USF Libraries
USF Digital Collections

ECIS assessment of cytotoxicity and trans-endothelial migration of metastatic cancer cells

MISSING IMAGE

Material Information

Title:
ECIS assessment of cytotoxicity and trans-endothelial migration of metastatic cancer cells
Physical Description:
Book
Language:
English
Creator:
Opp, Daniel
Publisher:
University of South Florida
Place of Publication:
Tampa, Fla
Publication Date:

Subjects

Subjects / Keywords:
Micromotion
Cytochalasin B
Protien kinase inhibitor H7
Intravasation
Extravasation
Dissertations, Academic -- Physics -- Doctoral -- USF   ( lcsh )
Genre:
non-fiction   ( marcgt )

Notes

Summary:
ABSTRACT: The investigations conducted within this dissertation centers around the use of electric cell-substrate impedance sensing (ECIS). This system is able to characterize in real-time analysis, the adhesion of cells to their substrate and neighboring cells. With this, valuable information can be gathered with in-vitro experiments regarding a tissue culture's response to physiological stimulation. This dissertation has taken advantage of ECIS' ability to analyze toxicology, barrier function, and cancer invasion on a tissue culture. With proper analysis modifications, trans-epethelial resistance (TER) can be used as a cytotoxicity assay with higher sensitivity than previously thought. In vitro assessment of cytotoxicity based on TER needs more quantitative methods to analyze the alteration of cell morphology and motility.Here, we applied ECIS to evaluate dose-dependent responses of human umbilical vein endothelial cells (HUVEC) and mouse embryonic fibroblasts (NIH 3T3) exposed to cytochalasin B and protein kinase inhibitor H7. To detect subtle changes in cell morphology, the frequency-dependent impedance data of the cell monolayer were measured and analyzed with a theoretical cell-electrode model. To detect the alternation of cell micromotion in response to cytochalasin B and H7 challenge, time-series impedance fluctuations of cell-covered electrodes were monitored and the values of power spectrum, variance, and variance of the increment were calculated to verify the difference. While a dose-dependent relationship was generally observed from the overall resistance of the cell monolayer, the analysis of frequency-dependent impedance and impedance fluctuations distinguished cytochalasin B levels as low as 0.1μM and H7 levels as low as 10 μM for HUVEC and 3T3 layers.Even though overall resistance values are relatively small for 3T3 layers, and frequency scan measurements are negligible, impedance fluctuation analysis reveals significant micromotion for cytotoxic detection. Our results show that cytochalasin B and H7 causes a decrease of junctional resistance between cells and an increase of membrane capacitance. Cigarette smoke is cytotoxic and tumorigenic. Initial studies were conducted to evaluate the cytotoxicity of cigarette smoke condensate (CSC) on HUVEC layers. The focus was then turned to investigations involving in vitro cancer invasion assays with CSC on HUVEC layers. ECIS is an excellent investigative device that can be utilized to observe cancer invasion on normal tissue cultures due to the significantly higher impedance signature of cancer cells.The investigation in this dissertation focused on cigarette smoke's influence on cellular mechanics of endothelial cells and the invasive potential of two ovarian cancer cell lines (ALST and OVCA429) against a fully active endothelium. The HUVEC cultures responded to CSC with an increase in junctional binding, where as ALST and OVCA429 relieved adhesion thereby providing an improved motility when evaluated in wound healing assays. Transmigration of the HUVEC layer by ALST cells exhibit a pre-CSC exposure time-dependence affecting the effectiveness of ALST transmigration. The HUVEC layer's decreased tight junction binding that resulted from CSC exposure, allowed for a more aggressive ALST layer formation that occurred during simulated intravasation. Increased HUVEC layer tight junction binding that occurred in the first five hours in response to CSC during extravasation contributes to impeding ALST transmigration at high concentrations of CSC.Overall, CSC has an impeding effect on ALST transmigration during extravasation while causing aggressive transmigration during intravasation.
Thesis:
Dissertation (Ph.D.)--University of South Florida, 2009.
Bibliography:
Includes bibliographical references.
System Details:
Mode of access: World Wide Web.
System Details:
System requirements: World Wide Web browser and PDF reader.
Statement of Responsibility:
by Daniel Opp.
General Note:
Title from PDF of title page.
General Note:
Document formatted into pages; contains 113 pages.
General Note:
Includes vita.

Record Information

Source Institution:
University of South Florida Library
Holding Location:
University of South Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
aleph - 002064131
oclc - 567549006
usfldc doi - E14-SFE0003064
usfldc handle - e14.3064
System ID:
SFS0027380:00001


This item is only available as the following downloads:


Full Text

PAGE 1

ECIS Assessment of Cytotoxicity and Trans-En dothelial Migration of Metastatic Cancer Cells by Daniel Opp A dissertation submitted in partial fulfillment of the requirements for the degree of Doctor of Philosophy Department of Physics College of Arts and Sciences University of South Florida Major professor: Chun-Min Lo, Ph.D. Dale Johnson, Ph.D. Myung Kim, Ph.D. Garrett Matthews, Ph.D. Kay-Pong Yip, Ph.D. Date of Approval: June 29, 2009 Keywords: micromotion, cytochalasin B, pr otein kinase inhibito r H7, intravasation, extravasation Copyright 2009, Daniel Opp

PAGE 2

ACKNOWLEDGMENTS I am grateful and appreciative of the opportunities given to me by my advisor Dr. Chun-min Lo. I consider myself very fortunate to have been able to spend my graduate dissertation work within his lab. I have benefited from his experience, knowledge, and demonstration of hard work and dedication. I would like to thank Dr. Dale Johnson, Dr. Myung Kim, Dr. Garrett Matthews, and Dr. Ka y-pong Yip for serving on my dissertation committee. I would also like to thank Ma ry Ann Prowant and Daisy Matos for their assistance with anything I needed. I am inde bted to my parents, br others, sister, and Dr. Kerriann Greenhalgh for their love and support.

PAGE 3

i TABLE OF CONTENTS LIST OF TABLES iv LIST OF FIGURES vii ABSTRACT xvi CHAPTER 1 INTRODUCTION A ND BACKGROUND Motivation and Aim 1 ECIS Historical Review 2 Dissertation Outline 3 CHAPTER 2 ECIS SYSTEM AND THEORY ECIS System 5 Theoretical Model 7 Frequency Scan 10 Micromotion 11 CHAPTER 3 MATERIALS AND METHODS Cell Culture Procedures 14 CSC 15 ECIS 15 Cytotoxicity Assay 16 Effects of CSC on Wound Healing (Gap Recovery) 17 Effects of CSC on Extravasation/Intr avasation 17

PAGE 4

ii CHAPTER 4 CELL SPECIFIC IMPEDANCE ANALYSIS Attach and Spreading 21 Frequency Scan 22 CHAPTER 5 USE OF ECIS TO ASSESS CYTOTOXICITY Introduction 31 HUVEC-Cytochalisin B Variance 33 HUVEC-H7 Variance 34 3T3-Cytochalisin B Variance 43 3T3-H7 Variance 48 HUVEC-Cytochalasin B Frequency Scan 53 HUVEC-Cytochalasin B micromotion 57 HUVEC-H7 micromotion 65 3T3-Cytochalasin B micromotion 72 3T3-H7 micromotion 73 Discussion 85 CHAPTER 6 EFFECT OF CSC ON CANCER INVASION Introduction 87 Wound Recovery 89 HUVEC-CSC 92 OVCA429-CSC & ALST-CSC 99 Invasion Assay 103

PAGE 5

iii Discussion 103 CHAPTER 7 SUMMARY AND FUTURE WORK Conclusion 108 Future Work 108 REFERENCES 136 ABOUT THE AUTHOR End Page

PAGE 6

iv LIST OF TABLES Table 3.1 An assay indicating the concentr ation of CSC within the culture medium before and after ALST invasion is initiated. 20 Table 4.1 From frequency scan measurements, the junctional resistance (Rb), cell-substrate separation (h), and cell membrane capacitance (Cm) were determined for the various cell types listed. 30 Table 5.1 Impedance analysis of HUVEC mo nolayer 20 hr after exposure to cytochalasin B. All values, Rb, h, and Cm, were obtained from fitting measured impedance with Eq. 1. Values shown are mean standard error (n = 8). Each value is also expressed as a percentage of control. 56 Table 5.2 ECIS micromotion data obtaine d from confluent HUVEC layers 20 hours after exposure to cytochalas in B. The column labeled Res. is the average resistance value of the cell layer over a 2048-second run. The column labeled power slope is the slope of the log-log plot of power versus frequency. The column labeled Var32 is the statistical variance for the 32-point intervals of the normalized 2048-second data set. The column labeled VoI32 is the variance of the increment for the 32-point sampling intervals of the 2048second data set. The column labeled Hurst is the Hurst exponent. Values shown are mean standard error ( n = 10 for each concentration of cytochalasin B). Each value is also expressed as a percentage of control. 64 Table 5.3 ECIS micromotion data obtained from confluent HUVEC cell layers 20 hours afte r exposure to H7. The column labeled Res. is the average resistance value of the cell layer over a 2048-second run. The column labeled Var. is the statistical variance for the 64-

PAGE 7

v point intervals of th e normalized 2048-second data set. The column labeled Var. of Inc is the variance of increments for the 64point sampling intervals of the 2048-second data set. Values shown are mean standard error ( n = 10). Each value is also expressed as a percentage of control.. 71 Table 5.4 ECIS micromotion data obtai ned from confluent 3T3 cell layers 20 hours after exposure to cytochalas in B. The column labeled Res. is the average resistance value of the cell layer over a 2048-second run. The column labeled Var. is the statistical variance for the 64point intervals of th e normalized 2048-second data set. The column labeled Var. of Inc is the variance of increments for the 64point sampling intervals of the 2048-second data set. Values shown are mean standard error ( n = 18). Each value is also expressed as a percentage of control. 78 Table 5.5 ECIS micromotion data obtaine d from confluent 3T3 cell layers 20 hours after exposure to H-7. The column labeled Res. is the average resistance value of the cell layer over a 2048-second run. The column labeled Var. is the statistical variance for the 64-point intervals of the normalized 2048-s econd data set. The column labeled Var. of Inc is the variance of increments for the 64-point sampling intervals of the 2048-second data set. Values shown are mean standard error ( n =15). Each value is also expressed as a percentage of control. 84 Table 6.1 The recovery rates for various wounded cell cultures against different concentrations of CSC. 91 Table 6.2 Values returned from frequency scan analysis of HUVEC layers. The first column indicates the concentration of CSC ( g/mL). The second column is the analysis of the increasing change in junctional resistance normalized ( Rb). The third column is the change in height normalized ( h).. 94

PAGE 8

vi Table 6.3 The junctional resistan ce reported for various cell types at time of 0 and 40 hours for each cell ty pe. The columns labeled 0 g/mL, 10 g/mL, 50 g/mL, and 100 g/mL note the CSC concentration within the culture medium. All cell types were allowed one day to reach confluency before frequency scan measurements were taken. 100 Table 6.4 The cell-substrate separation repor ted for various cell types at time of 0 and 40 hours for each cell type. The columns lableled 0 g/mL, 10 g/mL, 50 g/mL, and 100 g/mL note the CSC concentration within the culture medium. All cell types were allowed one day to reach confluency before frequency scan measurements were taken. 109

PAGE 9

vii LIST OF FIGURES Figure 2.1 An overview of a single well on the microelectrode array. Here it can be observed how much larger the counter electrode is as compared to the working electrode. (image from Applied Biophysics) 6 Figure 2.2 A diagram of the experimental set up used in the ECIS system. 6 Figure 2.3 Illustration of tissue culture and cell-substrate spacing. Current flows radially under the cells as we ll as through cell-cell junctions. The dashed lines represent membrane capaciatance. is the resistivity of the solution, Vn is the electrode potential Vm is the membrane potential, I is the current flow, and h is the cellsubstrate separation on the order of nanometers. (image from Lo, C.M. et al1) 9 Figure 2.4 A logarithmic graph of resi stance (a) and capacitance (b) measurements for both confluent cultures on electrodes and naked electrodes as a function of frequenc y. 12 Figure 2.5 A logarithmic graph of normali zed resistance (a) and normalized capacitance (b) measurements as observed through ECIS. The normalized values are determined by dividing the cell-covered and cell-free measurements for the corresponding frequency. 13 Figure 3.1 A confluent monolayer (a) was scratched with a micropipette tip to (b) create a gap across the HUVEC monolayer (c). (image from the text Molecular Biology of the Cell) 19 Figure 3.2 The resistance graph of a cancerous invasion of a normal endothelial cell layer can clearly indicate when the monolayer has been compromised. 20

PAGE 10

viii Figure 4.1 ECIS attachment and spreading data measured at 4 kHz showing resistance following inoculation of NIH 3T3 fibroblastic cells (1 105 cells/ cm2) at time zero in eight different electrodecontaining wells. 24 Figure 4.2 ECIS attachment and spreading data measured at 4 kHz showing resistance following in oculation of HUVECs (1 105 cells/cm2) in eight different electrode-containing wells. Cells were seeded 1 hour after the start of monitoring the impedances. 25 Figure 4.3 ECIS attachment and spreading data for ALST cells (84 cell/cm2) measured at 4 kHz. Eight electrodes were initiated one hour before inoculation with ALST cells. 26 Figure 4.4 Frequency scan data representing the changes in log resistance as a function of frequency and time following inoculation of HUVECs on an ECIS electrode well. Freq uency scan measurements were acquired every hour for 12 hours. 27 Figure 4.5 Frequency scan data representing the changes in log capacitance as a function of frequency and time following inoculation of HUVECs on an ECIS electrode well. Frequency scan measurements were acquired every hour for 12 hours. 28 Figure 4.6 Log-log frequency spectroscopy resistance plot of confluent OVCA429 cultures. The top traces are OVCA429 covered electrodes while the bottom traces are cel l free electrodes. 29 Figure 4.7 Log-log frequency spectroscopy re sistance plot of confluent ALST cultures. The top traces are ALST covered electrodes, the bottom traces are cell free electrodes. 29 Figure 4.8 Log-log frequency spectroscopy re sistance plot of confluent NIH 3T3 cultures. The top traces are 3T3 covered electrodes while the bottom traces are cell free electrodes. 30 Figure 5.1 Normalized resistance meas urements of a confluent HUVEC monolayer upon addition of different concentrations of

PAGE 11

ix cytochalasin B. Cells were inoculated into electrode-containing wells and allowed to develop into confluent layers for approximately 24 h. At the time indicated by the arrow, cytochalasin B diluted in DMSO was added to give the final concentrations of 0.1 M (red), 0.5 M (blue), 1.0 M (green), 2.5 M (yellow), 5.0 M (violet), 10 M (cyan), and control (black), and the resultant changes in normalized resistance were followed. Data were collected every 2 minutes for 20 hours. 35 Figure 5.2 Variance data of a confluent HUVEC monolayer upon addition of different concentrations of cytochalasin B at 0.1 M (red), 0.5 M (blue), 1.0 M (green), 2.5 M (yellow), 5.0 M (violet), 10 M (cyan), and control (black). 36 Figure 5.3 Variance of increment analysis of the resistance data in Figure 4.5. The different concentrations of cytochalasin B are 0.1 M (red), 0.5 M (blue), 1.0 M (green), 2.5 M (yellow), 5.0 M (violet), 10 M (cyan), and control (black). 37 Figure 5.4 Average normalized resistance at 0, 5, 10, 15, and 20 h after addition of cytochalasin B ( n = 10 for each concentration) against HUVEC layers. The different concen trations of cytochalasin B are 0.1 M (red), 0.5 M (blue), 1.0 M (green), 2.5 M (yellow), 5.0 M (violet), 10 M (cyan), and control (black). 37 Figure 5.5 Normalized resistance measur ements of confluent HUVEC cell layers upon addition of different con centrations of H7. Cells were inoculated into electrode-containi ng wells and allowed to develop into confluent layers for approximately 24 h. H7 diluted in DMSO was added to give the final concentrations of 10 M (red), 25 M (blue), 50 M (green), 100 M (yellow), 500 M (violet), and control (black), and the resultant changes in normalized resistance were followed. 39

PAGE 12

x Figure 5.6 Variance data of a confluent HUVEC monolayer upon addition of different concentrations of H7 at 10 M (red), 25 M (blue), 50 M (green), 100 M (yellow), 500 M (violet), and control (black). 40 Figure 5.7 Variance of increment analysis of the resistance data in Figure 5.4. The different concentrations of H7 are 10 M (red), 25 M (blue), 50 M (green), 100 M (yellow), 500 M (violet), and control (black). 41 Figure 5.8 Average normalized resistance at 0, 5, 10, 15, and 20 h after addition of H-7 ( n = 10 for each concentration) against HUVEC layers. The different con centrations of H7 are 10 M (red), 25 M (blue), 50 M (green), 100 M (yellow), 500 M (violet), and control (black). 42 Figure 5.9 Normalized resistance measurem ents of confluent NIH 3T3 cell layers upon addition of different con centrations of cytochalasin B. Cells were inoculated into elec trode-containing wells and allowed to develop into confluent layers for approximately 24 h. At the time indicated by the arrow, cytochalasin B diluted in DMSO was added to give the final concentrations of 0.1 M (red), 1.0 M (blue), 2.5 M (green), 5.0 M (yellow), 10 M (violet), and control (black), and the resultant changes in normalized resistance were followed. 44 Figure 5.10 Variance data of a confluent NI H 3T3 monolayer upon addition of different concentrations of cytochalasin B at 0.1 M (red), 0.5 M (blue), 1.0 M (green), 2.5 M (yellow), 5.0 M (violet), and control (black). 45 Figure 5.11 Variance of increment analysis of the resistance data in Figure 5.8. The different concentrations of cytochalasin B are 0.1 M (red), 0.5 M (blue), 1.0 M (green), 2.5 M (yellow), 5.0 M (violet), and control (black). 46

PAGE 13

xi Figure 5.12 Average normalized resistance at 0, 5, 10, 15, and 20 h after addition of cytochalasin B (n = 14 for each concentration) against 3T3 layers. The different concentr ations of cytochalasin B are 0.1 M (red), 0.5 M (blue), 1.0 M (green), 2.5 M (yellow), 5.0 M (violet), and control (b lack). 47 Figure 5.13 Normalized resistance measurem ents of confluent NIH 3T3 cell layers upon addition of different co ncentrations of a protein kinase inhibitor H-7. At the time indicat ed by the arrow, H-7 diluted in complete medium or medium alone was added to give the final concentrations of 1.0 M (red), 10 M (blue), 25 M (green), 50 M (yellow), 100 M (violet), 1000 M (cyan), and control (black), and the resultant changes in normalized resistance were followed. Data were collected every 2 minutes for 20 hours. 49 Figure 5.14 Variance data of confluent NIH 3T3 monolayers upon the addition of H-7 at various concentrations of 1.0 M (red), 10 M (blue), 25 M (green), 50 M (yellow), 100 M (violet), 1000 M (cyan), and control (black). 50 Figure 5.15 Variance of increment analysis for the resistance data of Figure 4.11 The different concentrations of H-7 are 1.0 M (red), 10 M (blue), 25 M (green), 50 M (yellow), 100 M (violet), 1000 M (cyan), and control (black). 51 Figure 5.16 Average normalized resistance at 0, 5, 10, 15, and 20 h after addition of H7 ( n = 15 for each concentration) against 3T3 layers. The different concentrations of cytochalasin B are 1.0 M (red), 10 M (blue), 25 M (green), 50 M (yellow), 100 M (violet), 1000 M (cyan), and control (black). 52 Figure 5.17 Normalized resistance as a function of log10(frequency) obtained from a frequency scan measurement 20 hours after the addition of medium containing cytochalasin B or medium alone to a confluent HUVEC layer to give the fi nal concentrations of 0.1 M (red), 0.5

PAGE 14

xii M (blue), 1.0 M (green), 5.0 M (yellow), 10 M (violet), and control (black). 54 Figure 5.18 Normalized capacitance as a function of log10(frequency) obtained from a frequency scan measurement 20 hours after the addition of medium containing cytochalasin B or medium alone to a confluent HUVEC layer to give the fi nal concentrations of 0.1 M (red), 0.5 M (blue), 1.0 M (green), 5.0 M (yellow), 10 M (violet), and control (black). 55 Figure 5.19 Normalized resistance data reco rded 20 hours after the addition of medium containing cytochalasin B or medium alone to a confluent HUVEC layer to give the fi nal concentrations of 1 M (red), 5 M (blue), 10 M (green), and control (bla ck). Overall, 2048 data points taken at one-second intervals. 58 Figure 5.20 Fast Fourier transform of conf luent HUVEC layers exposed to cytochalisin B concentrations of 1 M (red), 5 M (blue), 10 M (green), and control (black). 59 Figure 5.21 Variance analysis of confluent HUVEC layers exposed to cytochalisin B concentrations of 1 M (red), 5 M (blue), 10 M (green), and control (black). 60 Figure 5.22 Variance of the increment analysis of confluent HUVEC layers exposed to cytochalisin B concentrations of 1 M (red), 5 M (blue), 10 M (green), and control (black). 61 Figure 5.23 Normalized resistance recorded 20 hours after the addition of medium containing H7 or medi um alone to a confluent HUVEC cell layer to give the fi nal concentrations of 50 M (red), 100 M (blue), 500 M (green), and control (black). Each curve consists of 2048 data points taken at one-second intervals. 67 Figure 5.24 Fast Fourier transform of confluent HUVEC cell layers exposed to H7 concentrations of 50 M (red), 100 M (blue), 500 M (green),

PAGE 15

xiii and control (black). Each curve consists of 2048 data points taken at one-second intervals. 68 Figure 5.25 Variance analysis of confluent HUVEC cell layers exposed to H7 concentrations of 50 M (red), 100 M (blue), 500 M (green), and control (black). Each curve consists of 2048 data points taken at one-second intervals. 69 Figure 5.26 Variance of the increment analysis of confluent HUVEC cell layers exposed to H7 concentrations of 50 M (red), 100 M (blue), 500 M (green, and control (black). Each curve consists of 2048 data points taken at one-second intervals. 70 Figure 5.27 Normalized resistance recorded 20 hours after the addition of medium containing cytochalasin B or medium alone to a confluent 3T3 cell layer to give the final concentrations of 5 M (red), 10 M (blue), and control (black). Each curve consists of 2048 data points taken at one-second intervals. 74 Figure 5.28 Fast Fourier transform of conf luent 3T3 cell laye rs exposed to cytochalisin B concentrations of 5 M (red), 10 M (blue), and control (black). 75 Figure 5.29 Variance analysis of confluen t 3T3 cell laye rs exposed to cytochalisin B concentrations of 5 M (red), 10 M (blue), and control (black). 76 Figure 5.30 Variance of the increment analysis of confluent 3T3 cell layers exposed to cytoch alisin B concentrations of 5 M (red), 10 M (blue), and control (black). 77 Figure 5.31 Normalized resistance recorded 20 hours after the addition of medium containing H-7 or medium alone to a confluent 3T3 cell layer to give the fina l concentrations of 10 M (red), 100 M (blue), 1000 M (green), and control (black). Each curve consists of 2048 data points taken at one-second intervals. 80

PAGE 16

xiv Figure 5.32 Fast Fourier transform of conflu ent 3T3 cell layers exposed to H-7 concentrations of 10 M (red), 100 M (blue), 1000 M (green), and control (black). 81 Figure 5.33 Variance analysis of confluent 3T3 cell layers exposed to H-7 concentrations of 10 M (red), 100 M (blue), 1000 M (green), and control (black). 82 Figure 5.34 Variance of the increment analysis of confluent 3T3 cell layers exposed to H-7 concentrations of 10 M (red), 100 M (blue), 1000 M (green), and control (black). 83 Figure 6.1 Images of HUVEC cultures rec overing a gap. The bottom images are an example of a culture that was influenced by 10 g/mL of CSC as compared to a control culture. 90 Figure 6.2 40 hour observation of various co ncentrations of CSC affecting HUVEC layers. The resistance values were normalized for better graphical representation of the behavior CSC induces on HUVECs adhesive properties. 93 Figure 6.3 Images of confluent ALST (top) and HUVEC (bottom) layers. The images on the left are not expos ed to CSC while the images on the right are exposed to CS C concentrations of 250 g/mL. 94 Figure 6.4 30 hour continual frequency s can of a single HUVEC covered electrode exposed to 50 g/mL of CSC. The graph presents the time series of the log-log capacitance frequency plot. 95 Figure 6.5 30 hour continual frequency s can of a single HUVEC covered electrode expos ed to 100 g/mL of CSC. The graph presents the time series of the log-log capacitance frequency plot. 96 Figure 6.6 30 hour continual frequency s can of a single HUVEC covered electrode exposed to 50 g/mL of CSC. The graph presents the time series of the log-log resistance frequency plot. 97

PAGE 17

xv Figure 6.7 30 hour continual frequency s can of a single HUVEC covered electrode exposed to 50 g/mL of CSC. The graph presents the time series of the log-log resistance frequency plot. 98 Figure 6.8 Frequency scan data of juncti onal resistance and cell-substrate separation (number above each column) for ALST cells attaching from a suspended state to form layers. 104 Figure 6.9 Cell-substrate separation (top) and junctional resistance (bottom) values gathered from frequency scans for ALST invasion of HUVEC layers under the influence of 50 g/mL of CSC. 105 Figure 6.10 Resistance traces starting off initially as a HUVEC layer, then being invaded upon by suspended ALST cells. HUVECs (dotted and red traces) were initially exposed to 50 g/mL of CSC for the first twenty hours. At twenty hours, CSC exposure changed (red and green traces), and ALST cells were added. The total observation occurred for 132 hours. 106 Figure 6.11 Fluorescence microscopy images of confluent HUVEC (green) layers being invaded by suspended ALST (red) cells under the influence of 50 g/mL of CSC. (a) HUVEC layer invaded by ALST cells under no CSC influence. (b) HUVEC layer invaded by ALST cells under CSC influence. (c) HUVEC layer exposed to CSC for twenty hours then invaded by ALST. 107

PAGE 18

xvi ECIS ASSESSMENT OF CYTOTOXI CITY AND TRANS-ENDOTHELIAL MIGRATION OF METASTATIC CANCER CELLS Daniel Opp ABSTRACT The investigations conducted within th is dissertation centers around the use of electric cell-substrate impedance sensing (ECIS) This system is able to characterize in real-time analysis, the adhesi on of cells to their substrat e and neighboring cells. With this, valuable information can be gathered with in-vitro experiments regarding a tissue cultures response to physiological stimulation. This dissertation has taken advantage of ECIS ability to analyze toxicology, barri er function, and cancer invasion on a tissue culture. With proper analysis modifications, trans-epethelial resistance (TER) can be used as a cytotoxicity assay with higher sensitivity than pr eviously thought. In vitro assessment of cytotoxicity based on TER needs more quantitative methods to analyze the alteration of cell morphology and motility. Here, we applied ECIS to evaluate dose-dependent responses of human umbilical ve in endothelial cells (HUVEC) and mouse embryonic fibroblasts (NIH 3T3) exposed to cytochalasin B and protein kinase inhibitor H7. To detect subtle changes in cell morphology, the frequencydependent impedance data of the cell monolayer were measured and analyzed with a theoretical cell-electrode mode l. To detect the alternation of cell micromotion in response to cytochalasin B and H7 challenge, time-series impedance fluctuations of cellcovered electrodes were monitored and the values of power spectrum, variance, and variance of the increment were calculated to verify the difference. While a dosedependent relationship was generally observe d from the overall resistance of the cell monolayer, the analysis of frequency-depende nt impedance and impedance fluctuations distinguished cytochalasin B levels as low as 0.1 and H7 levels as low as 10 for

PAGE 19

xvii HUVEC and 3T3 layers. Even though overall re sistance values are relatively small for 3T3 layers, and frequency scan measuremen ts are negligible, impedance fluctuation analysis reveals signif icant micromotion for cytotoxic de tection. Our results show that cytochalasin B and H7 causes a decrease of junctional resistance between cells and an increase of membrane capacitance. Cigarette smoke is cytotoxic and tumorigenic. Initial studies were conducted to evaluate the cytotoxicity of cigarette sm oke condensate (CSC) on HUVEC layers. The focus was then turned to investigations i nvolving in vitro cancer invasion assays with CSC on HUVEC layers. ECIS is an excellent in vestigative device that can be utilized to observe cancer invasion on normal tissue cultures due to the significantly higher impedance signature of cancer cells. The i nvestigation in this dissertation focused on cigarette smokes influence on cellular mechanics of endothe lial cells and the invasive potential of two ovarian can cer cell lines (ALST and OVC A429) against a fully active endothelium. The HUVEC cultures responded to CSC with an in crease in junctional binding, where as ALST and OVCA429 reli eved adhesion thereby providing an improved motility when evaluated in wound h ealing assays. Transmigration of the HUVEC layer by ALST cells exhibit a pre-CS C exposure time-dependence affecting the effectiveness of ALST transmigration. The HUVEC layers decreased tight junction binding that resulted from CSC exposure, a llowed for a more aggressive ALST layer formation that occurred during simulated in travasation. Increased HUVEC layer tight junction binding that occurred in the fi rst five hours in response to CSC during extravasation contributes to impeding ALST transmigration at high concentrations of CSC. Overall, CSC has an impeding effect on ALST transmigration during extravasation while causing aggressive transmig ration during intravasation.

PAGE 20

1 CHAPTER 1 INTRODUCTION AND BACKGROUND Motivation and Aim Biological systems have been investigated in vivo and in vitro In vivo behavior involves highly complex systems with a larg e degree of difficulty in manipulation. Experimental results are qualitative at best as in vivo behavior can easily be misinterpreted or too subtle to notice. The vague nature of in vivo results makes experimental repeatability an issue. In vitro tissue cultures on th e other hand allow for complete environmental manipulation, with ability for quantif ication, as well as experimental repeatability. This allows for better experiment al evaluations with conclusive results. Normal cell tissue cultures observe gr owth inhibition when initiated by a physiological or chemical factor. Endotheli al and fibroblast cel l types continue to proliferate and migrate until i nhibited by contact from other endothelial or fibroblast cells. Once contact inhibition takes place, th e proliferation and motility mechanism stops as the cells form a junction. In culture these mechanisms cause endothelial and fibroblast cells to form c onfluent monolayers. Conflu ent endothelial monolayers are particularly useful to inves tigate cell response to chemical environment due to endothelial interaction with the circulatory system. Transformed, or cancer cells do not behave similarly to normal cells. Missing or over e xpressed molecules in the cell signal cascade result in transformed cells forming tightly bound multi-layers within culture. Discrepancies such as this between norma l and transformed cu ltures along with environmental manipulation make in vitro experimental assays ideal for can cer research.

PAGE 21

2 Experimental evaluation of a tissue cu lture is done by observing cell behavior often through a form of microscopy or imm unoprecipitation. Phase contrast microscopy can be used for basic qualitative analysis of motility, proliferation, and wound healing3. Fluorescence imaging can be used to quant ify specific molecule present to observe markers for apoptosis and cytotoxic reactions4. Protein markers for certain cellular activities can be observed with immunoprecipitation by comparing relative protein expression5. A method for tissue culture evaluation involves observing the tissues response to electrical current, this is known as bioimpedance. Phle bography correlates bioimpedance measurements to volume of blood change in detection of venous th rombosis. A more popular method of bioimpedance measurement is its use in calculating body composition. In this dissertation, bioimpedan ce measurements are scaled dow n as the areas of interest are in micro to nanometer range with more sensitive and rigid analysis. This is particularly useful in tissue culture to evaluate cell-cell a nd cell-substrate interactions as measurements can be applied in a time seri es. A real-time analytic bioimpedance measuring system was developed by Giaever and Keese6. ECIS Historical Review With the commissioning of electric cell-s ubstrate sensing (ECIS) in 1984, Giaever and Keese created micro-sized gold electrodes wh ich could be used as substrates for cell cultures6. Initial experiments involved time series impedance analysis of attachment and spreading for lung fibroblasts WI-38 a nd transformed fibroblasts WI-38 VA 136. Soon thereafter, ECIS was successfully evaluated as toxicology assay analyzing the effects of Tween 20, benzalkonium chloride, Triton X100 and sodium lauryl sulfate on WI-38 fibroblasts and MDCK epithelial cells7. Giaever and Keese then utilized ECIS to analyze the fractal Brownian motion of tissue cultures. They were able to quantify the motion with the Hurst exponent and differentiate WI-38 and WI-38 VA 13 cell lines8.

PAGE 22

3 In 1991, Giaever and Keese further improve d ECIS analysis by developing a theoretical model calcu lating the morphological propertie s of cells based on frequency dependent impedance measurements1. From this model, junc tional resistance and cellsubstrate separation can be derived by comp aring calculated impedance values with frequency scan measurements. Along with overall resistance measurements, junctional resistance and cell-substrate measurements are useful indicators for cancer invasion assays where the junctional resistance is si gnificantly higher and cel l-substrate separation lower for cancer layers as compared to normal cell layers9. Similarly, normal cell layers effectiveness as a barrier agai nst invasion can be evaluated10. Data analysis from this system has also revealed the behavior of endothelial response to medium induced shear stress 11,12, proliferation13, and angiogenesis14. Dissertation Outline Chapter 1 briefly describes the history as well as the abilities of ECIS. The significance of the projects focused on within this dissertation is discussed within this chapter. Chapter 2 focuses on the ECIS system and concept. The electrode to PC system interface has different steps in order to measure tissue culture impedance. The interpretation of tissue culture impedance measurements is based of the theoretical model Giaever and Keese developed. From the m odel, the frequency scan and micromotion ECIS technique become im portant analytic tools. Chapter 3 describes the experimental me thods and materials utilized in this dissertation. An endotheli al, fibroblast, and carcinoma cell lines were used for experiments. A cytotoxic, CSC, and intravas ation/extravasation ECIS assay was used for experimental data acquisition. A wound healing assay and reactive oxygen species measurements were also used for CSC investigations.

PAGE 23

4 Chapter 4 presents the data for ECIS ch aracterization of the various cell types used within this di ssertation. Attachment and spread ing reveals the typical overall resistance values for specific cell types. Frequency scan was used to characterize and emphasize the difference of junctional resist ance and cell-substrate separation between endothelium, fibroblasts, and car cinoma cells. Chapter 5 is the results for the cytot oxicity study. HUVEC and NIH 3T3 cell lines were challenged by cytochalasin B and pr otein kinase inhibitor H7 were the focus of this chapter. The primary emphasis is the e ffectiveness of ECIS micromotion analysis as a tool for cytotoxic detection. Chapter 6 is the ECIS inve stigation into CSC effects on HUVEC layers as well as cancer invasion. The metastat ic cancer lines used were ALST and OVCA429. ALST extravasation/intravas ation assays against HUVEC layers in the presence of CSC were created and observed by ECIS. To verify EC IS measurements of transmigration assays, fluorescent microscopy was used. Chapter 7 summarizes the dissertation with presented conclusions. Future work is also suggested based of preliminary studies conducted.

PAGE 24

5 CHAPTER 2 ECIS SYSTEM AND THEORY ECIS system. ECIS observes impedance changes that correlate with changes in junctional resistance and cell substrate separation of ti ssue cultures. The tissue samples are cultured into gold film microelectrodes wells. As seen in Figure 2.1, the obse rvable electrode is 250 m diameter which sits in the center of the wells, a larger coun ter electrode sits on the edge of the wells. The significantly larg er counter electrode allows for the impedance measurements on the central electrodes to dominate the measurements. The leads from the electrodes connect to an 830 lock-in amplifier from Stanford Research Systems. Impedance analysis was conducted with a 1V 4000 Hz AC signal through a 1M resistor providing a current of 1 A. The ECIS expe rimental set up with tissue cultures on the microelectrodes is depicted in Figure 2.2. The tissue cultures are maintained in an incubator at 37 C with 5% CO2 with a coaxial cable connecting the electrodes to the lock-in amplifier. Each elec trode well used for the ECIS me asurements is well explained by Keese et.al.15. All eight electrode contains sm all working electrode (area= 5-4 cm2) and a large counter electrode (area=0.15 cm2). Because of the diffe rence in surface area, the impedance of the small electrode determines the total impedance of the system. This is due to the impedances inverse dependence on the electrode area. The active electrode is delineated by circular openings (diamete r = 250 m) in a photoresist overlayer that insulates the rest of the deposited gold film from bulk electrolyte. As cells in the culture attach to the electrode, the cell behavior ca n be correlated to the impedance measured by this device. The size of the electrode was inte ntionally selected to be this small as cell related signals are more easily detectable with smaller electrodes. This is because solution resistance in the culture dish is much larger than the electrodes impedance. The

PAGE 25

Figure 2.1 An overview of a singlewell on the microelectrode array. Here it can be observed how much larger the counter electrode is as compared to the working electrode. ( ima g e from A pp lied Bio p h y sics ) Figure 2.2 A diagram of the experimental se t up used in the ECIS system. 6

PAGE 26

7 impedance of the electrode electrolyte interf ace (Faradaic impedance) is proportional to the inverse of the electrode ar ea, but constriction resistance (spreading resistance) for the circular disk electrode in a conducting medium of infi nite extent varies as /2d, where is the resistivity of the medium and d is the diameter of the electrode. The electrode impedance can be made to always domi nate the constriction resistance by using sufficiently small electrodes. When the electrode area is reduced to 10-4 cm2, the Faradaic resistance of the electrode-electrolyte inte rface at 4000 Hz is many times larger than the constriction resistance so the motility of the cells can be easily studied. Theoretical Model Giaever and Keese in 1991 1 developed a model that describes the various impedance measurements observed by the ECIS and the mathematical analysis needed to determine the junctional resist ance and cell-substrate separa tion. The resistance of the culture medium acts in series with the impe dance of both electrodes thereby causing the medium to dominate the impedance measurements. The solution resistance or constriction resistance inherits a dependenc y on electrode size. The impedance of the electrode-electrolyte in terface has to be proportional to th e inverse of the area of the circular electrode, 4/d2. Therefore, the electrode-electro lyte interface can be forced to dominate the constriction resistance by maki ng the electrode sufficiently small. When measured under the proper frequency and proper ly sized electrode, the real part of the impedance (faradaic resistance) is several times larger than the constrictions resistance. As a result the binding and anchoring behavior of the tissue samples can be measured and calculated. The equa tions below demonstrate the math ematics of the physical model used to calculate the specific impedance (i mpedance for a unit area) of a cell covered electrode. A critical assumption is that the cells are approximated to be circular disks of radius rc. for calculation purposes. The frequency dependent specific impedance calculations are based off the model illustrated in Figure 2.3. The model is based on the specific impedance of a cell-free electrode Zn( v ) and the cell layer Zm( v ) as well as the resistivity of the medium. Both Zn( v ) and Zm( v ) exhibit frequency v dependence. The assumption has been made that the current flows radially between the ventral surface of

PAGE 27

the cell and substrate and that the current de nsity in this region is unvarying in the z direction. This equation can be combined to yield Vc is the potential of the electrode, and h is the height of the space between the ventral surface of the cell and the substratum. The equation can be solv ed with the sum of modified Bessel functions of first and second kind. Usi ng proper boundary conditions, the specific impedance for a cell-covere d electrode is obtained as follows: 8

PAGE 28

Figure 2.3 Illustration of tissue culture and cel l-substrate spacing. Current flows radially under the cells as well as through cell-cell junctions. Th e dashed lines represen t membrane capacitance. is the resistivity of the solution, Vn is the electrode potential Vm is the membrane potential, I is the current flow, and h is the cell-substrate separation on the order of nanometers. (ima ge from Giaever and Keese et al1,2) 9

PAGE 29

Where I0 and I1 are modified Bessel functions of the first kind with orde r 0 and 1. The solution exhibits a dependency on two variable parameters which are Rb (the resistance between the cells for a unit area), and (related to cell substrate separation) which can be given as: With the specific impedance values being evaluated as two cell membranes in series, and Rb are the only adjustable parameters in the frequency depende nt equation stated above. Frequency Scan A unique feature of the ECIS system is the application of frequency spectroscopy to the tissue-electrode interface model. In Figure 2.4, the resistance and capacitance are plotted versus frequency on a logarithmic graph. In order to observe impedance attributed to exclusively the cell-layer, im pedance measurements of cell-free electrodes as well as cell-covered electrodes must both be taken into account. Notice the two distinct traces in each graph as being labeled without any ce lls (no cells), and confluent with a HUVEC cell layer (HUVECs). The resistance and capac itance values are normalized by dividing resistance and capac itance values of the cell-free impedance measurements, with the cell-layer impeda nce measurements for the corresponding frequencies. In Figure 2.5 the normalized valu es are plotted as a function of logarithmic frequency. The solid lines in Figure 2.4 repres ent the calculated values of the resistance (Figure 2.4a) and capacitance (Figure 2.4b), as determined from the measured values of impedance. The points represent the calculate d values based off the equations stated in previous section developed for the cell-layer model. The traces the points follow are forced to mimic the normalized resistance and capacitance measurements by the adjustment of the parameters and Rb within the theoretical model. This method of 10

PAGE 30

11 model fitting is used to determine and Rb, which directly relate to the cell-substrate separation and binding of the tight junctions between cells. Micromotion As cells move around on electrodes, the measured impedance fluctuates as the adhesion sites relax and bind. This action continues even wh en cell layers have reached confluency. This behavior is quantified with micromotion data which is analyzed through Fast Fourier transform, variance, a nd variance of the incr ement. Due to the random nature and oscillations of the data, fast Fourier transforms were applied which is the first method used in micromotion analysis The 2048 points are divided into 4 subsets of 512 seconds each, with the square of th e Fourier coefficients (power spectra) calculated with the average power spectrum presented. The second method utilized is calculating the variance for different time peri ods. The 2048 points is split into 512 equal groups of 4 points with the variance calcu lated for each group then averaged for the entire set. The analysis is then repeat ed for groups of 8, 16, 32, 64, etc. points then presented as a log-log plot of average va riance versus number of points (sampling period). Another method for micromotion analys is is the calculation of variance of the increments for the time samples. For 2048 points analyzed with 1 second sampling times, 2047 increments are calculated by su btracting successive data points then calculating the variance for these steps. The sampling time is then increased to 2 seconds giving 1024 points with 1023 increments. This process is continued for sampling times of 4, 8, 16, etc. seconds. If a log-log plot of the variance versus the sampling interval is a straight line, this satisfies the condition of the data being fractal. The Hurst exponent can be calculated by the slope of the straight line.

PAGE 31

(a) (b) igure 2.4 A logarithmic graph of resistance (a) and capacitance (b) measurements for F both confluent cultures on electrodes and nake d electrodes as a function of frequency. 12

PAGE 32

(a) (b) Figure 2.5 A logarithmic graph of normalized re sistance (a) and normalized capacitance (b) measurements as observed through ECIS. The normalized values are determined by dividing the cell-covered and cellfree measurements for the corresponding frequency. 13

PAGE 33

14 CHAPTER 3 MATERIAL AND METHODS Cell Culture Procedure. Samuel Mok at Harvard Medical School supplied the OVCA429 (a standard ovarian cancer cell line used in studies) and ALST (another ovarian cancer cell line) cells in frozen ampoules. They were thawed a nd grown in medium 199 and MCDB 105 (1:1) (Sigma) supplemented with 10% fetal calf se rum (Sigma), 100 units/ml penicillin, and 100 g/mL streptomycin under 5% CO2, and 370C, high humidity atmosphere. The Human Umbilical Vein Endothelial Cells (HUVECs) (Clonetics Corp., San Diego, CA) were cultured at 37 C and 5% CO2 in endothelial cell grow th medium (EGM; Clonetics Corp.) which is supplemented with the following: 10 ng/ml human recombinant epidermal growth factor, 1 g/ml hydr ocortisone, 50 g/ml getamicin, 50 ng/ml amphotericin B, 12 g/ml bovine brain extr act, and 2% fetal bovine serum (amounts indicate final concentration). HUVECs were subculture when they were 70% confluent, and the medium was changed every 48 hours th ereafter. HUVECs passaged less than six times were used in experiments. Confluent HUVEC layers were formed in electrode wells using an inocul ation density of 84 cell/cm2. The 3T3 fibroblas ts, obtained from the American Type Culture Collection (Manassas, VA),were grown in DMEM(4.5 g/L dglucose) (Mediatech, Manassas, VA) supplemented with 10% FBS (Mediatech), 50 ng/mL streptomycin, 50 units/mL penicill in, and 250 ng/mL amphotericin B under 5% CO2, and a 37 C, high-humidity atmosphere. For ECIS micromotion measurements, cells were harvested and grown to confluence 24 h be fore addition of cytochalasin B into the electrode wells, resulting in a cell density that was controlled at 105 cells/cm2. Cytochalasin B (Sigma, St. Louis, MO) was diluted in DMSO as a 10-mM solution before use.

PAGE 34

15 CSC Cigarette smoking condensate (CSC) was purchased from Murty Pharmaceuticals (Lexington, KY) and was prepared by smoking University of Kentucky's 1R3F Standard Research Cigarettes on an FTC Smoke Machine. The stock solution of CSC was then prepared in dimethyl sulfoxide (DMSO) at a concentration of 40 mg/mL and stored in a small aliquot at -70C until used. CSC was diluted with the culture medium to prepare solutions for the experiments. Concentratio ns of 10 and 50 g/mL of CSC were used which correlated to equivalent CSC concentr ations within a light and heavy smokers blood stream. 100 g/mL of CSC was also used to demonstrate an extreme smoke exposure. Out of the total of 16 wells per experimental trial, each concentration of control, 10, 50, and 100 g/mL of CSC were distributed into 4 wells with 0.1 mL of the diluted CSC solutions. Cancer invasion assays were either implemented immediately or 10 hours after the HUVEC monolay ers were exposed to the va rious concentrations of CSC ECIS The software, electrode arrays, and lock in amplifier comprising the ECIS system were acquired from Applied Biophysics (Troy, NY). The electrode arrays contain eight wells that are roughly 1 cm in height and 0.9 cm2 in basal area with a total volume of 0.7 mL. Within each well there is a gold elec trode with a diameter of 250m accompanied by a counter electrode significantly larger. HUVEC, OVCA429 and ALST were deposited into the wells and allowed to reach confluency. Cells atta ched to the electrode surfaces demonstrate insulating characteristics therefore causing current to flow radially between the cell-electrode interface towards cellular junctions. The endothelial layer is evaluated as an RC circuit in series. In all work involving HUVEC monolayer s reported, the electrodes were precoated with 0.2 mg/ml gelatin for 20 minutes. For cell attachment and spreading assay, cells were plated into electrode wells at 105 cells/cm2 density and impedance changes

PAGE 35

16 were measured immediately. For impeda nce measurements of HUVEC monolayers upon addition of cytochalasin B, cells were allowe d to attach and spread for at least 24 hr before impedance was measured. After 24 hour s in culture, the conf luency and viability of the cell monolayer was confirmed by li ght microscopy and electrically by the resistance values. Cytochalasin B (Sigma-A ldrich, ST. Louis, MO) in DMSO or DMSO alone as a control was added to each cell -covered electrode well. The electrical impedance of each well was measured every 2 min and up to sixteen individual wells were followed successively. For detection of cell micromotion, impedance data of each well were taken every second with exquisite sensitivity until 2048 points had been acquired and then another well was measured. The time series data were normalized and numerically analyzed by calcula ting power spectrum, variance (the square of the standard deviation), and variance of the increment as we previously described 24. Experimental frequency dependent impe dance measurements compared to the calculated values of the cell-electrode model allow for the determination of the junctional resistance and cell substrate separation. With a fre quency range of 25 Hz to 60 kHz, the impedance was measured for an electrode covered with cells as well as the same electrode without cells. The measurements acquired from the cel l free electrode scans were used to subtract the background data from culture medium resistance in the cell covered electrode measurements. Cytotoxicity Assay Electrode wells were cultured with confluent layers of HUVEC and NIH 3T3 cells. Both cell types were then exposed to various concentrations of cytochalasin B ranging from 0.1 M to 10 M and various concentrations of protein kinase inhibitor H7 ranging from 1 M to 1000 M. Both cytotoxic chemicals adversely affect the adhesive properties of cells. Before and after cytochal isin B or H7 exposure, impedance data was collected through ECIS.

PAGE 36

17 Effects of CSC on wounding healing (gap recovery) HUVEC, OVCA429, and ALST were cultu red to confluency in 35mm x 10mm polystyrene culture dishes. The intent of this assay was to evaluate the effect CSC has on the motility of the various cell cultures us ed. Two lines were drawn on the outside bottom portion of the dish to be used as reference points when acquiring images from a microscope. Gaps were created by gently dragging a 200L pipette tip across the bottom of the cultured dishes. Figure 3.1 demonstrates how the micropipette tip is used to drag cells cleanly away to form a wound in th e confluent monolayer. This is a common technique utilized in wound reco very assays. Three gaps pe rpendicular to the reference lines per dish were created to allow for six observable sections. Images were captured in two hour time increments by a CCD camera attached to a Zeiss Axiovert 200M inverted microscope. Through pixel analysis and time signature imaging the gaps could be measured and recovery rates calculated. When calculating recovery rates, images taken within the first 2 hours were excluded due to misleading accelerated recovery rates of OVCA429 and ALST layers. The accelerated r ecovery was attributed to overcrowded OVCA429 and ALST populations being able to push cells bordering the gap. This occurrence had to be omitted in order to analyze cancerous cells natural motility mechanism. After two hours, it was observed that the recovery ra te stabilized and was consistent for the subsequent time and imag es. For the HUVEC cultures the experiment can be considered a reendothelialization experiment, where as the OVCA429 and ALST cultures are observed as a collective cell motility experiment. Reendothelialization refers to endothelial cells forming a monolayer after being wounded. Upon contact the two opposing groups of cells on either side of the wounds will cease movement and proliferation. With OVCA429 and ALST, cells will continue to proliferate despite forming a monolayer, thus forming cultures with stacked layers. Effects of CSC on Extravasation/Intravasation HUVECs were cultured to confluency th en exposed to 50 g/mL of CSC. After 20 hours, HUVEC layers were then ch allenged by ALST invasion. Table 3.1 shows the

PAGE 37

18 CSC exposure assay implemented before and after ALST invasion, as well as the label reference used in the impedance analysis data presented later. The CSC exposure exclusively during ALST invasion was intended to simulate the extravasation process of metastatic cancer cells. The exclusive CSC exposure before invasion is intended to mimic the HUVEC and ALST interaction with the CSC during the intravasation process. HUVEC layers have lower impedance values than ALST layers, therefore ALST invasion was interpreted through impedance meas urements monitored by ECIS. Figure 3.2 shows how ECIS is able to observe invasi on of cancerous cells upon a normal cell monolayer. The decreasing resistan ces as compared with control indicate a breaking of junctions of the normal monolayer by the invading cancerous cells. In the case of OVCA429, the cells have been able to penetrate to the bo ttom on the electrode surface and being forming an OVCA429 layer. A continually increasing resistance trace (not shown) indicates OVCA429 beginning to proliferate upon th e initial monolayer.

PAGE 38

(a) (b) (c) Figure 3.1 A confluent monolayer (a) was scratche d with a micropipette tip to (b) create a gap across the HUVEC monolayer (c). (image from the text, Molecular Biology of the Cell ) 19

PAGE 39

Table 3.1 An assay indicating the concentration of CSC within the culture medium before and after ALST invasion is initiated. Cancer Invasion 6000 9000 12000 15000 0481 21 6 Time (hrs)Resistance (ohm)2 0 HOSE15 OVCA429 SKOV3 Control Figure 3.2 The resistance graph of a cancerous i nvasion of a normal endothelial cell layer can clearly indicate when th e monolayer has been compromised. 20

PAGE 40

21 CHAPTER 4 CELL SPECIFIC IMPEDANCE ANALYSIS The data presented in this chapter is fo r purposes of character izing cell types used for the various experiments within this dissert ation. In order to ma ke effective use of ECIS, preliminary data must be acquired showing distinct impedance signatures distinguishing the different cell types. Endot helium (HUVEC), fibr oblast (NIH 3T3), and ovarian carcinoma cells (ALST and OVCA429) were the cell types focused on for the following research. Single frequency impedance measurements and frequency spectroscopy data was used for cell type characterization. Attach and Spreading Resistance traces acquired from ECIS measurements of eight electrode wells inoculated with NIH 3T3 cells are displaye d in Figure 4.1 and HUVEC cells displayed in Figure 4.2. After ECIS measurements were initiated, NIH 3T3 cells were added immediately to the electrode wells while HUVE Cs were added an hour after. Both the resistance traces experienced a temporar y increase from suspended NIH 3T3 and HUVEC cells attaching and spread ing to confluency. At conf luency the resistance for NIH 3T3 stabilizes at ~2.5 k with HUVEC cultures at ~10 k Figures 4.1 and 4.2 verify that despite the eight individual electr odes, the overall behavior of the traces is similar. Note the difference in the overall resistance between HUVEC and NIH3T3, this is due to endothelial cells commonly having st ronger cell-cell junctions as compared with fibroblasts. Small changes in the cell-s ubstrate space due to cell motions caused the impedance to fluctuate with time. Figure 4.3 is the resistance measurements at 4 kHz acquired for electrode wells inoculated with ALST cells. Through the fi rst thirty hours of ALST attachment and

PAGE 41

22 spreading, there is a wide range of resistance values. A reason for this occurrence is that after an ALST culture is trypsinized from a dish to be added to the electrode wells, the centrifuge process causes ALST cells to bind together. Even after thorough dilution with fresh medium, ALST cells can sometimes form a mass of small cluttered cells. This can be seen in the Figure 4.3 as some of the tr aces are quicker to reach an equilibrium resistance of about 25 k Despite the wide range of resistance values within the first thirty hours of attachment and spreading, al l eight wells experience an equilibrium around 25 k It should be noted the large difference between HUVEC (10 k ) and ALST (25 k ) make these two cell types easily distinguishable. Frequency Scan Impedance values for time series data is acquired at a single frequency, but impedance measurements can also be detected using a wide range of frequencies. Figure 4.4 and 4.5 shows frequency sc ans of the attachment and spreading of HUVECs, where the cells were seeded into the electrode well at time zero and the data were continuously recorded for 12 hours. At each sampling time the impedance was measured with an AC signal from 25 Hz to 60 kHz and reported as that due to a series RC circuit. Resistance and capacitance measured at time zero were the values of the cell-free electrode, Rn and Cn. As evidently seen in Figure 4.4, Rn significantly decreased as the applied AC frequency increased. In Figure 4.5, Cn slightly decreased as the applied AC frequency increased. However, the measured resistance and capacitance after time zero varied in different ways as cells attached and spread on the electrodes. At the high frequency range, when cells covered up some of the elect rode area, the resistance increased and the capacitance decreased because the cells impaired the movement of ions, resulting in less current coming out of the electrode. At lo w frequency, both resist ance and capacitance did not change much even when there were cells on the electrode because the impedance from the electrode-electrolyte interf ace dominated the measured impedance 14. Note that the observed impedance alterations are due to cell morphological changes and cannot be explained by changes in medium conductivity.

PAGE 42

23 As shown in Figure 4.4, for HUVECs th e largest change of the measured resistance curve between cell-free and cell-c onfluent electrodes appears at 4 kHz (peak frequency). This frequency is quite different from that for epithelial cells such as MDCK cells, for which the largest change is at 700 Hz 28. However, it is quite similar to fibroblastic cells such as WI-38 VA13 and HGF cells, where the largest change are respectively at about 4 kHz and 6 kHz 14, 45. We have previously modeled the cellelectrode system to understand the type and the magnitude of fitting parameters that could account for the frequency-dependent impedance 28. Basically, when parameter or Cm increases, the peak frequency shifts toward the low frequency side; when parameter Rb increases, the size of the largest resist ance change increases. For the time course measurement of HUVECs, therefore, the AC signa l is usually set at 4 kHz (or at 40 kHz sometimes) to obtain the substantial respons es of resistance (o r capacitive reactance) variations to cellular activities 46. Figures 4.6, 4.7, and 4.8 display the fre quency scan data for OVCA429, ALST, and NIH 3T3 cells, respectively. All sets are compared against cell free electrodes in which that data is used to help determ ine the junctional resistance, cell-substrate separation, and cell membrane capacitance ( list by cell type in Table 4.1). Figure 4.1 demonstrates how fibroblast la yers have a relatively low resi stance value. In Figure 4.8, it is clearly visible how closely the 3T3 frequency scan follows the cell free electrode scan. Due to this, the 3T3 values returned from frequency scan an alysis are negligible and therefore not presented in Table 4.1. Despite that, 3T3 la yers can still be effectively evaluated in a cytotoxicity a ssay through micromotion analysis described in the following chapter.

PAGE 43

0 5 101520Time (hrs) 1500 2000 2500 3000 3500Resistance () Figure 4.1 ECIS attachment and spreading data measured at 4 kHz showing resistance following inoculation of NI H 3T3 fibroblastic cells (1 105 cells/ cm2) at time zero i n eight different electr ode-containing wells. 24

PAGE 44

0 5 10 15 20Time (hrs) 2000 4000 6000 8000 10000 12000 14000Resistance () Figure 4.2 ECIS attachment and spreading data measured at 4 kHz showing resistance following inoculation of HUVECs (1 105 cells/cm2) in eight different electrodecontaining wells. Cells were seeded 1 hour after the start of monitoring the impedances. 25

PAGE 45

Figure 4.3 ECIS attachment and spreading data for ALST cells (84 cell/cm2) measured at 4 kHz. Eight electrodes were initiated one hour before inoculation with ALST cells. 26

PAGE 46

Figure 4.4 Frequency scan data representing the changes in log resistance as a functio n of frequency and time following inoculat ion of HUVECs on an ECIS electrode well. Frequency scan measurements were acquired every hour for 12 hours. 27

PAGE 47

Figure 4.5 Frequency scan data repr esenting the changes in l og capacitance as a functio n of frequency and time following inoculat ion of HUVECs on an ECIS electrode well. Frequency scan measurements were acquired every hour for 12 hours. 28

PAGE 48

Figure 4.6 Log-log frequency spectroscopy resist ance plot of confluent OVCA429 cultures. The top traces are OVCA429 cove red electrodes while the bottom traces are cell free electrodes. Figure 4.7 Log-log frequency spectroscopy resistan ce plot of confluent ALST cultures. The top traces are ALST covered electrodes, the bottom traces are cell free electrodes. 29

PAGE 49

30 Cell type Rb ( cm2) h (nm) Cm ( F/cm2) HUVEC 1.73 0.15 78.7 3.4 2.46 0.09 ALST 10.2 1.40 16.3 1.8 2.51 0.07 OVCA429 56.2 7.91 0.74 0.18 3.72 0.24 Figure 4.8 Log-log frequency spectroscopy resi stance plot of confluent NIH 3T3 cultures. The top traces are 3T3 covered electrodes while the botto m traces are cell free electrodes. Table 4.1 From frequency scan measurements, the junctional resistance (Rb), cellsubstrate separation (h), and cell membrane capacitance (Cm) were determined for the various cell types listed.

PAGE 50

31 CHAPTER 5 USE OF ECIS TO ASSESS CYTOTOXICITY Introduction A promising and unique feature of the ECIS assay is the noise analysis of the impedance time series. A few published ECIS results have demonstrated that cell motion may reveal itself as a fluctuation in the measured impedance, which is always associated with living cells and persists even when the cells grow into a confluent layer 1,16, 2,8. The impedance fluctuation is attributed to incessa nt changes in the size of the cell-substrate space as cells persistently rearrang e their cell-substrate adhesion sites 17. Furthermore, the magnitude of this sort of vertical mo tion detected by ECIS is of the order of nanometers and referred to as micromotion 1. Note that the observed impedance alterations are due to cell morphological changes and cannot be explained by changes in medium conductivity. If sensitive and accurate in vitro cyto toxicity assays are properly developed, animal testing can be greatly reduced 18 Through more sophisticated in vitro analysis, eliminating animal tests stands to benefit anim al welfare and researcher costs. To predict the response to toxins from the entire biological system of the animal subject, methods of quantifying that toxic response on a more basic cellular level are bei ng researched. Cellbased biosensors are able to quantify cellula r response by incorporating viable cells as part of the sensory system 19. When cultured on to electr odes, healthy cells present a unique electrical signature due to their ability to restrict ion transport. Motion of these cells can also be detected relative to their interaction with the cell substrate and is represented as fluctuations in the impedance time series. A method taking advantage of this is known as electrochemical impedance spectroscopy (EIS). Using microelectrode

PAGE 51

32 arrays, EIS is able to detect tissue culture re sponse to various toxins by analyzing the cell motility and morphological properties suggested by the impedance measurements7,20,21. With the commissioning of electric cell-s ubstrate sensing (ECIS) in 1984, Giaever and Keese created micro-sized gold electrodes wh ich could be used as substrates for cell cultures 6. This system has the ability to uniquely measure frequency dependent impedance values of cell cultures. Each indi vidual well of the electrode array has a small electrode and a large counter electrode with an alternate current (AC) signal applied. A lock-in amplifier monitors the voltage as the cu lture medium serves as an electrolyte. A PC stores and processes the in-phase and out-of-phase voltage while regulating between switching measurements among the various elec trodes. ECIS demons trates incredible sensitivity detecting slight cha nges within the cell-substrate in terface. Data analysis form this system has revealed the behavior of cell attachment and spreading22, cell motility1,2,8, barrier function of cell layers1, and in vitro toxicology7,23,24. Traditionally cytotoxicity is quantified with ECIS or EIS through impedance time series analysis derived from response functions half-inhibition concentration. Cells are commonly exposed in two different ways in determining the half-inhibition concentration. Toxin concentrations and e xposure times can be varied to disturb the ability of cells to attach and spread23. Toxins are added to the medium of the electrode wells before introduction of cells that are at tempting to attach and proliferate. Another method involves observing the effect toxins have on cultured monolayers 20,25,26. After confluency is reached, the toxic compounds are added and impedance measurements are observed. EIS assays agreed with natural red uptake (NRU ), 3-(4,5-dimethylthiazol-2yl)-2,5-diphenyltetrazolium bromide (MTT) test, lactate dehydrogenase (LDH) measurement, colony forming efficiency (CFE) growth assay, and standard luminescence-based methods when measuring half-inhibition concentrations20,27. EIS for the most part has identified cytotoxicity through impedance analysis as a time series in which the frequency parameter is held constant. When frequency

PAGE 52

33 spectroscopy is utilized and applied to a theoretical cell-electr ode model, junctional resistance and cell-substrate separation can be calculated from the data 24,28. Cell adhesion to the substrate and neighboring ce lls restricts current thereby making the electrical impedance dependent on intercellular bi nding and cell-substr ate separation. Noise analysis techniques have been applied to the impedance time series to reveal that fluctuations indicate cell moti on. Contact inhibited cells th at have grown to confluent monolayers continue to exhibit noise in im pedance measurements suggesting vertical fluctuations in the cell-substrate separation 1,2,8,16. The fluctuating cell-substrate separation has been detected by the ECIS system on the nanometer scale 1 and has been attributed to the activity of adhesion sites 17. This chapter focuses on the effectiveness of ECIS in detecting a cytotoxic response from HUVEC and NIH3T3 cultures. Specifically, varying concentrations of cytochalisin B were exposed to HUVEC an d NIH3T3 cultures as well as varying concentrations of protein ki nase inhibitor H-7 were expos ed to tissue cultures of NIH 3T3. The response was quantified ECIS impedance measurements of the cell-substrate interactions through analysis of micromotion and frequenc y spectroscopy. NIH 3T3 was shown to be sensitive to 25 M of prot ein kinase inhibitor H-7 and 2.5 M of cytochalisin B while HUVECs were sensitiv e to 0.1 M of cytochalisin B. HUVEC-Cytochalisin B Variance The effect of varying cytochalasin B con centration on the overall resistance of the HUVEC monolayer was monitored for 20 hours. Sixteen electrodes were followed one after another with each wells data point requiring a few seconds. Although the impedance of each electrode well was set to be measured every 2 minutes, fluctuations were observed on each curve at different level. The data were first presented as the measure resistance normalized to its value at the start of each run (Figure 5.1). At high concentrations of cytochalasin B such as 2.5, 5.0, or 10 M with cells, a drastic drop of resistance was observed almost immediately following the addition. At low concentrations such as 0, 0.1, 0.5, and 1.0 M the resistance drop was less evident.

PAGE 53

34 Noise analysis was also applied to charac terize the normalized resistance time courses shown in Figure 5.1. The variance and the variance of the increment of the data were calculated and plotted agains t time (Figures 5.2 and 5.3). Here the variance was calculated for the first 60 points of the da ta (data points 1 thr ough 60) and this was plotted as one point. Next the 60 point da ta was shifted over one point (data points 2 through 61) and the variance was calculated and plotted. This process was continued until the last 60 points of data was read. To determine the variance of the increment, first all the increments of the successive data were calculated. Their variance was then calculated and plotted in the same way as described above. Decreased variance and variance of the increment for the cytochalasin B treated cells, as compared with controls, indicated reduced fluctuations in the resist ance time course. Although overall resistance and its fluctuations indicate different aspect s of cell activities, together these results suggest that while the three hi gher concentrations were easily distinguis hed from the four lowest, it was difficult to te ll the differences among the three lowest concentrations. Figure 5.4 plots the normalized resistance of several trials (n = 10) to show this segregation across the various concentrations of cytochalasin B was consistent and repeatable. HUVEC-H7 Variance The normalized resistance (Figure 5.5), va riance (Figure 5.6), and variance of increment (Figure 5.7) of HUVEC cell layers challenged by various concentrations of protein kinase inhibitor H-7 we re measured and calculated. At the higher concentrations of 25, 50, 100 and 1000 M the cell layer integrity is immediately degraded as the normalized resistance drops. The HUVEC layer does exhibit an ability to recover cell layer integrity as the layers exposed to 25, 50, and 100 M of protein kinase inhibitor H-7 experienced an increase in resistance afte r the initial drop o ff followed by another resistance decrease. Concentration of 10 M showed no visible affect on the cell layers normalized resistance as they acted similarl y to the control. Th e variance (Figure 5.6) plot reveals no significant di fference between concentrations. The variance of increment (Figure 5.7) plots reveals that the concentration of 100 and 500 M significantly reduced

PAGE 54

0 5 10 15 20Time (hrs) 0.2 0.4 0.6 0.8 1.0 1.2 1.4No r malized r esistance Control 0.1 M 0.5 M 1.0 M 2.5 M 5.0 M 10 M Figure 5.1 Normalized resistance measurements of a confluent HUVEC monolayer upon addition of different concentrations of cyto chalasin B. Cells were inoculated into electrode-containing wells and allowed to develop into confluent layers for approximately 24 h. At the time indicated by the arrow, cytochalasin B diluted in DMSO was added to give the fi nal concentrations of 0.1 M (red), 0.5 M (blue), 1.0 M (green), 2.5 M (yellow), 5.0 M (violet), 10 M (cyan), and control (black), and the resultant changes in normalized resistance were followed. Data were collected every 2 minutes for 20 hours. 35

PAGE 55

0 5 10 15 20Time (hrs) -7.0 -6.0 -5.0 -4.0 -3.0 -2.0 -1.0Log Variance Figure 5.2 Variance data of a confluent HUVEC monolayer upon addition of different concentrations of cytochalasin B at 0.1 M (red), 0.5 M (blue), 1.0 M (green), 2.5 M (yellow), 5.0 M (violet), 10 M (cyan), and control (black). 36

PAGE 56

0 5 10 15 20Time (hrs) -8.0 -7.0 -6.0 -5.0 -4.0 -3.0 -2.0Log Var. of Inc. Figure 5.3 Variance of increment analysis of th e resistance data in Figure 4.5. The different concentrations of cytochalasin B are 0.1 M (red), 0.5 M (blue), 1.0 M (green), 2.5 M (yellow), 5.0 M (violet), 10 M (cyan), and control (black). 37

PAGE 57

0 5 10 15 20Time (hrs) 0.2 0.4 0.6 0.8 1.0 1.2No r malized r esistance Figure 5.4 Average normalized resistance at 0, 5, 10, 15, and 20 h after addition o f cytochalasin B ( n = 10 for each concentration) against HUVEC layers. The differen t concentrations of cytochalasin B are 0.1 M (red), 0.5 M (blue), 1.0 M (green), 2.5 M (yellow), 5.0 M (violet), 10 M (cyan), and control (black). 38

PAGE 58

0 5 10 15 20Time (hrs) 0.2 0.4 0.6 0.8 1.0 1.2 1.4No r malized r esistance Figure 5.5 Normalized resistance measurements of confluent HUVEC cell layers upon addition of different concentrations of H 7. Cells were inoculated into electrodecontaining wells and allowed to develop into confluent layers for approximately 24 h. H7 diluted in DMSO was added to gi ve the final concentrations of 10 M (red), 25 M (blue), 50 M (green), 100 M (yellow), 500 M (violet), and control (black), and the resultant changes in normali zed resistance were followed. 39

PAGE 59

0 5 10 15 20Time (hrs) -5.0 -4.0 -3.0 -2.0 -1.0Log Variance Figure 5.6 Variance data of a confluent HUVEC monolayer upon addition of different concentrations of H7 at 10 M (red), 25 M (blue), 50 M (green), 100 M (yellow), 500 M (violet), and control (black). 40

PAGE 60

0 5 10 15 20Time (hrs) -6.0 -5.0 -4.0 -3.0Log Var. of Inc. Figure 5.7 Variance of increment analysis of th e resistance data in Figure 5.4. The different concentrations of H7 are 10 M (red), 25 M (blue), 50 M (green), 100 M (yellow), 500 M (violet), and control (black). 41

PAGE 61

0 5 10 15 20Time (hrs) 0.4 0.6 0.8 1.0 1.2No r malized r esistance Figure 5.8 Average normalized resistance at 0, 5, 10, 15, and 20 h after addition of H-7 ( n = 10 for each concentrati on) against HUVEC layers. The different concentrations o f H7 are 10 M (red), 25 M (blue), 50 M (green), 100 M (yellow), 500 M (violet), and control (black). 42

PAGE 62

43 the cell activity of the HUVEC cell layers. The variance and variance of increment plots do not segregate the activity of the lesser concentrations well. To better display the overall behavior of the cell layer integrity across several trials (n=10), the normalized resistance was averaged and the standard deviation of error was applied to show significant differences despite fluctuations (Figure5.8). The trials were only conducted for 20 hours, with 10 M being the only concentration not significantly affecting cell layer resistance. At the twenty hour time period, all normalized resistance values are around 60% for H7 concentrations other than 10 M. 3T3-Cytochalisin B Variance NIH 3T3 monolayers were challenged by va rious concentration of cytochalasin B and observed for 20 hours. Before the cytoch alasin B was added to the wells, NIH 3T3 cells were allowed to attach and spread across the electrode surface for 24 hours. In Figure 5.9 the resistance measurements were normalized against initial measurements to better compare the behavior occurring as a result from the various concentrations of cytochalisin B. At 5 and 10 M the cell layers were imme diately compromised causing an instant decrease in normalized resistance with the 10 M cytochalisin B exposure being more severe. For both concentrations there are no fluctuations alluding to cell death or detachment. With a concentration of 2.5 M the normalized resistance slowly decreased while fluctuations still remained. The cell layer was still active despite the layer being weakened by the cytochalisin B. The behavior of the NIH 3T3 layers exposed to 1 and 0.1 M of cytochalisin B acted simila rly to the control group. The overall trend for all three groups neither increased or decrease d and fluctuations maintained for the 20 hour duration. Fluctuat ion analysis of the resistance traces of Figure 5.9 resulted in the variance and variance of increment calculations (Figures 5.10 and 5.11). The variance and variance of increm ent data reveal that the fluctuations resulting from the cell activity of the 2.5 M cytochalisin B exposed NIH 3T3 cell layers, are more similar to the lower concentrations as compared to the higher. To better display the overall behavior of the cell layer integrity across several trials (n=14), the normalized resistance was averaged and the standard deviation of error was applied to show

PAGE 63

0 5 101520Time (hrs) 0.4 0.6 0.8 1.0 1.2 1.4No r malized r esistance Control 0.1 M 1.0 M 2.5 M 5.0 M 10 M Figure 5.9 Normalized resistance measurements of confluent NIH 3T3 cell layers upon addition of different concentrations of cyto chalasin B. Cells were inoculated into electrode-containing wells and allowed to develop into confluent layers fo r approximately 24 h. At the time indicated by the arrow, cytochalasin B diluted in DMSO was added to give the fi nal concentrations of 0.1 M (red), 1.0 M (blue), 2.5 M (green), 5.0 M (yellow), 10 M (violet), and control (black), and the resultant changes in normalized resistance were followed. 44

PAGE 64

0 5 101520Time (hrs) -7.0 -6.0 -5.0 -4.0 -3.0 -2.0 -1.0Log Variance Figure 5.10 Variance data of a confluent NIH 3T3monolayer upon addition of differen t concentrations of cytochalasin B at 0.1 M (red), 0.5 M (blue), 1.0 M (green), 2.5 M (yellow), 5.0 M (violet), and control (black). 45

PAGE 65

0 5 101520Time (hrs) -8.0 -7.0 -6.0 -5.0 -4.0 -3.0 -2.0Log Var. of Inc. Figure 5.11 Variance of increment analysis of th e resistance data in Figure 5.8. The different concentrations of cytochalasin B are 0.1 M (red), 0.5 M (blue), 1.0 M (green), 2.5 M (yellow), 5.0 M (violet), and control (black). 46

PAGE 66

0 5 101520Time (hrs) 0.6 0.8 1.0 1.2 1.4No r malized r esistance Control 0.1 M 1.0 M 2.5 M 5.0 M 10 M Figure 5.12 Average normalized resistance at 0, 5, 10, 15, and 20 h after addition o f cytochalasin B ( n = 14 for each concentration) against 3T3 layers. The differen t concentrations of cytochalasin B are 0.1 M (red), 0.5 M (blue), 1.0 M (green), 2.5 M (yellow), 5.0 M (violet), and control (black). 47

PAGE 67

48 significant differences despite fluctuations (F igure 5.12). The trials were only conducted for 20 hours. At the twenty hour time peri od, the normalized re sistance values are distinguishable at 1 M. After an initial decrease in resistance, overall values for 2.5, 5, and 10 M of cytochalasin B remain cons istent with little variation. 3T3-H7 Variance The normalized resistance (Figure 5.13), va riance (Figure 5.14), and variance of increment (Figure 5.15) of NIH 3T3 cell layers challenged by various concentrations of protein kinase inhibitor H-7 we re measured and calculated. At the higher concentrations of 25, 50, 100 and 1000 M the cell layer integrity is immediately degraded as the normalized resistance drops. The NIH 3T3 layer does exhibit an ability to recover cell layer integrity as the layers exposed to 25 and 50 M of protein kinase inhibitor H-7 experienced an increase in resistance after the initial drop off. Concentrations of 1 and 10 M showed no visible affect on the cell layers normalized resistance as they acted similarly to the control. The variance (Figure 5.14) and variance of increment (Figure 5.15) plots reveal that the concentration of 1000 M significantly reduced the cell activity of the NIH 3T3 cell layer. Concentrations of 50 and 100 M slightly reduced cell activity when compared with the lesse r concentrations including the control group. Once again to better display the overall beha vior of the cell layer integr ity across several trials, the normalized resistance was averaged and the st andard deviation of error was applied to show significant differences despite fluctuati ons. The trials were only conducted for 20 hours, but 50 and 100 M concentrations ended the trials in an increasing trend and could possibly completely recover cell integrity if trials were conducted longer. The overall behavior of the cell layer integrity across se veral trials (n=15) was displayed as the normalized resistance was averaged and the st andard deviation of error was applied to show significant differences despite fluctuatio ns (Figure 5.16). The trials were only conducted for 20 hours. At the twenty hour time period, the nor malized resistance values are not distinguishable for concentrations up to 25 M of H7. After an initial decrease in resistance, overall resistance values increase and recover for 25 and 50 M of H7.

PAGE 68

0 5 101520Time (hrs) 0.7 0.8 0.9 1.0 1.1 1.2No r malized r esistance Contorl 1.0 M 10 M 50 M 100 M 1000 M Figure 5.13 Normalized resistance measurements of confluent NIH 3T3 cell layers upon addition of different concentrations of a protein kinase inhibitor H-7. At the time indicated by the arrow, H-7 diluted in comp lete medium or medium alone was added to give the final concentrations of 1.0 M (red), 10 M (blue), 25 M (green), 50 M (yellow), 100 M (violet), 1000 M (cyan),and control (black), and the resultant changes in normalized resistance were followed. Data were collected every 2 minutes fo r 20 hours. 49

PAGE 69

0 5 101520Time (hrs) -6.0 -5.0 -4.0 -3.0 -2.0Log Va r iance Figure 5.14 Variance data of confluent NIH 3T3 monolayers upon the addition of H-7 a t various concentrations of 1.0 M (red), 10 M (blue), 25 M (green), 50 M (yellow), 100 M (violet), 1000 M (cyan), and control (black). 50

PAGE 70

0 5 101520Time (hrs) -7.0 -6.0 -5.0 -4.0 -3.0Log Var. of Inc. Figure 5.15 Variance of increment analysis fo r the resistance data of Figure 5.12 The different concentrations of H-7 are 1.0 M (red), 10 M (blue), 25 M (green), 50 M (yellow), 100 M (violet), 1000 M (cyan), and control (black). 51

PAGE 71

0 5 101520Time (hrs) 0.7 0.8 0.9 1.0 1.1 1.2No r malized r esistance Control 1.0 M 10 M 25 M 50 M 100 M 1000 M Figure 5.16 Average normalized resistance at 0, 5, 10, 15, and 20 h after addition of H7 ( n = 15 for each concentration) against 3T3 layers. The different concentrations o f cytochalasin B are 1.0 M (red), 10 M (blue), 25 M (green), 50 M (yellow), 100 M (violet), 1000 M (cyan), and control (black). 52

PAGE 72

53 HUVEC-Cytochalisin B frequency scan Cytochalasin B is known to disrupt actin filaments and rapidly affect cellular morphology 26. To further understand the effect of cytochalasin B addition on morphological changes of a HUVEC monolayer, frequency scan measurements were taken from cell-covered electr odes 20 hours after exposure to di fferent concentrations of cytochalasin B (Figures 5.17 and 5.18). To analyze differences in impe dance curves, it is helpful to use normalized values, where th e impedance values from the electrode confluent with HUVECs were divided by the corresponding quantitie s for the cell-free electrode. For the control impedance spectrum (black lines in Figures 5.17 and 5.18), the normalized resistance starts from ~1.0 at 25 Hz, increases with increasing frequency to the highest value, approximately ~7.0 at 4 kHz, and then decr eases with increasing frequency until ~2.5 at 60 kHz. The reason for th e peak is that the constriction resistance masks the resistance of the cell-cov ered electrode at high frequency 1. Normalized capacitance on the other hand remains 0.95 from 25 Hz to 200 Hz and then decreases with increase in frequency until approxi mately 0.15 at 60 kHz. After fitting the experimental data with Eq. 1, the result in dicated that upon challenge of 0, 0.1, 0.5, and 1 M cytochalasin B, the junctional resistance between cells ( Rb) dose-dependently decreased, indicating the alte rnation of cell morphology due to the perturbation of the actin filaments (Table 5.1). Furthermore, while the distance between the basolateral cell surface and substratum (h) slightly decreased, the membrane capacitance ( Cm) dosedependently increased, implying an increase of membrane folding. Notice that model analysis was unable to fit frequency scan data of the HUVEC monolayer exposed to 5.0 and 10 M cytochalasin B since the cell monolayer under these conditions developed noticeabl e holes and was no longer conf luent. Although the mean values for Rb at every concentration in Table 5.1 ar e separated from those at every other concentration by at least a standard error, Students t-Test for two-sample unequal variance show an insignificant ( p = 0.09) separation between the control and 0.1 M populations. All other pairs of popula tions are significantly separated ( p < 0.009). As for the mean values of Cm, not only do they display a clear tr end, they also show a significant

PAGE 73

1.02.03.04.05.0Log frequency, Hz 0.0 2.0 4.0 6.0 8.0Normalized resistance Control 0.1 M 0.5 M 1.0 M 5.0 M 10 M Figure 5.17 Normalized resistance as a function of log10(frequency) obtained from a frequency scan measurement 20 hours afte r the addition of medium containing cytochalasin B or medium alone to a confluent HUVEC layer to give the final concentrations of 0.1 M (red), 0.5 M (blue), 1.0 M (green), 5.0 M (yellow), 10 M (violet), and control (black). 54

PAGE 74

1.02.03.04.05.0Log frequency, Hz 0.0 0.2 0.4 0.6 0.8 1.0Normalized capacitance Control 0.1 M 0.5 M 1.0 M 5.0 M 10 M Figure 5.18 Normalized capacitance as a function of log10(frequency) obtained from a frequency scan measurement 20 hours afte r the addition of medium containing cytochalasin B or medium alone to a confluent HUVEC layer to give the final concentrations of 0.1 M (red), 0.5 M (blue), 1.0 M (green), 5.0 M (yellow), 10 M (violet), and control (black). 55

PAGE 75

Table 5.1 Impedance analysis of HUVEC monolayer 20 hr after exposure to cytochalasin B. All values, Rb, h, and Cm, were obtained from fitting measure d impedance with Eq. 1. Values shown are mean standard error ( n = 8). Each value is also expressed as a percentage of control. 56

PAGE 76

57 ( p < 0.01) separation between any pairs. In general, the specific capacitance of cell membranes is approximately 1 F/cm2 but can appear to be much larger if the membrane wrinkles and its surface roughness becomes higher. Using frequency scan measurement the effect of cytochalasin B was discernibl e at levels as low as 0.1 M, as Table 5.1 brought out. HUVEC-Cytochalisin B micromotion Cytochalasin B is also known to interfer e with cell motility and can substantially reduce the amplitude of the fluctu ation in the measured impedance 6. By looking at the variance analysis of the resist ance time course shown in Figur es 5.1, there is an intriguing hint that noise analysis coul d be a better assessment of cytotoxicity if data acquisition was faster and more sensitive. To th is end, micromotion measurements of HUVEC covered electrodes were taken 20 hours after exposure to different concentrations of cytochalasin B. At this moment cultures ha d reached some sort of equilibrium and did not exhibit a large downward or upward trend in impedance. Due to the selection of 4 kHz AC signal for the micromotion measurement, the fluctuations in the resistance were larger than those in the capacitive reactance. Thus, we focused on the resistance time course for data analysis. Figure 5.19 shows a graph of typical resistance data normalized by dividing with the averag e resistance for the whole period, where dose-dependently smaller fluctuations were evidently observed for cytochalasin B treated cells as compared with controls. Here only four resistance curv es are shown, as the to tal seven data curves in a single graph would be too crowded to see their differences. Three methods of numerical analysis, fast Fourier transform (FFT), variance, and variance of the increments were used to charac terize the level of fluctuations 2. As seen in Figures 5.20, 5.21, and 5.22, for all frequencies and times, there is a clear difference in the level of fluctuations obtained under the four different conditions. The scale is logarithmic; thus, the difference is significant.

PAGE 77

01 02 03 0Time (min)4 0 0.90 0.95 1.00 1.05 1.10Normalized resistance Figure 5.19 Normalized resistance data recorded 20 hours after the addition of medium containing cytochalasin B or medium alone to a confluent HUVEC layer to give the final concentrations of 1 M (red), 5 M (blue), 10 M (green), and control (black). Overall, 2048 data points taken at one-second intervals. 58

PAGE 78

-2.5-2.0-1.5-1.0-0.50.0Log Frequency (Hz) -13 -11 -9 -7 -5Log Power Figure 5.20 Fast Fourier transform of confluent HUVEC layers exposed to cytochalisin B concentrations of 1 M (red), 5 M (blue), 10 M (green), and control (black). 59

PAGE 79

0.51.01.52.02.53.03.5Log Time sequence (sec) -9 -8 -7 -6 -5 -4 -3 -2Log Variance Figure 5.21 Variance analysis of confluent HUVE C layers exposed to cytochalisin B concentrations of 1 M (red), 5 M (blue), 10 M (green), and control (black). 60

PAGE 80

0.00.51.01.52.02.53.0Log Sampling time (sec) -4.5 -4.0 -3.5 -3.0 -2.5 -2.0 -1.5 -1.0Log Variance of increments Figure 5.22 Variance of the increment analysis of confluent HUVEC layers exposed to cytochalisin B concentrations of 1 M (red), 5 M (blue), 10 M (green), and control (black). 61

PAGE 81

62 A summary of micromotion data is show n in Table 5.2. In order to represent the resistance fluctuations with a single number for each analysis, the table includes the average resistance value of the cell layer over a 2048-second run, the slope of the log-log plot of power spectrum (power slope), and both variance and variance of the increment for the 32-point intervals of the normalized 2048-second data set (Var32 and VoI32). The power slope characterizing fnoise was estimated with least-squares straight-line fits of power at the lowest 100 frequencies (excluding the fi rst and the last). The Var32 was obtained by calculating the statistical variance for each 32-point segment of the 2048 point data set and then taking the average of these 64 values. The VoI32 was obtained in the same way except that all the increments of the successive data were calculated at first. To easily evaluate the sensitivity of different numerical methods, analyzed value for each treatment group was also expressed as a rati o obtained by dividing its average value by that of the control group. The first thing we notice in Table 5.2 is that even averaging the resistance on the order of ten independent 2048-second runs can only distinguish the concentrations as low as 1 M ( p = 0.03). This result confirms the incapacity of overall resistance measurement to distinguish the three lowest concentrations of cytochalasin B as described in Figure 5.1. Similarly, the lo west concentration that the power-slope averages can significantly distinguis h from the control is also 1 M ( p = 0.04). A log-log plot of power spectrum against frequency f indicates an inte nsity varying as f Brownian noise displays a f 2 power law and a nonzero shows signs of long-time (or fractal) correlations 29. Giaever and Keese have analy zed the resistance fluctuations obtained from both normal WI-38 and transf ormed WI-38 VA13 cells using a digital filtering technique 30. The power spectrum for the WI-38 VA13 cells varies as f -2.6, and the WI-38 cells have a slightly smaller exponent The higher (~0.9) value suggests a stronger temporal correlation in the HUVECs activities. Of special interest is the apparent drop of the from 2.8 to 2.0, as concentrati on increases from 1 M to 10 M (Figure 5.20 and Table 5.2). It is likely that cytoskeletal structures are involved in cellular responses to cytotoxic stimulation. The initial cyto toxic event at the cell surface

PAGE 82

63 may cause cytoskeleton reorganization, via generating a variety of second messengers and activating diverse signal tr ansduction pathways. As actin filaments are disrupted by addition of the cytochalasin B, cellular netw orks no longer act as integrated units to transduce cytotoxic stimulation into coordinated cellular responses, leading to the loss of long-time correlation in cell behavior. Other than the average resistance and power slope analysis, variance and variance of the increments give us efficient ways of quantifying resistance fluctuations. By simply inspecting the log-log plots in Figures 5.21 and 5.22, it is apparent that as cytochalasin B concentration increases, calculated variance decreases for all time sequences and calculated variance of the increments decr eases for all sampling times. Both Var32 and VoI32 values show a dose-de pendent drop and are capable of detecting effects of low levels of cytochalasin B on the HUVEC monolayer (Table 5.2), however, the reduction of the percentage of control is less apparent for VoI32 values (88% for 0.1 M, p = 0.06) than for Var32 values (71% for 0.1 M, p = 0.02). Values for the IC50 (half-inhibition concentration) were interpreted through the micromotion analysis data of Table 5.2. The a ssumption is made that the response versus the log of concentration is linear between the two concentrations suggesting the halfinhibition region. IC50 reported from Var32 and VoI32 were 0.6 M (~0.3 g/ml) and 1.3 M (~0.6 g/ml), respectively. Previous values of IC50 for cytochalisin B reported have been larger for various other methods such as MTT assay (~1.5-6.3 g/ml)31, DNA synthesis measurement (~2.5 g/ml)32, and whole cell patch clamp technique to measure the cytochalasin B-induced inhi bition of hKv1.5 currents (~4.2 M)33. Overall this leads to the suggestion that ECIS micromotion anal ysis proves to be a better technique for cytotoxic evaluation. In general, the variance is similar to th e power obtained from FFT, and the inverse of the number of sampling points is similar to a frequency. On the other hand, if the data is fractal, the variance of the increments ma y yield information about the Hurst exponent

PAGE 83

Table 5.2 ECIS micromotion data obtained from confluent HUVEC layers 20 hours after exposure to cytochalasin B. The column labeled Res. is the average resistance value o f the cell layer over a 2048-second run. Th e column labeled power slope is the slope of the log-log plot of power versus frequency. The column labeled Var32 is the statistical variance for the 32-point intervals of the nor malized 2048-second data set. The column labeled VoI32 is the variance of the increment for the 32-point sampling intervals of the 2048-second data set. The column labeled Hurst is the Hurst exponen t. Values shown are mean standard error (n = 10for each concentration of cytochalasin B). Each value is also expressed as a percentage of control. 64

PAGE 84

65 which equals to the slope of the line 2. The Hurst exponents for the HUVECs in response to different concentr ations of cytochalasin B are in the range of 0.6 to 0.8, suggesting that the cells have a persistent behavior (Table 5.2). Though Hu rst exponents are not dose-dependently different, before a decline is observed from 2.5 M to 10 M, there is a noticeable rise of Hurst exponent from 0.1 M to 1.0 M when compared with the control. One obvious result is that th e three higher concentrations were easily distinguished from the four lowest. This dist inction is also clearly seen from the average resistance and power slope analysis (Table 5.2). The reason for the increase in Hurst exponents upon exposure to cytochalasin B shown by the HUVECs is not known. In these experiments, we deal with the average ac tivities of 20-30 cells that interact with each other and with the electrode. Cellular networks connected by both cell-cell and cellsubstrate adhesion complexes may represent an effective means for communication and play a crucial role in the persistent responses to cytotoxic stimuli. HUVEC-H7 micromotion Micromotion measurements of conf luent HUVEC covered electrodes were acquired 20 hours after H7 exposure. Figur e 5.23 shows a graph of typical resistance data normalized by dividing with the average resistance for 50 M, 100 M, and 500 M of H7 and control. Results for micromotion analysis are presented for resistance. For purposes of clear graphical representation, only the 50 M (red), 100 M (blue), 500 M (green), and control (black) data were presented. The resistance traces demonstrate the difference in fluctuations between the H7 expos ed layers as compared with the control. Cell motility fluctuations were noticeabl e for the HUVEC layers exposed to 50 M of H7 (Figure 5.23). Fast Fourier transform (FFT), variance, and variance of the increments were used to better character ize the level of fluctuations. As seen in Figures 5.24, 5.25, and 5.26, for all frequencies and times, there is a clear difference in the level of fluctuations between the thr ee situations presented. Micromotion data for HUVEC H7 is presen ted in Table 5.3 for ten experimental runs. The table includes the average resistance value of the cell layer over a 2048-second

PAGE 85

66 run, the variance, and variance of the incr ement for the 64-point intervals of the normalized 2048-second data set. The statistical variance for each 64-point segment of the 2048 point data set was aver aged for all 32 values. The variance of increment was obtained in successive 64-point segments then averaged. Values were also presented as a ratio compared with the control group. The average resistances presented in Table 5.3 show a slight decrease with increasing H7 concentration, but the only signi ficant difference occurs at a concentration of 2.5 M. The log-log plots in Figures 5.25 and 5.26 reveal that as H7 concentration increases, calculated variance initially increases for 10 M before decreasing for all time sequences. Similarly, the calculated variance of the increments temporarily increases for 10 and 25 M of H7 before decreasing for all sampling times Both variance and the variance of increment values show a deviat ion from control behavior depending on dose and can detect the lowest co ncentration of H7 agitating th e HUVEC layers (Table 5.3). The increase of the percentage of control fo r the variance (133% for 10 M) and variance of the increment (122% for 10 M) demonstrat es a strong segregation from control. The Hurst exponents for the HUVECs in response to different concentrations of H7 demonstrate dose dependence (Table 5.3). The IC50 values for H7 challenged HUVEC la yers were interpreted through the micromotion analysis data of Table 5.3. IC50 reported from Var and Var. of Inc. were 53 M (~26.5 g/ml) and 75.7 M (~37.9 g/ml), respectively. Due to the unique dose response from HUVEC to H7, va riance and variance of increm ent is not necessarily the best method for IC50 calculation. The reported IC50 for the resistance is 33 M (~16.5 g/ml) which is less than Var an d Var. of Inc. Even though IC50 is a standardized method for comparing different cy totoxic assays, Var. and Var. of Inc. is a more sensitive mean for detecting the agitation H7 has on HUVE C as compared with overall resistance measurements.

PAGE 86

01 02 03 0Time (min)4 0 0.90 0.95 1.00 1.05 1.10Normalized resistance Figure 5.23 Normalized resistance recorded 20 hours after the addition of mediu m containing H7 or medium alone to a conf luent HUVECcell layer to give the final concentrations of 50 M (red), 100 M (blue), 500 M (green), and control (black). Each curve consists of 2048 data poi nts taken at one-second intervals. 67

PAGE 87

-2.5-2.0-1.5-1.0-0.50.0Log Frequency (Hz) -13 -11 -9 -7 -5Log Power spectrum Figure 5.24 Fast Fourier transfor m of confluent HUVEC cell layers exposed to H7 concentrations of 50 M (red), 100 M (blue), 500 M (green), and control (black). Each curve consists of 2048 data poi nts taken at one-second intervals. 68

PAGE 88

0.51.01.52.02.53.03.5Log Time sequence (sec) -9 -8 -7 -6 -5 -4 -3 -2Log Va r iance Figure 5.25 Variance analysis of confluen t HUVEC cell layers exposed to H7 concentrations of 50 M (red), 100 M (blue), 500 M (green), and control (black). Each curve consists of 2048 data poi nts taken at one-second intervals. 69

PAGE 89

0.00.51.01.52.02.53.0Log Sampling time (sec) -4.5 -4.0 -3.5 -3.0 -2.5 -2.0 -1.5 -1.0Log Variance of increments Figure 5.26 Variance of the increment analysis of confluent HUVEC cell layers expose d to H7 concentrations of 50 M (red), 100 M (blue), 500 M (green, and control (black). Each curve consists of 2048 data poi nts taken at one-second intervals. 70

PAGE 90

Table 5.3 ECIS micromotion data obtained fr om confluent HUVEC cell layers 20 hours after exposure to H7. The column labeled Res. is the average resistance value of the cell layer over a 2048-second run. The column labeled Var. is the statistical variance for the 64-point intervals of the normalized 2048-s econd data set. The column labeled Var.o f I nc is the variance of increments for the 64-point sampling intervals of the 2048-secon d data set. Values shown are mean standard error (n = 10). Each value is also expresse d as a percentage of control. 71

PAGE 91

72 3T3-Cytochalisin B micromotion Micromotion measurements of confluent 3T3 fibroblast covered electrodes were acquired 20 hours after cytochalasin B exposur e. Figure 5.27 shows a graph of typical resistance data normalized by dividing with the average resistance for the lowest two concentrations of cytochalisin B and control. Fluctuations in resistance were greater than capacitance due to the frequency se lection, as a result micromotion analysis is presented for resistance. For purposes of clea r graphical representation, only the 5 M (red), 10 M (blue), and control (bl ack) data were presented. The resistance traces demonstrate the difference in fluctuations between the cy tochalasin B exposed layers as compared with the control. Cell motility fluctuations were minimal for the 3T3 layer exposed to any concentration of cytoch alisin B (Figure 5.27). Fast Fourier transform (FFT), variance, and variance of the increments were used to better characterize the level of fluctuations. As seen in Figures 5.28, 5.29, and 5.30, for all frequencies and times, there is a clear difference in the level of fluc tuations between the control, 5, and 10 M cytochalasin B exposed HUVEC layers. Micromotion data for 3T3 cytochalisin B is presented in Ta ble 5.4 for eighteen experimental runs. The table includes the aver age resistance value of the cell layer over a 2048-second run, the variance, and variance of the increment for the 64-point intervals of the normalized 2048-second data set. The stat istical variance for each 64-point segment of the 2048 point data set was av eraged for all 32 values. Th e variance of increment was obtained in successive 64-point segments then averaged. Values were also presented as a ratio compared with the control group. The average resistances presented in Table 5.4 show a slight decrease with increasing cytochalisin B c oncentration, but the only signifi cant difference occurs at a concentration of 2.5 M. The log-log plot s in Figures 5.29 and 5. 30 reveal that as cytochalasin B concentration increases, calculated variance decreases for all time sequences and calculated variance of the increments decr eases for all sampling times. Both variance and the variance of increment values show a dose-dependent drop and can

PAGE 92

73 detect the lowest concentration of cytochal asin B affecting the 3T 3 layers(Table 5.4). The reduction of the percentage of cont rol for the variance (60% for 0.1 M) and variance of the increment (81% for 0.1 M) demonstrates a strong segregation from control. The Hurst exponents for the 3T3s in response to different concentrations of cytochalasin B demonstrate dose dependence (T able 5.4). Starting from the control (H = 0.73.02), the Hurst exponent is ch aracterized as strongly pe rsistent but degrades as cytochalasin B increases. The behavior of th e 3T3 becomes random in nature with higher concentrations of cytochalasin B. Though the Hurst exponent isnt as sensitive as variance in distinguishing cytochalasin B concen trations, it is interesting to note that the dose dependent behavior of the Hurst expo nent as it approaches Brownian motion (H=0.5). The IC50 values for cytochalasin B challenged NIN 3T3s were interpreted through the micromotion analysis data of Table 5.4. IC50 reported from Var and Var. of Inc. were 3.6 M (~1.8 g/ml) and 3.8 M (~1.9 g/ml), respectively. The dose response from 3T3 decreases in variance at 0.1 M (60% of control), then increases variance at 1 and 2.5 M of cytoB before dose response drops off and IC50 can be determined. The dose response reported from the VoI plateaus at 0.1, 1, and 2.5 M (~80% of control) with the following dose increment containing the IC50 region. 3T3-H7 micromotion Measured impedance data has been observed to have a decrease in fluctuation due to the ability of H7 to decrease cell motility. Micromotion analysis was utilized to further investigate the cytotoxic effects protein kinase inhibitor H7 had on the 3T3 cell layers motility and activity. Fluctuations in resistan ce were greater than capacitance due to the frequency selection, as a result resistance da ta is present for micromotion. The average resistance for the entire time course was used in normalizing typical resistance traces then displayed in Figure 5.31. For purposes of clear graphical repres entation, only the 10 M (red), 100 M (blue), 1000 M (green), and control (black) data were presented. 100 M of protein kinase inhibitor H-7 induced fluctu ations similar to the control group, while 10

PAGE 93

74 01020304Time (min)0 0.97 0.98 0.99 1.00 1.01No1.02 1.03 r malized r esistance Figure 5.27 Normalized resistance recorded 20 hours after the addition of mediu m containing cytochalasin B or medium alone to a confluent 3T3 cell layer to give the final concentrations of 5 M (red), 10 M (blue), and control (black). Each curve consists o f 2048 data points taken at one-second intervals.

PAGE 94

-2.5-2.0-1.5-1.0-0.50.0Log Frequency (Hz) -14 -12 -10 -8Log Power-6 Figure 5.28 Fast Fourier transform of confluent 3T3 cell layers exposed to cytochalisi n B concentrations of 5 M (red), 10 M (blue), and control (black). Each curve consists of 2048 data points taken at one-second intervals. 75

PAGE 95

0.51.01.52.02.53.03.5Log Time sequence (sec) 76 -10 -9 -8 -7 -6 -5Log Variance-4 -3 Figure 5.29 Variance analysis of confluent 3T3 ce ll layers exposed to cytochalisin B concentrations of 5 M (red), 10 M (blue), and control (black). Each curve consists o f 2048 data points taken at one-second intervals.

PAGE 96

77 0.00.51.01.52.02.53Log Sampling time (sec).0 -4.5 -4.0 -3.5 -3.0Log Va-2.5 -2.0 r iance of inc r emen ts Figure 5.30 Variance of the increment analysis of confluent 3T3 cell layers exposed to cytochalisin B c oncentrations of 5 M (red), 10 M (blue), and control (black). Each curve consists of 2048 data points taken at one-second intervals.

PAGE 97

Table 5.4 ECIS micromotion data obtained from confluent 3T3 cell layers 20 hours after exposure to cytochalasin B. The column labeled Res. is the average resistance value o f the cell layer over a 2048-sec ond run. The column labeled Var. is the statistical variance for the 64-point inte rvals of the normalized 2048-second data set. The column labele d Var. of Inc is the variance of increments for th e 64-point sampling intervals of the 2048second data set. Values shown are mean standard error (n = 18). Each value is also expressed as a percen tage of control. 78

PAGE 98

79 M generated fluctuations significantly greater than the control group. To better analyze the fluctuating data Fast Fourier transfor m (FFT) (Figure 5.32), variance (Figure 5.33), and variance of the increments (Figure 5.34) were utilized. Figure 5.33 and 5.34 reveal a more complicated nature of the variance and variance of increments of NIH 3T3 layers affected by protein kinase i nhibitor H-7. The variance of increments (Figure 5.34) is shown to be greater for 100 M than 10 M at the very end of the sampling period but still being less than the control group. The variance (Figure 5.33) is shown to be greatest for 10 M exposed cell layers towards the end of the sampling period, which return variance values even higher than the contro l group. Cell motility fluctuations appeared significant for the two lowest concentrations as well as the control 3T3 layer (Figure 5.31). 3T3-H7 micromotion Micromotion data for 3T3 H7 is presente d in Table 5.5 for fifteen experimental runs. The table includes the average resistance value of the cell layer over a 2048-second run, the variance, and variance of the incr ement for the 64-point intervals of the normalized 2048-second data set. The statistical variance for each 64-point segment of the 2048 point data set was aver aged for all 32 values. The variance of increment was obtained in successive 64-point segments then averaged. Values were also presented as a ratio compared with the control group. The average resistances presented in Ta ble 5,5 show a stea dy decrease with increasing H7 concentration, but the only signi ficant difference occurs at a concentration of 25 M. The log-log plots in Figure 5.34 rev eal that as H7 concentration increases, calculated variance decreases for all time se quences and calculated variance of the increments decreases for all sampling times. The variance plot (Figure 5.33) does not make a clear distinction graphica lly between control(black) and 10 M (red). Both variance and the variance of increment valu es show a dose-dependent drop making a distinction at 10 M of H7 affecting the 3T3 layers (Table 5.5). The reduction of the percentage of control for th e variance (58% for 10 M) and variance of the increment

PAGE 99

80 01020304Time (min)0 0.98 0.99 1.00 1.01 1.02 1.03No r malized r esistance Figure 5.31 Normalized resistance recorded 20 hours after the addition of medium containing H-7 or medium alone to a confluent 3T3 cell layer to give the final concentrations of 10 M (red), 100 M (blue), 1000 M (green), and control (black). Each curve consists of 2048 data points taken at one-second intervals.

PAGE 100

-2.5-2.0-1.5-1.0-0.50.0Log Frequency (Hz) -14 -12 -10 -8Log Powe-6 r Figure 5.32 Fast Fourier transform of confluent 3T3 cell layers exposed to H-7 concentrations of 10 M (red), 100 M (blue), 1000 M (green), and control (black). Each curve consists of 2048 data points taken at one-second intervals. 81

PAGE 101

0.51.01.52.02.53.03.5Log Time sequence (sec) 82 -10 -9 -8 -7 -6 -5Log Va-4 -3 r iance Figure 5.33 Variance analysis of confluen t 3T3 cell layers exposed to H-7 concentrations of 10 M (red), 100 M (blue), 1000 M (green), and control (black). Each curve consists of 2048 data points taken at one-second intervals.

PAGE 102

83 0.00.51.01.52.02.53Log Sampling time (sec).0 -4.5 -4.0 -3.5 -3.0Log Va-2.5 -2.0 r iance of inc r emen ts Figure 5.34 Variance of the increment analysis of confluent 3T3 cell layers exposed to H-7 concentrations of 10 M (red), 100 M (blue), 1000 M (green), and control (black). Each curve consists of 2048 data points taken at one-second intervals.

PAGE 103

Table 5.5 ECIS micromotion data obtained from confluent 3T3 cell layers 20 hours after exposure to H-7. The column labeled Res. is the average resistan ce value of the cell laye r over a 2048-second run. The column labeled Var. is the statistical variance for the 64p oint intervals of the normalized 2048-second data set. The column labeled Var. of Inc is the variance of increments for the 64-point sampling intervals of the 2048-second dat a set. Values shown are mean standard error ( n =15). Each value is also expressed as a p ercentage of control. 84

PAGE 104

85 (79% for 10 M) demonstrates a strong segreg ation from control. The Hurst exponents for the 3T3s in response to different concentrations of H7 demonstrate dose dependence that isnt apparent until 200 M (Table 5.5). Starting from the control (H = 0.72.01), the Hurst exponent is characterized as strongly persistent for the lowest concentrations of H7. Once again, the most sensitive analytic method for distinguishing cytotoxicity has been variance and the variance of the increment. The IC50 values for protein kinase inhibi tor H7 challenged NIN 3T3s were interpreted through the micromotion analysis data of Table 5.5. IC50 reported from Var and Var. of Inc. were 16.7 M (~8.3 g/ml) and 112.5 M (~56.3 g/ml), respectively. The difference in dose response between Var and Var. of Inc. is large in this instance, with Var being the more sensitive calculations. Discussion We have applied various ECIS assays to follow the activities of HUVECs in response to different concentrations of cytoch alasin B and protein kinase inhibitor H7. Frequency scan have been used to det ect subtle changes in cell morphology and micromotion measurement have been used to monitor time-series impedance fluctuations. Although these two measurements provide differe nt profiles of data, they can sensitively distinguish toxin leve ls as low as 0.1 of cytochalasin B and 10 of H7. The cytotoxic detection demonstrated a unique example of how micromotion analysis and overall resistance measurements can be used together. For the HUVEC cytochalasin B assay, variance and variance of increment prove d to be the best method for detection. Overall resistance measurements of the HUVEC H7 assay were the more sensitive method for cytotoxic detection. The calculated variance of th e fluctuations may be taken as an indication of the general cel l health and can be used to follow the cells in real time. Similarly, ECIS assays were also used to evaluate the cytotoxic nature of cytochalisin B and protein ki nase inhibitor H7 against c onfluent NIH 3T3 fibroblast cell layers. Even though overall resistance values are relatively small for 3T3 layers, and

PAGE 105

86 frequency scan measurements are negligible, impedance fluctuation analysis reveals significant micromotion fo r cytotoxic dete ction. Toxin levels of 0.1 of cytochalisin B and 10 of H7 were detected. The analytical methods us ed in this study can serve as a model approach for ECIS and other EIS bi osensors to investigate various aspects of cellular responses to toxins in general.

PAGE 106

87 CHAPTER 6 EFFECT OF CSC ON CANCER INVASION Introduction The endothelium serves as the interface between blood s upply and the rest of the tissue within the human body as it lines the enti re circulatory system. Through this layer, water and other molecules are allowed to exit the bloodstream and into the surrounding tissue. A secondary purpose of this conflu ent layer is acting as a protective barrier against vascular diseases. Cancerous cells drifting through blood vessels and lymphatic ducts that have become independent of their original growths attach on to the endothelium in order to initiate new growth s. A complete understa nding of the cellular mechanics involved in transmigration is neces sary in limiting and possibly nullifying the metastatic potential of ca ncer through blood vessels. Cellular motility and proliferation is a vital function in reendothelialization. A confluent endothelial cell monolayer demons trates contact-inhibited latency. When the endothelium is confluent and latent, it allows for an optimized interaction with the blood stream. During the loss of confluency, ce llular motility and proliferation increases to recover compromised regions. Motility continues to remain highly active until contact inhibition restricts the monolay er back to latency. The endothelium serves as the initial barrier for metastatic cancer cells through the blood stream searching for new sites to initiate new tumor growths thro ugh extravasation. It is also the last barrier invading cancer cells must breach to access the lumen of blood vessels through a process called intravasation. Primary tumors account for about 10% of deaths from cancer while the remaining 90% is a result of transient tumors spawned from primary growth 34. Metastasis is the ability of can cer cells to detach from primary

PAGE 107

88 tumors, travel the body through blood or ly mphatic vessels, and initiate new tumor growths in different regions of the body. Current techniques involving cancer penetration utilize polycarbonate membranes containing 8m pores covered in a thin layer of a protein matrix to simulate an endothelial layer 35, afterwards the penetrated layer is observed through cell staining. The poly carbonate membranes are static and morphologically constant as they do not respond to stimulus in their environment. This utilization of the polycarbonate membrane doe s not truly observe ca ncer penetration of the endothelium as it is more of a cellular motility experiment With ECIS, an active fully responsive endothelium can serve as a barrier against cancer penetration which is quantifiable into electrical impedance measurements. Interactions of cancer cells with the vascular endotheli um and basement membrane are crucial in both intravasati on and extravasation processes 36, involving the regulation and expression of many genes involved in such diverse activities as endothelial binding 37, cell signaling resulti ng in endothelial retraction 38, the synthesis and secretion of proteolytic enzymes 39, and cell locomotion 40. Cigarette smoke-induc ed disruption of the endothelial integrity may facilitate cancer penetration, but the unde rlying mechanism is unclear. Cigarette smoke has primarily been utilized as a tumorigenic factor in both in vitro experiments and with live animals. It has been used to induce cancer in samples for investigation of genetic cha nges and regulations that occu r due to the cigarette smoke influence 41. The interactions between normal cells and cancerous cells within the same smoke exposed environment have not been in vestigated. This is important because normal and cancer cells respond differently to environmental changes. Endothelial cells and metastatic cancer cells may exhibit differe nt sensitivities and reactive mechanisms to cigarette smoke. Cigarette smoke contains over 4000 different chemicals with 69 of these constituents being individu ally identified as carcinogens 42. Carcinogenesis occurs due to oxidative stress induced by the many reactive oxygen species (ROS) contained within cigarette smoke as well as the biolog ical ROS production in re sponse to cigarette

PAGE 108

89 smoke exposure43. ROS initiated cell signaling cascades cause loss of cell adhesion44, cell motility45, proliferation46, and apoptotic induction 47. ROS has also been shown to increase neutrophil and colon cancer cell a dhesion to endothelial layers when observed for an hour48. Similarly, cigarette smoke condensate (CSC) has been reported to have adverse effects on endothelial cells, including inhibition of migration 49, induction of apoptosis 50, cellular contraction 51, and cancer cell transmigra tion across polycarbonate membranes simulating the endothelium 52. A novel molecular probe sensitive to specific ROS molecules was previously used to quantify ROS within cigarette smoke.53 This approach was not used in an in vitro setting though to evaluate any sort of biological ROS production. This chapter reports the real-time effects CSC have on the adhesive properties of endothelial cell layers through ECIS analys is in conjunction with ROS production. Endothelial cell motility, junctional binding, an d barrier function against transmigrating cancer cells under CSC exposure were also investigated. Wound Recovery. Previous studies indicate that endothelium cells contract54 and have inhibited migration abilities49 when exposed to CSC, while cert ain cancers experience stimulation in cell proliferation and angiogenic factor expression55. Wound healing assays with confluent cultures of HUVEC, OVCA429, and ALST were prepared and observed under time signature microscopic imaging. For all ce ll cultures, the firs t image was taken two hours after wounding the cultures. This wa s due to the cancer ous cells stacking upon each other where as normal endothelium cell types do not. Within the first 90 minutes the cancerous cultures experience an acceler ated recovery due simply to the over crowded cells being pushed into the gap. At 2 hours the cancerous cultures gap recovery reaches a steady rate. Figure 6.1 shows the images for the wounded HUVEC cultures control group in comparison to the 10 g/mL of CSC. Degradation in gap recovery can be observed when comparing the images. Th e images were analyzed with software

PAGE 109

Figure 6.1 Images of HUVEC cultures recovering a gap. The bottom images are an example of a culture that was influenced by 10 g/mL of CSC as compared to a control culture. 90

PAGE 110

Table 6.1 The recovery rates for various w ounded cell cultures against different concentrations of CSC. 91

PAGE 111

92 having the ability to count the pixels within a designated regi on. In this case, the region was specified as the gap between the encro aching groups of cells. With the known length and area of the gaps in pixels, the average width of the gaps could be calculated and converted to micrometers. Table 6.1 graphs the quantification of the gaps recovered in micrometers per hour as a function of CSC concentration. HUVECs experienced degradation in gap recovery with increasing CSC concentrations as expected. ALST cultures had a degraded gap recovery at 10 g /mL of CSC but then in creased almost back to normal at 50 g/mL of CSC. OVCA429 expe riences a simulated wound recovery at 10 g/mL of CSC before degrading back to normal at 50 g/mL of CSC. With CSCs unusual affect of increasing motility of cancerou s cells as opposed to decreased motility for normal cells, CSC could possibly stimulate ca ncers metastatic ability to transmigrate an endothelium. HUVEC-CSC Previous studies have shown results th at suggest a stronger tight junction formation within the 8 hours of CSC exposure, while 24 hour studies have shown a decrease in TER. With real-time observati on of impedance values from ECIS, Figure 6.2 demonstrates a trend that agrees with prev ious independent studies while revealing the varying time dependent nature. HUVEC layers exposed to 10 g/mL of CSC experience a slight degradation in resistance over a 40 hour exposure time. Layers exposed to 50 and 100 g/mL of CSC experience an increase in resistance that lasts about 4 and 5 hours respectively before decreasing to resistance values 40% the normalized resistance. No data was returned for concentrations of 250 g/mL as the HUVECs rounded off releasing their junctional binding as evident in Figure 6.3. The HUVEC layers exposed to CSC con centrations of 50 and 100 g/mL were further investigated with contin ual frequency scans for the firs t thirty hours of exposure. Figures 6.4 and 6.5 plots the log capacitanc e for a single HUVEC covered electrode exposed to 50 and 100 g/mL of CSC, respectiv ely. The plots reveal that after a certain amount of time after CSC exposure, the lo g-log capacitance frequency plot has a

PAGE 112

0 0.2 0.4 0.6 0.8 1 1.2 1.4 0510152025303540Time (hrs)Normalized Resistance control 10 g/mL 50 g/mL 100 g/mL Figure 6.2 40 hour observation of various co ncentrations of CSC affecting HUVEC layers. The resistance values were normalized for better graphical representation of the behavior CSC induces on HUVE Cs adhesive properties. 93

PAGE 113

Figure 6.3 Images of confluent ALST (top) and HUVEC (bottom) layers. The images on the left are not exposed to CSC while th e images on the right are exposed to CSC concentrations of 250 g/mL. Table 6.2 Values returned from frequency scan analysis of HUVEC layers. The first column indicates the concentra tion of CSC (g/mL). The s econd column is the analysis of the increasing change in junctional resistance normalized ( Rb). The third column is the change in height normalized ( h). 94

PAGE 114

Figure 6.4 30 hour continual frequency scan of a single HUVEC covered electrode exposed to 50 g/mL of CSC. The graph presents the time series of the log-log capacitance frequency plot. 95

PAGE 115

Figure 6.5 30 hour continual frequency scan of a single HUVEC covered electrode exposed to 100 g/mL of CSC. The graph presents the time series of the log-log capacitance frequency plot. 96

PAGE 116

Figure 6.6 30 hour continual frequency scan of a single HUVEC covered electrode exposed to 50 g/mL of CSC. The graph presents the time series of the log-log resistance frequency plot. 97

PAGE 117

Figure 6.7 30 hour continual frequency scan of a single HUVEC covered electrode exposed to 50 g/mL of CSC. The graph presents the time series of the log-log resistance frequency plot. 98

PAGE 118

99 regime where it behaves as a cell-free electr ode plot. For Figure 6.4 this time frame is around ten hours while Figure 6.5 occurs at twen ty hours. The same cell-free electrode behavior can be seen in th e corresponding log-log resistance frequency plots (Figure 6.6 and 6.7). A decrease in overall resistance in response to CSC occurs before the HUVEC layers recover normal resistance values. Figure 6.7 also reveals an increase in resistance within the first 5 hours, this resi stance is subtle in Figure 6.6. Based off the data in Figures 6.4 6.7, as well as other electrodes, the frequency scan data was summarized and presented in Table 6.2. Table 6.2 compares the increase in junctional resistance ( Rb) to the change in cell-substrate separation ( h) for 50 and 100 g/mL of CSC exposed HUVEC layers. Both Rb and h were normalized to initial values for a better comparison. Continual fr equency spectroscopic analysis was utilized in Table 6.2 to reveal that the initial resist ance increases can be attributed to greater change in junctional resist ance as opposed to cell-substr ate binding. The 15-20 hour regime of HUVEC exposed to CSC was expl ored for purposes of metastatic cancer transmigration. With a decrea sed junctional resistance, i nvading ALST cells should be able to transmigrate more aggressively than compared to control samples. OVCA429-CSC & ALST-CSC. Preliminary frequency scans revealed the CSC dose dependent effect on confluent layers of ALST and OVCA429 junctional re sistance (Table 6.3) and cell-substrate separation (Table 6.4). These data also serve to identify whether the cell layer primarily on the gold electrodes are HUVEC, ALST, or OVCA429 for the invasion assays. The large discrepancy between junctional resi stance and cell-substrate separation between HUVEC as compared with ALST and OVCA429 is largely noticeable regardless of CSC concentration. This discrepancy make s layer identificati on through frequency spectroscopy straight forward. The ALST and OVCA429 junctional re sistance is several times larger while the cell-substrate separation is several times smaller, than compared with HUVEC layers. Table 6.3 presents th e data supporting that ALST and OVCA429 can maintain a layer in the presence of CSC, while Figure 6.8 reveals that the ALST has

PAGE 119

Table 6.3 The junctional resistance reported for various cell types at time of0 and 40 hours for each cell type. The columns labeled 0 g/mL, 10 g/mL, 50 g/mL, and 100 g/mL note the CSC concentration within th e culture medium. All cell types were allowed one day to reach confluency before frequency scan measurements were taken. 100

PAGE 120

Table 6.4 The cell-substrate separation reported for various cell type s at time of 0 an d 40 hours for each cell type. The columns labele d 0 g/mL, 10 g/ml, 50 g/mL, and 100 g/mL note the CSC concentration within th e culture medium. All cell types were allowed one day to reach confluency before frequency scan measurements were taken. 101

PAGE 121

Figure 6.8 Frequency scan data of junctional resistance and cell-substrate separation (number above each column) for ALST cells attaching from a suspended state to form layers. 102

PAGE 122

103 the ability to attach and bind to a substrate in the presence of 50 g/mL of CSC. In the presence of CSC, OVCA429 could not conclu sively attach and bind to the electrode surface (data not shown). Invasion Assay HUVEC layers exposed to 50 g/mL of CSC for 20 hours showed to be advantageous for ALST cells invading under no CSC conditions. Frequency spectroscopy revealed the str ong layers of ALST that were able to form when CSC exposure was exclusive to HUVECs before invasion as indicat ed by columns /c in Figure 6.9. It should be specifically noted th e ALST junctional resistance of the /c assays is nearly three times higher than the control assay c/ c in the 140 hour set. The resistance traces in Figure 6.10 reveal the time-dependent nature of the over all resistance. Within the first 20 hours, the decrease in the HUVEC layer binding is observed in the /c and /50 trace. After the first twenty hours, the /c trace is observed to have the most aggressive increas ing behavior with frequency scans (Figure 6.9) revealing junctional resistance as the c ontribution causing the large discrepancy. For the c/50 data in the frequency scans (Fi gure 6.9), proper tight junc tion of ALST layers were not formed for the 140 hour set. Without proper tight junction resistance, cellsubstrate separation could not be interpolate d. To verify that ALST invasion was not completely inhibited during the c/50 assa ys, confocal fluorescence microscopy images were acquired (Figure 6.11b). Before fl uorescence image acquisi tion, ALST were allowed to invade HUVEC layers for 24 hours before inducing arrest with 4% formaldehyde. Discussion ALST and OVCA429 experienced differe nt reactive mechanisms towards CSC when compared with normal HUVEC cells. It is reasonable to assume that cancerous cell types are already missing components within the cell signaling pathways that would mediate a normal response towards CSC exposure. The HUVEC cultures responded to CSC with an increase in junctional bindi ng, where as ALST and OVCA429 relieved

PAGE 123

104 adhesion thereby providing an improved motility. ALST cultures demonstrated a nicely reactive unbinding in accordance to increasing concentrations of CSC while HUVEC junctional and substrate binding shows no vi sible CSC concentration dependence. Transmigration of the HUVEC layer by ALST cells exhibit a pre-CSC exposure time-dependence affecting the effectivene ss of ALST transmigration. The HUVEC layers decreased tight junction binding that resulted from CSC exposure, allowed for a more aggressive ALST layer formation that occurred during simula ted intravasation. Increased HUVEC layer tight junction binding th at occurred in the first five hours in response to CSC during extravasation contri butes to impeding ALST transmigration at high concentrations of CSC. Overall, CSC has an impeding effect on ALST transmigration during extravasation while causing aggressive transmigration during intravasation.

PAGE 124

Figure 6.9. Cell-substrate separation (top) and junctional resistance (bottom) values gathered from frequency scans for ALST i nvasion of HUVEC layers under the influence of 50 g/mL of CSC. 105

PAGE 125

Figure 6.10 Resistance traces starting off initi ally as a HUVEC layer, then being invaded upon by suspended ALST cells. HUVECs (dotted and red traces) were initially exposed to 50 g/mL of CSC for the first tw enty hours. At twenty hours, CSC exposure changed (red and green traces), and ALST cells were added. The total observatio n occurred for 132 hours. 106

PAGE 126

Figure 6.11 Fluorescence microscopy images of confluent HUVEC (green) layers being invaded by suspended ALST (red) cells under the influence of 50 g/mL of CSC. (a) HUVEC layer invaded by ALST cells under no CSC influence. ( b ) HUVEC laye r invaded by ALST cells under CSC influence. (c) HUVEC layer exposed to CSC fo r twenty hours then invaded by ALST. 107

PAGE 127

108 CHAPTER 7 SUMMARY AND FUTURE WORK Conclusion ECIS was used to monitor human umb ilical vein endothe lial cells and 3T3 fibroblasts in response to the challenge of cytochalasin B an d protein kinase inhibitor H7, with a wide range of concentrations. By performing both micromotion and frequency scan experiments and analyzing the impeda nce data, we were able to distinguish cytochalasin B levels as low as 0.1 M and protein kinase inhib itor H7 concentration of 10 M for both cell lines. This suggests that these two methods provide a more sensitive assessment of cytotoxicity than the measuremen t of time course of overall impedance. Utilizing frequency spectroscopy and noise an alysis with impedance measurements makes ECIS a stronger analytic sy stem than traditional TER. CSCs cytotoxic nature against HUVEC cultures was characte rized with ECIS analysis. From that characterization, a cancer invasion assay was developed with HUVEC layers challenged by 50 g/mL of CSC. The HUVEC layers decreased tight junction binding that resulted from after about five hours of CSC exposure, allowed for a more aggressive ALST layer formation. Fewer than five hours of CSC exposure increased HUVEC layer tight junction binding that contributes to impeding ALST transmigration at 50 g/mL of CSC. Future Work Previous work utilizing ECIS to observe fluid flow has taken advantage of the system to measure TER 12,56. ECIS was utilized to show increased TER correlated with increased focal adhesion protein expression in human pulmonary artery endothelial cells 12. Even though the overall impedance was meas ured and characterized in these studies,

PAGE 128

109 micromotion noise analysis and frequency spec troscopy could be used to further reveal the affects of fluid flow on the cell-substrat e interface and junctional binding. A future project will involve ECIS analys is of fluid flow induced shear stress effects on endothelial barriers, cancer invasi on, and CSC challenged endothelial layers under cancer invasion. Preliminary studies ha ve been conducted on ECIS analysis of dynamic shear stress on HUVEC monolayers from low flow rates simulating the microcirculatory system to high flow rates simulating aortic vessels. The low flow rate studies will set up the investig ations of metastatic cancer cells under fluid flow. The ability of cancer cells to a ttach to the endothelium and ex travasate occurs within the microcirculation system where shear stress le vels are low. Depending on the results returned, the fluid flow apparatus may be utilized for previous experiments where tissue cultures were exposed to static conditions. The idea is to further simulate in vivo conditions. Specifically, a st udy will be conducted with how cigarette smoke condensate affects HUVEC layers under fluid flow conditi ons as well as ALST invasion in these same conditions.

PAGE 129

110 REFERENCES 1 I. Giaever and C. R. Keese, Proceedings of the National Academy of Sciences of the United States of America 88 (17), 7896 (1991). 2 C. M. Lo, C. R. Keese, and I. Giaever, Experimental Cell Research 204 (1), 102 (1993). 3 A. B. Zimrin, B. Villeponteau, and T. Maciag, Biochemical and Biophysical Research Communications 213 (2), 630 (1995). 4 B. Bodaghi, M. E. P. Slobbe-van Drunen, A. Topilko et al., Investigative Ophthalmology & Visual Science 40 (11), 2598 (1999). 5 C. L. Myers, S. N. Desai, J. Schembri king et al., American Journal of Physiology 262 (2), C365 (1992). 6 I. Giaever and C. R. Keese, Proceedings of the National Academy of Sciences of the United States of America 81 (12), 3761 (1984). 7 C. R. Keese, N. Karra, B. Dillon et al., In Vitro & Mo lecular Toxicology-a Journal of Basic and Applied Research 11 (2), 183 (1998). 8 I. Giaever and C. R. Keese, Physica D 38 (1-3), 128 (1989). 9 C. R. Keese, K. Bhawe, J. Wegener et al., Biotechniques 33 (4), 842 (2002). 10 C. Tiruppathi, A. B. Malik, P. J. Delvecchio et al., Proceedings of the National Academy of Sciences of the United States of America 89 (17), 7919 (1992). 11 N. DePaola, J. E. Phelps, L. Florez et al., Annals of Biomedical Engineering 29 (8), 648 (2001); J. Seebach, P. Dieterich, F. Luo et al., Laboratory Investigation 80 (12), 1819 (2000). 12 Y. Shikata, A. Rios, K. Kawkitinaro ng et al., Experimental Cell Research 304 (1), 40 (2005). 13 M. C. Lundien, K. A. Mohammed, N. Nasreen et al., Journal of Clinical Immunology 22 (3), 144 (2002). 14 Y. Shi, B. Reitmaier, J. Regenbogen et al., American Journal of Pathology 166 (1), 303 (2005).

PAGE 130

111 15 P. M. Ghosh, C. R. Keese, and I. Giaever, Biophysical Journal 64 (5), 1602 (1993). 16 A. Sapper, J. Wegener, and A. Janshoff, Analytical Chemistry 78 (14), 5184 (2006). 17 P. Geggier and G. Fuhr, Applied Phys ics a-Materials Science & Processing 68 (5), 505 (1999); L. B. Smilenov, A. Mikha ilov, R. J. Pelham et al., Science 286 (5442), 1172 (1999). 18 A. M. Goldberg and J. M. Frazier, Scientific American 261 (2), 24 (1989). 19 J. J. Pancrazio, J. P. Whelan, D. A. Borkholder et al., Annals of Biomedical Engineering 27 (6), 697 (1999). 20 L. Ceriotti, A. Kob, S. Drechsler et al., Analytical Biochemistry 371 (1), 92 (2007); J. Z. Xing, L. J. Zhu, J. A. J ackson et al., Chemical Research in Toxicology 18 (2), 154 (2005). 21 G. E. Slaughter, E. Bieberic h, G. E. Wnek et al., Langmuir 20 (17), 7189 (2004); F. Asphahani and M. Zhang, Analyst 132 (9), 835 (2007). 22 P. Mitra, C. R. Keese, and I. Giaever, Biotechniques 11 (4), 504 (1991). 23 C. Xiao, B. Lachance, G. Sunahara et al., Analytical Chemistry 74 (6), 1333 (2002). 24 K. S. C. Ko, C. M. Lo, J. Ferrier et al., Journal of Micr obiological Methods 34 (2), 125 (1998). 25 J. H. Yeon and J. K. Park, Analytical Biochemistry 341 (2), 308 (2005). 26 R. Ehret, W. Baumann, M. Brischwein et al., Fresenius Journal of Analytical Chemistry 369 (1), 30 (2001). 27 C. D. Xiao, B. Lachance, G. Sunahara et al., Analytical Chemistry 74 (22), 5748 (2002). 28 Y. L. Qiu, R. L. Liao, and X. Zhang, Analytical Chemistry 80 (4), 990 (2008). 29 J.B. Bassingthwaighte, L.S. Liebovitch, and B.J. Weat, Fractal Physiology (Oxford University Press, New York, 1994). 30 I. Giaever and C. R. Keese, IEEE Tr ansactions on Biomedical Engineering 33 (2), 242 (1986). 31 W. Berger, M. Micksche, and L. El bling, Experimental Cell Research 237 (2), 307 (1997).

PAGE 131

112 32 B. L. Brownstein, E. Rozengurt, L. Jimenezdeasua et al., Journal of Cellular Physiology 85 (3), 579 (1975). 33 B. H. Choi, J. A. Park, K. R. Kim et al., American Journal of Physiology-Cell Physiology 289 (2), C425 (2005). 34 R A Weinberg, The Biology of Cancer (Garland Science, New York, 2007). 35 H. et al. Ichikawa, Journal of Immunology 174 (11), 7383 (2005); L. J. Xu and X. M. Deng, Journal of Biological Chemistry 281 (7), 4457 (2006). 36 G. L. Nicolson, Current Opinion in Cell Biology 1 (5), 1009 (1989); F. W. Orr, H. H. Wang, R. M. Lafrenie et al., Journal of Pathology 190 (3), 310 (2000). 37 R. M. Lafrenie, T. J. Podor, M. R. Buchanan et al., Cancer Research 52 (8), 2202 (1992). 38 K. V. Honn, D. G. Tang, I. Grossi et al., Cancer Research 54 (2), 565 (1994). 39 P. C. Brooks, S. Stromblad, L. C. Sanders et al., Cell 85 (5), 683 (1996). 40 D. E. Epner, A. W. Partin, J. A. Schalken et al., Cancer Research 53 (9), 1995 (1993). 41 H. Ichikawa, Y. Takada, S. Shishodia et al., Molecular Cancer Therapeutics 5 (6), 1434 (2006); K. S. Ahn, G. Sethi, A. K. Jain et al., Journal of Biological Chemistry 281 (29), 19798 (2006). 42 S. S. Hecht, Environmenta l and Molecular Mutagenesis 39 (2-3), 119 (2002). 43 John C. Stavridis, in Oxidation: The Cornerstone of Carcinogenesis (Springer Netherlands, 2008), pp. 211. 44 S. van Wetering, J. D. van Buul, S. Quik et al., Journal of Cell Science 115 (9), 1837 (2002). 45 L. Moldovan, K. Mythreye, P. J. Goldschmidt-Clermont et al., Cardiovascular Research 71 (2), 236 (2006). 46 E. Giannoni, F. Buricchi, G. Raugei et al., Molecular and Cellular Biology 25 (15), 6391 (2005). 47 Y. Tagawa, N. Hiramatsu, A. Kasai et al., Free Radical Biology and Medicine 45 (1), 50 (2008). 48 K. Suzuki, H. Eguchi, Y. H. Koh et al., Biochemical and Biophysical Research Communications 257 (1), 214 (1999).

PAGE 132

113 49 R. M. Snajdar, S. J. Busuttil, A. Av erbook et al., Journal of Surgical Research 96 (1), 10 (2001). 50 J. Wang, D. E. L. Wilcken, and X. L. Wang, Molecular Genetics and Metabolism 72 (1), 82 (2001). 51 J. Barnoya and S. A. Glantz, Circulation 111 (20), 2684 (2005). 52 V. K. Kalra, Y. Ying, K. Deemer et al., Journal of Cellular Physiology 160 (1), 154 (1994); Y. M. Shen, V. Rattan, C. Sultana et al., American Journal of Physiology-Heart and Circulatory Physiology 39 (5), H1624 (1996). 53 B. X. Ou and D. J. Huang, Analytical Chemistry 78 (9), 3097 (2006). 54 D. Bernhard, A. Csordas, B. Henderson et al., Faseb Journal 19 (9), 1096 (2005). 55 Y. N. Ye, W. K. K. Wu, V. Y. Shin et al., European Journal of Pharmacology 519 (1-2), 52 (2005). 56 N. DePaola, J. E. Phelps, L. Florez et al., Annals of Biomedical Engineering 29 (8), 648 (2001); J. E. Phelps and N. De Paola, American Journal of PhysiologyHeart and Circulatory Physiology 278 (2), H469 (2000).

PAGE 133

ABOUT THE AUTHOR Daniel Winston Opp received a B.S. in Physics from the University of South Florida in 2004. Under the guidance of Dr. Chun-min Lo, he also received his M.S. in Physics from the University of South Fl orida in 2007. From 2007-2008, he was awarded an IGERT fellowship which focused on interdis ciplinary graduate training. His research with Dr. Lo and the IGERT group at USF has resulted in several c onference presentation as well as journal publications From fall 2008 to spring 2009, Daniel has been receiving formal training at the Moffitt Cancer Center as a medical physicist. His post-graduate work will involve research fo r the Moffitt Cancer Centers Radiation Oncology Physics group which will serve as his prerequisite experience in becoming a board certified medical physicist.


xml version 1.0 encoding UTF-8 standalone no
record xmlns http:www.loc.govMARC21slim xmlns:xsi http:www.w3.org2001XMLSchema-instance xsi:schemaLocation http:www.loc.govstandardsmarcxmlschemaMARC21slim.xsd
leader nam 2200409Ka 4500
controlfield tag 001 002064131
005 20100323123212.0
007 cr mnu|||uuuuu
008 100323s2009 flu s 000 0 eng d
datafield ind1 8 ind2 024
subfield code a E14-SFE0003064
035
(OCoLC)567549006
040
FHM
c FHM
049
FHMM
090
QC21.2 (Online)
1 100
Opp, Daniel.
0 245
ECIS assessment of cytotoxicity and trans-endothelial migration of metastatic cancer cells
h [electronic resource] /
by Daniel Opp.
260
[Tampa, Fla] :
b University of South Florida,
2009.
500
Title from PDF of title page.
Document formatted into pages; contains 113 pages.
Includes vita.
502
Dissertation (Ph.D.)--University of South Florida, 2009.
504
Includes bibliographical references.
516
Text (Electronic dissertation) in PDF format.
538
Mode of access: World Wide Web.
System requirements: World Wide Web browser and PDF reader.
590
Advisor: Chun-Min Lo, Ph.D.
520
ABSTRACT: The investigations conducted within this dissertation centers around the use of electric cell-substrate impedance sensing (ECIS). This system is able to characterize in real-time analysis, the adhesion of cells to their substrate and neighboring cells. With this, valuable information can be gathered with in-vitro experiments regarding a tissue culture's response to physiological stimulation. This dissertation has taken advantage of ECIS' ability to analyze toxicology, barrier function, and cancer invasion on a tissue culture. With proper analysis modifications, trans-epethelial resistance (TER) can be used as a cytotoxicity assay with higher sensitivity than previously thought. In vitro assessment of cytotoxicity based on TER needs more quantitative methods to analyze the alteration of cell morphology and motility.Here, we applied ECIS to evaluate dose-dependent responses of human umbilical vein endothelial cells (HUVEC) and mouse embryonic fibroblasts (NIH 3T3) exposed to cytochalasin B and protein kinase inhibitor H7. To detect subtle changes in cell morphology, the frequency-dependent impedance data of the cell monolayer were measured and analyzed with a theoretical cell-electrode model. To detect the alternation of cell micromotion in response to cytochalasin B and H7 challenge, time-series impedance fluctuations of cell-covered electrodes were monitored and the values of power spectrum, variance, and variance of the increment were calculated to verify the difference. While a dose-dependent relationship was generally observed from the overall resistance of the cell monolayer, the analysis of frequency-dependent impedance and impedance fluctuations distinguished cytochalasin B levels as low as 0.1M and H7 levels as low as 10 M for HUVEC and 3T3 layers.Even though overall resistance values are relatively small for 3T3 layers, and frequency scan measurements are negligible, impedance fluctuation analysis reveals significant micromotion for cytotoxic detection. Our results show that cytochalasin B and H7 causes a decrease of junctional resistance between cells and an increase of membrane capacitance. Cigarette smoke is cytotoxic and tumorigenic. Initial studies were conducted to evaluate the cytotoxicity of cigarette smoke condensate (CSC) on HUVEC layers. The focus was then turned to investigations involving in vitro cancer invasion assays with CSC on HUVEC layers. ECIS is an excellent investigative device that can be utilized to observe cancer invasion on normal tissue cultures due to the significantly higher impedance signature of cancer cells.The investigation in this dissertation focused on cigarette smoke's influence on cellular mechanics of endothelial cells and the invasive potential of two ovarian cancer cell lines (ALST and OVCA429) against a fully active endothelium. The HUVEC cultures responded to CSC with an increase in junctional binding, where as ALST and OVCA429 relieved adhesion thereby providing an improved motility when evaluated in wound healing assays. Transmigration of the HUVEC layer by ALST cells exhibit a pre-CSC exposure time-dependence affecting the effectiveness of ALST transmigration. The HUVEC layer's decreased tight junction binding that resulted from CSC exposure, allowed for a more aggressive ALST layer formation that occurred during simulated intravasation. Increased HUVEC layer tight junction binding that occurred in the first five hours in response to CSC during extravasation contributes to impeding ALST transmigration at high concentrations of CSC.Overall, CSC has an impeding effect on ALST transmigration during extravasation while causing aggressive transmigration during intravasation.
653
Micromotion
Cytochalasin B
Protien kinase inhibitor H7
Intravasation
Extravasation
690
Dissertations, Academic
z USF
x Physics
Doctoral.
773
t USF Electronic Theses and Dissertations.
4 856
u http://digital.lib.usf.edu/?e14.3064