University of South Florida

Digital Commons @ University of

South Florida

USF Tampa Graduate Theses and Dissertations USF Graduate Theses and Dissertations

April 2020

Manipulation and Patterning of Mammalian Cells Using Vibrations

and Acoustic Forces
Joel Cooper
University of South Florida

Follow this and additional works at: https://digitalcommons.usf.edu/etd

Part of the Acoustics, Dynamics, and Controls Commons, and the Biomedical Engineering and
Bioengineering Commons

Scholar Commons Citation

Cooper, Joel, "Manipulation and Patterning of Mammalian Cells Using Vibrations and Acoustic Forces"
(2020). USF Tampa Graduate Theses and Dissertations.
https://digitalcommons.usf.edu/etd/8178

This Dissertation is brought to you for free and open access by the USF Graduate Theses and Dissertations at
Digital Commons @ University of South Florida. It has been accepted for inclusion in USF Tampa Graduate Theses
and Dissertations by an authorized administrator of Digital Commons @ University of South Florida. For more
information, please contact [email protected].
 Manipulation and Patterning of Mammalian Cells Using

Vibrations and Acoustic Forces

by

Joel Cooper

A dissertation submitted in partial fulfillment

of the requirements for the degree of
Doctor of Philosophy
Department of Mechanical Engineering
College of Engineering
University of South Florida

Co-Major Professor: Nathan Gallant, Ph.D.

Co-Major Professor: Rasim Guldiken, Ph.D.
Daniel Hess, Ph.D.
Garrett Matthews, Ph.D.
Ryan Toomey, Ph.D.

Date of Approval:
April 12, 2020

Keywords: Faraday Waves, Resonance, Fluorescent Microscopy, Atomic Force Microscopy,

Tissue Engineering

Copyright © 2020, Joel Cooper

 Dedication
I would like to dedicate this dissertation to my amazing family, friends, and loved ones.

Their love and support throughout this journey made this work possible. They were always there

when I needed them, laughing with me through good times and helping me push through the bad

times.
 Acknowledgments
I would like to express my deepest appreciation to my advisors, Dr. Nathan Gallant and

Dr. Rasim Guldiken. This dissertation would not have been possible without their guidance,

support, and endless patience over the years. I am forever grateful to both of them for everything

they have done.

I would also like to thank the members of my committee for their invaluable advice

throughout my time here at USF. Dr. Daniel Hess who was always willing to share his insights

and understanding of vibrations. Dr. Ryan Toomey who helped broaden my knowledge of

polymers. Dr. Garrett Matthews who provided valuable knowledge of AFM.

I acknowledge and am grateful for the National Science Foundation. During my degree I

received support under NSF GRFP, GROW, and EAPSI. Through GROW and EAPSI, I was

able to live in Nagoya, Japan and conduct international collaborative research. I learned so much

through this experience and am truly grateful.

To my fellow students from the Cellular Mechanotransduction and Biomaterials Lab and

the Microfluidics and Acoustics Lab, both past and present, thank you all for your support,

collaboration, and friendship.

Finally I would like to thank my family and friends for their unconditional love and

constant support. They were with me every step of the way.

 Table of Contents

List of Tables ................................................................................................................................. iii

List of Figures ................................................................................................................................ iv

Abstract ......................................................................................................................................... vii

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

Chapter 2: Manipulation of Cell Attachment and Viability via Insonification .............................. 4

2.1 Background ................................................................................................................... 4
2.2 Experimental Design ..................................................................................................... 6
2.2.1 Cell Culture and Reagents ............................................................................. 6
2.2.2 Automated Fluorescent Microscopy Analyses .............................................. 6
2.2.3 Acoustic Manipulation Experiment Protocol................................................. 7
2.2.4 Statistical Analysis ....................................................................................... 10
2.3 Results ......................................................................................................................... 10
2.3.1 Effects of Insonification on Cell Attachment .............................................. 10
2.3.2 Effects of Insonification on Cell Viability ................................................... 14
2.3.3 Effects of Distance from Source of Acoustic Excitation on Cell
Attachment ............................................................................................................ 16
2.4 Conclusions ................................................................................................................. 18

Chapter 3: Manipulation and Patterning of Cells using Acoustic Forces ..................................... 20

3.1 Background ................................................................................................................. 20
3.2 Computer Modeling and Experimental Validation ..................................................... 21
3.3 Methods and Experimental Setup ............................................................................... 23
3.3.1 Cell Culture and Reagents ........................................................................... 23
3.3.2 Cell Manipulation ........................................................................................ 24
3.3.3 Automated Fluorescent Microscopy Analyses ............................................ 25
3.3.4 Macro Fluorescent Image Analyses ............................................................. 26
3.4 Results and Discussion ............................................................................................... 26
3.4.1 Simulation Validation .................................................................................. 26
3.4.2 Cell Pattern Characterization using Macro Images ..................................... 29
3.5 Conclusion and Future Work ...................................................................................... 54

i
Chapter 4: Key Findings and Future Work ................................................................................... 56
4.1 Incomplete and Ongoing Work: Characterization of Cellular Mechanical
Properties using AFM ....................................................................................................... 57
4.1.1 Background .................................................................................................. 57
4.1.1.1 Cell Mechanical Properties: Elasticity and Natural
Frequency.................................................................................................. 58
4.1.1.2 Cytoskeletal Composition and Cell Shape .................................... 59
4.1.1.3 Objectives ..................................................................................... 60
4.1.2 Methods and Materials ................................................................................. 61
4.1.2.1 Cell Culture and Reagents ............................................................ 61
4.1.2.2 Fabrication of Microcontact Printing Stamp................................. 62
4.1.2.3 Microcontact Printing Procedure .................................................. 63
4.1.2.4 Experimental Procedure ................................................................ 65
4.2 Limited Results ........................................................................................................... 67
4.2.1 Microcontact Printing Direct Proteins ......................................................... 67
4.2.2 ESEM Analysis of Mammalian Cells .......................................................... 68
4.3 Conclusion and Future Work ...................................................................................... 69
4.3.1 Future Work: AFM and Cytoskeletal Inhibitors .......................................... 69

References ..................................................................................................................................... 71

Appendix A Copyright Permissions ............................................................................................. 76

ii
 List of Tables

Table 1 SolidWorks vs Ansys modal simulation results for a 100mm polystyrene tissue
culture dish ...................................................................................................................... 27

Table 2 Longitudinal measurements of the distance between the center of the bands of
cells for 40Hz on Day 1 ................................................................................................... 37

Table 3 Longitudinal measurements of the distance between the center of the bands of
cells for 40Hz on Day 2 ................................................................................................... 37

Table 4 Longitudinal measurements of the distance between the center of the bands of
cells for 40Hz on Day 3 ................................................................................................... 37

Table 5 Longitudinal measurements of the distance between the center of the bands of
cells for 75Hz on Day 1 ................................................................................................... 37

Table 6 Longitudinal measurements of the distance between the center of the bands of
cells for 75Hz on Day 2 ................................................................................................... 38

Table 7 Longitudinal measurements of the distance between the center of the bands of
cells for 75Hz on Day 3 ................................................................................................... 38

iii
 List of Figures

Figure 1. Schematic representation of experimental setup. ............................................................ 7

Figure 2 Representative macro images of fluorescently labeled cells before and after
insonification with maximum or minimum energy for high and low power. ................ 10

Figure 3 The percent of cells remaining attached (a) 1 h or (b) 24 h after sonication at
various power and energy values. .................................................................................. 11

Figure 4. The percent of viable cells (a) 1 h or (b) 24 h after insonification at various
power and energy levels. ............................................................................................... 14

Figure 5. Relation between acoustic intensity and cell detachment after exposure to a
power of 3 watts and energy of 3 joules (3W3J). .......................................................... 16

Figure 6 SolidWorks modal frequency simulation results compared to experimental

modal frequencies. ......................................................................................................... 22

Figure 7 10µm polystyrene spheres in 5ml of DI H2O at 75Hz. .................................................. 28

Figure 8 Macro image of green fluorescent NIH3T3 fibroblasts during vibrational

patterning at 40Hz. ......................................................................................................... 29

Figure 9 Macro image of NIH3T3 fibroblasts after patterning at 40Hz. ...................................... 29

Figure 10 Macro image of NIH3T3 fibroblasts after patterning at 40Hz. .................................... 30

Figure 11 Macro image of NIH3T3 fibroblasts at 40 Hz with increased amplitude. ................... 31

Figure 12 Representative image of NIH3T3 Fibroblasts patterned at 40Hz. ............................... 33

Figure 13 Representative image of NIH3T3 Fibroblasts patterned at 75Hz. ............................... 34

Figure 14 Representative image of distance measurements ......................................................... 36

Figure 15 Average distances and standard deviations between bands of cells over time............. 38

Figure 16 Representative fluorescent intensity vs location for 40Hz. .......................................... 39

Figure 17 Combined fluorescent intensity data for 40Hz 4 hours after patterning ...................... 41

Figure 18 Combined fluorescent intensity data for 75Hz 4 hours after patterning ...................... 41

iv
Figure 19 40Hz 4 hours after patterning. ...................................................................................... 42

Figure 20 40Hz 4 hours after patterning. ...................................................................................... 43

Figure 21 40Hz 4 hours after patterning. ...................................................................................... 43

Figure 22 40Hz 4 hours after patterning. ...................................................................................... 44

Figure 23 40Hz 4 hours after patterning. ...................................................................................... 44

Figure 24 40Hz 24 hours after patterning. .................................................................................... 44

Figure 25 40Hz 24 hours after patterning. .................................................................................... 45

Figure 26 40Hz 24 hours after patterning. .................................................................................... 45

Figure 27 40Hz 24 hours after patterning. .................................................................................... 45

Figure 28 40Hz 24 hours after patterning. .................................................................................... 46

Figure 29 40Hz 48 hours after patterning. .................................................................................... 46

Figure 30 40Hz 48 hours after patterning. .................................................................................... 46

Figure 31 40Hz 48 hours after patterning. .................................................................................... 47

Figure 32 40Hz 48 hours after patterning. .................................................................................... 47

Figure 33 40Hz 48 hours after patterning. .................................................................................... 48

Figure 34 75Hz 4 hours after patterning. ...................................................................................... 48

Figure 35 75Hz 4 hours after patterning. ...................................................................................... 48

Figure 36 75Hz 4 hours after patterning. ...................................................................................... 49

Figure 37 75Hz 4 hours after patterning. ...................................................................................... 49

Figure 38 75Hz 4 hours after patterning. ...................................................................................... 49

Figure 39 75Hz 24 hours after patterning. .................................................................................... 50

Figure 40 75Hz 24 hours after patterning. .................................................................................... 50

Figure 41 75Hz 24 hours after patterning. .................................................................................... 50

Figure 42 75Hz 24 hours after patterning. .................................................................................... 51

Figure 43 75Hz 24 hours after patterning. .................................................................................... 51

v
Figure 44 75Hz 48 hours after patterning. .................................................................................... 51

Figure 45 75Hz 48 hours after patterning. .................................................................................... 52

Figure 46 75Hz 48 hours after patterning. .................................................................................... 52

Figure 47 75Hz 48 hours after patterning. .................................................................................... 52

Figure 48 75Hz 48 hours after patterning. .................................................................................... 53

Figure 49 Fabrication of microcontact printing stamp and stamping procedure. ......................... 62

Figure 50 Surfcorder ET200 used to verify etch depth of photoresist .......................................... 63

Figure 51 Nano-robotic manipulation system inside ESEM ........................................................ 65

Figure 52 ESEM images of AFM cantilever calibration. ............................................................. 65

Figure 53 Green fluorescent fibronectin stamped onto a clean glass slide. .................................. 67

Figure 54 Successfull cell attachment to fibronectin islands ........................................................ 67

Figure 55 NIH3T3 fibroblast imaged using ESEM. ..................................................................... 68

Figure 56 ESEM cell elasticity using tipless AFM cantilevers. ................................................... 69

vi
 Abstract

Recently, there has been a surge in researchers and scientists investigating different

methods which move, manipulate, and pattern biological cells. Multiple different mechanisms

can be used for cellular manipulation, microfluidics, biochemical queues, and even optics, just to

name a few. However, all techniques have their downsides. A majority of these methods require

expensive equipment or reagents and can only manipulate a small number of cells at a time.

Some of the most common cell manipulation devices utilize acoustic pressure waves to

move the cells to desired locations. Currently, it is unknown what level of force from these types

of devices a biological cell can withstand before irreparable damage occurs. The first section of

this dissertation investigates this issue. Briefly, this study found that the power into the acoustic

device, cell exposure time to the acoustic waves and the distance from the source of the acoustic

wave generator all effect cell attachment and viability. These results aid in providing a better

understanding of the acoustic pressures a cell can withstand before permanent damage occurs.

The core of the research described in this dissertation investigates a new method of cell

manipulation and patterning. This new method utilizes standing waves, generated by an

inexpensive mechanical vibrator, to manipulate cells into the mode shapes of the container the

cells are within.

NIH3T3 Fibroblasts were successfully patterned using the frequencies 40Hz and 75Hz

and the overall pattern persisted for over 48 hours. Further investigations confirmed that the

vii
patterning method described here had no noticeable negative effects on cell viability,

proliferation, or migration. The results of this research provide a fast, flexible, and inexpensive

method for manipulating and patterning large numbers of cells.

viii
 Chapter 1: Introduction
One of the biggest dilemmas in tissue engineering currently is the development of proper

3D vasculature throughout engineered tissues [1]. Vasculature is critical to the success of an

implanted constructed tissue as it is required for sufficient nutrient, oxygen and waste transport

to and from cells within the structure. Without access to a vascular network, cells outside the

diffusion limit [1, 2] will develop a buildup of metabolic waste and be starved of nutrients; this

combination can lead to cell death and eventual failure of the implant [1]. Previously it has been

shown that the body responds to an implanted tissue by developing its own vascular network

throughout; however, this is a slow process and cannot be relied upon for successful implant

integration.

Optimal integration of this network must mimic the function and structure of vasculature

in native tissues an implant is attempting to replace. This means the engineered vascular network

must be distributed throughout the tissue similar to capillaries, connect to larger vessels acting as

veins and arteries, and have locations where the host’s native vasculature can be connected

through surgical means [1-3]. Knowing this, tissue engineers have begun developing methods for

creating vasculature in an engineered tissue in vitro. These methods include biodegradable and

decellularized scaffolds, cell self-assembly, perfusable microvascular networks, and many others

[3]. Each of these techniques has their disadvantages. For example biodegradable scaffolds with

polymer remnants can degrade the mechanics of tissue [4] and cell self-assembly has been shown

to unroll during high blood flow [5]. While these methods of developing vasculature do have

their benefits, it is clear that new techniques should still be investigated.

1
 This research focuses on manipulating the spatial organization of cells to rapidly form

patterned layers of cells. Ultimately this novel method could enable the formation of complex

structures such as vasculature without the use of biomaterials or expensive specialized

equipment. The cell manipulation method discussed in this paper stems from cymatics, a subset

of modal vibrations. Vibrational and acoustic forces, induced by a simple mechanical wave

driver and signal generator, act on the cells suspended in media and force the cells to move to

nodal locations. However, the interaction between acoustic forces and biological cells is a

relatively recent research topic and there are multiple gaps in the current literature. In order to

successfully utilize acoustic forces in cell manipulation, these gaps must be investigated.

First, there has been little investigation into both the immediate and long-term effects of

acoustic forces on cell viability. Previous acoustic particle manipulation studies have shown that

small forces generated by interdigital transducers can move particles and cells in suspension but

cannot generate enough force to detach adherent cells [6-9]. It is also known that very large

acoustic forces can cause cavitation which destroys cell membranes, effectively killing the cell.

The research described in this section delves into the intermediate range of acoustic pressure

waves and attempts to shed light on two main questions: (1) how much acoustic force cells can

handle without obtaining permanent damage and (2) how much acoustic force is required to

detach cells from a normal tissue culture dish.

The second section, and main portion of this research, focuses on the manipulation and

patterning of cells in common polystyrene tissue culture dishes using only vibrational and

acoustic forces. Preliminary work has proven the importance of computer modeling for accurate

prediction of the parameters required to achieve specific patterns. Insights gained from the

previous section aids in the development of these computer models and experimental data is be

2
used to verify accuracy. This section delves into successful cell patterning and analyses long

term cell viability, pattern retention over time; ensuring this new method of cellular manipulation

is practical for researchers worldwide.

Finally, it is known that the acoustic force acting on a particle in suspension varies based

on the compressibility of the particle [10].Unlike microbeads used in a majority of acoustic

manipulation studies, biological cells are heterogeneous in structure and mechanical properties of

the cell can change depending on a multitude of factors. This is due to the ever-changing internal

components of the cell’s cytoskeleton and it’s response to stimuli. Current literature discussing

the cell’s mechanical properties, including compressibility measurements, only probe minimally

into the cell’s surface correlating to local stiffness values [11-14]; yet quantification of acoustic

force on a cell requires the overall or global stiffness. This section discusses ongoing and future

work which will lead to the measurement of two physical properties of cells: (1) global cellular

compressibility and (2) global cellular natural frequency, a property which has previously been

ignored in the literature but is important when working with acoustics. Additionally, this section

discusses the possible contribution of the shape of the cell as well as specific cytoskeletal

elements to its overall, or global, mechanical properties.

3
 Chapter 2: Manipulation of Cell Attachment and Viability via Insonification
2.1 Background

Manipulation of bio-particles and live cells is crucial to a variety of biomedical

applications, including diagnostics and fundamental studies of cell-cell interactions [7, 15-19]. In

recent decades, interest has grown in the use of acoustic devices, mostly in the ultrasound range,

to manipulate, pattern, or alter biological cells. While there have been several successful

approaches [7, 9, 20-28], only a few investigated cell viability after ultrasonic exposure [25, 29-

31]. Additionally, these in vitro studies only examined cells in suspension, and none attempted to

use acoustic devices to manipulate or alter cells adhered to a substrate. Currently, there is a gap

in the understanding of how adhered cells are affected by exposure to acoustic excitations large

enough to detach the cell from the surface.

Therapeutic ultrasound has a long history [24, 27, 32, 33], however the use of acoustics

for cellular manipulation was not reported until the 1970s by Dyson et al. [34]. This work

revealed that a stationary wave field excited from an ultrasonic transducer could suspend blood

cells flowing through the vessels of chick embryos. Notably, further investigation revealed that

the plasma membranes of the endothelial cells lining these vessels had been damaged, producing

a similar effect to thrombosis [35]. These studies helped spark an interest in using ultrasound for

bio-manipulation techniques.

Modern ultrasonic transducers used in bio-manipulation operate at relatively low input

power, ranging from 0.5W to 1.5 W, resulting in 100 pN scale forces applied to the particle or

cell [7, 20, 36]. These sub-nN scale forces have proven to be suitable for manipulating particles

4
in static fluid; however, studies have shown that as flow rates increase, acoustic trapping forces

are overcome, and guided movement efficiency decreases [20, 37]. This decrease in efficiency

can be addressed by increasing the input power to the transducer, correlating to a larger force

restraining the particle [20, 37]. Large forces can distort, damage, or even destroy a cell;

therefore, it is important to determine the range of acoustic excitations that expose cells to forces

which they can withstand without suffering permanent damage.

Current acoustic manipulation techniques are typically only used on cells suspended in

fluid media and not applied to cells adhered onto a surface. We have previously showed that an

approximately 200 nN adhesive force must be overcome to detach individual fibroblasts [38, 39].

This is orders of magnitude higher than previously reported acoustic forces on cells in

suspension [6, 7, 36]. It is clear that acoustic transducers must generate at least 1000 times as

much force compared to those in current use, while not irreversibly disrupting cellular integrity if

they are to be used to manipulate adherent cells. The development of more powerful therapeutic

devices and cell manipulation techniques utilizing acoustic transducers requires further

investigation of the effects of insonification on cell attachment and viability.

In this study, these relationships were investigated by using a sonication probe, a device

commonly used for cell lysis, which operates at the low end of ultrasonic frequencies, as a

simple tool for acoustic excitation of adherent cells. Adherent cells were insonified over a range

of 3 to 48 J while varying the power and duration of acoustic excitation. The findings presented

here demonstrate that the power output from the sonicator, the exposure time, and the acoustic

intensity all play a role in cell detachment and regulating viability.

5
2.2 Experimental Design

2.2.1 Cell Culture and Reagents

NIH3T3 murine embryonic fibroblasts (CRL-1658, ATCC) were cultured in complete

growth medium (CGM) containing 10% newborn calf serum and 1% penicillin-streptomycin in

Dulbecco’s modified Eagle’s medium. Cells were passaged every three days, using 0.25%

trypsin-EDTA and used between passages 5 and 20. Cells were labeled with 3µM calcein-AM

suspended in Dulbecco’s phosphate-buffered saline containing calcium and magnesium (DPBS).

The initially non-fluorescent calcein AM penetrates the cellular membrane where intracellular

esterases within living cells hydrolyze acetoxymethyl ester, producing a green fluorescence in

the cytoplasm. This fluorescence is only retained in live cells and quickly dissipates if the cell’s

membrane is compromised [40, 41]. Cell culture reagents were purchased from Invitrogen unless

otherwise noted.

2.2.2 Automated Fluorescent Microscopy Analyses

Cell attachment and spatial distribution analyses in each sample were performed using an

Eclipse Ti-U (Nikon Instruments) inverted microscope equipped with a CCD camera

(CoolSNAP HQ2 Photometrics) and appropriate fluorescent filter sets. Automated fluorescence

microscopy, controlled by NIS-Elements software (version 4.20, Nikon Instruments), was used

to quantify cell number and distribution in each experiment. This automation used a computer-

controlled stage, in conjunction with the microscope setup described above, to capture

fluorescent images at 73 different x-y locations on each dish for a total analyzed area of 87.5

mm2 (9% of total dish area). The number of cells at each location was recorded.

6
2.2.3 Acoustic Manipulation Experiment Protocol

Cells were seeded 16 hours before the experiment at 100 cells/mm2 onto 35mm sterile

tissue culture polystyrene (TCP) dishes (Corning). This time was chosen so cells would attach

and fully spread but not proliferate appreciably before insonification [38, 42]. During this time,

the cells were incubated at 75% relative humidity, 5% CO2, and 37°C. The next day the media

was aspirated, and the cells were rinsed with 2ml DPBS. Fluorescent labeling of live cells was

achieved by adding 0.75ml of 3µM calcein AM to each dish, which were then incubated for 30

minutes. Following labeling, the cells were washed with DPBS, and 2ml of fresh CGM was

added to each dish. Samples were individually imaged and analyzed using the previously

described automated fluorescent microscopy technique to determine the average cell density and

cell distribution information prior to insonification.

Figure 1. Schematic representation of experimental setup. (a) Sonicator probe with microtip horn attachment
inside biosafety cabinet with close up of microtip probe submerged in a 35mm diameter tissue culture
polystyrene (TCP) dish. (b) Before excitation, the microtip was submerged in the TCP dish containing growth
media and living cells attached to the surface of the dish. (c) During excitation, the sonicator is activated and
the microtip insonifies the region. Green and red particles represent alive and dead cells, respectively.

7
 After initial imaging, 4ml of additional fresh CGM was added to each dish. This was

done per operation manual recommendations to operate below the cavitation region [43]. Cells

were then exposed to one of four power settings (3W, 6W, 9W, or 15W) for varied time intervals

(1s to 15s) to obtain a range of total energy inputs (3 joules to 48 joules) as calculated by

equation (1) below. A Misonix Sonicator 3000 (Cole-Parmer), equipped with a 1/8 inch micro-

tip horn that operates at 20 kHz and vibrates in the longitudinal direction [43], was used to

generate the desired acoustic excitation in each experiment from a fixed position 2mm above the

dish surface Figure 1. This unfocused sonication probe is capable of generating an intermediate

range of acoustic intensities representing the gap between microtransducers, generating 0.05

W/cm2 [7] to 1 W/cm2 [44], and high-intensity therapeutic devices, which typically generate

several W/cm2 to orders of magnitude more [45, 46]. Cellular damage due to cavitation was

considered; however, this effect was minimized by maintaining a 2mm gap between the fully

submerged probe tip and the cells on the surface of the dish. Sonication devices are known to

generate heat, so the temperature was recorded using a thermal probe (SD-3007 Reed

Instruments) to ensure the dish never exceeded 40 °C, which is generally regarded as safe for

mammalian cells [47].

Energy (J) = Power (W) * Time (s) (1)

Immediately after insonification, the medium, containing cells which had detached from

the plate, was carefully transferred from the original dish to a new sterile 35mm tissue culture

dish, and 2ml of fresh CGM was carefully dispensed into the original dish. This allowed

differentiation between cells which remained attached, cells destroyed during insonification, and

cells detached from the plate but survived. Each sample, now consisting of an original dish and a

re-plated dish, was incubated at 37°C for 1 hour to allow the detached cells to adhere to the new

8
dish [38, 42] before the second round of microscopic imaging and analysis. After imaging, 4ml

of media was removed from the re-plated dishes to allow for proper O2 diffusion. Each dish was

then incubated overnight under standard culture conditions. The medium was aspirated 24 hours

after the initial imaging, and each dish was rinsed with 2ml DPBS before re-labeling with

calcein-AM for 30 min. Following this final incubation, samples were washed with DPBS, and

2ml of fresh CGM was added to each dish. Each sample was then imaged and quantified a third

and final time.

Following this procedure, the data consisted of the number and distribution of cells in

each dish at three unique time points: before exposure (BE), immediately after exposure (AE),

and 24 hours after initial imaging. Considering the latter two time points contain data from both

the (1) original and (2) re-plated dishes, this analysis provides a total of five unique sets of data

for each sample. The number of cells contained in each data set was averaged and divided by the

analysis area to determine cell density in cells/mm2. To account for variations in cell seeding

density, attachment, and viability data were normalized to respective “before exposure” cell

densities. Percent cell attachment and viability were determined immediately after exposure by

the following equations.

AE (1)
% Cells Attached = ∗ 100 (2)
BE

[AE (1)+AE (2)]

% Viable Cells = ∗ 100 (3)
BE

Similarly, cell attachment and viability 24 hours after exposure were determined by

substituting data gathered 24 hours after initial imaging into the above equations for AE.

9
2.2.4 Statistical Analysis

Data are reported as mean values ± standard deviation of three independent experiments,

each performed in triplicate. Statistical comparisons, performed using SigmaPlot 11.2 software,

were based on analysis of variance and the Holm-Sidak post hoc test for multiple pairwise

comparisons, with a p-value < 0.05 considered significant.

2.3 Results

2.3.1 Effects of Insonification on Cell Attachment

Significant forces are required to detach adherent and well-spread cells [16]. The effect of

acoustic excitation from a sonic transducer on cell adhesion was investigated using a sonication

horn adjusted to various power and total energy levels. Figure 2 contains representative macro

images of four different 35mm TCP dishes before and after insonification. Before exposure, cells

were fully attached and uniformly distributed over each dish surface. After exposure, the

distribution of cells that remained attached varied depending on power, energy, and distance

from the excitation source.

3W3J 3W45J 15W15J 15W45J

a b c d
Before

e f g h
After

Figure 2 Representative macro images of fluorescently labeled cells before and after insonification with
maximum or minimum energy for high and low power. Before excitation, cells were fully attached and
uniformly distributed over each dish surface (a-d). Unique distributions arose for cells remaining attached after
exposure to (e) 3 watts and 3 joules (3W3J), (f) 3 watts and 45 joules (3W45J), (g) 15 watts and 15 joules
(15W15J), or (h) 15 watts and 45 joules (15W45J). Live cells fluorescently labelled with calcein AM were
adhered in 35mm diameter tissue culture polystyrene (TCP) dishes.

10
 Cells insonified at the lowest power setting, 3 watts, coupled with the lowest energy, 3

joules, resulted in a distinct radial cell distribution. The lack of fluorescence at the center of the

dish indicates significant cell detachment from the surface at radial distances near the excitation

source, while cells located further from the center remained adhered to. Using the same power

but increasing energy exposure to 45 joules yielded a very different result. At this combination of

power and energy dose, a majority of cells were uniformly detached from the surface. The only

difference was an increase in energy (extended duration), suggesting a correlation between

energy levels and cell attachment. On the higher end of the power spectrum (15 watts), energy

levels affected cell adhesion to a lesser degree; exposure to either 15 or 45 joules caused

significant cell detachment. The representative images in Figure 2 can also be used to compare

samples with similar energy levels but different power inputs. Cells exposed to 45 joules using

powers 3 or 15 watts were similar in that a majority of cells detached.

Analyzing a range of energy and power inputs further suggests that the power and energy

of the acoustic excitation strongly modulate cell attachment Figure 3. Using equation (2), the

a 70
b 70
3W 3W
Percentage of Cells Attached (24h)
Percentage of Cells Attached (1h)

60 6W 60 6W
9W 9W
50 15W 50 15W

40 40

30 30

20 20

10 10

0 0

3 6 9 12 15 18 21 24 27 30 33 36 39 42 45 48 3 6 9 12 15 18 21 24 27 30 33 36 39 42 45 48
Energy (J) Energy (J)

Figure 3 The percent of cells remaining attached (a) 1 h or (b) 24 h after sonication at various power and
energy values. Data points grouped within dashed boxes show no significant differences among each other
(p>0.05). Groups of data with significant differences (p<0.05) among power series for similar energy ranges
are indicated with dashed arrows.

11
percent of cells remaining attached 1 hour after being insonified and 24 hours after initial

imaging were determined. Data points grouped inside of dashed boxes in Figure 3 and Figure 4

showed no significant differences among each other (p>0.05), while groups of data with

significant differences (p<0.05) for similar energy levels are denoted using dashed arrows.

One hour after insonification, the lowest power setting, 3 watts, had a significantly higher

percentage of cells remain attached compared to the other power settings up to 45 joules. At 45

joules, cell adhesion was reduced to below 10% for all power levels. The 3-watt input also

induced the largest decrease in the attachment as energy was increased, ranging from an average

of 39% remaining attached at 3 joules and decreasing down to an average of 8% remaining

attached at 45 joules. In contrast, the highest power setting, 15 watts, detached nearly all the cells

for all the energy levels tested. These results are consistent with a previous report which

concluded that as the power of the acoustic excitation increases, the percent of cells detached

from the surface also increases [48].

The 3, 6, and 9-watt power settings had significant differences in the percent of cells

remaining attached at different energy levels within each respective series; however, the 15-watt

power setting shows no significant differences as few cells remained at all energy levels. The 3-

watt series had significant differences in the percent of cells remaining attached at the following

energy values: 3 joules and 27 joules (p=0.041), 3 joules and 45 joules (p=0.008), 6 joules and

27 joules (p=0.038), 6 joules and 45 joules (p=0.008), 9 joules and 45 joules (p=0.024). The 6-

watt series had significant differences in the percent of cells remaining attached at 6 joules and

48 joules (p=0.008). The 9-watt series had significant differences (p<0.05) in the percent of cells

remaining attached at all energy values except between 27 joules and 45 joules. These significant

trends demonstrate that relatively low power acoustic excitations (3 – 9 W) can detach more cells

12
from the surface by merely increasing the energy into the system. The facts that few cells

remained attached and that there were no significant differences within the 15W series suggests

that there is a threshold acoustic power which detaches most of the cells from the surface

regardless of the energy input.

Similar trends were observed after allowing cells to recover for 24 hours. That is,

following the lowest power exposure, 3 watts, a significantly higher percentage of cells remain

attached when compared to the other power series. Again, this difference disappeared when the

total energy input reached 45 joules, where none of the power settings were significantly

different from each other in terms of cell adhesion.

Data collected 24 hours after the initial imaging showed there were significant

differences in the fraction of cells attached at different energy levels within the 3, 9, and 15-watt

power settings. The 3-watt series had significant differences in the fraction of cells remaining

attached at 3 joules and 45 joules (p<0.001) and 9 joules and 45 joules (p=0.014). The 9-watt

series had significant differences (p<0.05) in the fraction of cells remaining attached at all energy

values except between 27 joules and 45 joules. The 15-watt series had a significant difference

only between the 15 joules and 30 joules (p=0.031) settings.

Healthy NIH3T3 fibroblasts have a doubling time of approximately 20 hours. Comparing

the cell count with the corresponding data 24 h later showed that there were no statistically

significant increases or decreases in the number of cells remaining attached over 24 hours;

therefore, these populations did not follow this standard doubling time. This suggests that the

acoustic excitation either disrupted internal cellular mechanisms responsible for the proliferation,

or some cells were damaged beyond repair, dying at a rate similar to those proliferating. While

not statistically significant, there were minimal increases and decreases in the percent of cells

13
over 24 hours. Specifically, the 3-watt series tended to increase, whereas the 6, 9, and 15-watt

series all tended to decrease, suggesting that long term survival and proliferation of cells which

remained attached after insonification largely depends on the power of the excitation.

2.3.2 Effects of Insonification on Cell Viability

In addition to insonification-related cell detachment, effects on cell viability were also

investigated. The percent of viable cells was determined using equation (3) whereby the sum of

viable cells which remained attached and those that re-attached was divided by the initial cell

count. The viable fraction 1 hour after insonification and 24 hours after the initial imaging are

displayed in Figure 4a and Figure 4b, respectively. One hour after excitation, the 6-watt power

setting at 6 joules had the highest percentage of viable cells (approximately 60%). The lowest

viability belonged to the 9-watt power series at 45 joules, where only 4% of the cells were still

viable. The 15-watt power series showed a consistently low survival rate regardless of the energy

into the system (<20%).

The 6-watt power series had a significant difference between 6 joules and 48 joules

(p=0.023). The 9-watt power series had significant differences between 9 joules and 45 joules

a b
80 3W 80 3W
Percentage of Viable Cells (24h)

6W 6W
Percentage of Viable Cells (1h)

70 9W 70 9W
15W 15W
60 60

50 50

40 40

30 30

20 20

10 10

0 0

3 6 9 12 15 18 21 24 27 30 33 36 39 42 45 48 3 6 9 12 15 18 21 24 27 30 33 36 39 42 45 48
Energy (J) Energy (J)
Figure 4. The percent of viable cells (a) 1 h or (b) 24 h after insonification at various power and energy levels.
Data points grouped within dashed boxes show no significant differences among each other (p>0.05). Groups
of data with significant differences (p<0.05) among different power series for similar energy ranges are
indicated with dashed arrows.

14
(p<0.001), and between 18 joules and 48 joules (p<0.001). There were no significant differences

in the percent of viable cells between any energy levels in the 3-watt or 15-watt power series.

Interestingly there were no significant differences among the different power series at similar

energy levels. For example, exposing adhered cells to 3 watts 9 joules gives similar results as

exposing them to 9 watts 9 joules. There were many significant differences between the 6 and 9-

watt power series at different energy levels. This could indicate that initial viability is determined

more by energy exposure than the power of the excitation.

At 24 hours after initial imaging, the 3-watt series retained the highest percent of viable

cells over 24 hours. At all energy levels tested, the 3-watt series had significantly more cells

remain viable than any other power input tested. For each power series, there were no significant

differences in the percent of viable cells for all energy levels. However, there were significant

differences between the percent of viable cells analyzed 1 hour after insonification and the

percent of viable cells 24 hours after initial imaging. Overall, every point on the 3-watt series

showed an increase in viable cells over 24 hours, indicating cell proliferation; however, only the

samples exposed to an energy of 27 joules showed a statistically significant increase (p=0.003).

Both the 6 and 9-watt power series trended towards a decrease in viable cells over 24 hours, with

a few of the points significantly different. The 6-watt series at 6 joules showed a significant

decrease (p=0.026). At 9 watts, both the 9 and 18 joule energy settings showed a significant

decrease (p=0.033 and p=0.003 respectively). The 15-watt series had no significant increase or

decrease in the percent of viable cells over 24 hours.

Overall, these findings showed that long term viability of cells after insonification

depends on both power of the acoustic transducer and the energy into the system. It also

demonstrated that the initial assessment of cell viability does not indicate long-term survival.

15
Cells exposed to 6 watts and 6 joules initially had the highest percent survive, but over 24 hours,

these cells exhibited marked decreases in viability. Furthermore, while many cells initially

survived detachment and were able to re-attach to a new surface, a majority of those (especially

>3W) must have sustained damage since long-term viability was diminished (≥24 h).

2.3.3 Effects of Distance from Source of Acoustic Excitation on Cell Attachment

Cell detachment was nearly uniform for insonification greater than 6 J; however, a

distinct biphasic spatial distribution resulted in lower energy trials Figure 2e. The distribution

comprised of overall higher cell survival and decreased cell detachment beyond a few

millimeters from the source of excitation. A spherical pressure wave at the center of the dish,

such as that generated by the sonicator microtip, would cause a radial gradient of insonification

with an acoustic intensity that diminishes as it travels farther from the probe [49, 50].

Figure 5. Relation between acoustic intensity and cell detachment

after exposure to a power of 3 watts and energy of 3 joules (3W3J).
Radial values are relative to the center of the dish, closest to the tip
of the excitation source. A sigmoidal curve was fit to attached cells
for 3 independent experiments.

16
 In Figure 5, acoustic intensity for a spherical pressure wave emanating from the

transducer tip, calculated using equation (4), is compared to the percent of cells remaining

attached after insonification; both of these values are plotted from the center of the tissue culture

dish to the outside edge in the radial direction. The power output from the transducer was

divided by the cross-section area of the spherical pressure wave as it interacts with the surface of

the dish to calculate the radial dependence of intensity. The calculated acoustic intensity was

compared to the percent of cells remaining attached after exposure at corresponding radial

locations. We observed that cell detachment had an inverse relationship with each cell’s relative

distance from the sonicator, and was directly related to the acoustic intensity incident on the

surface.

Concerning established cell adhesion characterization methods, this result resembles the

detachment trend in a radial flow assay where shear stress is inversely proportional to the radial

position in the laminar flow [16]. Based on a sigmoidal fit of adherent cells after insonification

with 3W for 1s, 50% of cells were removed at approximately 9.6 mm radially from the center,

corresponding to a predicted 0.26 W/cm2. Only the lowest power and energy settings available

for these experiments had the range of intensity to allow the visualization of this distinct radial

detachment profile. Correspondingly, this prospective threshold is exceeded at 14mm for 6W,

17mm for 9W and 22mm for 15 W. Since we were only able to measure out to 17mm from the

center of the 35mm dishes used here, the mean adhesion strength of cells was presumably

exceeded as far as the outer edge for all but the 3-watt series. Together, these results demonstrate

the ability to fine-tune a range of acoustic excitation values which spatially control cell

detachment while maintaining acceptable levels of cell viability. Further research into this

technique is required to fully understand detachment profiles generated by the acoustic

17
transducer and to deduce a direct correlation to existing well-characterized detachment force

models.

Pavg (W)
Acoustic Intensity = 4πr2 (cm2) (4)

Finally, cellular damage and detachment due to cavitation effects are common concerns

for sonication probes, such as the one used in this study. Cell damage caused by the shock of

cavitation requires the microbubble to be within micrometers of the cell [51]. Cavitation

microbubbles are initially formed at the tip of the probe[52]. Then the microbubbles will either

implode, rise due to buoyancy forces, or be forced downwards towards the cells by the generated

acoustic excitation or microstreams within the fluid [52]. In this study, the relatively large

distance between the probe tip and the cell surface (2mm) minimized the probability of a

microbubble coming within micrometers of a cell before imploding. It has been shown that as

exposure time increases, cavitation microbubbles travel further down into the fluid medium

before imploding; however, these times greatly exceed the length of exposure used within this

study [52]. Therefore, it is unlikely that cavitation is the primary mechanism of acoustic-based

cell detachment or destruction in this study. Instead, these results suggest that the excitation from

the ultrasonic transducer is responsible for the mechanical manipulation and its impacts on cell

adhesion distribution and viability observed here.

2.4 Conclusions

In summary, this study demonstrates that the power, energy, and intensity of an acoustic

pressure wave have profound impacts on both cell attachment and viability. At similar energy

values, higher power inputs corresponded to a higher percentage of cells detaching from the

surface in shorter exposure times. However, as power and energy increased, long-term viability

18
of fibroblasts was compromised, with some samples retaining less than 10% of the original cells

after 24 hours. This research also demonstrates that as the distance from the transducer increased,

causing the relative acoustic intensity to decrease, there was a corresponding decrease in

detachment of cells. Acoustic transducers generating pN scale forces have previously been used

to trap or move suspended cells and high-intensity acoustic transducers are commonly used for

cell destruction; these results indicate that the application of intermediate levels of excitation

from acoustic transducers may be useful in manipulating or measuring the adhesion strength of

stably adhered cells, or even in selectively patterning cell populations. Further investigation is

warranted to determine the mechanism by which insonification negatively impacts viability such

that manipulation can be fine-tuned, and stimulation protocols can be developed. Additionally,

native tissue consists of multiple different cell types; thus, cells other than fibroblasts should also

be investigated to determine how similar ranges of acoustic excitation affect different types of

cells.

19
 Chapter 3: Manipulation and Patterning of Cells using Acoustic Forces
3.1 Background

Vibration is defined as the oscillation of particles within an elastic body or fluid about an

equilibrium point. This natural mechanical phenomenon affects every physical system in the

universe as long as that system has mass and stiffness. The relationship between a system’s mass

and stiffness defines its natural frequency, or the frequency a system vibrates at in the absence of

outside forces. Resonance occurs when a system is driven at its natural frequency; producing

large amplitude oscillations which often cause catastrophic failure.

In 1787 German physicist Ernst Chladni published a technique which allowed for the

visualization of various modes of vibration on a rigid surface. This was accomplished by

drawing a bow over a flat metal plate until the plate reached resonance, effectively creating a

standing wave. Sand on the surface of the plate would move towards locations of zero amplitude,

known as nodal points, and away from points of maximum amplitude, known as antinodal

points.

Michael Faraday, fascinated by Chladni’s work, discovered what is now known as

Faraday instability back in 1831 [53]. Faraday instability is the terminology used to describe

nonlinear standing waves which appear in liquids enclosed within a vibrating container. Similar

to Chladni patterns, Faraday waves can take many forms depending on multiple factors

including, but not limited to, frequency and amplitude of oscillation, depth of the liquid, and

material properties of both the liquid and its container.

20
 Recently, scientists have begun taking advantage of Faraday waves to assemble

biological materials such as cells, cell spheroids, and cell-seeded microcarrier beads.

This central theme of this research focuses on the manipulation and patterning of cells in

a common tissue culture dish using vibrational and acoustic forces. Preliminary work has proven

the importance of computer modeling for accurate prediction of the parameters required to

achieve specific patterns. Insights gained from the previous section aids in the development of

these computer models and experimental data is used to verify accuracy. Successful cell

patterning is performed and both long term cell viability and pattern retention over time are

investigated; ensuring this new method of cellular manipulation is practical.

3.2 Computer Modeling and Experimental Validation

Physical systems theoretically have an infinite number of natural frequencies; however,

only a limited number of specific frequencies will generate a desired pattern. Thus,

experimentally identifying the parameters to generate specific patterns is impractical. This type

of problem can be resolved using computer modeling to predict potential useful frequencies

which are verified experimentally. Both SolidWorks and Ansys were used to find natural

frequencies of a 100mm polystyrene tissue culture dish and results from each were compared to

ensure accuracy.

Table 1 shows the first 10 modes and corresponding frequency in Hz for the same tissue

culture dish analyzed in SolidWorks and Ansys. Figure 6 illustrates results of SolidWorks

simulation results compared with images from sand inside a tissue culture dish being stimulated

at the given frequency. Initially the sand clung to the tissue culture dish resulting in uneven

pattern formation. This was remedied by rubbing the dish with a simple dryer sheet to remove

any static electricity, results shown in Figure 6.

21
 Figure 6 SolidWorks modal frequency simulation results compared to experimental modal frequencies.

Modeling natural frequencies and mode shapes in air is relatively simple; however, this

does not accurately represent the environment of a mammalian cell in culture. Mammalian cells

die rapidly in air and require medium in culture to survive. This medium is mostly water with

dissolved minerals, proteins, growth factors, and antibiotics. Knowing this, the medium must be

22
accounted for in order to generate an accurate model of the experiment. The liquid will add mass,

dampen resulting vibrations, and introduce streaming effects which were not a factor previously.

Adding mass to the system and dampening vibrations dramatically lowers the frequency these

mode shapes take place.

3.3 Methods and Experimental Setup

3.3.1 Cell Culture and Reagents

NIH3T3 murine embryonic fibroblasts (CRL-1658, ATCC) were cultured in complete

growth medium (CGM) containing 10% new born calf serum and 1% penicillin-streptomycin in

Dulbecco’s modified Eagle’s medium. Cells were passaged every three days using 0.25%

trypsin-EDTA and used between passages 5 and 20. Cell culture reagents were purchased from

Invitrogen unless otherwise noted.

CellTracker Green CMFDA was diluted to a working solution of 20 µM in DMEM

following the protocol from Thermo Fisher [54]. Culture media was removed from adhered

fibroblasts and the dish was rinsed with DPBS (++) before adding 2ml of the CMFDA solution.

Cells were then incubated at 75% relative humidity, 5% CO2 and 37°C for 45 minutes.

Calcein AM Green was diluted to a working solution of 2 µM in DMEM following the

protocol from Thermo Fisher [55]. Culture media was removed from adhered fibroblasts and the

dish was rinsed with DPBS (++) before adding 2ml of the calcein AM solution. Cells were then

incubated at 75% relative humidity, 5% CO2 and 37°C for 45 minutes.

Cell manipulation experiments were performed on NIH3T3 fibroblasts stained with

either CellTracker Green CMFDA, CellTracker Orange CMTMR, or Green Calcein AM to

visually verify pattern formation. CMFDA and CMTMR were chosen when imaging the cells

over multiple days as they remain fluorescent for over 3 days and transfer to daughter cells when

23
cell division occurs [54]. CMTMR was chosen to differentiate between cell patterns when

multiple pattern experiments were performed. Calcein AM was chosen when assessing cell

viability and when short term fluorescence was required as fluorescence from calcein AM is only

retained in live cells and quickly dissipates if the cell’s membrane is compromised [40, 41].

3.3.2 Cell Manipulation

A mechanical vibrator (Pasco Scientific SF-9324), controlled using a 12MHz DDS sweep

function generator (BK Precision model 4013DDS), was used to manipulate cells in suspension.

In order to manipulate cells using a mechanical vibrator, a sterile tissue culture dish was affixed

with a banana plug which fit into the vibrator. A guide was drawn up in SolidWorks and 3D

printed to ensure accurate placement of the banana plug in the center of each tissue culture dish.

Originally the banana plug was directly adhered to the dish using cyanoacrylate (super glue);

however, this posed an issue with automated fluorescent microscopy imaging as the banana plug

prevented the objective from getting within working distance of the dish. This was remedied by

sandwiching a small glass coverslip between the banana plug and the dish. After 3 hours in the

incubator, the glue between the banana plug and the glass cover slip weakened enough to be

easily removed by hand without damaging the tissue culture dish.

80% confluent dishes containing adhered fibroblasts were stained using one of the

protocols above, rinsed of any remaining staining solution using Dulbecco’s phosphate buffered

saline without calcium or magnesium (DPBS --), and incubated at 75% relative humidity, 5%

CO2 and 37°C with 2ml of 0.25% trypsin-EDTA for 5 minutes. 8ml of fresh CGM was then

added to the dish and gently pipetted 20 times to dislodge any remaining adhered cells. The

suspended cell solution was then moved to a 15ml conical tube and spun down in a centrifuge at

24
300 rcf for 5 minutes to form a pellet. The media was then aspirated off of the top and the pellet

was suspended in 5ml of fresh CGM to remove any remaining trypsin-EDTA.

The 5ml suspended cell solution was then added to a tissue culture dish with a banana

plug attached to its center and moved to a mechanical vibrator. Here the cells were exposed to a

vibration at a specified frequency and amplitude for 1 minute. After the minute was up the

mechanical vibrator was turned off and the dish was left undisturbed for 1 hour to allow for cell

adhesion [38]. After 1 hour the dish was carefully moved to the incubator at 75% relative

humidity, 5% CO2 and 37°C for 3 additional hours. Once 4 total hours were up the media was

gently aspirated off and 10ml of fresh CGM was added. Pattern formation and fidelity was then

analyzed using automated fluorescent microscopy.

3.3.3 Automated Fluorescent Microscopy Analyses

Samples were imaged and pattern fidelity analyses were performed using an Eclipse Ti-U

(Nikon Instruments) inverted microscope equipped with a CCD camera (CoolSNAP HQ2

Photometrics) and appropriate fluorescent filter sets. Automated fluorescence microscopy,

controlled by NIS-Elements software (version 4.20, Nikon Instruments), was used to capture and

stitch together 41x10 images for each sample on each day. This automation used a computer

controlled stage, in conjunction with the microscope setup described above, to capture

fluorescent images at 410 different locations, 10 columns of 41 images, traveling radially on

each dish for a total analysed area of 439 mm2 (8% of total dish area). The images were saved

and stitched together with a 5% overlap, resulting in a final image which was 12669 pixels wide

by 38480 pixels tall, or 10895.34 µm wide by 33092.8 µm tall. This final image covered an area

of 360.5mm2 (6.5% of total dish area). Fluorescent intensity data over the radial direction,

normalized to the background fluorescence, was recorded for each sample.

25
3.3.4 Macro Fluorescent Image Analyses

Pattern formation and fidelity was also confirmed using macro images of the dish.

Fluorescence was expressed using blue 470nm (Thorlabs LIU103) or green 525nm (Thorlabs

LIU525A) LEDs with appropriate glass filters 495nm LongPass (Thorlabs FGL495S) or 590nm

LongPass (Thorlabs FGL590S). A Nikon d3200 digital SLR equipped with magnification lenses

ranging from 1x-10x was used to capture images corresponding to t=0 (before excitation), 1

minute (after excitation), 4 hours, 24 hours, and 48 hours. These images were used to empirically

confirm initial pattern formation and retention over time.

3.4 Results and Discussion

3.4.1 Simulation Validation

Previous experimental work has shown sand particles can be patterned inside a common

tissue culture dish. This was accomplished by rigidly fixing the dish to a simple mechanical

shaker and controlling the amplitude and frequency of actuation with a function generator. Initial

modal analysis results of a normal 100 mm tissue culture dish was performed using SolidWorks

and Ansys to ensure accuracy of the model, briefly displayed in Table 3.4.1. These results match

up with the experimental results in air as seen in Figure 6.

The next step was to see if particle size impacted the particle’s movement to nodal

locations. Similarly, polystyrene microspheres were patterned at the same frequencies; again

showing excellent results when compared to the simulations; indicating that the size of the

particles being patterned does not play a large role in pattern formation in air.

Since mammalian cells must be suspended in liquid the obvious next step was to add

liquid to the simulation and verify via experimentation. However, addition of a liquid to the

26
simulation renders modal analysis moot since liquids do not have a Young’s Modulus. Future

computer modelling of the system must be performed using vibrational or acoustic forces and

solid-fluid interactions.

Table 1 SolidWorks vs Ansys modal simulation results for a 100mm polystyrene tissue culture dish

SolidWorks Ansys
Mode Frequency [Hz] Frequency [Hz]
1 720.05 727.52
2 720.12 727.75
3 736.1 740.29
4 736.14 740.72
5 950.43 959.43
6 950.44 959.55
7 1109.1 1114.5
8 1109.1 1114.8
9 1210.7 1221.1
10 1210.7 1221.4

Introducing a liquid medium makes the system more complex. Liquids, such as water or

CGM, introduce mass, damping, and fluid forces on particles becomes a factor. Additional mass

is easy to model and simply shifts the modal frequencies down such that the same modes appear

at lower frequencies. However, damping and fluid forces on particles must also be taken into

consideration. Liquids move and shift with the vibration; this happens in air as well but the

overall fluid dynamic effects on the particles are significantly lower than the fluid dynamic

effects of particles suspended in a liquid.

27
 A brief experiment, example results shown in Figure 7, was performed with 10µm

diameter polystyrene spheres suspended in 5ml of DI H20. The dish was vibrated at all of the

frequencies previously explored but no movement of the microspheres was recorded. The

frequency was then dramatically lowered in order to find the mode shapes experimentally. The

dish was swept through a range of frequencies and still no pattern formed. However, at certain

frequencies the flow of the water became turbulent which suggests the dish was being vibrated

near one of its modal frequencies. The amplitude of the vibration was then adjusted. Eventually

this led to the microspheres moving within the dish to the nodal locations and successful pattern

formation. This finding further suggests that fluid forces play a pivotal role when the particles

are suspended in liquid rather than in air.

a) b)
Figure 7 10µm polystyrene spheres in 5ml of DI H2O at 75Hz. a) full amplitude b) half amplitude

28
Figure 8 Macro image of green fluorescent NIH3T3 fibroblasts during vibrational patterning at 40Hz. Cells are
moving in streamlines from the inner nodes towards the outer nodes.

3.4.2 Cell Pattern Characterization using Macro Images

Suspended NIH3T3 fibroblasts were subjected to one of two frequencies, 40Hz or 75Hz,

for 1 minute. Each sample was run in triplicate to ensure the accuracy of the data collected.

Macro images were taken in order to empirically verify pattern formation. Figure 8 was captured

during vibrational patterning. The lines are made up of NIH3T3 fibroblasts fluorescing green.

a) b)
Figure 9 Macro image of NIH3T3 fibroblasts after patterning at 40Hz. a) immediately after vibrational force ends
cells no longer streamlining between nodes and starting to settle into nodal locations. b) A 1 hour after vibrational
force ends

29
This image shows that the cells are being forced towards nodal locations on the dish. The

zoomed in portion of the figure shows the cells seem to be moving in streamlines.

Figure 9a shows the same dish immediately after the vibrational force was concluded. A

majority of the cells have been forced to the nodes. Some cells are still suspended in the fluid

column and will float down to the surface of the dish. Figure 9b was taken 1 hour after the

vibrational force was concluded to allow precipitation of the suspended cells and adhesion to the

surface. This shows that a majority of the cells in the dish have been successfully patterned and 1

hour after stimulation the cells remain in the nodal locations.

Pattern reproducibility was also assessed using macro images. Figure 10 shows another

dish that was patterned at 40Hz. Comparing this with Figure 9, it is evident that a 40Hz vibration

forces the cells into a pattern with 6 defined bands and very few cells in between.

Figure 10 Macro image of NIH3T3 fibroblasts after patterning at 40Hz. This shows an almost completely
symmetric pattern formation over the majority of the dish

30
 Pattern symmetry was also assessed using macro images. While each pattern could be

reproduced reliably, not all of the patterns turned out symmetric. Figure 10 is one example with

around three-quarters of the dish was successfully patterned. Partial patterning was a consistent

problem during this study. The single point of contact between the tissue culture dish and the

mechanical vibrator was most likely the cause of these partial patterns. One point of contact

leaves multiple degrees of freedom where the dish could move. In some trials the vibration

caused the banana plug-dish assembly to lean to one side, creating a non-symmetric pattern. In

order to obtain a perfectly symmetric pattern, the dish would have to remain perfectly level both

during excitation and until the cells attached to the surface.

As previously stated, amplitude of vibration also became a factor once liquid was

introduced. Figure 11 shows fibroblasts, fluorescing green, during excitation, 40Hz, at an

Figure 11 Macro image of NIH3T3 fibroblasts at 40 Hz with increased amplitude.

Cells cannot be forced to nodal locations due to turbulent flow of the media

31
increased amplitude than what was found adequate for patterning. The increased amplitude

causes turbulence in the media which in turn overrides the vibrational force placed on the cell.

With this additional force, the cells are not able to settle at nodal locations.

3.4.3 Cell Pattern Characterization using Automated Fluorescent Microscopy

Cell pattern formation was characterized using automated fluorescent microscopy.

Representative images, Figure 12 and Figure 13 empirically demonstrate pattern formation and

retention over time. These images are composed of 10x41 individual fluorescent images stitched

together with a 5% overlap.

Figure 12 represents results taken from a pattern formed using a frequency of 40Hz while

Figure 13 represents results taken from a pattern formed using a frequency of 75Hz. During

excitation, suspended cells were forced into nodal locations and clustered together to form

defined bands of cells. The number of bands correlates to the frequency of excitation and can be

seen in these images; 5 bands were formed at 40Hz and 8 bands were formed at 75Hz. The

distance between each band appears to be constant, with the exception of the furthest two bands

created from excitation at 40Hz. This could be due to a reflection of the vibrational waves at the

wall of the tissue culture dish creating an additional node but further investigation is needed to

verify this hypothesis. The bands created from excitation at 75Hz all appear equidistant from

each other.

It should be noted that while a majority of cells were successfully patterned during

excitation, these images show that some cells attached outside of nodal locations. This was

expected for two reasons. First, cells were suspended in CGM during excitation. Forces caused

by the vibration of the culture dish vary throughout the height of the fluid column; therefore,

cells floating near the surface of the CGM will experience different forces than cells floating

32
 a) b) c)
Figure 12 Representative image of NIH3T3 Fibroblasts patterned at 40Hz. a) 4 hours after patterning. b) 24
hours after patterning. c) 48 hours after patterning

near the surface of the dish. This difference in force could cause some cells to remain suspended

over non nodal locations, eventually sinking down and attaching after excitation is concluded.

Second, cells were allowed to adhere to the surface for 1 hour after excitation, enough time to

develop around 40% of their normal adhesion strength [42]. The media was then aspirated off

and new CGM was added to the dish. It is possible that some of the attached cells were detached

and carried to a different location by the flow of new media.

Figure 12 and Figure 13 also show that the pattern formed during excitation is mostly

retained even after 48 hours under normal culture conditions. Parts b) and c) of each figure, 24

and 48 hours after patterning, display an increase in both pattern band width and number of cells

in between bands when compared to the initial pattern shown in a) of each figure. This is most

33
 a) b) c)
Figure 13 Representative image of NIH3T3 Fibroblasts patterned at 75Hz. a) 4 hours after patterning. b) 24
hours after patterning. c) 48 hours after patterning

likely caused by normal fibroblast migration and proliferation. Previous studies in wound healing

show that, under normal conditions, NIH3T3 fibroblasts which are in contact with one another

begin to migrate to locations void of fibroblasts within 24 hours. Therefore, the increase in

pattern band width over time can be partially contributed to the clusters of fibroblasts migrating

away from each other and into less dense areas outside of the original pattern. Additionally, parts

b) and c) of each figure appear to have a higher number of cells contained within the same

34
imaging area, with c) containing the largest number of cells. Normal, healthy NIH3T3 fibroblasts

have a doubling time of 20 hours so in both figures b) should have double the amount of cells

that were in a) and c) should have four times as many as a). While the effects on long term cell

proliferation was not explicitly investigated, data gathered from these images suggests that there

are few, if any, harmful effects. Overall, these figures show that while some migration does

occur and the cells seem to proliferate normally, the overall pattern is retained up to 48 hours

after initial patterning.

The length between each band of cells was calculated by measuring from the midpoint of

each band to the bottom of each image as shown in Figure 14. Patterns created were circular, so

measurements were taken at five different locations to ensure accuracy of the distance between

each band. The data is shown in Table 2-7; respective averages and standard deviations are

plotted on Figure 15.

Patterns created using a frequency of 40Hz displayed 5 bands of cells clustered together.

Distance between the first four bands on days 1, 2, and 3 averaged 4777 ± 56.1µm. The distance

between bands 4 and 5 averaged 5142.3 ± 55.6µm for the first two days and 4972.1 ± 34.2µm on

day 3. This discrepancy between the centers of cell bands 1-4 and 4-5 is not significant, therefore

it is still a question as to why the 5th band is seemingly much closer to the 4th than the other

bands.

Patterns created using a frequency of 75Hz displayed extremely consistent distances

between the center of each band of cells. Days 1 and 2 had an average distance of 3315.3 ±

224.5µm and 3316 ± 235.4µm respectively while Day 3 had an average distance of 3287.1 ±

164.5µm. While these distances are not significantly different, it does show the center of the

band of cells shifts slightly after 48 hours.

35
Figure 14 Representative image of distance measurements

36
Table 2 Longitudinal measurements of the distance between the center of the bands of cells for
40Hz on Day 1

Cell Cluster Distance between bands of cells (µm)

First to Second 4689.62 4597.12 4874.65 5090.08 4929.99
Second to Third 4875.36 4566.76 4659.3 4659.31 5091
Third to Fourth 4905 4874.16 4843.32 4843.31 4936.11
Fourth to Fifth 5275.79 5121.53 4967.27 4936.43 5214.11

Table 3 Longitudinal measurements of the distance between the center of the bands of cells for
40Hz on Day 2

Cell Cluster Distance between bands of cells (µm)

First to Second 4508.75 4601.3 4976.84 5066.12 5126.54
Second to Third 4350.28 4597.23 4442.45 4746.45 4871.79
Third to Fourth 4843.84 5121.39 4886 4832.51 4796.43
Fourth to Fifth 5275.85 5244.99 5141.13 5051.57 5194.78

Table 4 Longitudinal measurements of the distance between the center of the bands of cells for
40Hz on Day 3

Cell Cluster Distance between bands of cells (µm)

First to Second 5361.09 5392.11 4953.06 5240.84 4761.36
Second to Third 4308.28 4018.54 4254.33 3998.65 4637.77
Third to Fourth 4882.55 4978.28 4754.55 4850.54 4371.1
Fourth to Fifth 5074.55 5234.11 5010.39 4754.67 4786.71

Table 5 Longitudinal measurements of the distance between the center of the bands of cells for
75Hz on Day 1

Cell Cluster Distance between bands of cells (µm)

First to Second 3661.34 3040.8 2877.15 2808.96 3988.63
Second to Third 3061.36 3336.25 3461.96 3506.46 3145.05
Third to Fourth 3441.02 3306.84 3177.21 3155.44 3291.85
Fourth to Fifth 3297.52 3196.74 3302.8 3209.8 3427.98
Fifth to Sixth 3275.02 3264.8 3264.79 3306.63 3346.39
Sixth to Seventh 3281.84 3186.18 3349.29 3378.01 3399.49
Seventh to Eighth 3209.29 3347.58 3575.42 3405.69 3161.21

37
Table 6 Longitudinal measurements of the distance between the center of the bands of cells for
75Hz on Day 2

Cell Cluster Distance between bands of cells (µm)

First to Second 3851.42 3210 2913.98 3209.4 4055.31
Second to Third 3128.19 3251.97 3083.23 3421.08 2998.13
Third to Fourth 3434.76 3293.57 3378.2 2913.72 3589.17
Fourth to Fifth 3253.45 3251.64 3336.75 3463.13 3124.94
Fifth to Sixth 3276.62 3209.93 3294.67 3421.24 3462.98
Sixth to Seventh 3133.14 3209.25 3294.8 3294.3 3462.98
Seventh to Eighth 3225.57 3336.39 3418.99 3335.37 2956.2

Table 7 Longitudinal measurements of the distance between the center of the bands of cells for
75Hz on Day 3

Cell Cluster Distance between bands of cells in µm

First to Second 3547.97 3031.72 3294.18 3417.37 3209.6
Second to Third 3209.17 3008.04 3209.82 2913.68 3378.53
Third to Fourth 3378.5 3300.11 3378.25 3506.05 3716.44
Fourth to Fifth 3167.36 3325.09 3082.9 3250.44 3294.05
Fifth to Sixth 3378.87 3138.64 3336.28 3336.48 3251.82
Sixth to Seventh 3040.31 3220.59 3167.48 3251.59 3590.25
Seventh to Eighth 3294.16 3210.09 3420.73 3462.97 3462.38

6000
Average distance between centers

5000
of cell clusters (µm)

4000

3000 75 Hz
40Hz
2000

1000

0
4 hours 24 hours 48 hours
Figure 15 Average distances and standard deviations between bands of cells over time

38
 39
a)
b)
Figure 16 Representative fluorescent intensity vs location for 40Hz. a) Data before subtracting background fluorescence. b) Data after subtracting background
fluorescence.
 Fluorescent intensity over location was also analyzed, as shown in the representative

image Figure 16. High fluorescent intensity directly correlates to locations where the green

fluorescence stain, CMFDA, is expressed within the cells. Low fluorescent intensity correlates to

locations where there are no cells, only background fluorescence is captured. This data was

gathered in a 1 pixel wide measurement along the length of the stitched images. A total of five, 1

pixel wide measurements were taken for each sample on each day of analysis. The line

measurements analyzed were located at: (1) left side of the first column of images, (2) 3rd

column from the left, (3) 5th column from the left nearest to center of stitched image, (4) 3rd

column from the right, and (5) right side of the final column. It is important to note that, due to

the 1 pixel wide measurements, the data contains intensity spikes between the patterned bands of

cells. These spikes are most likely due to individual cells on the culture dish which were not

patterned properly and happened to be within the field analyzed. Figure 16a) represents the data

gathered directly from the fluorescent microscope. The graph’s baseline intensity is not zero due

to the background fluorescence picked up by the microscope. This is normal in fluorescent

microscopy and can be eliminated while preserving the integrity of the data. Figure 16b)

represents the data when background fluorescence is eliminated by subtracting the background

intensity from each measurement. Background intensity was calculated individually for each data

set by taking the average intensity in an area which contained no cells.

40
 40Hz 4 hours after patterning combined profiles
8000
7000
6000
Intensity

5000
4000
3000
2000
1000
0
0 5000 10000 15000 20000
Location µm

Figure 17 Combined fluorescent intensity data for 40Hz 4 hours after patterning

75Hz 4 hours after patterning combined profiles

6000
5000
Intensity

4000
3000
2000
1000
0
0 5000 10000 15000 20000 25000 30000
Location µm

Figure 18 Combined fluorescent intensity data for 75Hz 4 hours after patterning

The original intent of this analysis was to take the average of each intensity value and

plot it vs location. This would aid in gaining a better understanding of the location of patterned

bands of cells, cell migration over time, and cell proliferation over time. Unfortunately this type

of analysis did not prove fruitful. As seen Figure 17 Combined fluorescent intensity data for

40Hz 4 hours after patterning and Figure 18, when averaged the data becomes jumbled and does

not represent the overall pattern created; likely due to a combination of the curvature of the

pattern formed and intensity spikes between bands of cells due to individual cells outside of the

patterned bands but within measurement fields. Thus, the idea of averaging intensity measured

41
was dismissed. Instead, the five 1 pixel wide fluorescent intensity measurements were compared

intensity measurements over each day the samples were imaged.

In Figure 16, locations on the graph where the fluorescent intensity remains consistently

high directly correlate to the patterned bands of clustered cells. Locations on the graph where the

fluorescent intensity remains consistently low directly correlates to areas between the bands

where cells were not patterned. As briefly stated before, there are spikes of high intensity

measurements between the expected bands of cells, likely due to errant individual cells which

were not successfully patterned.

Figure 19 40Hz 4 hours after patterning. Location (1)

3.4.4 Cell pattern fluorescence data over length of the images

42
 Figure 21 40Hz 4 hours after patterning. Location (3)

The following figures contain fluorescent intensity data over the length of the stitched

images for cell patterns created using 40Hz and 75Hz. Data was normalized to remove

background fluorescence, similar to Figure 16b.

Figure 20 40Hz 4 hours after patterning. Location (2)

43
Figure 24 40Hz 24 hours after patterning. Location (1)

Figure 22 40Hz 4 hours after patterning. Location (4)

Figure 23 40Hz 4 hours after patterning. Location (5)

44
Figure 27 40Hz 24 hours after patterning. Location (4)

Figure 25 40Hz 24 hours after patterning. Location (2)

Figure 26 40Hz 24 hours after patterning. Location (3)

45
Figure 30 40Hz 48 hours after patterning. Location (2)

Figure 28 40Hz 24 hours after patterning. Location (5)

Figure 29 40Hz 48 hours after patterning. Location (1)

46
Figure 31 40Hz 48 hours after patterning. Location (3)

Figure 32 40Hz 48 hours after patterning. Location (4)

47
Figure 33 40Hz 48 hours after patterning. Location (5)

Figure 34 75Hz 4 hours after patterning. Location (1)

Figure 35 75Hz 4 hours after patterning. Location (2)

48
Figure 36 75Hz 4 hours after patterning. Location (3)

Figure 37 75Hz 4 hours after patterning. Location (4)

Figure 38 75Hz 4 hours after patterning. Location (5)

49
Figure 39 75Hz 24 hours after patterning. Location (1)

Figure 40 75Hz 24 hours after patterning. Location (2)

Figure 41 75Hz 24 hours after patterning. Location (3)

50
Figure 42 75Hz 24 hours after patterning. Location (4)

Figure 43 75Hz 24 hours after patterning. Location (5)

Figure 44 75Hz 48 hours after patterning. Location (1)

51
Figure 45 75Hz 48 hours after patterning. Location (2)

Figure 46 75Hz 48 hours after patterning. Location (3)

Figure 47 75Hz 48 hours after patterning. Location (4)

52
 Figure 48 75Hz 48 hours after patterning. Location (5)

Figure 19 through Figure 33 represent the fluorescent image data gathered for 40Hz

patterns using automated fluorescent microscopy. Figure 19 - Figure 23, 4 hours after patterning,

all display 5 main peaks in fluorescent intensity. The location of each peak varies which is due to

the curvature of the pattern formed. If the pattern created were truly symmetric and imaging was

done perfectly then location pairs, such as location 1 and 5 or 2 and 4, should have intensity

peaks at relatively similar locations. From these graphs, location 1 has peaks around 2000-

5000µm, 6000-8000µm, 11000-13000 µm, 16500-19000 µm, and 22000-23000µm. Location 5

has peaks between 1000-4000 µm, 6250-6750 µm, 11000-13000 µm, 16000-18500 µm, and

21000-23000 µm. The overall locations where intensity peaks is similar for location pair 1 and 5.

Location pair 2 and 4 share peak locations around 3000-4000 µm, 7500-9000 µm, 12000-13000

µm, 17000-18000 µm, and 22000-23000 µm. As expected, these results become harder to

interpret as time passes. Over time the cells migrate and proliferate, leading to a blending of

intensity peaks and making it more difficult to accurately determine where intensity peaks begin

and end, as seen in 48 hours after patterning, Figure 29-Figure 33.

Interestingly, intensity profiles for 75Hz, all have distinct peaks with little noise. That

being said, it still follows the trend of intensity peaks becoming wider as cells have time to

53
migrate and proliferate. This could be due to more accurate matching of the desired excitation

amplitude at 75Hz than the excitation amplitude used at 40Hz. Neither input or output amplitude

were directly measured for this study, future work should remedy this by measuring the force

output at each frequency.

Initial band width differed in 40Hz and 75Hz, specifically 75Hz displayed a thinner

initial band width than 40Hz. This can be visualized when comparing the representative Figure

12 and Figure 13 can be analyzed using the intensity profiles for each frequency. This coincides

with one of the basics of vibrations: if frequency increases, wavelength must decrease.

3.5 Conclusion and Future Work

In conclusion, this work demonstrates the feasibility of using vibrational forces to

manipulate and pattern large numbers of cells at once. The patterns discussed, corresponding to

the frequencies of 40Hz and 75Hz, here have no deleterious effects on NIH3T3 fibroblast

viability, proliferation, or migration. Once cells are patterned in suspension and allowed to

adhere to the surface of the tissue culture dish, the cells overall remain in that pattern for up to 48

hours. This work is the tip of the iceberg, it demonstrates two of the multitude of possible

patterns and future work is highly encouraged. Further research into cellular patterning can be

expanded by adjusting different factors such as:

 Frequency and/or amplitude of the vibrational force

 Changing the application point of the vibrational force to different locations on

the dish

 Adding multiple contact points of the vibrational force

 Forcing nodal locations by damping specific points om the dish

 Size, shape, and/or density of the tissue culture dish

54
 Overall, the experimental data gathered and presented within this section appears to fit a

sinusoidal waveform; an expected result due to the sinusoidal vibrational force used during

cellular patterning. While not within the scope of this study, a natural extension would be to

analyze the data further and develop a mathematical model which matches the experimental

results. This type of analysis could benefit an expansion of the computer modeling performed

here.

Modeling and simulations performed for this study were used to predict the frequencies

of mode shapes in air. While the model was not developed to predict modal frequencies with the

addition of liquid, it could predict the types of mode shapes expected from experimental results.

In order to further the development of cellular patterning and manipulation using vibrational

forces, this model requires further refinement. Specifically, the model would greatly benefit from

the addition of fluid-solid interactions, particle tracking, and a tunable sinusoidal force applied to

the bottom of the tissue culture dish.

Further analysis of the cellular patterns developed in this work should be performed.

Band width, the thickness of each band of cells that is produced, should be further characterized

and specific boundaries of where a band begins and ends could be defined. This would assist in

determining the relationship between band width and excitation frequency. Additionally,

defining boundaries of the produced cell bands would aid in analyzing band-band encroachment

due to cell migration and/or proliferation over time.

55
 Chapter 4: Key Findings and Future Work
The key findings of this dissertation are as follows

 Pressure waves generated by acoustic devices have an intermediate range which is

acceptable for manipulating and patterning adherent cells.

 Power, energy, and intensity of acoustic pressure waves all have impacts on cell

attachment and viability.

 An increase in power to an acoustic device will detach more cells from the surface

in shorter exposure times at the cost of long term cell viability.

 Attached mammalian cells which are detached from a surface by sonication do

not necessarily lyse or go through apoptosis.

 Simple vibrations can be utilized to pattern microscopic particles, including living

mammalian cells in suspension.

 The addition of liquid to a vibrating dish causes pattern formation to become

dependent on both frequency and amplitude of the vibration.

 Fibroblasts congregate at nodal locations when exposed to appropriate vibrational

and acoustic forces.

 Once placed in a particular pattern while in suspension, adherent cells retain the

patterned formation for up to 48 hours.

 Forces involved with the patterning described were low enough to not cause

permanent damage to the cell as there was no significant cell death or loss of

population.

56
  Cell population continued to grow, migrate, and proliferate once patterned.

4.1 Incomplete and Ongoing Work: Characterization of Cellular Mechanical Properties

using AFM

4.1.1 Background

Mechanical properties of cells are extremely important to many fields including tissue

engineering, disease detection, and acoustic manipulation. Tissue engineers strive to develop

new ways to repair or replace damaged tissues. To do this, the mechanical properties of

individual cells and complete tissues must be known. Similarly, diseases, such as cancer, affect

the mechanical properties of native cells [18, 56]. Studying these changes can lead to the

development of early detection devices and new drug delivery mechanisms.

Currently there are multiple techniques to determine the physical characteristics of

individual cells. Micropipette aspiration, first published by MITCHISON and SWANN [57],

uses suction pressure to suck part of a cell into a micropipette. Mechanical properties are

ascertained by observing the change in shape of the cell as a result of the suction pressure.

Furthermore, this technique has been used to characterize membrane stiffness of both red blood

cells [58] and leukocytes [59]. Optical traps, more commonly known as optical tweezers or laser

tweezers, were developed in 1970 [60]. This technique uses a laser beam to exert small attractive

or repulsive forces on dielectric particles. Since its development, optical tweezers have been used

to trap viruses and bacteria [61], test optical deformability of red blood cells [62], and even

observe protein folding in response to mechanical force [63]. Magnetic twisting cytometry

evaluates cytoskeletal stiffness by measuring the rotation of cell membrane bound magnetic

microbeads [64]. Atomic Force Microscopy (AFM), first developed by Binnig, et al. [65], has

proven to be an extremely useful tool in the study of cellular mechanics. While AFM is a

57
relatively new technology, many scientists worldwide are utilizing its precise measurement

capabilities to study processes never before possible.

Researchers have also used AFM to develop specific techniques in order to investigate

specific cellular properties. One of these new techniques, dubbed single cell force spectroscopy

(SCFS) [66, 67] utilizes AFM’s precise measurement capabilities to investigate the adhesive

force of cells between various surfaces and even other cells [66-68]. This is done by attaching a

cell to the end of a tip-less AFM cantilever and using it as a living biological probe. However

conventional SCFS does have its drawbacks. Because of the precise measurement capabilities,

AFM is unable to generate large forces such as those required to remove cells after long term

adhesion has taken place, restricting this technique to adhesion times within a few minutes.

4.1.1.1 Cell Mechanical Properties: Elasticity and Natural Frequency

Utilizing AFM, researchers have begun to investigate cellular elasticity [11-14].

However, these studies probed cells with sharp tipped cantilevers and reported varying stiffness

values at different locations of the cell [14]. Data reported by Mathur, et al. [14] demonstrates

that measurements made over the nucleus are significantly stiffer than those made over the cell

body. The inconsistencies of reported stiffness values along with indentation ranges of only a

few hundreds of nanometers give poor insight into the overall stiffness of the cell.

Natural frequency of an object can be defined as the frequency, or sets of frequencies, an

object vibrates at when disturbed. This basic physical principal applies to all structures including

biological organisms. Excitation of an object at one of its natural frequencies causes that object

to accumulate energy; if the energy is not dissipated the object’s structure can become unstable

and permanent damage can ensue. While this could be advantageous for targeted disease

treatments, it could also be disastrous for tissue construction. Multiple studies have used models

58
to predict cell natural frequencies [69-73] but there is insufficient research into the experimental

verification of the natural frequencies of cells.

This research proposes a rethinking of cellular stiffness into both global, over the entire

cell, and local, at a specified location, stiffness. Under this definition previous research using

tipped AFM cantilevers would fall under local stiffness measurements. The work proposed here

would fall under global stiffness which could depend on multiple factors including cytoskeletal

composition and cell shape. Certainly if global cell stiffness is an amalgamation of its internal

components and overall shape, it stands within reason that a cell’s natural frequency would also

depend on these factors.

4.1.1.2 Cytoskeletal Composition and Cell Shape

The cytoskeleton is vital to eukaryotic cells and assists in multiple cellular activities

including cell division, intracellular nutrient transport, and even helps regulate cell shape and

mechanical properties[42]. It does this through the dynamic disassembly and reassembly of three

main protein structures known as microfilaments, intermediate filaments and microtubules.

Microfilaments, also known as actin filaments, generate contractile forces within a cell through a

myosin motor moving along actomyosin fibers. Actin filaments can also create protrusions

against the cell membrane; these protrusions have been reported as being stiffer than surrounding

locations most likely due to the high concentration of actomyosin bundles in the area [13].

Intermediate filaments are mainly used within the cell to anchor organelles within the cell’s

structure and to structurally support the nucleus. Finally, microtubules, the largest of the three

cytoskeletal structures, assist the cell in counterbalancing actin filaments by resisting

compression.

59
 As previously discussed, studies of cellular stiffness utilizing AFM have made small

indentions into the cell [11-14]. These local measurements likely correspond to the cell

membrane and the cytoskeletal configuration immediately below. These reports, combined with

the current gap in literature on cellular global stiffness, fail to investigate how the major

cytoskeletal elements contribute to the cell’s overall properties.

Along with cytoskeletal composition, cell shape plays a major role in regulating the cell’s

mechanical properties [11, 42]. Mammalian cells naturally take different shapes depending on

their function. For example epithelial cells will restrict their shape by tightly packing together to

form a protective barrier while fibroblast cells generally spread out and relocate to produce

extracellular matrix components where needed. Analogous to cytoskeletal composition, current

literature does not delve into how a cell’s shape contributes to its global properties.

4.1.1.3 Objectives

The goal of this research was to cover three main objectives. The first objective was to

experimentally measure individual cell stiffness using both a technique developed in Dr.

Hasegawa’s Micro-Nano Control and Biorobotics Laboratory [74, 75] and a new technique

similar to SCFS using AFM. Stiffness values obtained through the nano-indentation technique

would have been used to verify the validity of measurements made using AFM. Once verified,

the AFM will also be used to investigate a cell’s natural frequency, a mechanical property which

has previously not been experimentally investigated.

The second objective was to use the developed measurement technique to further

characterize a cell’s stiffness and natural frequency. Currently it is unknown how a cell’s shape

or internal structural composition affect its global stiffness. To test these factors, individual cells

would either be exposed to one of four biochemical factors which influence specific properties of

60
the cytoskeleton: serum-free media, Y-27632, nocodazole, or sodium azide, or be confined to a

specified adhesion area developed through microcontact printing [42, 76, 77].

The final objective of this research was to characterize a second cell type’s stiffness and

natural frequencies using the developed AFM technique. It is important to ensure the developed

technique is applicable to multiple cell types as the mechanical properties of cells differ

corresponding to the role and shape of the cell. Additionally, this second cell type’s stiffness and

natural frequency were going to be examined after either exposure to the biochemical factors

listed above or confined via controlled protein binding sites.

4.1.2 Methods and Materials

4.1.2.1 Cell Culture and Reagents

NIH3T3 murine embryonic fibroblasts (CRL-1658, ATCC) cultured in complete growth

medium (CGM) containing 10% new born calf serum and 1% penicillin-streptomycin in

Dulbecco’s modified Eagle’s medium. Cells will be passaged every three days using 0.25%

trypsin-EDTA and used between passages 5 and 20.

Cell fixation was performed by adding 500uL of 4% paraformaldehyde in PBS to the

growth media for 2 minutes, preventing cell damage from abrupt change in osmolarity. The

media was then aspirated and rinsed with PBS before covering the cells in a 4%

paraformaldehyde in PBS solution and incubated for 15 minutes at room temperature. A final

aspiration and rinse in PBS was performed before experimentation.

61
 4.1.2.2 Fabrication of Microcontact Printing Stamp

First a glass slide was cleaned and sterilized by sonicating in acetone, methanol,

isopropanol, and DI H2O for 10 minutes each followed by rinsing in flowing DI H2O for 30

seconds and blown dry using compressed N2. The slide was then baked on a hot plate at 250°C

for 30 minutes to remove any excess water from the surface. Following this, the slide was placed

under a glass container on a hot plate at 150°C. 3ml of HMDS was evaporated under the glass

container for 15 minutes; forming a thin hydrophobic layer on the slide which improves

photoresist adhesion. AZ 5214E was spun onto the slide at 2000 rpm to achieve a thickness of

2µm and soft baked at 105°C for 50 seconds before exposure. A table-top micro pattern

generator (Heidelberg instruments µPG 101) was used to expose the desired pattern directly.

This micro pattern generator acts similar to a printer. The UV light is mounted on a movable

head which exposes the photoresist based on a specified pattern; this is advantageous because

multiple patterns can be fabricated without the requirement of a mask. The photoresist was then

submerged in NMD 3 developer for 50 seconds under gentle agitation to etch away the exposed

resist. Etching was neutralized by submerging in DI H2O for 30 seconds and blown dry. Pattern

Figure 49 Fabrication of microcontact printing stamp and stamping procedure.

62
 Figure 50 Surfcorder ET200 used to verify etch depth of photoresist

fidelity was assessed via optical microscopy and surface measurements were taken using a

Surfcorder ET200. PDMS was mixed at a 10:1 ratio, poured over the photoresist template to

achieve a thickness of 5mm, degassed under vacuum, and cured at 65°C for 2 hours. Once curing

was complete and the PDMS cooled down to room temperature it was carefully peeled off of the

photoresist template and trimmed appropriately.

4.1.2.3 Microcontact Printing Procedure

The microcontact printing procedure took place within a biosafety cabinet and all

materials were sterilized prior to starting. The PDMS stamp was sterilized via sonication in

ethanol for 10 minutes, followed by second sonication in DI H2O for 10 minutes, rinsed in

flowing DI H2O for 30 seconds, blown dry with N2, and finally exposed to UV light for 15

minutes. Once sterilization was complete, one drop of 50µg/ml FN in PBS was placed onto the

stamp and gently spread over the surface. The protein coated stamp was covered for 30 minutes,

allowing the FN to adsorb from the solution onto the surface of the PDMS stamp. After, the

63
remaining FN solution was aspirated off, the stamp was immediately triple rinsed with PBS, and

immediately dried using compressed N2. Using tweezers, the stamp was inverted and the protein

covered microstructured surface was placed in contact with the desired substrate. Gentle pressure

was applied with the tweezers to the back of the stamp to ensure conformal contact between the

stamp and the surface. The stamp was left in contact with the surface for 1 minute to ensure FN

transfer before carefully removing, making sure not to cause roof collapse upon removal. Once

the stamp was removed it was be submerged in H2O to keep the microstructures intact. The

patterned substrate, in this case a clean and sterilized glass slide, was backfilled with either 0.2%

(w/v) Pluronic F-127 in PBS for 1 hour or 1% heat denatured BSA in PBS for 1 hour.

Backfilling is done to block non-specific binding sites on the surface of the glass, ensuring

pattern fidelity. The surface was then rinsed with PBS, cells seeded onto the surface at 10,000

cells/cm2, and incubated for 2 hours. After 2 hours the excess unattached cells were removed

using a gentle flow of fresh media from one side of the dish and aspirated off from the opposite

side.

64
 Figure 51 Nano-robotic manipulation system inside ESEM

4.1.2.4 Experimental Procedure

Initial characterization of the tip-less AFM cantilevers was performed using the nano-

robotic manipulation system, shown in Figure 51, developed in the Micro-Nano Control and Bio-

Robotics Lab [74, 75]. This process was performed in real time using an environmental scanning

electron microscope (ESEM) and a reference cantilever of known stiffness. The two cantilevers

b) c)
c

a)
a d) e)
e
Figure 52 ESEM images of AFM cantilever calibration. a) reference cantilever in bending with tipless
cantilever. b) reference cantilever making contact c) beginning of bending d) continuing bending e) final point
for measurement

65
were brought into initial contact Figure 52b and carefully pressed against each other until elastic

deformation occurs Figure 52e. In this configuration the cantilevers are in equilibrium and can be

used to calculate the unknown cantilever’s stiffness.

Characterization of the tip-less AFM cantilevers was also performed using AFM. While

ESEM setup can only be used to measure stiffness, AFM can easily derive both cantilever

stiffness and natural frequency. Cantilever natural frequency was found from a fast, simple, and

convenient technique commonly used in AFM called a frequency sweep. A frequency sweep is

when the AFM cantilever is excited through a range of frequencies; cantilever natural frequency

was easily identified as the frequency with the greatest amplitude. Previous work has shown that

the stiffness of the cantilever can be calculated from thermal data generated by thermal tuning

[78, 79]. This method of calculating beam stiffness has multiple advantages over other methods

as demonstrated by Burnham, et al. [78]. Recent work has increased the accuracy of these beam

stiffness measurements to within 3-4% for beams with spring constants up to 50 N/m [80].

After cantilever characterization, limited cell measurement experiments were performed

on the AFM. Cells initially were patterned using the microcontact printing stamps to control cell

contact area and fixed for ease of use. Previously in Dr. Hasegawa’s lab this nano-robotic

manipulation system inside ESEM has been used to study yeast cell stiffness [81]. A similar

procedure was attempted to measure fibroblasts cell stiffness. Mammalian cells cannot survive

for very long in the harsh environments created inside the ESEM. To circumvent this issue, cell

elasticity measurements were performed on dead cells which were fixed in a 4%

paraformaldehyde PBS solution. The fixation process likely increases cellular stiffness due to the

paraformaldehyde crosslinking proteins together and were noted when beginning AFM stiffness

measurements\

66
4.2 Limited Results

Results from this research are limited which is why this is ongoing and future work.

4.2.1 Microcontact Printing Direct Proteins

Fabrication of stamps for microcontact printing was integral in controlling cell adhesion

sites. Multiple different molds were successfully fabricated, one of which is shown in Figure 50.

Although roof collapse was an ongoing issue, eventually a successful mold and stamp was

created. Figure 53 shows successful direct fibronectin transfer from the PDMS stamp to a clean

glass slide. After the protein was printed onto a surface it was blocked with Pluronic F127 to

Figure 53 Green fluorescent fibronectin stamped onto a clean glass slide. a) 6µm b) 10µm

Figure 54 Successfull cell attachment to fibronectin islands

67
prevent non-specific protein adhesion. Figure 54 displays an example of cells adhering to the

fibronectin islands.

4.2.2 ESEM Analysis of Mammalian Cells

Normally spread NIH3T3 Fibroblasts were fixed and prepared for imaging using ESEM.

After fixation cells were triple rinsed with DIH2O to ensure no salt crystals would form in the

vacuum environment of the ESEM. Figure 55 shows the results of ESEM imaging on a fixed

fibroblast as the pressure decreases. In part a) the cell is barely visible at a pressure of 700 Pa.

Decreasing the pressure greatly increases image quality of the fibroblast until Figure 55e) when

the pressure reaches 210 Pa and the cell becomes damaged. Based on these limited results it is

possible to image mammalian cells using ESEM if the pressure is kept above 350 Pa with good

results achievable at 500 Pa.

Using these results, the next step was to attempt and deform a fibroblast using a tipless

AFM cantilever. Figure 56 shows one attempt at bringing a cantilever in contact with a cell. Due

a) b) c)

d) e) f)

Figure 55 NIH3T3 fibroblast imaged using ESEM. a) 799 Pa b) 651 Pa c) 500Pa d) 351 Pa e) 210 Pa f)
160 Pa

68
 Figure 56 ESEM cell elasticity using tipless AFM cantilevers. a) approaching contact b) in contact

to either the fixation process, dehydration from the ESEM atmosphere, or a combination of the

two, the cells would either refuse to deform or pop like a bubble immediately. Unfortunately

there were not enough trials to gather reliable data from this research.

4.3 Conclusion and Future Work

In conclusion, even though this research was not finished it was still a valuable learning

experience. This project granted hands on experience with AFM, ESEM, and microcontact

printing, important tools in modern research.

4.3.1 Future Work: AFM and Cytoskeletal Inhibitors

Future work will continue investigating the objectives and experiments outlined above.

Specific interests lie in atomic force microscopy measurements of cells under four different

factors, serum free media, Y-27632, nocodazole, and sodium azide. These four different factors

have been shown to affect the adhesion force of cells; hypothesizing that if adhesion force is

modified, elasticity of the cell should change as well. The first factor, serum-free media, chosen

because serum contains growth factors required by the cell to assemble internal mechanisms

which control adhesion [38, 77]. Studying cells in serum free conditions will show if these

69
growth factors affect cell stiffness as well. Additionally, treating cells with serum-free media is

simple and does not require additional materials to be purchased. The second factor, Y-27632, is

an actin-myosin contractility inhibitor. Specifically, Y-27632 inhibits the formation of

cytoskeletal fibers which regulate cell shape and propulsive force [77, 82]. This factor has been

shown to reduce adhesion strength [77] and inhibit microtissue release from swelling

microbeams [83]; both of which could correspond to a decrease in stiffness. The third factor,

nocodazole, depolymerizes microtubules within the cytoskeleton. Microtubules are thought to

balance contractile elements within the cell and have been shown to regulate cell adhesion

strength via cell shape [42]. If microtubules balance contractile elements and provide resistance

to compression, then exposing the cell to nocodazole should increase the cell’s compressibility,

resulting in a decrease in cellular stiffness. The final factor, sodium azide, is a potent ATP

production inhibitor [84]. Specifically, sodium azide inhibits cytochrome C oxidase in

mitochondria to block ATP production. ATP is known as the energy unit of the cell and is used

by many cellular processes including the formation of structural proteins. Without ATP

production many of the cell’s structural components will cease to function, leading to the

hypothesis that cellular stiffness will decrease when exposed to this factor.

70
 References
[1] J. Rouwkema, N. C. Rivron, and C. A. van Blitterswijk, "Vascularization in tissue engineering,"
Trends in Biotechnology, vol. 26, pp. 434-441, 2008/08/01/ 2008.
[2] R. K. Jain, P. Au, J. Tam, D. G. Duda, and D. Fukumura, "Engineering vascularized tissue," Nature
Biotechnology, vol. 23, pp. 821-823, 2005/07/01 2005.
[3] L. Gui and L. E. Niklason, "Vascular Tissue Engineering: Building Perfusable Vasculature for
Implantation," Current opinion in chemical engineering, vol. 3, pp. 68-74, 2014.
[4] S. L. M. Dahl, C. Rhim, Y. C. Song, and L. E. Niklason, "Mechanical properties and compositions of
tissue engineered and native arteries," Annals of biomedical engineering, vol. 35, pp. 348-355,
2007.
[5] T. N. McAllister, M. Maruszewski, S. A. Garrido, W. Wystrychowski, N. Dusserre, A. Marini, et al.,
"Effectiveness of haemodialysis access with an autologous tissue-engineered vascular graft: a
multicentre cohort study," Lancet, vol. 373, pp. 1440-6, Apr 25 2009.
[6] M. Evander, L. Johansson, T. Lilliehorn, J. Piskur, M. Lindvall, S. Johansson, et al., "Noninvasive
Acoustic Cell Trapping in a Microfluidic Perfusion System for Online Bioassays," Analytical
Chemistry, vol. 79, pp. 2984-2991, 2007/04/01 2007.
[7] X. Ding, S.-C. S. Lin, B. Kiraly, H. Yue, S. Li, I.-K. Chiang, et al., "On-chip manipulation of single
microparticles, cells, and organisms using surface acoustic waves," Proceedings of the National
Academy of Sciences, vol. 109, pp. 11105-11109, July 10, 2012 2012.
[8] A. Haake, A. Neild, D.-H. Kim, J.-E. Ihm, Y. Sun, J. Dual, et al., "Manipulation of cells using an
ultrasonic pressure field," Ultrasound in Medicine & Biology, vol. 31, pp. 857-864, 6// 2005.
[9] M. C. Jo and R. Guldiken, "Particle manipulation by phase-shifting of surface acoustic waves,"
Sensors and Actuators A: Physical, vol. 207, pp. 39-42, 3/1/ 2014.
[10] K. Yosioka and Y. Kawasima, "Acoustic radiation pressure on a compressible sphere," Acta
Acustica united with Acustica, vol. 5, pp. 167-173, // 1955.
[11] S.-Y. Tee, J. Fu, Christopher S. Chen, and Paul A. Janmey, "Cell Shape and Substrate Rigidity Both
Regulate Cell Stiffness," Biophysical Journal, vol. 100, pp. L25-L27, 3/2/ 2011.
[12] T. G. Kuznetsova, M. N. Starodubtseva, N. I. Yegorenkov, S. A. Chizhik, and R. I. Zhdanov, "Atomic
force microscopy probing of cell elasticity," Micron, vol. 38, pp. 824-833, 12// 2007.
[13] K. D. Costa, A. J. Sim, and F. C. P. Yin, "Non-Hertzian Approach to Analyzing Mechanical
Properties of Endothelial Cells Probed by Atomic Force Microscopy," Journal of Biomechanical
Engineering, vol. 128, pp. 176-184, 2005.
[14] A. B. Mathur, G. A. Truskey, and W. M. Reichert, "Atomic force and total internal reflection
fluorescence microscopy for the study of force transmission in endothelial cells," Biophysical
Journal, vol. 78, pp. 1725-1735, 2000.
[15] P. Gascoyne, J. Satayavivad, and M. Ruchirawat, "Microfluidic approaches to malaria detection,"
Acta Trop, vol. 89, pp. 357-69, Feb 2004.
[16] A. Garcia and N. Gallant, "Stick and grip," Cell Biochemistry and Biophysics, vol. 39, pp. 61-73,
2003/08/01 2003.
[17] S. Moon, U. A. Gurkan, J. Blander, W. W. Fawzi, S. Aboud, F. Mugusi, et al., "Enumeration of
CD4+ T-cells using a portable microchip count platform in Tanzanian HIV-infected patients," PLoS
One, vol. 6, p. e21409, 2011.

71
[18] H. Mohamed, M. Murray, J. N. Turner, and M. Caggana, "Isolation of tumor cells using size and
deformation," Journal of Chromatography A, vol. 1216, pp. 8289-8295, 11/20/ 2009.
[19] E. E. Hui and S. N. Bhatia, "Micromechanical control of cell–cell interactions," Proceedings of the
National Academy of Sciences, vol. 104, pp. 5722-5726, April 3, 2007 2007.
[20] M. C. Jo and R. Guldiken, "Active density-based separation using standing surface acoustic
waves," Sensors and Actuators A: Physical, vol. 187, pp. 22-28, 11// 2012.
[21] M. C. Jo and R. Guldiken, "Dual surface acoustic wave-based active mixing in a microfluidic
channel," Sensors and Actuators A: Physical, vol. 196, pp. 1-7, 7/1/ 2013.
[22] F. Guo, P. Li, J. B. French, Z. M. Mao, H. Zhao, S. X. Li, et al., "Controlling cell-cell interactions
using surface acoustic waves," Proceedings of the National Academy of Sciences of the United
States of America, vol. 112, pp. 43-48, Jan 6 2015.
[23] C. Poon, D. McMahon, and K. Hynynen, "Noninvasive and targeted delivery of therapeutics to
the brain using focused ultrasound," Neuropharmacology, vol. 120, pp. 20-37, 2017/07/01/
2017.
[24] Y.-J. R. Chang, J. Perry, and K. Cross, "Low-Frequency Ultrasound Debridement in Chronic Wound
Healing: A Systematic Review of Current Evidence," Plastic surgery (Oakville, Ont.), vol. 25, pp.
21-26, 2017.
[25] T. van Rooij, I. Skachkov, I. Beekers, K. R. Lattwein, J. D. Voorneveld, T. J. A. Kokhuis, et al.,
"Viability of endothelial cells after ultrasound-mediated sonoporation: Influence of targeting,
oscillation, and displacement of microbubbles," Journal of Controlled Release, vol. 238, pp. 197-
211, 2016/09/28/ 2016.
[26] D. Bani, A. Quattrini Li, G. Freschi, and G. L. Russo, "Histological and Ultrastructural Effects of
Ultrasound-induced Cavitation on Human Skin Adipose Tissue," Plastic and reconstructive
surgery. Global open, vol. 1, pp. e41-e41, 2013.
[27] D. L. Miller, N. B. Smith, M. R. Bailey, G. J. Czarnota, K. Hynynen, I. R. S. Makin, et al., "Overview
of therapeutic ultrasound applications and safety considerations," Journal of ultrasound in
medicine : official journal of the American Institute of Ultrasound in Medicine, vol. 31, pp. 623-
634, 2012.
[28] J. Voigt, M. Wendelken, V. Driver, and O. M. Alvarez, "Low-frequency ultrasound (20-40 kHz) as
an adjunctive therapy for chronic wound healing: a systematic review of the literature and
meta-analysis of eight randomized controlled trials," Int J Low Extrem Wounds, vol. 10, pp. 190-
9, Dec 2011.
[29] O. Doblhoff-Dier, T. Gaida, H. Katinger, W. Burger, M. Groschl, and E. Benes, "A novel ultrasonic
resonance field device for the retention of animal cells," Biotechnol Prog, vol. 10, pp. 428-32,
Jul-Aug 1994.
[30] S. Radel, A. J. McLoughlin, L. Gherardini, O. Doblhoff-Dier, and E. Benes, "Viability of yeast cells
in well controlled propagating and standing ultrasonic plane waves," Ultrasonics, vol. 38, pp.
633-637, 3// 2000.
[31] J. Hultström, O. Manneberg, K. Dopf, H. M. Hertz, H. Brismar, and M. Wiklund, "Proliferation and
viability of adherent cells manipulated by standing-wave ultrasound in a microfluidic chip,"
Ultrasound in Medicine & Biology, vol. 33, pp. 145-151, 1// 2007.
[32] F. W. Kremkau, "Cancer therapy with ultrasound: A historical review," Journal of Clinical
Ultrasound, vol. 7, pp. 287-300, 1979.
[33] J. A. Newell, "Ultrasonics in Medicine," Physics in Medicine and Biology, vol. 8, pp. 241-264,
1963/09/01 1963.
[34] M. Dyson, B. Woodward, and J. B. Pond, "Flow of Red Blood Cells stopped by Ultrasound,"
Nature, vol. 232, pp. 572-573, 08/20/print 1971.

72
[35] M. Dyson, J. B. Pond, B. Woodward, and J. Broadbent, "The production of blood cell stasis and
endothelial damage in the blood vessels of chick embryos treated with ultrasound in a
stationary wave field," Ultrasound in Medicine & Biology, vol. 1, pp. 133-148, 3// 1974.
[36] R. Guldiken, M. C. Jo, N. D. Gallant, U. Demirci, and J. Zhe, "Sheathless Size-Based Acoustic
Particle Separation," Sensors, vol. 12, pp. 905-922, 2012.
[37] F. Petersson, A. Nilsson, C. Holm, H. Jonsson, and T. Laurell, "Separation of lipids from blood
utilizing ultrasonic standing waves in microfluidic channels," Analyst, vol. 129, pp. 938-943,
2004.
[38] N. D. Gallant, K. E. Michael, and A. J. García, "Cell Adhesion Strengthening: Contributions of
Adhesive Area, Integrin Binding, and Focal Adhesion Assembly," Molecular Biology of the Cell,
vol. 16, pp. 4329-4340, September 1, 2005 2005.
[39] N. D. Gallant and A. J. García, "Model of integrin-mediated cell adhesion strengthening," Journal
of Biomechanics, vol. 40, pp. 1301-1309, // 2007.
[40] J. C. Bischof, J. Padanilam, W. H. Holmes, R. M. Ezzell, R. C. Lee, R. G. Tompkins, et al., "Dynamics
of cell membrane permeability changes at supraphysiological temperatures," Biophys J, vol. 68,
pp. 2608-14, Jun 1995.
[41] M. M. Roden, K.-H. Lee, M. C. Panelli, and F. M. Marincola, "A novel cytolysis assay using
fluorescent labeling and quantitative fluorescent scanning technology," Journal of
Immunological Methods, vol. 226, pp. 29-41, 6/24/ 1999.
[42] K. Elineni and N. Gallant, "Microtubules Mechanically Regulate Cell Adhesion Strengthening Via
Cell Shape," Cellular and Molecular Bioengineering, vol. 7, pp. 136-144, 2014/03/01 2014.
[43] M. Incorporated. (04/03/2020). Sonicator 3000 Ultrasonic Liquid Processor Operation Manual.
Available: https://archive-resources.coleparmer.com/Manual_pdfs/04711-xxMisonix3000.pdf
[44] P. Y. Chiou, A. T. Ohta, and M. C. Wu, "Massively parallel manipulation of single cells and
microparticles using optical images," Nature, vol. 436, pp. 370-2, Jul 21 2005.
[45] M. L. Jewell, N. J. Solish, and C. S. Desilets, "Noninvasive body sculpting technologies with an
emphasis on high-intensity focused ultrasound," Aesthetic Plast Surg, vol. 35, pp. 901-12, Oct
2011.
[46] R. A. Wong, B. Schumann, R. Townsend, and C. A. Phelps, "A survey of therapeutic ultrasound
use by physical therapists who are orthopaedic certified specialists," Phys Ther, vol. 87, pp. 986-
94, Aug 2007.
[47] M. Wiklund, "Acoustofluidics 12: Biocompatibility and cell viability in microfluidic acoustic
resonators," Lab Chip, vol. 12, pp. 2018-28, May 8 2012.
[48] D. Debavelaere-Callens, L. Peyre, P. Campistron, and H. F. Hildebrand, "On the use of
ultrasounds to quantify the longitudinal threshold force to detach osteoblastic cells from a
conditioned glass substrate," Biomolecular Engineering, vol. 24, pp. 521-525, 11// 2007.
[49] N. J. Hangiandreou, "AAPM/RSNA physics tutorial for residents: Topics in US - B-mode US: Basic
concepts and new technology," Radiographics, vol. 23, pp. 1019-1033, Jul-Aug 2003.
[50] F. J. Trujillo and K. Knoerzer, "A computational modeling approach of the jet-like acoustic
streaming and heat generation induced by low frequency high power ultrasonic horn reactors,"
Ultrasonics Sonochemistry, vol. 18, pp. 1263-1273, 2011/11/01/ 2011.
[51] I. Lentacker, I. De Cock, R. Deckers, S. C. De Smedt, and C. T. Moonen, "Understanding
ultrasound induced sonoporation: definitions and underlying mechanisms," Adv Drug Deliv Rev,
vol. 72, pp. 49-64, Jun 2014.
[52] I. Tzanakis, G. S. B. Lebon, D. G. Eskin, and K. A. Pericleous, "Characterizing the cavitation
development and acoustic spectrum in various liquids," Ultrasonics Sonochemistry, vol. 34, pp.
651-662, 2017/01/01/ 2017.
[53] R. S. (London), Philosophical Transactions of the Roday Society of London Volume 121, 1831.

73
[54] Thermofisher. (2020). CellTracker fluorescent probes Product Information Sheet. Available:
https://assets.thermofisher.com/TFS-
Assets/LSG/manuals/MAN0001826_CellTracker_Probes_PI.pdf
[55] Thermofisher. LIVE/DEAD® Viability/Cytotoxicity Kit. Available:
https://assets.thermofisher.com/TFS-Assets/LSG/manuals/mp03224.pdf
[56] W. Xu, R. Mezencev, B. Kim, L. Wang, J. McDonald, and T. Sulchek, "Cell Stiffness Is a Biomarker
of the Metastatic Potential of Ovarian Cancer Cells," PLoS ONE, vol. 7, p. e46609, 2012.
[57] J. M. MITCHISON and M. M. SWANN, "The Mechanical Properties of the cell Surface," II. The
Unfertilized Sea-Urchin Egg, vol. 31, pp. 461-472, 1954.
[58] R. P. Rand and A. C. Burton, "Mechanical Properties of the Red Cell Membrane," Biophysical
Journal, vol. 4, pp. 115-135, 1964/03/01 1964.
[59] C. Dong, R. Skalak, K. L. Sung, G. W. Schmid-Schonbein, and S. Chien, "Passive deformation
analysis of human leukocytes," Journal of Biomechanical Engineering, vol. 110, pp. 27-41, 1988.
[60] A. Ashkin, "Acceleration and Trapping of Particles by Radiation Pressure," Physical Review
Letters, vol. 24, pp. 156-159, 01/26/ 1970.
[61] A. Ashkin and J. Dziedzic, "Optical trapping and manipulation of viruses and bacteria," Science,
vol. 235, pp. 1517-1520, 1987.
[62] J. Guck, R. Ananthakrishnan, T. J. Moon, C. C. Cunningham, and J. Käs, "Optical Deformability of
Soft Biological Dielectrics," Physical Review Letters, vol. 84, pp. 5451-5454, 06/05/ 2000.
[63] B. Jagannathan and S. Marqusee, "Protein Folding and Unfolding Under Force," Biopolymers, vol.
99, pp. 860-869, 2013.
[64] N. Wang, J. P. Butler, and D. E. Ingber, "Mechanotransduction across the cell surface and
through the cytoskeleton," Science, vol. 260, pp. 1124-7, May 21 1993.
[65] G. Binnig, C. F. Quate, and C. Gerber, "Atomic force microscope," Physical review letters, vol. 56,
p. 930, 1986.
[66] J. Helenius, C.-P. Heisenberg, H. E. Gaub, and D. J. Muller, "Single-cell force spectroscopy,"
Journal of Cell Science, vol. 121, pp. 1785-1791, June 1, 2008 2008.
[67] J. Friedrichs, K. R. Legate, R. Schubert, M. Bharadwaj, C. Werner, D. J. Muller, et al., "A practical
guide to quantify cell adhesion using single-cell force spectroscopy," Methods, vol. 60, pp. 169-
78, Apr 1 2013.
[68] H. Xie, M. Yin, W. Rong, and L. Sun, "In Situ Quantification of Living Cell Adhesion Forces: Single
Cell Force Spectroscopy with a Nanotweezer," Langmuir, vol. 30, pp. 2952-2959, 2014/03/18
2014.
[69] P. V. Zinin, J. S. Allen, and V. M. Levin, "Mechanical resonances of bacteria cells," Physical Review
E, vol. 72, p. 061907, 12/12/ 2005.
[70] M. M. Zarandi, A. Bonakdar, and I. Stiharu, "Investigations on natural frequencies of individual
spherical and ellipsoidal bakery yeast cells."
[71] H. Wee and A. Voloshin, "Modal analysis of a spreading osteoblast cell in culturing," in 2012
38th Annual Northeast Bioengineering Conference (NEBEC), 2012, pp. 167-168.
[72] S. K. Jaganathan, A. P. Subramanian, M. V. Vellayappan, A. Balaji, A. A. John, A. K. Jaganathan, et
al., "Natural frequency of cancer cells as a starting point in cancer treatment," Current Science,
vol. 110, pp. 1828-1832, 2016.
[73] A. Geltmeier, B. Rinner, D. Bade, K. Meditz, R. Witt, U. Bicker, et al., "Characterization of
Dynamic Behaviour of MCF7 and MCF10A Cells in Ultrasonic Field Using Modal and Harmonic
Analyses," PLOS ONE, vol. 10, p. e0134999, 2015.
[74] T. Fukuda, M. Nakajima, H. Tajima, S. Yajing, and Y. Tao, "Micro-nanomanipulation system
toward biological cell analysis and assembly," in Innovative Engineering Systems (ICIES), 2012
First International Conference on, 2012, pp. 31-36.

74
[75] T. Fukuda, M. Nakajima, L. Pou, and M. R. Ahmad, "Bringing the nanolaboratory inside electron
microscopes," Nanotechnology Magazine, IEEE, vol. 2, pp. 18-31, 2008.
[76] Kranthi K. Elineni and Nathan D. Gallant, "Regulation of Cell Adhesion Strength by Peripheral
Focal Adhesion Distribution," Biophysical Journal, vol. 101, pp. 2903-2911, 05/30/received
11/09/accepted 2011.
[77] D. W. Dumbauld, H. Shin, N. D. Gallant, K. E. Michael, H. Radhakrishna, and A. J. Garcia,
"Contractility modulates cell adhesion strengthening through focal adhesion kinase and
assembly of vinculin-containing focal adhesions," J Cell Physiol, vol. 223, pp. 746-56, Jun 2010.
[78] N. A. Burnham, X. Chen, C. S. Hodges, G. A. Matei, E. J. Thoreson, C. J. Roberts, et al.,
"Comparison of calibration methods for atomic-force microscopy cantilevers," Nanotechnology,
vol. 14, p. 1, 2003.
[79] S. M. Cook, T. E. Schäffer, K. M. Chynoweth, M. Wigton, R. W. Simmonds, and K. M. Lang,
"Practical implementation of dynamic methods for measuring atomic force microscope
cantilever spring constants," Nanotechnology, vol. 17, p. 2135, 2006.
[80] I. M. Malovichko, "Measuring AFM cantilever stiffness from a thermal noise spectrum," Bulletin
of the Russian Academy of Sciences: Physics, vol. 77, pp. 972-974, 2013/08/01 2013.
[81] Y. Shen, M. Nakajima, Z. Yang, H. Tajima, Z. Najdovski, M. Homma, et al., "Single cell stiffness
measurement at various humidity conditions by nanomanipulation of a nano-needle,"
Nanotechnology, vol. 24, p. 145703, Apr 12 2013.
[82] T. Ishizaki, M. Uehata, I. Tamechika, J. Keel, K. Nonomura, M. Maekawa, et al., "Pharmacological
properties of Y-27632, a specific inhibitor of rho-associated kinases," Mol Pharmacol, vol. 57, pp.
976-83, May 2000.
[83] O. O. Akintewe, S. J. DuPont, K. K. Elineni, M. C. Cross, R. G. Toomey, and N. D. Gallant, "Shape-
changing hydrogel surfaces trigger rapid release of patterned tissue modules," Acta
Biomaterialia.
[84] F. Palmieri and M. Klingenberg, "Inhibition of Respiration under the Control of Azide Uptake by
Mitochondria," in European Journal of Biochemistry, C. Liébecq, Ed., ed: Springer Berlin
Heidelberg, 1967, pp. 439-446.

75
 Appendix A Copyright Permissions
There are no copyright permissions needed.

76