pooled, untreated RBCs was 2,374ng/g and was subtracted in all calculations of exogenous Rb uptake.

Figure 2.3: Rb taken up by RBCs in the absence and presence of 1mM ouabain, measured by ICP-MS and referred to Fe based mass. The RSD for each group was 1.6%.

2.4 Conclusion

Accurate measurement of cellular levels, distribution, and flux of metal ions and metal- containing proteins and metabolites is crucial to understanding physiology in both health and disease. Here we describe a method to measure metal ion transport by the Na+, K+ ATPase in small biological samples using ICP-MS. The method was validated in mouse skeletal muscle, a multicellular tissue, and in suspensions of human red blood cells. The capability of ICP-MS to measure multiple metal ions in the same sample allowed us to use the S content of muscle or the Fe content of RBCs as an elemental index of tissue mass, to increase precision and statistical power.

The K transport rate of Na+, K+ ATPase measured by ICP-MS in resting mouse EDL muscle was 506 nMolK/g-min at 32 °C, measured with tracer Rb using muscle S content as a mass index. For mouse EDL muscles, which weigh less than 12 mg, conventional normalization to wet weight overestimates muscle mass and thereby underestimates Na+, K+ ATPase transport rate by about 20%.

A direct comparison of Rb content measured by ICP-MS and tracer 86Rb in human RBCs further validated the method in a cell suspension without interstitial spaces. For microliter volumes of RBCs, the highest precision and accuracy was obtained using ICP-MS with Fe-based mass.

The Na+, K+ ATPase transport rate of human RBCs under our conditions was 18.6 nMolK/g-min at 37 °C. The low transport rate of human RBCs reflects the low density of Na+, K+

ATPase in RBC membranes. The specific membrane density of Na+, K+ ATPase in human RBCs is

1–2 per μm2 [30], compared to 200–900 μm2 in mammalian skeletal muscle sarcolemma [27]

(assuming a 4–5-fold higher t-tubule membrane area than outer sarcolemma). This transport rate is equivalent to a specific pump turnover rate of 1300–3000/min. The major Na+, K+ ATPase isoform of human RBCs is the alpha1 isoform. This turnover rate is comparable to measurements in other isolated or cultured cell types which predominantly express the same alpha isoform [31,32].

Collectively, these results validate the use of ICP-MS to measure ion transport by the

Na+, K+ ATPase, a vital metal ion membrane transport protein, in small biological samples. Key advantages of ICP-MS are that it allows a broader range of experimental designs and

physiological contexts than possible using radioisotopes, and the capability of ICP-MS to measure multiple metal ions in the same sample provides a more accurate index of cell mass for normalization. In future studies, this ability can be exploited further to study the movement of multiple metal ions whose transport is inter-related by coupled exchangers or secondary transporters.

Investigations into the influence of metals on molecular mechanisms remains an unexplored area of research. The ICP-MS method described here for quantifying metal ion transport by the Na+, K+ ATPase is easily extended to measurement of other metal ions and transporters and is widely applicable to the analysis of other metal ion-dependent physiological processes. It is applicable to a range of cell types with appropriate adaptation of sample preparation.

Acknowledgments

This work was supported by the National Institutes of Health, NIAMS grant RO1-

AR063710. The authors thank Dr. Joseph Caruso for helpful critiques of the work and manuscript.

Chapter 3: ICP-MS-MS Analysis of Biological Micro Samples with Heteroatoms as Internal Tag

for Mass-Free Quantification of Selected Elements

3.1 Introduction

Inductively coupled plasma mass spectrometry (ICP-MS) has evolved as a sophisticated technique for elemental analysis at trace and ultra-trace levels[33, 34]. The source of its remarkable advantages over other elemental detectors is the high-energy plasma ionization source, which offers robustness and low matrix-dependent effects. This advantage coupled with an efficient sample introduction system and a mass spectrometer provides excellent detection power for complex matrices[34].

There is a need to measure trace elements and metals in small biological samples [35,

36]. Trace elements are ubiquitous in cells and are required for diverse cell functions. However, the application of ICP-MS to elemental analysis of biological samples for biomedical research remains a mostly unexplored area of investigation. This application poses unique challenges.

Spectral interferences from polyatomic and doubly charged ions affect the detection of elements of interest like P, S, Se, Fe, and Cr; and most biological samples are too small for trace level detection using conventional approaches.

Most commercial ICP-MS instruments are equipped with a quadrupole as a mass analyzer and use liquid sampling introduction systems. A dissolution or mineralization of solid samples is therefore needed to make the sample homogeneous, reduce the matrix concentration, and facilitate calibration with liquid standards.

A quadrupole as mass analyzer delivers the ion transmission and ion capacity required for a broad and linear quantification range, and its use, calibration, and maintenance is relatively simple and robust. The major spectral interferences in a low-resolution mass analyzer like a quadrupole are polyatomic ions and doubly charged ions. These interferences are mainly formed from atmospheric gasses like nitrogen, oxygen, and carbon dioxide, as well as from abundant elements within the sample like carbon, chlorine, and sodium. Different instrumental approaches have been developed to overcome these spectral interferences. The addition of a collision reaction cell before the mass analyzer is now standard practice [37, 38]. While this is useful for the analysis of many metallic elements, metalloids, and non-metals suffer from poor ionization and are more challenging to detect in trace, and ultra-trace levels under high helium flow in the collision reaction cell. Another approach to overcoming interferences is the use of a high-resolution mass analyzer that can distinguish the small mass differences between monoatomic ions and polyatomic interferences[39]. These instruments are large, expensive, and require more specialized training while offering less robustness for routine analysis. Such high-resolution instruments are primarily used for isotope distribution assessments in larger sample sizes, where precision and mass accuracy are required[40].

A typical ICP-MS instrument is designed to analyze liquid samples of 0.5 to 2 ml of total sample volume in a continuous flow mode. The addition of flow injection sample introduction devices (FIA) can decrease the sample volume required to 0.01 ml; however, sample preparation needs to be improved to maintain the accuracy of the analysis for samples in the microliter final dilution range[41].

The analysis of trace element concentrations in biological samples by ICP-MS poses additional, unique problems. The sample size needed to quantify base on the sample mass at the trace level is often not available. For example, a typical cell number required for mass- based concentration analysis of cultured cells is in the range of 106 – 109 cells. Many tissue types, biopsies, and forensic or clinical specimens are significantly below this range. Water content is another confounding variable, especially in tissues and cell preparations that require freezing before analysis or contain a variable amount of rinsing volume.

For these reasons, many biomedical methods rely on the use of radioactive isotopes instead of spectroscopic methods to track changes in elemental composition or transport[42].

Scintillation counters that detect particle emissions from the radioactive isotopes give good comparative measurements between sample groups. While the advantages of radiolabeled elements include high signal-to-noise ratio and specificity without the need of complex sample processing, the health hazards associated with the use of radioactive materials impose severe restrictions in experiment design. The possibility of using naturally abundant isotopes or non- radioactive labeled isotopes by ICP-MS can eliminate these restrictions. Also, ICP-MS can be used in a broader range of physiological contexts and can detect metal ions for which no feasible radioactive isotopes exist, with multi-isotope and multi-element capabilities.

In order to perform elemental analysis by ICP-MS on small tissues or cell cultures, two alternatives are commonly used: a solid sampling interface, with laser ablation or laser microdissection as sample introduction in front of the ICP-MS[43-45]; and microfluidic sample introduction systems for small discrete sampling with direct ICP-MS analysis[46,47]. Both

approaches offer the possibility of semi-quantitative analysis and have been successfully applied to many sample types. Metal profile mapping of thin slices of brain tissue obtained by laser ablation, LA-ICP-MS, has been used successfully to visualize colocalization of elements and their distribution across the sections ([44,45]). However, the ablation mechanisms and efficiencies are matrix-dependent, and the use of an internal standard is challenging.

Moreover, this approach is not suitable for cell culture analysis as the cell density homogeneity becomes a significant source of instrumental noise.

Microfluidic devices for analyzing micro biological samples has been used to analyze single cells by mass spectrometry. Typically, a diluted sample is introduced into micro-channels and pushed to the ionization source by pneumatic “micro cannons” to ensure the introduction of cells in a single or low number mode [48-50]. The interface of a microfluidic device for cell analysis with ICP-MS requires modifications not available commercially as a standard package and often requires fine adjustments for a particular sample application. Also, the instrument detector needs to operate at a maximum speed to catch and process individual sample events, often degrading performance. Beyond the technical difficulties of a prototype design for a microfluidic sample introduction system, inherent problems like cell bursting at the ejection device (causing incomplete sample introduction and memory effects), lack of standards or internal standards for quantification, the reduction of multi-elemental capabilities of the detector to a single isotope of a particular element in order to obtain useful information, and the necessity of data processing to evaluate results all pose significant obstacles yet to be overcome.

To take advantage of the multi-elemental and matrix-independent capabilities of ICP-

MS, we previously explored the use of internal elemental tags as a mass index. Although the number of candidate elements for this is not large, phosphorus, sulfur, and iron are abundant and stable in concentration and were successfully used for large amounts of sample in DNA, muscle and red blood cells ([51,52]). This approach, combined with the implementation of a triple quad ICP-MS (ICP-QQQ or ICP-MS-MS), brought the analysis of elements typically problematic by ICP-MS to a new level of accuracy and robustness.

In the present study, we extend this approach to trace element detection in micro-scale biological samples after micro mineralization and liquid sampling analysis. Sulfur, phosphorus, and iron provided a reliable index for normalization of element content. An accurate mass- based concentration was possible when enough material was available to obtain an element-to- mass ratio for the particular cell type. The method retains the advantages of standard in- solution sample digestion while extending the application to microscale sample sizes. We could quantify Zn in culture samples of H. capsulatum yeast containing as little as 103 cells, and Rb in samples of 1-5 freshly isolated skeletal muscle cells.

Significant advantages are gained from the use of smaller sample sizes. For the case of primary cell cultures or live cells or tissues, a lower number of cells per sample allows more biological replicates with fewer organisms or rare transgenic animals, even down to a single organism, thereby reducing animal-to-animal variation and decreasing the cost of the study.

This is particularly relevant when transgenic animals are used or when a purified cell population is too small for bulk analysis.

3.2 Experimental

3.2.1 Chemicals and Materials

All solutions were made using double deionized water of > 18MΩ per cm and trace metal grade chemicals from Fisher Scientific (Hampton, NH). Nitric acid used for micro digestions was optima grade, Fisher Scientific (Hampton, NH). All vials used were premium grade clear polypropylene, Fisher Scientific (Hampton, NH), acid-washed (10% nitric acid) and dried before use.

3.2.2 Sample Preparation

3.2.2.1 Culture & Quantification of RAW Macrophages

RAW 264.7 macrophages (ATTC) were cultured at 37°C in a CO2 incubator to confluent monolayers in RPMI (BioWhittaker, MD) complete media containing 10% fetal bovine serum

(HyClone Laboratories, Utah), and gentamycin sulfate (10 mg/L). Cell monolayers were gently scraped using a rubber policeman and transferred to tubes, followed by centrifugation at 1600 rpm for 5 min. The media supernatant was discarded, and the pellet was washed twice in 0.9% saline prepared with double deionized (DDI) water. RAW 264.7 cells were resuspended in 0.9% saline, counted using a hemocytometer, and a specified number of cells were transferred to

Eppendorf tubes that were previously washed with 10% nitric acid, rinsed with DDI water and dried. Cells were centrifuged at 1600 rpm for 5 min; the supernatant was gently removed by pipetting, and macrophage pellets were frozen at -80°C before metal analysis.

3.2.2.2 Culture & Quantification of Yeasts

Histoplasma capsulatum strain G217B yeasts were grown in HAM’s F12 media (Sigma-

Aldrich) to log phase at 37°C in a shaker. Yeasts were harvested by centrifugation at a low speed of 300 rpm for 5 min to remove cell clumps. The pellet was discarded, and yeasts in the supernatant were centrifuged at 1600 rpm for 5 min. The supernatant was discarded, and the yeast pellet was washed twice with 0.9% saline in DDI water. The pellet was resuspended in

0.9% saline, and H. capsulatum yeasts were counted using a hemocytometer. A specified number of yeast cells were then transferred to acid-washed Eppendorf tubes and pelleted using a microcentrifuge at 13,000 g for 5 min. We used Eppendorf tubes that were previously washed with 10% nitric acid and dried. Supernatant saline was removed gently by pipetting, and yeast pellets were frozen at -80°C before metal analysis.

Blastomyces dermatitidis strain 26199 yeasts were cultured in HAM’s F12 media to log phase at 37°C in a shaker. Yeasts were harvested by centrifugation at a low speed of 200 rpm to remove cell clumps. The pellet was discarded, and the supernatant was passed through a 40

µM nylon filter to eliminate aggregated yeasts. The filtrate was centrifuged at 1600 rpm for 5 min; the supernatant was discarded, and yeast pellet was washed twice in 0.9% saline in DDI water. After washing, B. dermatitidis yeasts were prepared for metal analysis using the same procedure as described for H. capsulatum.

3.2.2.3 Dissociation & Quantification of Single Muscle Cells (Fibers)

Animals

C57/B6 mice (The Jackson Laboratory, USA) were used as a source of muscle tissue.

Mice were anesthetized (2.5% Avertin, 17 mL/kg) during tissue extraction and euthanized after tissue removal. All procedures were performed in accordance with the Guide For the Care and

Use of Laboratory Animals (National Research Council of the National Academies, USA) and were approved by the Institutional Animal Care and Use Committee of the University of

Cincinnati

Single fiber dissociation

The extensor digitorum longus (EDL) muscle was surgically removed from an anesthetized mouse and incubated for 90 – 110 min at 37 C under 5% CO2 in a collagenase solution (MEM media containing 1.5 mg/mL Type I collagenase (Sigma), 4.5 mg/mL Type II collagenase (Worthington), and 10% v/v horse serum). Following collagenase treatment, muscle fibers were washed in standard Tyrode’s solution (mM: 140 NaCl, 5.5 KCl, 10 HEPES free acid,

5.5 D-Glucose, 2 MgCl2, 2.5 CaCl2, pH 7.2) and further dissociated by trituration with fire- polished glass Pasteur pipettes.

Rb Uptake

Two to four dissociated muscle fibers were transferred to a multi-well plate containing standard Tyrode’s solution and equilibrated for 10 min at 37°C. The criteria used to select viable fibers were: (a) A secure attachment of a single cell to the well surface. (b) A straight, translucent, continuous, and smooth surface devoid of kinks. After equilibration, ouabain (final concentration: 2mM) or Salbutamol (final concentration: 10 M) was added to the control and test wells, respectively. After 5 minutes, RbCl (final concentration: 0.5 mM) was added to the wells to start the uptake of Rb by myofibers. After 10 minutes of incubation in Rb uptake solution, cells were washed 10 times with wash solution containing (mM): 15 Tris-Cl, 2.5 CaCl2,

1.2 MgCl2, 263 Sucrose, pH 6.8, at 0 C.

3.2.2.4 Sample Preparation of Total Metal Analysis

The use of an internal elemental tag for normalization allows a flexible sample preparation protocol if high purity reagents are used. Nevertheless, the dilution factor applied and the final acid content needs to be within the compatible range for ICP-MS analysis. This study analyzed two groups of samples - small cultured cells and relatively large myofibers. The dilution factors were calculated and prepared to give an instrument signal above the quantification limit for the elements of interest and reflected the difference in size of the cells analyzed.

For macrophage and yeast cells, the mineralization design was based on the cell number of each sample. These samples were placed in metal-free 1.5 ml centrifuge tubes with 1:1

HNO3: H2O and internal standard mix. Given the extended range of cell numbers used in each sample (104 - 109 cells) the amount of nitric acid and internal standard mix was scaled to a dilution factor range of 400 – 200,000 cells per microliter for H. capsulatum, 50 – 100,000 cells per microliter for B. dermatitidis and 40 – 4,000 cells for raw macrophages. After addition of acid and internal standard, the samples were placed on a heating block for 2 h at 95 °C. Once the samples were digested completely, two final volumes were used, 250 µl or 500 µl, depending on the cell count and cell type by adding doubly deionized water.

For myofibers, from 1 to 10 dissociated myofibers were transferred to acid-washed, metal-free 1.5 ml clear centrifuge tubes or metal-free polypropylene 96-well plates (Agilent

Technologies) containing 200 L 1:1 HNO3: H2O and 20 L of the internal standard mix. Samples were digested by heating in an oven overnight at 60 °C. After digestion, sample volumes were brought to a final volume of 500 µl with DDI water. This protocol yields a dilution factor of

0.006 to 0.02 cells per microliter.

3.2.3 Instrumentation

The microliter volumes of digested samples precluded the use of a conventional auto sampler. Two approaches were evaluated for sample introduction into the ICP-MS nebulizer - a manual handling of the intake tubing without varying the uptake speed rate, and the use of a micro well plate reader probe in the auto sampler under traditional uptake and rinse program.

For manual sample uptake, the samples were introduced by using a 20 cm tubing made of inert perfluoroalkoy (PFA) with an internal diameter of 0.50 mm connected to the peristaltic

pump tubing and the microconcentric micro mist nebulizer to the Agilent 8800 ICP-MS-QQQ.

The peristaltic pump was used at analysis speed (0.1 rps) equivalent to approximately 0.3 ml min-1 for sample uptake, analysis, and rinsing. The rinsing program was manually performed by dipping the uptake tube into DDI water followed by 10% nitric acid and finally DDI water before analyzing the next sample. Under this conditions, the sample uptake time was approximately 15 seconds, after which the instrumental signal was stable. 220 l was the minimum sample volume required for this method of introduction.

For the second method, a well-plate tray adaptor kit for the SPS 4 auto sampler (Agilent

Technologies) was used. This accessory allows the sampling from 96 well plates of 500 l volume per well. The kit includes well plates, well plate adapter and a sample probe of 0.25 mm of internal diameter and 70 cm in length. The peristaltic pump tubing had 0.25 mm internal diameter, and the total tubing volume was approximately 65 µL. A small volume nebulizer is recommended for this application; we used a Meinhard high-efficiency low-flow concentric nebulizer (Meinhard HEN-90-A0.03) rated for 30 l/min at 95 PSI. The nebulizer was used with the peristaltic pump at 0.25 rps for a flow of approximately 60 µL min-1. The uptake and rinsing program was carried out under the same flow rate. A pre-emptive rinsing program was used to start the rinsing 40 secs before the end of the acquisition to decrease the analysis time. Under this configuration, the volume of sample used for each analysis was less than 180 µl, with the added advantage of a fully automated sampling and rinsing.

The selected elements – 23Na, 31→47P, 32→48S, 39K, 43, 44Ca, 56Fe, 63Cu, 66Zn, & 85Rb were determined by ICP-MS using the external calibration method from 0.01 to 250 ppb, on the

Agilent 8800 inductively coupled plasma mass spectrometer triple quadrupole (ICP-MS QQQ,

Agilent Technologies, USA) equipped with a micro concentric nebulizer or an high-efficiency low-flow nebulizer (Meinhard MicroMist, Meinhard, USA), a Peltier cooled double pass spray chamber and standard torch. Data acquisition and analysis were performed using Mass Hunter workstation version 4.1 of ICP-MS QQQ software (Agilent Technologies, USA) to quantify elemental composition in the micro-samples after mineralization. The instrument was tuned using a solution containing a mixture of Li, Co, Y, Ce and Tl at 1 ppb each. For the two used nebulizers, the carrier and uptake gas were optimized, while the parameters for the lenses were consistent in both modes. Three tune modes were used sequentially to ensure proper ionization and interference removal (Table 3.1 & 3.2). As internal standard Sc, Y, In, & Ce were used to represent the full mass range. Data acquisition was performed in spectrum mode, with

0.15 s of integration time, four replicate readings, and one-point sampling.

Table 3.1: Instrument tune parameters for ICP-MS-QQQ using micro concentric nebulizer.

No Gas mode He mode O2 mode

Forward Power 1550 W 1550 W 1600 W

Nebulizer Gas Flow 1.00 L/min 1.00 L/min 1.00 L/min

Extract 1 2.0 V 0.5 V 0 V

Extract 2 -180.0 V -165.0 V -200.0 V

Isotopes Monitored 23Na, 39K, 45Sc, 43Ca, 44Ca, 56Fe, 63Cu, 66Zn, 31→47P, 32→48S, &

& 85Rb 45Sc & 115In 89→105Y

-1 Cell Gas Flow He: 3 ml min O2 gas flow: 30%

OctP Bias -8.0 V -18.0 V -5.0 V

Typical counts for 1 59Co: 7,500 59Co: 3,800 59Co: 5,400 ppb at 0.1s integration 89Y: 14,500 89Y: 4,500 89→105Y: 9,700 140Ce: 12,500 140Ce: 7,100 205Tl: 4,800 140→156Ce: 10,400 205Tl: 7,900 205Tl: 6,700

Table 3.2: Instrument tune parameters for ICP-MS-QQQ using a high-efficiency low-flow nebulizer.

No Gas mode He mode O2 mode

Forward Power 1550 W 1550 W 1550 W

Nebulizer Gas Flow 1.02 L/min 1.02 L/min 1.02 L/min

Extract 1 1.0 V -3.0 V -3.0 V

Extract 2 -175.0 V -175.0 V -200.0 V

Isotopes Monitored 23Na, 39K, 45Sc, 43Ca, 44Ca, 56Fe, 63Cu, 66Zn, 31→47P, 32→48S, &

& 85Rb 45Sc & 115In 89→105Y

-1 Gas Flow He: 3 ml min O2 gas flow: 30%

OctP Bias -8.0 V -18.0 V -5.0 V

Typical counts of 1 59Co: 4,240 59Co: 1,885 59Co: 4,860 ppb at 0.1 s 89Y: 6,998 89Y: 2,514 89→105Y: 6,886 integration 140Ce: 7,082 140Ce: 4,060 205Tl: 5,204 205Tl: 3,062 140→156Ce: 7,427 205Tl: 4,275

3.2.4 Statistical Analysis

To evaluate the differences in the results obtained from the two sample introduction systems used, T-test was used (Origin 7.0, Originlab Corporation MA, USA) to compare values obtained for individual elements.

3.3 Results and Discussion

Instrument optimization

Two strategies are commonly used for analysis of small sample volumes by liquid sampling ICP-MS – the use of flow injection accessories, consisting of a pumping device such as a peristaltic or syringe pumps, a sample loop and a switching port valve to discreetly introduce a fixed amount of standards and samples; and the use of small flow nebulizers, with or without spray chambers to achieve a better mass transference to the torch. For our application, removal of the spray chamber was not needed, and only minor adjustments to the carrier and makeup gases were used.

The tune performance comparison between the standard sample introduction and the micro-nebulizer are shown in Table 3.1 & 3.2. The analysis time per sample for the traditional sample introduction system under manual sampling was 4 min per sample, while the automated well plate sampler needed 2.5 minutes per sample. In addition to the increased analysis time, more consistent uptake and rinse times reduced the residual standard deviation

RSD on the instrumental replicates from 6-9% to 3-4%. This reduced the instrumental limit of detection and increased the reproducibility of technical replicates from 8% to 5%. For these

reasons, all presented results were obtained by using the micro well plate sampler and

Meinhard micro-nebulizer as sample introduction. Besides the increased accuracy and lower detection limit, there were no statistical differences between the results obtained from the two sample introduction systems used.

Digestion optimization

The use of high purity acid and water was essential for the quantification strategy, and an optimized digestion protocol was necessary for accuracy and reproducibility. The value of blank equivalent concentrations in digestion blanks was considered in choosing the maximum allowed dilutions. We avoided final digestion volumes below 250 µl, to reduce the risk of sample drying and charring. The smallest sample size was limited to that which gave an ICP-MS response seven times above the average digestion blank level after being mineralized in 250 µl.

This threshold was different for different cell types, due mainly to their wide range of cell sizes.

For example, the lowest number of H. capsulatum cells that gave an acceptable signal for S, P,

Zn and Fe was 1E4 cells for a minimum of three replicates. The larger yeast cells of B. dermatitidis allowed for a 5E3 cell count as the minimum sample size for the same elements.

Macrophages are larger than yeast cells and the lowest cell number that showed reproducible results was 5E3 cells for the studied metals. In contrast, myofibers could be analyzed with as low as two cells per sample with consistent results, due to their larger size and high content of both the target elements (Fe, Rb, and K) and the heteroatoms used for normalization (S and P).

Calibration

A typical example of the calibration results for the studied elements is summarized in Table 3.3.

Both the instrument detection limits and the blank equivalent concentrations were consistent across different experimental days.

Table 3.3 Calibration Summary: Summary of common ICP-MS-MS calibration parameters.

3.3.1 S or P content provides better index of cell mass than cell number

To evaluate the precision with which metal ion concentrations can be measured in small samples of cultured cells based on the cell count, we measured the total P, Zn, and S content of

H. capsulatum, B. dermatitidis and RAW Macrophages cells by ICP-MS-MS and plotted total elemental content (ng) versus the calculated number of cells in each sample (Figure 3.1). The correlation coefficients for each element, although 0.95 - 0.97, fail to reproduce the individual slopes from separate experiments. Under the identical conditions of these experiments, the elemental content per cell is expected to be identical across all samples in all experiments. The likely source of the discrepancy between different batches is the significant error in obtaining accurate cell counts using a hemocytometer. Other sources of discrepancies between different batches can be caused by rinsing the cells to remove them from a plate of media and cell growth rates.

In contrast, when the content of these elements is plotted by pairs, the correlations between sulfur vs. phosphorous, sulfur vs. zinc, and phosphorous vs. zinc are consistent over five orders of magnitude in cell count. The element correlations for P and S are consistent for samples of 105 to 1010 cells (Figure 3.2). Phosphorous and Sulfur are potentially useful ions for mass index. Because they are abundant at the percentage level in biological tissues and are part of intrinsic cell structures and molecules, their concentrations do not change under different experimental conditions. Phosphorus is abundant in most tissues, where it is present in phospholipids at the cell membrane and genetic material. Sulfur is present in proteins as cysteine and methionine and metabolites. Sulfur is abundant in muscle fibers at percentage

levels and was recently used as an index of cell mass in whole muscle samples measure metal ion transport by the Na+, K+ ATPase ([51]).

Fig. 3.1: A) P and B) S content (total ng per sample) plotted versus the number of H. capsulatum cells in each sample. Four independent batches of cells ( 1 2 3 4 ) were aliquoted into samples containing 1.0E+05, 5.0E+05, 1.0E+06, 5.0E+06, 1.0E+07, 5.0E+07, 1.0E+08, 5.0E+08 and 1.0E+09 cells. Cell counting was performed using a standard hemocytometer. The fitted line for each element was obtained using measurements from all batches. A magnification of the lower part of the plot is shown to the right for each element.

Fig. 3.2: P versus S content (total ng per sample) plotted for different cell numbers of H. capsulatum cells. Four independent batches of cells ( 1 2 3 4 ) were aliquoted into samples containing 1.0E+05, 5.0E+05, 1.0E+06, 5.0E+06, 1.0E+07, 5.0E+07, 1.0E+08, 5.0E+08 and 1.0E+09 cells. The fitted line for each element was obtained using measurements from all batches. A magnification of the lower part of the plot is shown to the right for each element.

3.3.2 Prediction of cell mass from the S or P content of H. capsulatum

When enough material is available for accurate mass measurement, it is possible to precisely quantify the sulfur and/or phosphorous concentration ([S] or [P]) in a cell type, and use the concentration to estimate a sulfur- or phosphorous-based mass accordingly to the following equation:

풏품 풐풇 푺 풏품 풐풇 푺 [푆] = Therefore 푔 표푓 푠푎푚푝푙푒 = 푔 표푓 푠푎푚푝푙푒 [푆]

After obtaining a reproducible total P (or S) content in ng by ICP-MS, this can be used to recalculate the sample mass (“P-based mass,” or S-based mass”). We applied this approach to cell cultures of H. capsulatum. 1010 yeast cells were weighed and analyzed for sulfur and phosphorous concentration (ng S/g or ng P/g). This concentration was then used to calculate the mass of aliquots of cells containing 109 to 5x107 cells from the total content of phosphorous. The computed mass versus the measured mass was statistically the same (t-test, p>0.5) (Figure 3.3).

Figure 3.3: Cell mass computed from P content obtained by ICP-MS-MS versus experimental cell mass obtained in an analytical balance. Two batches of H. capsulatum containing 1010 yeast cells were analyzed, and P was quantified based on sample mass. The P concentration was used to calculate the P-Based mass (recalculated mass) and after the analysis of total P, batches of cells of 1E+09, 5E+08, 1E+08 and 5E+07 yeasts per sample. The analysis was performed four times, and the graph represents the average of the four measured samples and the calculated mass based on P content.

3.3.3 Normalization by S or P provides accurate quantification of metal elements in small cell numbers

A systematic quantification of any element of interest can be obtained in small samples of cells or tissue by normalizing to the total P or S content of the sample and following the total content ratio across different experimental treatment. For example, when Zn is to be measured, the total ng of Zn / total ng of P can be followed. To evaluate this normalization, the

Zn/P and Zn/S ratios were evaluated in samples of H. capsulatum and B. dermatitidis having cell counts of 103 to109 cells and maintained under constant conditions. Under these conditions, the Zn/P ratio is expected to remain constant regardless of sample size. Indeed, the Zn/P and

Zn/S ratios calculated from the total ng of each element in the samples were invariant over six orders of magnitude difference in cell count (Figure 3.4).

Figure 3.4: Ratio of Zinc/Sulfur calculated from the total ng of each element in different samples of H. capsulatum and B. dermatiti. Four batches of H. capsulatum ( 1 2 3 4 ) containing 1.0E+05, 5.0E+05, 1.0E+06, 5.0E+06, 1.0E+07, 5.0E+07, 1.0E+08, 5.0E+08 and 1.0E+09 yeast cells were analyzed and the ratio of total Zn/total P was found to be stable across the studied range.

3.3.4 Quantification of metal ion transport in single skeletal muscle cells

Skeletal muscles carry out important metal ion-dependent biological processes and metal ion transport. There is a need to measure these processes in isolated, single muscle cells

(fibers) because whole skeletal muscles are composed of multiple fiber subtypes, with each subtype having distinct metalloproteins and metal-dependent processes. We previously developed a method to measure Na+, K+ ATPase transport in whole mouse skeletal muscles by

ICP-MS-MS using S as a mass index ([51]). However, isolated single skeletal muscle cells present special challenges for quantification due to their great variations in cell size. For example, isolated mouse EDL fibers vary from 40 – 80 m in diameter and 8 – 20 mm in length and have wet weights in the g range.

To evaluate whether the reported ratiometric normalization can be extended to small samples of freshly dissociated skeletal muscle fibers, we measured the total endogenous contents of sulfur, phosphorous and rubidium (ng) in samples of 1, 2, 3, 5, 7 and 10 mouse EDL fibers. The abundance of sulfur in whole skeletal muscles make it a useful mass index; while the very low background levels of Rb (ppb) and its ability to substitute for K transport by the Na+, K+

ATPase, make it a useful metal ion for measurement of enzyme transport.

In contrast to cultured cells which have a more homogeneous cell size, the elemental content versus cell number correlations in non-cultured, freshly dissociated muscle cells show only a quasi-linear relationship and lower correlation coefficients (Figure 3.5). On the other hand, when the element-element correlations are plotted, a much better correlation is observed. An overlap of elemental content is evident between some samples of different fiber

numbers (Figure 3.6). This overlap occurs because a group of two muscle cells, for example, can have the same total mass as three smaller cells. The positive assessment of the elemental correlations was further explored by evaluating the Rb/S and Rb/P ratios computed from these measurements. Constant values of 0.32 ± 0.01 and 0.28 ± 0.01, respectively, for all 120 samples analyzed (twenty samples per group). This result validates the use of S or P as a mass index, and use of the element/S or element/P ratio for quantification of metal ion processes in single muscle cells.

Figure 3.5: Correlations between the K, S, P and Rb content (ng) and a number of muscle fibers in each sample. Data represent results of measurements from 120 total samples. 1, 2, 3, 5, 7 and 10 individual fibers were analyzed.

Figure 3.6: Correlations among the Rb, P, and S content (total ng) measured by ICP-MS-MS in samples of 1, 2, 3, and 5, 7 and 10 EDL fibers. Enzymatically dissociated, fresh mouse EDL fibers were obtained and processed as described in Methods.

As proof of concept, we measured Na+, K+ ATPase transport activity in single muscle cells using a tracer amount of Rb (500 µM) as a congener for K. The Na+, K+ ATPase is highly expressed in muscle cells, and its transport rate has significant physiological and clinical importance. To evaluate the use of sulfur-based normalization for rubidium quantification, four sets of three fibers each was treated as follow: A control group was not treated; a positive control was treated with Ouabain an established Na+, K+ ATPase blocker, and a third group was treated with salbutamol which stimulates the activity of the Na+, K+ ATPase. Groups of three muscle fibers were exposed to a physiological solution containing the tracer RbCl and allowed to take up Rb for ten minutes. After that, they were washed, digested and analyzed by ICP-MS-

MS. Results are summarized in a bar graph (Figure 3.7).

Figure 3.7: Transport activity of Na+, K+ ATPase in dissociated mouse EDL fibers. Transport activity of the Na+, K+ ATPase was measured in four groups of samples with 16-18 samples per group and 3 fibers per sample. Rb uptake in the test groups was measured using 500 M RbCl in the uptake buffer (control), uptake buffer with 1 mM ouabain (antagonist), or uptake buffer plus 10 M salbutamol (agonist). The endogenous Rb content of each sample was measured in buffer without added RbCl. A) The amount of Rb measured in each sample (ppb). B) The Rb/P ratio of each sample. C) The Rb/S ratio of each sample. * indicates statistical difference at p < .05; **** indicates statistical difference at p < .0001

Possible Source of Phosphorus and Sulphur contaminations in solutions and cell containers.

This quantification strategy requires the absence of significant phosphorous and sulfur contamination from the experimental solutions and materials in contact with the samples. For the case of H. capsulatum, Macrophages and B. dermatitidis the use of PBS saline buffer for rinsing can contaminate measurement of phosphorous for normalization; for this reason, a physiological sodium chloride solution was used as rinsing solution. For the myofibers, the concentration of S and P in the solutions used for washing and digesting samples as well as four different containers tested for washing were comparable to the endogenous P and S concentrations in 1 - 3 myofibers. Nitric acid and water may contain S and P although it was below our detection limit (0.03 and 0.01 ng/ml respectively). The least background contamination from vessels was obtained using clear metal-free 1.5 ml microcentrifuge propylene tubes and Polypropylene multi-well plates. Samples containing at least 3 myofibers were used to measure the rate of Rb uptake.

3.4 Conclusion

By exploiting the multi-element capabilities and detection power of ICP-MS-MS instrumentation, it was possible to extend the application of this technique to biological samples of sub-microgram mass. Total elemental composition proved to be a more reliable normalization tool than the cell number counted by hemocytometer for cell cultures of H. capsulatum, B. dermatitidis, and RAW macrophage cells, or fiber number for single skeletal muscle cells.

Fully quantitative results can be obtained for populations of cultured cells using the P- based or S-based calculated mass after a large batch of cells is analyzed in the traditional mass- based protocol to obtain the average P- or S-to-mass factor. A semi-quantitative analysis with low RSD can be obtained using the elemental composition ratio (Zn/S or P, Fe/S or P, etc.) when a sufficient sample size is not available to obtain an accurate element-to-mass average. The ability to normalize to cell P or S content is particularly useful for small samples of cells from rare or clinical specimens for which a cell count may not even be possible.

The proposed protocol is applicable to many cell cultures and tissues at microgram levels without concern of sample size variability among groups. To obtain reproducible results, the experimental protocols need to be revised to avoid contamination with low-quality reagents, buffers containing S or P, or vessels which can leech S or P.

Acknowledgments

This work was funded by NIH grants RO1 AR063710 and R01 AI 106269. We thank

Agilent Technologies for the great collaboration that provides instrumental and technical support with the 8800 ICP-MS/MS to the University of Cincinnati-Agilent Technologies

Metallomics Center of the Americas.

Chapter 4: Isolation & Purification of Novel Endogenous Cardiotonic Like Steroids from Sus

domesticas Skeletal Muscle Tissue

4.1 Introduction

Cardiotonic steroids are known for biologically targeting the Na+, K+ ATPase to cause inhibition or stop the transport of Na+ and K+ ions [53]. These small compounds can be extracted from a diverse range of living organisms and used in medicines to treat various medical conditions such as hypertension, congestive heart failure, and arrhythmia [9,54]. The skeletal muscle of pigs was chosen for this study for a couple of reasons. The first reason being it is known that approximately 84% of cardiotonic steroids are stored in skeletal muscle [55,56].

The second reason is that cardiotonic steroids in mammals have not been heavily studied or characterized. Based on these circumstances it is of high interest to pursue cardiotonic steroids in mammalian tissues.

A variety of different methods have been used to separate these compounds from their organisms of origin, but these separations can be very time consuming and a lengthy process due to the presence of many other large and small molecules. A purification process called the batch affinity extraction (BAE) was designed to extract cardiotonic steroids from a complex sample matrix using purified Na+, K+ ATPase proteins [57]. The BAE eliminates all the large and small molecules that do not interact with the Na+, K+ ATPase leaving only the cardiotonic steroids or other compounds that interact with its binding site for further analysis. Without the batch affinity extraction, it would be very challenging to isolate the cardiotonic steroids just using chromatographic techniques.

By adding the BAE as a purification method before using chromatographic techniques, it makes the purification process more efficient and less complicated. Following the BAE, size exclusion, reverse phase, and hydrophilic interaction liquid chromatography are used to separate the cardiotonic steroids based on different properties. During the chromatographic separations, fractions were collected and tested for biological activity using a 3H ouabain competition binding assay to determine which compounds interact with the Na+, K+ ATPase.

The compounds that showed the most biological activity were scaled up and further characterized by weight, structure, and their physiological properties.

4.2 Experimental

4.2.1 Chemicals and Materials

Batch Affinity Extraction Solutions

The Binding Buffer contained (mM): 140 NaCl, 10 Hepes (free acid), 5 MgCl2, 5 Na2HPO4

(dibasic), 0.02% Saponin, nanopure H2O; pH 7.4, 37 °C. The Release Buffer contained (mM): 140

NaCl, 10 HEPES (free acid), 5 EDTA (free acid), 0.02% Saponin, nanopure H2O; pH 7.4, 37 °C. All materials for the two buffers were obtained from Thermo Fisher Scientific or Sigma-Aldrich

(USA). All buffers were prepared fresh and used within one week.

3H-Ouabain Competition Binding Assay Solutions

The Vanadate Buffer contained (mM): 500 sucrose, 50 tris solution (pH 7.4), 770 MgSO4,

92.8 NaVO4 (pH 7.5), nanopure water. All materials for the buffer were obtained from Thermo

Fisher Scientific or Sigma-Aldrich (USA). The vanadate buffer was used within a week of being

prepared. The 200 nM 3H-Ouabain was prepared fresh and stored in a 4 °C refrigerator for radioactive materials.

4.2.2 Sample Preparation

4.2.2.1 Homogenization of Pig Skeletal Muscle

A 50% MeOH extract containing semi-purified OLC is prepared from pig skeletal muscle.

Hindlimb skeletal muscle from a pig was surgically removed, pulverized in liquid nitrogen, and stored at -80°C. From the bulk of pig skeletal muscle collected, 96 grams of muscle tissue was weighed out and placed in a 6.72 L chilled solution of 0.1% trifluoroacetic acid and 50% methanol (Thermo Fisher Scientific, USA). The pig skeletal muscle is then homogenized in solution using a tissue tearor for 2 min 30 secs while trying to avoid foaming. Once the muscle tissue is homogenized in solution, the sample is placed on a shaker (C1 Platform Shaker, New

Brunswick Scientific, USA) for 30 min at 40 rpm. After shaking, the sample is divided into 10 x 60 mL Beckman ultracentrifuge tubes and balanced by weight using a Harvard trip balance (Ohaus

Scale Corp., USA).

The homogenized solution is placed in an Optima L-90K Ultracentrifuge (Beckman

Coulter, USA) and spun at 15,000 g (11,400 rpm) for 20 min at 4°C. Once the centrifugation is complete, the supernatant is collected, and the pellet is discarded. The collected supernatant is placed in a Büchi rotavap R-205 (Büchi, USA) to remove the methanol before lyophilization.

After the methanol is removed from the sample, it is reconstituted with 600 mL of nanopure water and placed on a shaker for 30 min at 40 rpm. The sample is then placed in the ultracentrifuge once more at 100,000 g (29,400 rpm) for 20 min at 4°C. The supernatant is

collected and frozen with liquid nitrogen. After freezing, the sample is lyophilized with a Sentry

Benchtop 3L freeze dryer (The Virtis Company Inc., USA) until it is completely dry. The sample is then reconstituted with 120 mL of nanopure water which prepares the sample for the 3H- ouabain competition binding assay to test for biological activity of the unknown endogenous

OLC. If the OLC sample shows biological activity, the batch affinity extraction can be performed to purify the unknown endogenous OLC. This protocol is designed to make 96-gram batches, but 8-gram batches were made starting out to ensure the active compounds were present before scaling up.

4.2.2.2 Batch Affinity Extraction (BAE)

The batch affinity extraction is performed to purify OLC compounds in a sample matrix before using chromatographic techniques. First, a reaction mix is made that consist of 20.4 mL of binding buffer, 120 mL of the OLC sample obtained from the homogenization process, and

3.6 mL of the Na+, K+ ATPase (NKA) protein which is used to bind to the OLC compounds in solution. The reaction mix is then incubated in an all stainless-steel water bath (Precision

Scientific Inc., USA) for 3 hours at 37°C to ensure the NKA protein and the OLC compounds are binding in solution. After 3 hours, the incubation is stopped by diluting the reaction mix with

540 mL of cold binding buffer. The sample is then divided into 60 mL Beckman ultracentrifuge tubes and balanced by weight.

The sample is placed in an ultracentrifuge and spun at 38,000 rpm for 60 min at 4°C. The pellet contains the NKA protein and the OLC compound which is collected, and the supernatant is discarded. The pellet is reconstituted with 720 mL of release buffer and homogenized with a

tissue tearor which will free the OLC from the NKA protein during incubation. The sample is then incubated for 16 hours or overnight at 37°C to ensure the OLC compound unbinds from the NKA protein. After incubation, the pellet is discarded, and the supernatant which now contains the unbound OLC is collected. The supernatant is frozen in liquid nitrogen and lyophilized until it is completely dry. The sample is then reconstituted with 50 mL nanopure water. Finally, the sample can be tested for activity by using the 3H-ouabain competition binding assay. If the OLC sample shows biological activity, it will be further purified using various chromatographic techniques.

4.2.2.3 3H-Ouabain Competition Binding Assay

The competition binding assay is performed to ensure that the unknown OLC has an activity or a strong affinity for the Na+, K+ ATPase. The affinity for the Na+, K+ ATPase is shown by how well the unknown OLC can compete off the bond 3H-Ouabain from the Na+, K+ ATPase. A master mix is prepared which contains vanadate buffer, 1% saponin, 200 nM 3H-Ouabain, and

0.375 µg/mL of protein. Eppendorf tubes with a volume of 500 µL are used for this reaction.

The tubes are placed in a DNA thermal cycler 480 (Perkin Elmer, USA) set to 4°C before the master mix is added. 135 µL of master mix is placed into each Eppendorf tube used for the experiment. Afterward, 15 µL of known or unknown sample is added to each Eppendorf tube. If dilutions of the samples are required, smaller volumes of sample are added to the Eppendorf tubes. Once the master mix and the sample are added to the tubes, the temperature of the thermal cycler is increased to 37°C and will incubate for 90 mins to allow the competitive binding to occur.

Once the incubation is complete, the thermal cycler is set back to 4°C to stop the reaction. The samples are loaded onto a glass fiber filter, 25 mm diameter inside of a stainless steel washing well on a filter apparatus. The filters are washed 3 times with cold vanadate buffer to remove the unbound 3H-Ouabain. These washings are done under low vacuum to suction out the wash buffer. After the final wash, the filters are dried by setting the vacuum to its maximum pressure for 10 mins. Once the filters are dried, they are placed in 6.5 mL scintillation vials that contain 4 mL of Filter-Count scintillation fluid (Perkin Elmer, USA). These vials are then shaken overnight to degrade the filters and release the 3H-Ouabain. Afterward, these samples are placed on the Beckman LS6500 scintillation counter (Beckman Coulter, USA) to detect the amount of 3H-Ouabain present which determines the activity of the unknown OLC compound.

4.2.3 Instrumentation

4.2.3.1 Size Exclusion Chromatography (Analytical)

To further purify the OLC sample, an Agilent 1100 series high-performance liquid chromatography system (HPLC, Agilent Technologies, USA) equipped with a solvent module, vacuum degasser, binary pump, autosampler, thermostatted column compartment, and the diode-array detector is used to perform various separations based on the type of column employed. After the batch affinity extraction, the HPLC system is utilized for the OLC sample which is passed through a superdex peptide 10/300 GL size exclusion column (GE Healthcare

Life Sciences, USA) which is a prepacked column for high-resolution SEC for small-scale preparative purification and analysis of peptides and other small biomolecules with molecular

weights between 100 and 7000 Da. First, the column is equilibrated with 50 mM Ammonium

Acetate in doubly deionized water with 5% MeOH at a flow rate of 0.5 mL/min for 2 hours and

50 mins; this solution is also used as the mobile phase for the analysis. After equilibration of the column, 30 µL of the OLC sample is injected and carried through the system for 50 mins. During the analysis time, 4 fractions were collected based on retention times of peaks in the chromatogram. The collected fractions were frozen in liquid nitrogen, lyophilized to dryness and reconstituted in nanopure water. All fractions were analyzed using the 3H-Ouabain competition binding assay to check each fraction for biological activity.

4.2.3.2 Reverse Phase Chromatography (Analytical)

As a result of processing the OLC sample through size exclusion chromatography, the

OLC sample is now further purified yielding 4 OLC fractions. Each of these four fractions are processed through reverse phase chromatography. The Agilent 1100 series high-performance liquid chromatography system was used along with an Agilent Zorbax 80Å Extend C-18 reverse phase column (4.6 mm x 150 mm, Agilent Technologies, USA) to separate molecules based on their polarity. First, the column is equilibrated with 80% acetonitrile and 20% nanopure water at a flow rate of 1 mL/min for 35 mins. After equilibration of the column, 30 µL of each OLC fraction collected from size exclusion were injected into the system; each OLC fraction was analyzed separately. The sample was carried by a mobile phase using a gradient of 3% to 55% acetonitrile for 30 mins. During the time of analysis for each fraction, several fractions were collected based on retention times of the peaks in the chromatogram. The collected fractions were frozen in liquid nitrogen, lyophilized to dryness, and reconstituted in nanopure water. All

fractions were analyzed using the 3H-Ouabain competition binding assay to check each fraction for biological activity.

4.2.3.3 Hydrophilic Interaction Liquid Chromatography (Analytical)

After using reverse phase chromatography, fraction OLC B1 was shown to have the most activity by the 3H-ouabain competition binding assay. Since fraction OLC B1 had the most biological activity, it was processed through hydrophilic interaction liquid chromatography

(HILIC) to further purify fraction OLC B1. The Agilent 1100 series high-performance liquid chromatography system was used along with a Luna 3µ HILIC column (2.0 mm x 150 mm,

Phenomenex, USA) to separate molecules based on their polarity. First, the column is equilibrated with 90% acetonitrile, 10% doubly deionized water, and 0.1% formic acid at a flow rate of 0.1 mL/min for 1 hour and 5 mins; this solution is also used as the mobile phase for the analysis. After equilibration of the column, 2 µL of fraction OLC B1 is injected and carried through the system for 25 min. During the analysis time, 3 fractions were collected based on retention times of peaks in the chromatogram. The collected fractions were frozen in liquid nitrogen, lyophilized to dryness, and reconstituted in nanopure water. All fractions were analyzed using the 3H-Ouabain competition binding assay to check each fraction for activity.

4.2.3.4 Size Exclusion Chromatography (Preparatory)

To obtain more OLC material, the process was increased to a larger scale. A preparatory

Varian high-performance liquid chromatography system (HPLC, Varian-Agilent Technologies,

USA) equipped with two solvent delivery modules and a UV-VIS detector was used to perform various separations based on the type of column installed. After using the scaled up version of the batch affinity protocol, the Varian HPLC system is used for the OLC sample which is passed

through a superdex 30 prep grade (PG) in Hi-load column (16 mm x 600 mm, GE Healthcare Life

Sciences, USA) designed for high resolution preparative SEC and analysis of peptides and other small biomolecules with molecular weights less than 10,000 Da. First, the column is equilibrated with 50 mM Ammonium Acetate in doubly deionized water with 5% MeOH at a flow rate of 1.5 mL/min for 4 hours and 41 mins; this solution is also used as the mobile phase for the analysis.

The flow rate of the instrument was set to 3 mL/min, but splitters and resistors were used which reduced the flow rate to 1.5 mL/min. After equilibration of the column, 5 mL of the OLC sample is injected and carried through the system for 4 hours and 16 mins. During the analysis time, 4 fractions were collected based on retention times of peaks in the chromatogram. The collected fractions were frozen in liquid nitrogen, lyophilized to dryness, and reconstituted in nanopure water. All fractions were analyzed using the 3H-Ouabain competition binding assay to check each fraction for activity. The active fractions were processed by reverse phase chromatography.

4.2.3.5 Reverse Phase Chromatography (Preparatory)

To obtain more OLC material, the process was increased to a larger scale. As a result of processing the large batch of OLC sample through size exclusion chromatography, the OLC sample was further purified yielding 4 OLC fractions. Each of these four fractions were processed by using preparatory reverse phase chromatography. A preparatory HPLC system

(Varian PrepStar) was used along with a Phenomenex Kinetex 5µ Phenyl-Hexyl 100Å, AXIA packed reverse phase column (21.2 mm x 250 mm, Phenomenex, USA) to separate molecules based on their polarity. First, the column is equilibrated with 80% acetonitrile and 20% nanopure water at a flow rate of 10 mL/min for 2 hours and 3 mins. After equilibration of the

column, 1 mL of each OLC fraction collected from size exclusion was injected into the system; each OLC fraction was analyzed separately. The sample was carried by a mobile phase using a gradient of 3% to 85% acetonitrile for 40 mins. During the time of analysis for each fraction, several fractions were collected based on retention times of the peaks in the chromatogram.

The collected fractions were frozen in liquid nitrogen, lyophilized to dryness, and reconstituted in nanopure water. All fractions were analyzed using the 3H-Ouabain competition binding assay to check each fraction for activity.

4.2.3.6 Hydrophilic Interaction Liquid Chromatography (Semi-Preparatory)

To obtain more OLC material, the process was increased to a larger scale. After using reverse phase chromatography, fraction OLC B1 was shown to have the most activity by the 3H- ouabain competition binding assay. Since fraction OLC B1 had the most activity, it was processed through hydrophilic interaction liquid chromatography (HILIC) to further purify fraction OLC B1. A preparatory HPLC system (Varian PrepStar) was used along with a semi- preparatory Luna 5 µm HILIC 200Å column (10 mm x 250 mm, Phenomenex, USA) to separate molecules based on their polarity. First, the column is equilibrated with 60% acetonitrile, 40% doubly deionized water, and 10 mM ammonium formate at a flow rate of 4.5 mL/min for 1 hour; this solution is also used as the mobile phase for the analysis. After equilibration of the column, 400 µL of the OLC sample is injected and carried through the system for 8 mins. During the analysis time, 3 fractions were collected based on retention times of peaks in the chromatogram. The collected fractions were frozen in liquid nitrogen, lyophilized to dryness, and reconstituted in nanopure water. All fractions were analyzed using the 3H-Ouabain competition binding assay to check each fraction for activity.

4.2.4 Data Analysis

Various software’s were used for this research project. Origin Pro 8.5 (OriginLab

Corporation, USA) was used to plot chromatograms generated from all HPLC experiments.

Prism was used to plot results generated from the 3H-ouabain competition binding assay. Excel was used to make graphs for the competition binding assay and to store most of the data collected for this research project.

4.3 Results and Discussion

4.3.1 HPLC Fraction Naming System

Throughout this project, many fractions were collected from three different types of columns. Figure 4.0 provides a roadmap to better understand how each fraction was collected and processed. Fractions that contain one letter on the roadmap were collected from size exclusion chromatography. If the fraction contains one letter and one number on the roadmap it was collected from reverse phase chromatography. Finally, fractions that contain one letter and one number with a decimal place on the roadmap were collected from hydrophilic interaction liquid chromatography.

Figure 4.0: Road Map of HPLC Fraction Collections.

4.3.2 OLC SEC-HPLC Separations Analyzed by Competition Binding Assay

After the OLC sample is processed through batch affinity extraction, size exclusion chromatography is used to separate the molecules by size and to remove larger particles. SEC is a filtration technique in which larger particles will elute from the SEC column first because they have less path to travel compared to smaller molecules which will elute much later within the runtime of the analysis. The superdex peptide 10/300 GL size exclusion column was used for this separation. First, an ouabain standard (molecular weight: 584.652 g/mol) was processed

through the SEC column which is shown in Figure 4.1. The chromatogram shows ouabain eluting from the column with a retention time between 25-28 mins. Figure 4.2 shows the chromatogram and isoabsorbance plot of all 4 fractions that were collected using size exclusion chromatography. After collection, the fractions were then frozen in liquid nitrogen, lyophilized to dryness, and reconstituted in nanopure water. These 4 fractions were then tested using the

3H-ouabain competition binding assay to determine if these analytes were biologically active. If the analytes are biologically active, the analytes will have a high affinity for the Na+, K+ ATPase and displace the 3H-ouabain from the enzyme.

The 3H-ouabain competition binding assay in Figure 4.3 shows how well each fraction can displace 3H-ouabain from the Na+, K+ ATPase. A range of known concentrations of ouabain were used to create a standard curve to show displacement of 3H-ouabain. Fraction A was not presented in Figure 4.3 because it did not show any activity in this assay. Fraction B was the most active fraction with counts per minute (CPM) of 349.40. Fraction C and D had CPMs of

456.80 and 2,935.67, respectively. Since fraction B was the most active compound, it will be used as a candidate for further purification and testing. All the active fractions were processed using reverse chromatography and tested for activity a second time.

Figure 4.1: Size Exclusion Chromatogram & Isoabsorbance Plot of Ouabain Standard.

600 kDa & 158 kDa kDa 158 & kDa 600

44 kDa kDa 44

17 kDa kDa 17 1.3 kDa kDa 1.3

Figure 4.2: Size Exclusion Chromatogram & Isoabsorbance Plot of OLC compound. Fractions A, B, C, & D were all collected. The collected fractions were frozen in liquid nitrogen, lyophilized to dryness and reconstituted in nanopure water. Size markers are listed on the chromatogram from an SEC standard to give an estimated size of the peaks present.

5000

D 4000

3000 C 2000

CPM Bound, B 1000

0 -12 -10 -8 -6 -4 -2

Log Final [Oub]

Figure 4.3: CB assay of SEC fractions. Fraction A is not shown because it did not have any activity or the ability to displace 3H ouabain from the Na+, K+ ATPase. Fraction B was the most active fraction. CPM: Fraction B 349.40, fraction C 456.80, and fraction D 2,935.67.

4.3.3 OLC RP-HPLC Separations Analyzed by Competition Binding Assay

Each of the active fractions collected from size exclusion chromatography was processed through reverse phase chromatography. The Agilent Zorbax 80Å Extend C-18 reverse phase column has a C-18 stationary phase which attracts non-polar analytes and causes them to retain in the column while polar analytes are not retained on the column and elute early.

Figure 4.4 shows the reverse phase chromatogram and isoabsorbance plot for fraction B.

Multiple fractions were collected at the following retention times: 1) 0.8-2.1 mins, 2) 2.1-2.6 mins, 3) 24-34 mins, and 4) after 34 mins. Of these four fractions, only 2 of them were active.

The analytes that were collected in the first and fourth fraction were the most active for the separation of fraction B.

Figure 4.5 shows the RP separations and isoabsorbance plot for fraction C. Fractions were collected at the following retention times: 1) 0.8-2 min, 2) 2-2.2 min, 3) 2.2-2.5 min, 4)

2.5-5 min, 5) 5-14 min, 6) 24-34 min, and 7) after 34 mins. Of all several fractions, only the first and seventh fractions had the most activity for the separation of fraction C. Finally, Figure 4.6 shows the RP separation and isoabsorbance plot for fraction D. Only 2 fractions were collected from fraction D, and they have the following retention times: 1) 1.3-3.5 min and 2) 24-34 min.

Both collections from fraction D were active. The most active fractions collected from fraction

B, C, and D were all compared against each other to determine which fraction contained the most activity. The compound with the most activity is shown in Figure 4.7. The most active fraction was selected as the candidate for further testing.

Figure 4.4: Reverse Phase Chromatogram & Isoabsorbance Plot of Fraction B.

Figure 4.5: Reverse Phase Chromatogram & Isoabsorbance Plot of Fraction C.

Figure 4.6: Reverse Phase Chromatogram & Isoabsorbance Plot of Fraction D.

Figure 4.7: Competition Binding Assay of Reverse Phase Fractions. The fractions tested show that B1, C1, and D1 are very active compounds as they fit on the lower portion of the competition binding curve (black filled circles on 10-3). Since fraction B1 is the most active with a CPM of 112.60, it will be further processed and tested.

4.3.4 Preparative SEC-HPLC fractions

Once it was determined that biologically active compounds were present in the small 8- gram OLC batches, the batches were scaled up starting with 96 grams of pig skeletal muscle.

Due to the increase of sample size, preparative HPLC must be used to process the larger sample volumes to make the separations more efficient. The superdex 30 prep grade (PG) in Hi-load column (16 mm x 600 mm, GE Healthcare Life Sciences, USA) was used to process large amounts of sample. Figure 4.8 shows the chromatogram produced by this column.

C B D

Figure 4.8: Preparatory Size Exclusion Chromatogram of OLC compound. Fractions A, B, C, & D were all collected. The collected fractions were frozen in liquid nitrogen, lyophilized to dryness and reconstituted in nanopure water.

4.3.5 Preparative RP-HPLC fractions

Each fraction that was collected from size exclusion was processed through a preparative reverse phase column. The Phenomenex Kinetex 5µ Phenyl-Hexyl 100Å, AXIA packed reverse phase column (21.2 mm x 250 mm, Phenomenex, USA) was used to conduct the separations. Figure 4.9 shows all the chromatograms produced from each size exclusion fraction that was processed through the preparatory grade RP column. The fractions collected after reverse phase chromatography were tested to determine biological activity and the results were similar to those shown in Section 4.3.3, Figure 4.7. Fraction B1 was the most active fraction, so it was further processed through HILIC chromatography to determine if further purification is needed.

Figure 4.9: Preparatory Reverse Phase Chromatograms of OLC compound. A) This chromatogram shows the separation for fraction B in the reverse phase column. The fractions collected from this separation are fraction B1 between 5 and 10 mins and fraction B2 after 34 min. B) This chromatogram shows the separation for fraction C. The fractions collected are fraction C1 between 5 and 8 min and fraction C2 after 34 min. C) This chromatogram shows the separation for fraction D. The fractions collected are fraction D1 between 5 and 10 min and fraction D2 between 28 and 35 min.

4.3.6 HILIC Separations

Hydrophilic interaction liquid chromatography (HILIC) is a separation based on polarity.

The stationary phase of a HILIC column is composed of crossed linked diol groups which are very polar, allowing polar analytes to retain in the column. The mobile phase is typically comprised of a higher percentage of acetonitrile and a lower percentage of water along with a buffer such as ammonium formate. This technique was used to determine if fraction B1 can be further separated. The reverse phase chromatograms of fraction B show that the active compound (fraction B1) elutes early from the column. The early elution time of fraction B1 is a strong indication that the compound is very polar compared to compounds that elute later from the reverse phase column. Due to the polar traits of this compound, a HILIC column was used to determine if the polar analyte could be further separated before characterization.

First, a small scale analytical HILIC separation was done to determine if more than one compound was present. Figure 4.10 shows the chromatogram obtained from the HILIC column.

The HILIC chromatogram shows a separation of fraction B1 which indicates there is more than one compound present. Since there is more than one compound present, the sample was scaled up and processed through a semi-preparative HILIC column. Figure 4.11 shows the chromatogram and isoabsorbance plot obtained from the semi-preparative HILIC column. From this HILIC chromatogram two fractions were collected, fraction B1.1 and B1.2. These two fractions were tested for biological activity.

Lyophilized Fraction B1 200

180

160

140

120

100

Abs at 254 nm (mAU)254 nm at Abs 40

0 2 4 6 8 10 12 14 16 18 20 22 24 Retention Time (min)

Figure 4.10: HILIC Separation of Fraction B1.

Figure 4.11: Semi-preparative HILIC Chromatogram & Isoabsorbance Plot of Fraction B1.

Oub Std. Curve and OLC HILIC Fractions

8000

7000 B1.1 B1.2 6000 5000

4000 3000

CPM Bound, 2000

1000

0 -12 -11 -10 -9 -8 -7 -6 -5 -4 -3 -2 Log Final [Oub]

Figure 4.12: Competition Binding Assay of HILIC Fractions.

Figure 4.12 shows the competition binding assay for the two HILIC fractions. Both fraction B1.1 and B1.2 are biologically active which indicates both compounds have a strong affinity for the Na+, K+ ATPase according to the 3H-ouabain competition binding assay. Fraction

B1.1 and B1.2 will be characterized to determine their molecular weight and structure by using mass spectrometry and NMR respectively. Multiple assays will also be used to determine the physiological functions of fraction B1.1 and B1.2 when interacting with the Na+, K+ ATPase in various cells that are derived from lower and higher order mammals.

4.4 Conclusion

The batch affinity extraction aided in making the purification process efficient by eliminating compounds that do not interact with the Na+, K+ ATPase. This process works well for cardiotonic steroids that have a strong affinity for the Na+, K+ ATPase, but if there are cardiotonic steroids that have weaker affinities for the enzyme, it may not be possible to purify those compounds well. If cardiotonic steroids with weaker affinities to the Na+, K+ ATPase are of interest in future studies, a method would need to be designed to extract them.

After the batch affinity extraction, a series of chromatographic techniques are used such as size exclusion, reverse phase, and HILIC to purify the endogenous cardiotonic steroids from pig skeletal muscle. After using these various methods, two fractions were collected at the end of the purification process. These two fractions are called B1.1 and B1.2 which are very pure, and they show a robust biological activity according to the competition binding assay. Now that these endogenous compounds are extracted they can be characterized, and the physiological functions can be determined.

Acknowledgments

This work was supported by the National Institutes of Health, NIAMS grant RO1-

AR063710. I would like to thank Dr. Heiny from the Department of Molecular and Cellular

Physiology (University of Cincinnati), Dr. Julio Landero and Jiawei Gong from the Department of

Chemistry (University of Cincinnati), and David Cowart from Ohio University for all their help and support with this project.

100

Chapter 5: The Physiological Effects of Novel Endogenous Cardiotonic Like Steroids Extracted from Sus domesticas Skeletal Muscle Tissue

5.1 Introduction The Na+, K+ ATPase is a transmembrane protein that is composed of an α, β, and γ subunit which can have different combinations of isoforms based on the location within the organism. Cardiotonic steroids are known for inhibiting the Na+, K+ ATPase upon binding to the

α subunit which is the catalytic moiety of this enzyme [2]. The beta subunit is used for the recruitment of the α subunit to the plasma membrane and the occlusion of K ions. The γ subunit functions as a regulator of the Na+, K+ ATPase to ensure appropriate tissue functions such as renal Na+ reabsorption, muscle contractibility, and neuronal excitability [3]. There is a wide range of cardiotonic steroids, and all of them have different inhibitory effects or affinities for the Na+, K+ ATPase. Some cardiotonic steroids like ouabain have such a high affinity that it does not unbind from the enzyme, quickly making it very toxic and less likely to be used for therapeutic benefits [58]. For drug development, it is vital to find cardiotonic steroids that do not bind so strongly to the Na+, K+ ATPase to have a better regulation or working range of the

CTS for therapeutic uses.

Cardiotonic steroids can be extracted from various living organisms, but they have not been heavily studied in mammalian tissues. It is known that cardiotonic steroids are synthesized in the adrenal glands and used as regulators for the Na+, K+ ATPase to maintain a

Na+ gradient for other transporters that depend on it [59]. Approximately 80% of cardiotonic steroids are stored in skeletal muscle which is an excellent source for extracting these compounds to study them further [55,56]. Since mammalian tissues aren’t examined as heavily

101

as other organisms, there is a high probability of discovering new endogenous cardiotonic steroids that can only be derived from mammalian tissues. Upon extracting these compounds, it is also essential to understand how they function in comparison to other known cardiotonic steroids such as ouabain which is a typical control for physiological experiments in this field.

The goal of this research project is to determine the physiological effects of the endogenous cardiotonic steroids that were extracted from pig skeletal muscle in chapter 4.

There are a few assays that will be used in this chapter to determine the functionality of the unknown samples such as the ATPase assay, red blood cell assay, and the reversibility red blood cell assay. Each of these assays will be described in more detail throughout this chapter.

5.2 Experimental 5.2.1 Chemicals and Materials ATPase Assay Solutions This protocol requires purified NKA protein and various concentrations of OLC compounds. The main buffer that is used for this assay is the ATPase activity buffer, and it is composed of (final concentrations in mM): NaCl: 110, KCl: 20, MgCl2: 5, HEPES: 30, EGTA: 2, Na

Azide: 0.5, phosphoenolpyruvate: 1, Beta Nicotinamide adenine dinucleotide hydrate: 0.2, lactate dehydrogenase: 30 U. mL-1, pyruvate kinase 10 U.mL-1, saponin:0.01%, pH: 7.4. All materials for the solutions were obtained from Thermo Fisher Scientific or Sigma-Aldrich (USA).

The buffers were prepared fresh weekly and stored in a 4 °C refrigerator.

102

Red Blood Cell Assay Solutions The 5K/Rb free/ 10 glucose wash solution contained (mM): 130 NaCl, 5 KCl, 0.5 Na

H2PO4, 10 glucose, 10 HEPES (free acid), 12 NaHCO3, 2 CaCl2 2H20, 1 of 4.9M MgCl2 solution, and pH is adjusted to 7.4 at 2-4°C with 0.09 mL 10M NaOH which is 290 mM. The 5K/500 RbCl uptake buffer contained 100 mL of 5K/Rb free/ 10 glucose buffer and 500 µL of 1 M RbCl. The pH of this solution was adjusted to 7.4 at 37°C with NaOH. A 10 mM ouabain solution was made to make a range of ouabain dilutions from 1 mM to 20 nM. The 10 mM ouabain solution was made with 7.3 mg of ouabain-octahydrate (MW 728.77) per 10 mL of H2O. All materials for the solutions were obtained from Thermo Fisher Scientific or Sigma-Aldrich (USA). The buffers were prepared fresh weekly and stored in a 4 °C refrigerator.

Reversibility Red Blood Cell Assay Solutions This protocol uses two of the solutions that were employed in the red blood cell assay which is the 5K/Rb free/ 10 glucose wash solution and the 5K/ 500 RbCl uptake buffer. The RbCl uptake buffer + 10 µM ouabain was made by adding 10 µL of 1 mM ouabain stock per 1 mL of uptake buffer. The RbCl uptake buffer + ~10 µM B1.1 or B1.2 (which is an ouabain equivalent concentration) was made by adding 50 µL of B1.1 or B1.2 to 800 µL of uptake buffer. B1.1 and

B1.2 are samples that were purified from pig skeletal muscle in Chapter 4.

103

5.2.2 Sample Preparation ATPase Assay The ATPase assay is a spectrophotometric assay used to measure the continuous rate of

ATP hydrolysis in the presence of various known or unknown CTS with different concentrations.

To start, 15 µL of NKA protein is incubated with 40 µL of OLCs with varying factors of dilution ranging from 1:1 to 1:10 or 40 µL of ouabain ranging from 1 mM to 10-4 mM. After 15 minutes of incubation in a 96 multi-well plate, 36 µL of ATPase activity buffer was added to each well.

Next the well plate is incubated for 5 more minutes, and afterward, the absorption of the samples is measured simultaneously using a spectrophotometer (PerkinElmer EnVision

Multimode Plate Reader 2100) for 10 minutes. The NKA phosphorylation reaction was started by adding 15 µL ATP (30 mM) to each well. The rate of decline in absorption at 340 nM - caused by NADH oxidation- is an index of the rate for ATPase hybridization. The ATPase activity of NKA is determined by calculating the rate of decline in absorption for various concentrations of ouabain and OLC samples. Apparent affinity constant K0.5 was determined based on ATPase activity of individual samples in the presence of multiple concentrations of ouabain and OLCs by nonlinear regression with a Hill equation in the following form:

푑[퐴푇푃] 푉max . 퐿표푔 [푂퐿퐶]푏 = 푏 푏 푑푡 퐿표푔 [푂퐿퐶] + 퐾0.5

Red Blood Cell Assay This assay uses human red blood cells to determine if the unknown OLC compound has any inhibitory effects on the Na+, K+ ATPase in a high order mammal. A pooled bottle of red blood cells is shaken gently 2 times to ensure that the red blood cells are mixed well before

104

they are washed to be used for experiments. The red blood cells must be washed 3 times with a cold Rb free wash solution before they are treated with known and unknown compounds to determine the biological activity of each. From the bottle of pooled blood, 2 mL of red blood cells are taken and placed into a 13-mL rounded bottom tube. When working with red blood cells, it is essential to use wide bore pipette tips when transferring to vials to prevent bursting of the cells. Next, 8 mL of isotonic, cold 5K/Rb-free buffer is added to the 13-mL tube, mixed gently, and spun in a centrifuge with a 500 g/RCF for 10 min at 6°C to sediment the RBCs for the first washing of the red blood cells. The supernatant which contains the plasma and buffer phase is pipetted into a bio-waste bottle that contains 15% bleach. The washings are completed two more times, and after the last wash, any residue solution that is leftover must be removed by using blotting paper strips. After removal of all supernatant, the hematocrit should equal 80-

90%. Next, add enough 5K/Rb-free buffer to have a 2.4 mL final volume and a final hematocrit of 50%. This solution is kept cold at 4C until it is time for use.

After the red blood cells are thoroughly washed, they are now ready to undergo various treatments. Reaction mixes are made which contain known concentrations of ouabain (20 nM to 10 mM) or an unknown sample, 800 µL of Rb-free uptake buffer or 500 µM RbCl uptake buffer, and 150 µL of the pooled RBCs that are in Rb-free buffer. Once these components are placed in a tube together, they are incubated for 2 hours at 37°C and mixed gently every 30 min during the incubation. After the 2-hour incubation, the samples are removed from the water bath, placed in a chilled tray, and 800 µL of cold Rb-free wash buffer is immediately pipetted into the chilled tubes to stop the reaction. Once the reaction is finished, the treated red blood cells must be rewashed to remove the extracellular Rb that is present. The washings are done

105

by adding 8 mL of Rb-free buffer to the sample and mixing gently. Then the samples are placed in a centrifuge and run at 500g for 10 min at 6°C. The pellet is kept, and the supernatant is discarded. The washing is done 3 times to ensure there is no extracellular Rb left behind.

Now that the samples are treated and washed they must be digested to homogenize and solubilize the red blood cells into the solution to measure the internal elemental content.

The treated red blood cells are placed into a 10-mL glass digestion tube along with 1 mL of concentrated trace grade nitric acid and 400 µL of internal standard mix (500 ppb of In & Sc and

5 ppm of Y & Ce). Once all 3 components are added to the vial, the samples are placed on a hot block for 1 hour at 100°C. After the digestion, the samples are filled up to 10 mL with doubly deionized water (DDI) to dilute the acid concentration in the sample. From the 10-mL sample, 1 ml is taken, placed in a new vial, and filled to 4 ml with DDI for further dilution. Now the samples are ready for ICP-MS analysis.

Reversibility Red Blood Cell Assay Ouabain is known to have a prolonged off-rate once it is bound to the Na+, K+ ATPase.

This assay is used to determine if the purified unknown compound is more reversible than ouabain. The preparation for this assay is very similar to the standard red blood cell assay regarding washing the blood before the experiment and preparing reaction mixes. The reaction mixes are made in 13 mL tubes and contain 150 µL of red blood cells in Rb-free buffer and 850

µL of 1 out of the 3 following solutions; (1) Rb-free wash buffer, (2) RbCl uptake buffer with 10

µM ouabain, or (3) RbCl uptake buffer with 10 µM B1.1 (unknown OLC). Once the reaction mix is prepared, a pre-incubation is performed for 1 hour at 37°C followed by 3 washes of the red blood cells and removal of the supernatant. After the pre-incubation steps, either 850 µL of Rb-

106

free wash buffer, RbCl uptake buffer with 10 µM ouabain, or RbCl uptake buffer with 10 µM

B1.1 are added to the sample for a Rb uptake incubation which is done for 1 hour at 37°C. After the second incubation, the red blood cells are washed three times, and the supernatant is removed. Once the treatments for the red blood cells are complete, they must go through the same digestion protocol that was used for the standard red blood cell assay. When the digestion protocol is completed, the sample can be analyzed using ICP-MS.

5.2.3 Instrumentation For the red blood cell assay and the reversibility red blood cell assay, the selected elements - 23Na, 31→47P, 32→48S, 39K, 56Fe, 66Zn, and 85Rb were determined by ICP-MS by the external calibration method, using an Agilent 8800 inductively coupled plasma mass spectrometer triple quadrupole (ICP-MS QQQ, Agilent Technologies, USA equipped with a micro concentric nebulizer, (Meinhard MircroMist, Meinhard, USA), a Peltier cooled double pass spray chamber, standard torch, and autosampler. Data analysis was performed using Mass Hunter workstation version 4.1 for ICP-MS QQQ software (Agilent Technologies, USA) to determine the total concentration of elements in the tissues and solutions. Two tune modes (helium and oxygen mode) were used sequentially to ensure proper ionization and interference removal.

Internal standards for Sc, Y, In, & Ce were used to represent the full mass range.

5.2.4 Statistical Analysis Data analysis was performed using Mass Hunter workstation version 4.1 for ICP-MS

QQQ software. Excel was used to conduct statistical analysis and make graphs for the assays presented in this chapter. All results for this project were processed and stored in Excel documents.

107

5.3 Results and Discussion 5.3.1 ATPase Assay The ATPase assay is a colorimetric assay where ATP is consumed by the Na+, K+ ATPase and converted to ADP + Pi which is released in solution. The stoichiometric reaction is shown below.

1 ATP → 1 ADP + 1 Pi

The inorganic phosphate that is released into solution reacts with a fluorescent dye which represents the functionality of the Na+, K+ ATPase. If a compound such as a cardiotonic steroid is exposed to the system, it will cause inhibition to the enzyme which will cause less fluorescence to occur. This happens because if the pump is inhibited, it would be able to consume ATP which will stop the production of inorganic phosphate, so there will be no fluorescence to monitor.

For this assay, we used ouabain as a control and compared it to the unknown compound that was purified from pig skeletal muscle in chapter 4. Ouabain solutions were made at various concentrations ranging from 10-3 to 10-8 M to create a standard curve that can be used to compare with the unknown compound. Figure 5.1 contains a plot that shows the ATPase activity being changed based on the ouabain concentration. The ouabain solution that has a concentration of 10-3 M causes complete inhibition of the enzyme, while the concentrations at

10-7 to 10-8 have 100% enzyme activity. For this assay curve, the IC50 is 5.75E-006, which is the amount need to achieve 50% inhibition.

108

Figure 5.1: ATPase Assay of Ouabain. This plot shows a decrease of enzyme activity as the concentration of ouabain increases. Ouabain has an IC50 of 5.751E-006, which is the amount required to reduce Na+, K+ ATPase activity by 50%.

Next, the unknown compounds were tested with the ATPase assay. Since the actual concentration is not known for the unknown compounds, we use an ouabain equivalent concentration that was obtained from the competition binding assay. The results for unknown samples, B1.1 and B1.2, are shown in Figure 5.2. When the results are compared to ouabain, it shows the same trend of the enzyme activity decreasing as the concentration of the sample increases. But, the results also show that a higher concentration is needed for the unknown compounds to cause inhibition compared to ouabain. Sample B1.1 requires an ouabain equivalent concenctration of 100 M to produce complete inhibition and has an IC50 of 0.0474.

Sample B1.2 needs an ouabain equivalent concentration closer to 101 M to cause complete

109

inhibition and has an IC50 of 0.2751. These concentrations are much higher compared to ouabain which requires 10-3 M to cause complete inhibition of the enzyme.

Since a higher concentration of samples B1.1 and B1.2 are needed to cause 100% inhibition of the enzyme, this shows that the unknown compounds are not as potent or strong as ouabain. These unknown compounds may have a weaker affinity for the enzyme compared to ouabain which could mean it may have a broader range of therapeutic uses. Further physiological studies must be done to characterize the functionality of these possible novel endogenous cardiotonic steroids. Next, the red blood cell assay is done to determine if these unknown compounds can inhibit enzyme activity in higher order mammals such as humans.

Afterward, another assay will be conducted to determine if the unknown compound can easily unbind from the enzyme in comparison to ouabain.

110

Figure 5.2: ATPase Assay of OLC B1.1 & B1.2. These plots show the decrease in enzyme activity as the concentration of the unknown compound increases. Sample B1.1 requires 100 M ouabain equivalent concentration to cause complete inhibition of the enzyme, and B1.2 needs close to 10-1 M. B1.1 has an IC50 of 0.0474 and B1.2 has an IC50 of 0.2751.

111

5.3.2 Red Blood Cell Assay The red blood cell assay is used to determine if the unknown compounds that were extracted from pig skeletal muscle will work in higher order mammals such as humans. For this assay, we measure the inhibition of the Na+, K+ ATPase transport activity by monitoring the 85Rb concentration using ICP-MS. First, a negative control was tested to determine if the mobile phase from the last HPLC separation would interfere with this experiment. The mobile phase contained a composition of 60% acetonitrile with 10 mM ammonium formate in doubly deionized water. The results for the negative control are shown in Figure 5.3. The mobile phase tested was spiked with 10 µM ouabain to determine if we would see an additional effect from the mobile phase. As seen in the results, there is no additional effect from the mobile phase itself; we only see the effect from the spike of 10 µM ouabain. Therefore, using this mobile phase for the HPLC separation does not affect the physiological experiment.

Next, various ouabain concentrations ranging from 20 nM to 10 mM were used as a standard curve for this method to obtain an ouabain equivalent concentration for the unknown samples. The unknown samples B1.1 and B1.2 were tested in this assay with two different dilution factors, a 1:1 dilution and 1:10 dilution to determine if the unknowns are dose- dependent. Figure 5.4 shows the results of this experiment. From the results, it is shown that

B1.1 and B1.2 have inhibitory effects on the Na+, K+ ATPase and they both have an ouabain equivalent concentration between 2 µM to 20 µM at a 1:1 dilution. When the dilution factor is changed for the unknown samples to a 1:10 dilution, the enzyme shows more activity which means that both unknowns are dose-dependent.

112

0.04

0.0342 0.0351 0.035

0.03 0.0292

0.0241 0.0250 0.025

0.02

0.0150

Rb/Fe Signal Rb/Fe 0.015

0.01

0.005

Figure 5.3: Red Blood Cell Assay of Negative Controls. It is shown that the mobile phase used for the final HPLC separation does not affect the enzyme.

113

0.016

0.014 0.0124 0.0120 0.0121 0.0116 0.012 0.0112 0.0099 0.01

0.0078 0.0080 0.008

0.0058 Rb/Fe Signal Rb/Fe 0.006

0.00410.00380.0041 0.004

0.002

Figure 5.4: Red Blood Cell Assay of OLC B1.1 & B1.2. OLC B1.1 & B1.2 both have a ouabain equivalent concentration between 2 µM to 20 µM ouabain. They are both also dose- dependent.

114

5.3.3 Reversibility Red Blood Cell Assay

The reversibility red blood cell assay is used to determine how well the unknown compounds bind and unbind from the Na+, K+ ATPase compared to ouabain. Ouabain is known to have a slow off-rate from the Na+, K+ ATPase, making it toxic and more difficult to use for therapeutic medicines. We want to determine if the unknown compound has the same or different off-rate from the enzyme in comparison to ouabain. The reversibility red blood cell assay is similar to the standard red blood cell assay because we use the 85Rb concentration to measure the activity of the Na+, K+ ATPase using ICP-MS. There are two incubation periods for this method. The pre-incubation is the first step, and it is used to determine how well the compounds bind to the Na+, K+ ATPase. The Rb-uptake incubation period is the second step, which is used to observe how well the compounds unbind from the enzyme. A series of samples were prepared for this experiment. First, a red blood cell sample was treated with only RbCl free Tyrode’s Solution for both incubation periods to determine the concentration of endogenous Rb present within the red blood cells. The level of endogenous Rb present in the blood cells will be used to subtract from any treated samples throughout this experiment. Next, the treated red blood cell samples were prepared by either exposing known or unknown cardiotonic steroids to the samples during the pre-incubation period or the Rb-uptake incubation period, but not both periods at the same time. Figure 5.5 shows the results of the various sample treatments with ouabain and the unknown OLC compounds, B1.1 and B1.2.

115

Figure 5.5: Reversibility Red Blood Cell Assay of OLC B1.1 & B1.2.

The data shows the Rb/Fe signal for the samples used in this experiment. Iron (Fe) is used to normalize the Rb signal because it is very abundant and stable in red blood cells due to the presence of hemoglobin. The first sample was treated with 850 µL of RbCl uptake buffer during the pre-incubation period and washed with 850 µL of RbCl free tyrods during the Rb uptake buffer incubation, which resulted in an Rb/Fe signal of 0.05912. The second sample was treated like the first sample except it was spiked with 10 µM ouabain during the primary incubation which shows a 51.1% decrease in the Rb/Fe signal compared to the sample that was not spiked. The third sample did not contain ouabain during the first incubation, but ouabain was present for the second incubation to determine if ouabain can easily unbind from the Na+,

K+ ATPase. After washing out the ouabain, we can see a 35.5% increase in the Rb/Fe signal

116

which shows the enzyme regained some of its biological activity after being washed to remove ouabain.

The next few samples were treated with either B1.1 or B1.2 using the same method that was used for ouabain. When sample B1.1 was used during the first incubation period, it caused an 11.4% decrease in enzyme activity. Comparing the reduction of enzyme activity of B1.1 to ouabain shows that sample B1.1 has a weaker binding affinity for the Na+, K+ ATPase compared to ouabain. Next, B1.1 was used in the second incubation period and not only was the original

Rb/Fe signal recovered, but we obtained a higher Rb/Fe signal with a 49.89% increase which means that the Na+, K+ ATPase is being stimulated. We see very similar effects compared to

B1.1 when the sample is treated with B1.2. This trend of seeing a stimulatory effect is shown throughout the literature for other studies, but it is not entirely understood why this stimulatory effect happens so further studies would need to be conducted to understand this mechanism better.

5.4 Conclusion

It is known that cardiotonic steroids can be extracted from various living organisms, mostly plants, but not many CTS have been extracted from mammalian tissues. The unknown endogenous compounds that were extracted in chapter 4 were tested in multiple assays to determine their physiological functions. Learning the physiological functions of these unknown compounds is essential because these molecules could potently be used in new medicines for the treatment of various medical conditions, and could lead to discoveries within the field. The assays used to obtain the physiological information was the ATPase assay, red blood cell assay,

117

and the reversibility red blood cell assay. More assays can be used in the future to determine kinetic information of the unknown compounds, but the assays that were used answered the preliminary questions for this study.

The ATPase assay showed that the unknown samples, B1.1 & B1.2, are capable of causing inhibition to the Na+, K+ ATPase, but it requires a higher ouabain equivalent concentration than ouabain itself. This information tells us these particular compounds could potently have more of a therapeutic use than ouabain which is considered to be toxic. The red blood cell assay was able to show that B1.1 and B1.2 can be used in higher order mammals such as humans to cause inhibition to the Na+, K+ ATPase. We were also able to see that the unknown compounds were dose-dependent based upon changing the concentration within the samples. By using the reversibility red blood cell assay, we were able to see that the unknown compounds have a faster off-rate from the enzyme compared to ouabain. We also found that they have the ability to cause stimulatory effects to the Na+, K+ ATPase, but this effect needs to be investigated in future studies.

Having established that the unknown compounds function in comparably to ouabain in important respects, the next step was to chemically characterize these compounds using various analytical techniques. The methods used were mass spectrometry, liquid chromatography coupled to mass spectrometry (LC-MS), and nuclear magnetic resonance spectroscopy (NMR). Obtaining information from these various techniques will help us to understand the structure and function relationship between the compound and the Na+, K+

ATPase. Chapter 6 will cover this topic in more detail.

118

Acknowledgments

This work was supported by the National Institutes of Health, NIAMS grant RO1-

AR063710. I would like to thank Dr. Heiny and Hesam Hakimjavadi from the Department of

Molecular and Cellular Physiology (University of Cincinnati), Dr. Julio Landero from the

Department of Chemistry (University of Cincinnati), and David Cowart from Ohio University for all their help and support with this project.

119

Chapter 6: Characterization of Novel Endogenous Cardiotonic Like Steroids Extracted &

Purified from Sus domesticas Skeletal Muscle Tissue

6.1 Introduction

In Chapter 5 unknown OLC compounds were tested and shown to have biological activity at concentrations that could be useful for therapeutic applications in mammalian tissues. Since the results are promising, it is important to characterize these unknown OLC compounds. For characterization of these unknown OLC compounds the molecular weight and structure must be determined by using mass spectrometry and nuclear magnetic resonance

(NMR) respectively. It has been shown throughout the literature that mass spectrometry and

NMR have both been used for structure elucidation of cardiotonic steroids. One example from the literature shows how a research group in Brazil was able to elucidate the structure of nine major bufadienolides extracted from parotoid gland secretions of the Cuban endemic toad

(Peltophryne fustiger) [60]. The nine bufadienolides were ψ-bufarenogin, gamabufotalin, bufarenogin, arenobufagin, 3-(N-suberoylargininyl) marinobufagin, bufotalinin, telocinobufagin, marinobufagin and bufalin [60].

For this project, the first task is to determine the molar weight of the unknown OLC compounds using mass spectrometry and LC-MS. Afterward, the OLC samples will be analyzed using NMR to obtain a 1H NMR spectra, to start off. If the 1H NMR spectra for the unknown OLC resemble other known cardiotonic steroids, other NMR experiments will be conducted such as

COSY-45, HSQC, HMBC, TOCSY, 2D NOSEY, 2D ADEQUATE, 2D INADEQUATE, and carbon dept to obtain more information to help solve the structure. Both of these techniques are very

120

powerful and useful for the aim of this project. Once the molecular weight and structure are determined for the unknown OLC samples, I can move forward to conduct more physiological studies and better understand the behavior of this novel molecule.

6.2 Experimental

A ThermoFinnigan LTQ XL mass spectrometry (ThermoFisher, USA) was used to determine the molar weight of the unknown compounds purified in chapter 4. The unknown samples were prepared in 50% acetonitrile, 50% water, and 0.1% of formic acid which were obtained from Thermo Fisher Scientific or Sigma-Aldrich (USA). Once all components were added together in a vial, it is placed on a vortex to thoroughly mix the sample before analysis.

After the sample is prepared, it is put in a 500 µL syringe which was obtained from Eppendorf

(USA). The syringe is used to introduce the sample to the instrument by direct injection. This method is a destructive technique, so the sample cannot be recovered after analysis. Xcalibur version 2.0 (2.0.7) was the software used to conduct data analysis.

After obtaining the molecular weight, it is of interest to determine if there is more than one compound present in the unknown sample that gives the fragmentation pattern observed.

To perform LC-MS, an HPLC was coupled to a G2 Waters mass spectrometer (Waters Corp.,

USA) to further separate the unknown sample and obtain masses for the analytes that separate out. The column used for LC-MS was a HILIC column that was obtained from Phenomenex

(USA). The mobile phase used for this protocol contained 90% acetonitrile, 10% water, and

0.1% formic acid. All the solutions were obtained from either Thermo Fisher Scientific or Sigma-

121

Aldrich (USA). The flow rate was 0.1 mL/min for this experiment. MassLynx version 4.1 was the software used to perform data analysis and generate graphs.

Next, structural information is needed after obtaining mass for the unknown sample. A

Bruker iVDr 600 MHz NMR (Bruker Corp., USA) is used to obtain 1H NMR information. The buffers needed for this method are a phosphate buffer and Trimethylsilylpropanoic Acid (TSP).

These solutions were obtained from (USA). NMR tubes are also required for this method, and they are obtained from (USA). The unknown samples are added to the NMR tubes along with the buffers and gently mixed for analysis. This technique is a non-destructive technique so the sample can be used for a different method after analysis, following special treatments. Various types of software were used to perform data analysis such as TopSpin 3.5 p17, CMC-assist 2.6,

MestreNova version 12, and Chem Sketch.

6.3 Results and Discussion

6.3.1 Mass Spectrometry

The ThermoFinnigan LTQ XL mass spectrometer was used to determine the molar weight of the unknown OLC compound. This instrument has the capability to measure the mass with up to 2 decimal places of accuracy and based on isotopic abundances it can provide a theoretical elemental composition. The samples were introduced to the instrument by direct injection in a 500 µL syringe. First, to make sure the instrument is properly calibrated, a known sample must be introduced. The known sample used was ouabain which is the same standard used in our physiological experiments in chapter 5. The elemental composition for ouabain is

C29 H44 O12, and the molar weight is 584.652 g/mol. When the ouabain standard is introduced to the mass spectrometer via direct injection, we receive a parent peak at 585.17 m/z (Figure

122

6.1). Next, an MSMS is taken of 585.17 m/z to see the fragmentation pattern of this mass to charge ratio (Figure 6.2). The fragmentation occurs by increasing the collision energy within the system.

Figure 6.1: Direct Infusion of Ouabain using a ThermoFinnigan LTQ XL Mass Spectrometer. Ouabain is known to have a molar weight of 584.652 g/mol. When ouabain is introduced to the mass spec, it gives a parent peak at 585.17 m/z which confirms that the instrument is properly calibrated.

123

Figure 6.2: MSMS of Ouabain at 585.17 m/z. This mass spectra shows that fragmentation pattern of 585.17 m/z of ouabain.

124

After analyzing ouabain, OLC B1 was introduced to the mass spectrometer via direct injection (Figure 6.3). The results show that OLC B1 has a molecular ion peak at 737 m/z and it has fragmentations of 499.75 and 262.17 m/z. This data was acquired in positive mode. When these mass-to-charge ratios are compared with other cardiotonic steroids, nothing else compares to what we see in our experiment. There could be a strong possibility a novel CTS could be discovered, but further investigation will be done by using LC-MS.

Fraction B1

Figure 6.3: Mass spectra of OLC B1. This mass spectra shows that parent peak as well as the fragmentation pattern for OLC B1.

125

6.3.2 LC-MS

LC-MS was used to determine if the unknown sample would separate any further and obtain masses of any analytes that may separate out. A G2 waters mass spectrometer was used for this method. This instrument can provide a high-resolution mass spectra, and it can obtain exact mass which allows the mass spectrometer to measure the mass up to 4 decimal places.

This instrument provides a more accurate mass compared to the ThermoFinnigan mass spectrometer. With this ability, this instrument can distinguish compounds of similar masses and give a better prediction of elemental compositions. The column used for this method was a

HILIC column. The stationary phase of this column is composed of cross-linked diol groups for polar separations. Non-polar molecules elute from the column first, and polar molecules retain in the column giving later elution times.

First, ouabain was introduced to the system to make sure the instrument is calibrated correctly. Figure 6.4 shows a total ion chromatogram (TIC) that was generated to determine where the analytes at different masses are being eluted. An extracted ion chromatogram (EIC) at 585.3017 m/z was generated from the TIC to determine where on the chromatogram this mass-to-charge ratio was eluted. Figure 6.5 shows the mass spectra generated from the peak in the EIC that represented the mass-to-charge ratio of 585.3017. This data confirms that the instrument is properly calibrated and can be used to analyze the unknown sample that was purified from pig skeletal muscle.

126

Figure 6.4: Total Ion and Extracted Ion Chromatogram of Ouabain.

127

Figure 6.5: Mass Spectra of Ouabain Obtained from the EIC.

128

The unknown sample OLC B1 was processed the same as ouabain. First, a TIC was obtained, then the EIC’s for different masses were generated. Figure 6.6 shows the respective chromatograms of unknown sample OLC B1. The TIC shows multiple peaks, but only a couple of areas contain the masses of interest. The masses of interest were 497.1688 and 237.0885 m/z because these masses are similar to what was seen in the unknown sample that was directly injected into the ThermoFinnigan mass spectrometer. When these two masses are observed in the EIC, we can see there are two peaks that contain the same mass-to-charge ratios. This can be an indication that the same 2 compounds (OLC B1.1 & B1.2) are isomers of each other. The

MSMS of each mass were also observed Figure 6.7 & 6.8. From the MSMS we can see how mass 497.1688 further fragments into 237.0885, and how 237.0885 fragments no further. This shows that we are losing a repeat unit, but to determine what it is, further structural analysis needs to be conducted.

129

Figure 6.6: Total Ion & Extracted Ion Chromatogram of OLC B1.

130

Figure 6.7: MSMS of mass 497.1688 from OLC B1.

131

Figure 6.8: MSMS of mass 237.0885 from OLC B1.

132

6.3.3 Structure Elucidation by NMR

To determine the structure of the unknown OLC compounds, a Bruker iVDr 600 MHz

NMR spectrometer was used. A 1H NMR was obtained for OLC samples B1.1 & B1.2 and compared to each other along with other compounds. The other compounds that were compared to the unknown OLC compounds were obtained by predictions from a software called MestreNova. Figure 6.9 shows the NMR spectra for sample OLC B1.1 and B1.2. The NMR spectra look very similar, but there are small differences between these 2 compounds. This is a strong indication that these unknown compounds are isomers of each other. Next, OLC B1.2 is compared to known cardiotonic steroids to see if there are any similarities.

B1.1

B1.2

Figure 6.9: NMR spectra of unknown OLC compounds.

133

First, the NMR spectra of OLC B1.2 was compared to ouabain which has a 5-membered lactone ring. The comparison can be seen in Figure 6.10. It appears that OLC B1.2 and ouabain do not compare to one another. OLC B1.2 seems to have a less complex structure compared to ouabain. Since OLC B1.2 is expected to contain a 6-membered lactone ring, it is compared to a cardiotonic steroid that includes the same structural feature.

Ouabain

B1.2

Figure 6.10: Comparison between Ouabain & B1.2.

Next, OLC B1.2 is compared to bufalin, which is a CTS that contains a 6-membered lactone ring. Figure 6.11 shows the comparison between these two molecules. It is shown that these 2 compounds have no similarities either and once again it seems that OLC B1.2 is less complex than the CTS it's being compared to. Since neither one of the cardiotonic steroids is like the unknown compound, the unknown sample is compared to other compounds to

134

determine if there are any sources of contamination from solutions that were used throughout the purification process.

Bufalin

B1.2

Figure 6.11: Comparison between Bufalin & B1.2.

To further review the NMR spectra for the unknown compounds (Figure 6.9), it is noticeable that the intensity of the peaks is very high around 110,000 arbitrary units which is an indication that the sample is very concentrated. These OLC compounds should not have a high concentration due to typically being in the microgram range upon extraction. Therefore, it is believed that there is possible contamination present. First, OLC B1.2 is compared to Tris which is used as a buffering agent to help maintain a physiological pH within the solution. Figure 6.12

135

shows the comparison between these two compounds. It is shown that Tris is not present in the OLC B1.2 sample because there are no overlapping peaks present within the NMR spectra.

Tris

B1.2

Figure 6.12: Comparison between Tris & B1.2.

The same comparison was made with ammonium formate, HEPES, EDTA, and saponin.

None of these compounds were like OLC B1.2, so it is known that none of the solutions used throughout the purification process contributed to contaminating the unknown sample. The comparisons for the last 4 compounds to OLC B1.2 are shown in Figure 6.13. Further investigation must be done to determine what is present in the OLC samples.

136

Ammonium

Formate

B1.2

HEPES

Figure 6.13: Comparison of B1.2 with possible contaminants.

137

EDTA

B1.2

Saponin

Figure 6.13 continued: Comparison of B1.2 with possible contaminants.

138

6.4 Conclusion

After comparing other cardiotonic steroids and possible contaminants to the OLC samples, it was shown that none of the possible contaminants was responsible for the NMR spectra. When looking at Figure 6.9, it is shown that the sample is very concentrated due to the intensity of the OLC samples, so it is believed that another compound is present along with the unknown OLC being studied. The other molecule that could be present is from a class of compounds called glycans. Glycans are simple and complex carbohydrates that are synthesized in the endoplasmic reticulum and end up at the plasma membrane where the glycans or glycoproteins are either secreted or embedded in the plasma membrane [61]. There is a strong possibility glycans can be present especially since they have masses in the same range as cardiotonic steroids. If this is the case, another separation would need to be done before structurally characterizing the unknown OLC compound.

A column with orthogonal separation principle to reverse phase with higher resolving power than HILIC is needed for this final separation in which a porous graphitic carbon column

(PGC) is recommended . The PGC column is an entirely robust and versatile column with the ability to separate mixtures across the whole pH range, and a variety of solvents may be used for separations. This column is more hydrophobic and chemically more stable than any reversed phase column [62]. It can also be used to separate oligosaccharide alditols, fluorescently labeled glycans, permethylated glycans, and it is also very efficient in separating structural isomers [62]. With the power of the PGC column, I can attempt to separate the unknown OLC compounds further to determine if another compound such as a glycan is present. If the sample does not further separate, then it is possible that the material shown in

139

Figure 6.9 is the end product and this compound is giving the biological activity seen in Chapter

Acknowledgments

This work was supported by the National Institutes of Health, NIAMS grant RO1-

AR063710. I would like to thank Dr. Judith Heiny from the Department of Molecular and

Cellular Physiology (University of Cincinnati), Dr. Julio Landero from the Department of

Chemistry (University of Cincinnati), Dr. Balu Addepalli & Dr. Robert Ross from the Department of Chemistry – Mass Spec Facility (University of Cincinnati), and Dr. Miki Watanabe from the

NMR Metabolomics Core – Division of Pathology, Cincinnati Children’s Hospital Medical Center for all their help and support with this project.

140

Chapter 7: Summary, Conclusions, and Future Work This last chapter gives a summary of the work done in chapters 2 through 6 with proposals for future work that can be executed.

In chapter 2, I discuss a method to quantify metal ion transport by using ICP-MS in intact cells. To understand physiology in both health and disease, it is essential to obtain an accurate measurement of cellular levels, distribution, and flux of metal ions and metal-containing proteins and metabolites. Mouse skeletal muscle, which is a multicellular tissue, and suspended human red blood cells were used to validate the method using ICP-MS and ICP-MS-MS to measure metal ion transport. ICP-MS can measure multiple metal ions in the same sample. This ability allows for the use of sulfur content in skeletal muscle and the iron content in red blood cells to be used for normalization and an elemental index of cell mass, to increase precision and statistical power. The K+ transport rate of Na+, K+ ATPase measured by ICP-MS in resting mouse

EDL muscle was 506 nMol K/g-min at 32 °C, measured with tracer Rb using muscle S content as a mass index. For mouse EDL muscles, which weigh less than 12mg, conventional normalization to wet weight overestimates muscle mass and thereby underestimates Na+, K+ ATPase transport rate by about 20%. A direct comparison of Rb content measured by ICP-MS and tracer

86Rb in human RBCs further validated the method in a cell suspension without interstitial spaces. For microliter volumes of RBCs, the highest precision and accuracy was obtained using

ICP-MS with Fe-based mass. The Na+, K+ ATPase transport rate of human RBCs under our conditions was 18.6 nMolK/g-min at 37 °C. The low transport rate of human RBCs reflects the low density of Na+, K+ ATPase in RBC membranes.

141

Collectively, these results validate the use of ICP-MS to measure ion transport by the

Na+, K+ ATPase, a vital transmembrane protein that transports metal ions, in small biological samples. Key advantages of ICP-MS are that it allows a broader range of experimental designs and physiological contexts than possible using non-radioisotopes, and the capability of ICP-MS to measure multiple metal ions in the same sample provides a more accurate index of cell mass for normalization. In future studies, this ability can be exploited further to study the movement of multiple metal ions whose transport is inter-related by coupled exchangers or secondary transporters.

In chapter 3, I discuss ICP-MS-MS analysis of micro biological samples with heteroatoms as an internal tag for mass-free quantification of selected elements. By exploiting the multi- elemental capabilities and detection power of ICP-MS-MS instrumentation, it was possible to extend the application of this technique to biological samples of sub-microgram mass. Total elemental composition proved to be a more reliable normalization tool than the cell number counted by a hemocytometer for cell cultures of H. capsulatum, B. dermatitidis, and RAW macrophage cells, or fiber number for single skeletal muscle cells.

RSD can be achieved using the elemental composition ratio (Zn/S or P, Fe/S or P, etc.) when sufficient sample size is not available to obtain an accurate element-to-mass average. The

142

ability to normalize to cell P or S content is particularly useful for small samples of cells from rare or clinical specimens for which a cell count may not even be possible.

The proposed protocol applies to many cell cultures and tissues at microgram levels without concern of sample size variability among groups. To obtain reproducible results, the experimental protocols need to be revised to avoid contamination with low-quality reagents, buffers containing S or P, or vessels which can leech S or P.

In chapter 4, I discuss the isolation and purification of novel endogenous cardiotonic steroids from Sus domesticas skeletal muscle tissue. Cardiotonic steroids are known to be extracted from various living organisms. The skeletal muscle of pigs was chosen for this study for several reasons. The first reason is that it is known that approximately 84% of cardiotonic steroids is stored in skeletal muscle. The second reason being that cardiotonic steroids in mammals have not been heavily studied or characterized. Based on these circumstances it was of high interest to pursue cardiotonic steroids in mammalian tissues.

The batch affinity extraction aided in making the purification process efficient by eliminating compounds that do not interact with the Na+, K+ ATPase. This method works well for cardiotonic steroids that have a high affinity for the Na+, K+ ATPase, but if there are cardiotonic steroids that have weaker affinities for the enzyme, it may not be possible to purify those compounds as well. If cardiotonic steroids with weaker affinities to the Na+, K+ ATPase are of interest in future studies, a method would need to be designed to extract them.

143

After the batch affinity extraction, a series of chromatographic techniques are used such as size exclusion, reverse phase, and HILIC to purify the endogenous cardiotonic steroids from pig skeletal muscle. After using these various methods, two fractions could be collected at the end of the purification process. These two fractions are called B1.1 and B1.2 which are very pure, and they show a robust biological activity according to the competition binding assay.

Now that these endogenous compounds are extracted they can be characterized, and the physiological functions can be determined, which is discussed in chapters 5 and 6.

In chapter 5, I discuss the physiological effects of novel endogenous cardiotonic steroids extracted from Sus domesticas skeletal muscle tissue. To determine the physiological effects of the unknown purified compound, I used three different assays that show how it affects or interacts with the Na+, K+ ATPase. The three assays used are the ATPase assay, red blood cell assay, and the reversibility red blood cell assay. Ouabain is used as a control in all three tests and used to compare to the activity of the unknown OLC. The ATPase assay is a colorimetric

+ + assay where 1 ATP is consumed by the Na , K ATPase, and 1 ADP & 1 Pi is released in solution.

The Pi reacts with a fluorescent dye and if the unknown purified compound causes inhibition to

+ + the Na , K ATPase, less Pi will be released which will show a decrease in ATPase activity. Since the concentration is not known for the unknown compound, I can only present an ouabain equivalent activity to describe the biological activity of the sample. The unknown compound does cause inhibition to the Na+, K+ ATPase according to the ATPase assay, but it requires a higher ouabain equivalent activity to do so.

144

The red blood cell assay shows that the unknown sample can cause inhibition to the pump, which has an ouabain equivalent concentration between 2 – 20 µM, and that the sample is dose-dependent. The reversibility of the red blood cell assay is used to determine how well the unknown sample unbinds from the pump compared to ouabain. Ouabain is known to have a high affinity for the Na+, K+ ATPase. Its high affinity is due largely to its slow off rate; once bound it is hard to remove, making it toxic. When ouabain is compared to the unknown compound, we see that the unknown is much more reversible than ouabain is, which means it has a much broader range of therapeutic uses compared to ouabain. With the information obtained from these three assays, we can see that the unknown compound can be a beneficial candidate to be used in future studies for various medical conditions. Since this unknown compound is showing promising data, we move on to chapter 6 which talks about the structural characterization of this molecule.

In chapter 6, I discuss the characterization of novel endogenous cardiotonic steroids extracted and purified from Sus domesticas skeletal muscle tissue. To determine the structural characteristics of the unknown compound, multiple methods were used such as mass spectrometry, LC-MS, and NMR. Mass spectrometry and LC-MS were used to determine the molecular weight of the unknown compound. From the mass spectra, it was decided that the molecular mass was 737 m/z with fragmentations of 499 m/z and 262 m/z for the unknown sample. These particular masses and fragmentation pattern did not match any other known cardiotonic steroids.

145

Future Directions

Proton NMR was used as a first attempt for a structure elucidation of the unknown OLC sample. An 1H NMR spectra was obtained for the unknown molecule, and surprisingly the spectra was not similar to any other known cardiotonic steroids. Typically cardiotonic steroids are in µM concentrations, but from analyzing the NMR spectra, whatever is present has a very high concentration. There is a strong possibility that there is another molecule present along with the unknown cardiotonic steroid. Given the mass spectra and the NMR signature, the other compound could be a glycan. Mainly because they are known to have interactions with the cell membrane and to have repeating sugar units which is consistent with the repeat fragmentation pattern observed in the mass spectral data. This finding brings about a couple of new research questions. Are the cardiotonic steroids causing inhibition to the Na+, K+ ATPase, or could it possibly be glycans causing inhibitory effects if they are present? Also, can I separate the molecule that is interfering with the cardiotonic steroids, if so, will the cardiotonic steroids still show the same biological activity? Answering these questions could lead to a new area of research and future projects.

To answer some of those questions, a PGC column will be used to try to further separate samples B1.1 and B1.2, and if separation occurs the fractions for each sample will be collected and tested using various physiological assays. If something novel is discovered a manuscript will be prepared to publish in a journal that fits the scope of this research topic.

146

Glossary: Abbreviations ADSOL: Adenine, Dextrose, Sorbitol, Sodium Chloride, and Mannitol BAE: Batch Affinity Extraction CBA: Competition Binding Assay CTS: Cardiotonic Steroids Da: Daltons DDI: Doubly Deionized Water DNA: Deoxyribonucleic Acid EDL: Extensor Digitorum Longus EDTA: Ethylenediaminetetraacetic Acid EGTA: Ethylene glycol-bis(β-aminoethyl ether)-N,N,N',N'-tetraacetic Acid EIC: Extracted Ion Chromatogram HEPES: 4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid HILIC: Hydrophilic Interaction Liquid Chromatography HPLC: High-Performance Liquid Chromatography ICP-MS: Inductively Coupled Plasma Mass Spectrometry ICP-MS QQQ: Inductively Coupled Plasma Mass Spectrometry Triple Quadrupole kDa: Kilodalton LA-ICP-MS: Laser Ablation Inductively Coupled Plasma Mass Spectrometry LC-MS: Liquid Chromatography coupled to Mass Spectrometry MS: Mass Spectrometry Na+, K+ ATPase: Sodium-Potassium Pump NADH: Nicotinamide adenine dinucleotide NIST: National Institute of Standards and Technology NKA: Na+, K+ ATPase NMR: Nuclear Magnetic Resonance Spectroscopy OLC: Ouabain Like Compound

147

PFA: Perfluoroalkoy PG: Preparative Grade PGC: Porous Graphitic Carbon Column RBCs: Red Blood Cells RP: Reverse Phase RSD: Relative Standard Deviation SEC: Size Exclusion Chromatography STD: Standard TA: Tibialis Anterior TIC: Total Ion Chromatogram TSP or TMSP: Trimethylsilylpropanoic Acid

148

References 1. Betts, J. Gordon, et al. Anatomy & Physiology. OpenStax College, Rice University, 2017.

2. Horisberger, J.D., et al., Structure-function relationship of Na,K-ATPase. Annu Rev

Physiol, 1991. 53: p. 565-84.

3. Geering, K."FXYD Proteins: New Regulators of Na-K-ATPase." AJP: Renal

Physiology 290.2 (2006)

4. Pierre, S. V. and Z. Xie (2006). "The Na, K-ATPase receptor complex." Cell biochemistry

and biophysics 46(3): 303-315.

5. Bagrov, Alexei Y., and Joseph I. Shapiro. "Endogenous Digitalis: Pathophysiologic Roles

and Therapeutic Applications." Nature Clinical Practice Nephrology 4.7 (2008): 378-92.

6. Berg, Jeremy Mark., John L. Tymoczko, and Lubert Stryer. "Chapter 13 Membrane

Channels and Pumps." Biochemistry. 6th ed. New York: W. H. Freeman and, 2007. 357.

7. Prassas, I. and E. P. Diamandis (2008). "Novel therapeutic applications of cardiac

glycosides." Nat Rev Drug Discov 7(11): 926-935.

8. Zhang, Yurong, Zuyi Yuan, Heng Ge, and Yanping Ren. "Effects of Long-term Ouabain

Treatment on Blood Pressure, Sodium Excretion, and Renal Dopamine D1 Receptor

Levels in Rats." Journal of Comparative Physiology B 180.1 (2010): 117-24.

9. Hamlyn, John M., and Paolo Manunta. "Endogenous Ouabain: A Link Between Sodium

Intake and Hypertension." Current Hypertension Reports 13.1 (2011): 14-20.

10. Jacobs, B. E., Y. Liu, M. V. Pulina, V. A. Golovina, and J. M. Hamlyn. "Normal Pregnancy:

Mechanisms Underlying the Paradox of a Ouabain-resistant State with Elevated

149

Endogenous Ouabain, Suppressed Arterial Sodium Calcium Exchange, and Low Blood

Pressure." AJP: Heart and Circulatory Physiology 302.6 (2012): H1317-1329.

11. Cheval, L. and A. Doucet (1990). "Measurement of Na-K-ATPase-mediated rubidium

influx in single segments of rat nephron." American Journal of Physiology-Renal

Physiology 259(1): F111-F121.

12. Curfman, G. D., et al. (1977). "Thyroid-induced alterations in myocardial sodium-

potassium-activated adenosine triphosphatase, monovalent cation active transport, and

cardiac glycoside binding." The Journal of Clinical Investigation 59(3): 586-590.

13. Cheng, X. J., et al. (1997). "PKA-mediated phosphorylation and inhibition of Na(+)-K(+)-

ATPase in response to beta-adrenergic hormone." American Journal of Physiology-Cell

Physiology 273(3): C893-C901.

14. Sharon, Maheshwar, and Madhuri Sharon. "Chapter 5 Scintillation Counter." Nuclear

Chemistry. Boca Raton, FL: CRC, 2009. 84-92.

15. Moysich, K. B., et al. (2002). "Chernobyl-related ionising radiation exposure and cancer

risk: an epidemiological review." The Lancet Oncology 3(5): 269-279.

16. Ammann, Adrian A. "Inductively Coupled Plasma Mass Spectrometry (ICP MS): A

Versatile Tool." Journal of Mass Spectrometry 42.4 (2007): 419-27.

17. "Journal of Analytical Atomic Spectrometry." Determination of Radioactive Cesium

Isotope Ratios by Triple Quadrupole ICP-MS and Its Application to Rainwater following

the Fukushima Daiichi Nuclear Power Plant Accident - (RSC Publishing).

18. May, Thomas W., and Ray H. Wiedmeyer. "A table of polyatomic interferences in ICP-

MS." ATOMIC SPECTROSCOPY-NORWALK CONNECTICUT- 19 (1998): 150-155.

150

19. Houk, R. S. (1986). "Mass Spectrometry of Inductively Coupled Plasmas." Anal. Chem.

20. Gray, A. L. D., A. R. (1983). "Inductively Coupled Plasma Source-Mass Spectrometry

using Continuum Flow Ion Extraction." Analyst.

21. Profrock, D. P., A. (2012). "Inductively Coupled Plasma-Mass Spectrometry (ICP-MS) for

Quantitative Analysis in Environmental and Life Sciences: A reveiw of Chanllenges,

Solutions, and Trends." Appl. Spectros.

22. Diez Fernandez, S., Sugishama, N., Ruiz Encinar, J. & Sanz-Medel (2012). "Triple Quad

ICPMS (ICPQQQ) as a New Tool for Absolute Quantitative Proteomics and

Phosphoproteomics." Anal. Chem.

23. Szpunar, J. (2005). "Advances in Analytical Methodology for Bioinorganic Speciation

Analysis: Metallomics, Metalloproteomics and Heteroatom-Tagged Proteomics and

Metabolomics." Analyst.

24. Sanz-Medel, A. (1998). "Trace Element Analytical Speciation in Biological Systems:

Importance, Challenges, and Trends." Spectrochim Acta Part B: At. Spectrosc.

25. Haraguchi, H. (2004). "Metallomics as integrated biometal science." Journal of Analytical

Atomic Spectrometry 19(1): 5-14.

26. Koppenaal, D. W., et al. (2004). "Collision and reaction cells in atomic mass

spectrometry: development, status, and applications." Journal of Analytical Atomic

Spectrometry 19(5): 561-570.

27. Clausen, T. (2003). "Na+-K+ pump regulation and skeletal muscle contractility."

Physiological reviews 83(4): 1269-1324.

151

28. Radzyukevich, T., et al. (2009). "The cardiac glycoside binding site on the Na, K-ATPase

α2 isoform plays a role in the dynamic regulation of active transport in skeletal muscle."

Proceedings of the National Academy of Sciences 106(8): 2565-2570.

29. Hoffman, J. F. (1966). "The red cell membrane and the transport of sodium and

potassium." The American journal of medicine 41(5): 666-680.

30. Dunham, P. B. and J. F. Hoffman (1970). "Partial purification of the ouabain-binding

component and of Na, K-ATPase from human red cell membranes." Proceedings of the

National Academy of Sciences 66(3): 936-943.

31. Liang, M., et al. (2007). "Identification of a pool of non-pumping Na/K-ATPase." Journal

of Biological Chemistry 282(14): 10585-10593.

32. Maixent, J. and I. Berrebi-Bertrand (1993). "Turnover rates of the canine cardiac Na, K-

ATPases." FEBS letters 330(3): 297-301.

33. Balcaen, L., et al. (2015). "Inductively coupled plasma–Tandem mass spectrometry (ICP-

MS/MS): A powerful and universal tool for the interference-free determination of (ultra)

trace elements–A tutorial review." Analytica chimica acta 894: 7-19.

34. Vanhoe, H. (1993). "A review of the capabilities of ICP-MS for trace element analysis in

body fluids and tissues." Journal of trace elements and electrolytes in health and disease

7(3): 131-139.

35. Ahmad, I. and F. J. Maathuis (2014). "Cellular and tissue distribution of potassium:

physiological relevance, mechanisms and regulation." Journal of plant physiology 171(9):

708-714.

152

36. Altura, B. and B. Altura (2016). "Importance of ionized magnesium measurements in

physiology and medicine and the need for ion-selective electrodes." J Clin Case Studies

1(2): 1-4.

37. Olufson, K. P. and G. Moran (2016). "Polyatomic interference removal using a collision

reaction interface for plutonium determination in the femtogram range by quadrupole

ICP-MS." Journal of Radioanalytical and Nuclear Chemistry 308(2): 639-647.

38. Jackson, B. P., et al. (2015). "Advantages of reaction cell ICP-MS on doubly charged

interferences for arsenic and selenium analysis in foods." Journal of Analytical Atomic

Spectrometry 30(5): 1179-1183.

39. Jakubowski, N., et al. (1998). "Sector field mass spectrometers in ICP-MS."

Spectrochimica Acta Part B: Atomic Spectroscopy 53(13): 1739-1763.

40. Rodushkin, I., et al. (2013). "Isotopic analyses by ICP-MS in clinical samples." Analytical

and bioanalytical chemistry 405(9): 2785-2797.

41. Avivar, J., et al. (2012). "Fully automated lab-on-valve-multisyringe flow injection

analysis-ICP-MS system: an effective tool for fast, sensitive and selective determination

of thorium and uranium at environmental levels exploiting solid phase extraction."

Journal of Analytical Atomic Spectrometry 27(2): 327-334.

42. Garcia, M. L. and G. J. Kaczorowski (2016). "Ion channels find a pathway for therapeutic

success." Proceedings of the National Academy of Sciences 113(20): 5472-5474.

43. Becker, J. S., et al. (2014). "Bioimaging mass spectrometry of trace elements–recent

advance and applications of LA-ICP-MS: A review." Analytica chimica acta 835: 1-18.

153

44. Wu, B. and J. S. Becker (2012). "Bioimaging of metals in rat brain hippocampus by laser

microdissection inductively coupled plasma mass spectrometry (LMD-ICP-MS) using

high-efficiency laser ablation chambers." International Journal of Mass Spectrometry

323: 34-40.

45. Sussulini, A. and J. S. Becker (2015). "Application of laser microdissection ICP–MS for

high resolution elemental mapping in mouse brain tissue: A comparative study with

laser ablation ICP–MS." Talanta 132: 579-582.

46. Rakszewska, A., et al. (2014). "One drop at a time: toward droplet microfluidics as a

versatile tool for single-cell analysis." Npg Asia Materials 6(10): e133.

47. Mueller, L., et al. (2014). "Trends in single-cell analysis by use of ICP-MS." Analytical and

bioanalytical chemistry 406(27): 6963-6977.

48. Fetita, C., et al. (2012). An automated approach for single-cell tracking in

epifluorescence microscopy applied to E. coli growth analysis on microfluidics biochips.

Medical Imaging 2012: Biomedical Applications in Molecular, Structural, and Functional

Imaging, International Society for Optics and Photonics.

49. Shigeta, K., et al. (2013). "Sample introduction of single selenized yeast cells

(Saccharomyces cerevisiae) by micro droplet generation into an ICP-sector field mass

spectrometer for label-free detection of trace elements." Journal of Analytical Atomic

Spectrometry 28(5): 637-645.

50. Todolí, J. L. and J. M. Mermet (2006). "Sample introduction systems for the analysis of

liquid microsamples by ICP-AES and ICP-MS." Spectrochimica Acta Part B: Atomic

Spectroscopy 61(3): 239-283.

154

51. Figueroa, J. A. L., et al. (2016). "Metal ion transport quantified by ICP-MS in intact cells."

Scientific reports 6: 20551.

52. Wrobel, K., et al. (2010). "Phosphorus and osmium as elemental tags for the

determination of global DNA methylation—A novel application of high performance

liquid chromatography inductively coupled plasma mass spectrometry in epigenetic

studies." Journal of Chromatography B 878(5-6): 609-614.

53. Cornelius, F., et al. (2013). "A structural view on the functional importance of the sugar

moiety and steroid hydroxyls of cardiotonic steroids in binding to Na,K-ATPase." J Biol

Chem 288(9): 6602-6616.

54. Blaustein, Mordecai P. "The Cellular Basis of Cardiotonic Steroid Action."Trends in

Pharmacological Sciences 6 (1985): 289-92.

55. Harashima, H., et al. (1992). "Significance of binding to Na, K-ATPase in the tissue

distribution of ouabain in guinea pigs." Pharmaceutical research 9(4): 474-479.

56. Kjeldsen, K., et al. (1985). "Significance of skeletal muscle digitalis receptors for [3H]

ouabain distribution in the guinea pig." Journal of Pharmacology and Experimental

Therapeutics 234(3): 720-727.

57. Mandel, F., A. Vasiliev, et al. (2003). "Using Na,K-ATPase Itself for Large-Scale Isolation

and Purification of Endogenous Digitalis-Like Factors." Annals of the New York Academy

of Sciences 986(1): 617-619.

58. Valente, R. C., et al. (2010). "Diverse actions of ouabain and its aglycone ouabagenin in

renal cells." Cell Biology and Toxicology 26(3): 201-213.

155

59. Schoner, W. (2002). "Endogenous cardiac glycosides, a new class of steroid hormones."

European Journal of Biochemistry 269(10): 2440-2448.

60. Perera Córdova, W. H., et al. (2016). "Bufadienolides from parotoid gland secretions of

Cuban toad Peltophryne fustiger (Bufonidae): Inhibition of human kidney Na+/K+-

ATPase activity." Toxicon 110: 27-34.

61. Helenius, A. and M. Aebi (2004). "Roles of N-Linked Glycans in the Endoplasmic

Reticulum." Annual Review of Biochemistry 73(1): 1019-1049.

62. Ruhaak, L., et al. (2010). "Glycan labeling strategies and their use in identification and

quantification." Analytical and bioanalytical chemistry 397(8): 3457-3481.

156